1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
10 HS/01 Handle trailing backslash gracefully. (CVE-2019-15846)
16 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917, OVE-20190718-0006)
22 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
23 buffer overrun for (non-chunking) other transports.
25 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
26 TLS1.3, means that a server rejecting a client certificate is not visible
27 to the client until the first read of encrypted data (typically the
28 response to EHLO). Add detection for that case and treat it as a failed
29 TLS connection attempt, so that the normal retry-in-clear can work (if
32 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
33 and/or domain. Found and fixed by Jason Betts.
35 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
36 configuration). If a CNAME target was not a wellformed name pattern, a
39 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
40 the OS reports them interleaved with other addresses.
42 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
43 used both for input and for a verify callout, both encrypted, SMTP
44 responses being sent by the server could be lost. This resulted in
45 dropped connections and sometimes bounces generated by a peer sending
48 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
49 string. Previously it was possible to cause undefined behaviour in a
50 library routine (usually a crash). Found by "zerons".
52 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
55 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
56 API was removed, so update to use the newer ones.
59 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
60 to changes apparently associated with TLS1.3 handling some of the APIs
61 previously used were either nonfunctional or inappropriate. Strings
62 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
63 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
64 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
65 This affects log line X= elements, the $tls_{in,out}_cipher variables,
66 and the use of specific cipher names in the encrypted= ACL condition.
68 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
69 verification result was not updated unless hosts_require_ocsp applied.
71 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
72 queue_list_requires_admin set to false, non-admin users were denied the
75 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
76 directory-of-certs mode. Previously they were advertised despite the
79 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
80 success-DSN messages. Previously the From: header was always the default
81 one for these; the option was ignored.
87 JH/01 Remove code calling the customisable local_scan function, unless a new
88 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
90 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
91 non-signal-safe functions being used.
93 JH/03 Bug 2269: When presented with a received message having a stupidly large
94 number of DKIM-Signature headers, disable DKIM verification to avoid
95 a resource-consumption attack. The limit is set at twenty.
97 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
98 report of oldest_pass in ${authres } in consequence, and separate out
99 some descriptions of reasons for verification fail.
101 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
102 files in the spool were present and unlocked. A queue-runner could spot
103 them, resulting in a duplicate delivery. Fix that by doing the unlock
104 after the unlink. Investigation by Tim Stewart. Take the opportunity to
105 add more error-checking on spoolfile handling while that code is being
108 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
109 No known attacks, no CVE, this is defensive hardening.
111 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
112 a queue-runner could start a delivery while other operations were ongoing.
113 Cutthrough delivery was a common victim, resulting in duplicate delivery.
114 Found and investigated by Tim Stewart. Fix by using the open message data
115 file handle rather than opening another, and not locally closing it (which
116 releases a lock) for that case, while creating the temporary .eml format
117 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
119 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
120 $sender_verify_failure/$recipient_verify_failure to "random".
122 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
125 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
126 Previously this would segfault.
128 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
131 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
132 like zero, since the resolver should be doing this for us, But we need one
133 as a CNAME but no MX presence gets the CNAME returned; we need to check
134 that doesn't point to an MX to declare it "no MX returned" rather than
135 "error, loop". A new main option is added so the older capability of
136 following some limited number of chain links is maintained.
138 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
140 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
141 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
142 now-deprecated earlier definitions used only specified the range up to TLS
143 1.2 (in the older-version library docs).
145 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
147 JH/15 Rework TLS client-side context management. Stop using a global, and
148 explicitly pass a context around. This enables future use of TLS for
149 connections to service-daemons (eg. malware scanning) while a client smtp
150 connection is using TLS; with cutthrough connections this is quite likely.
152 JH/16 Fix ARC verification to do AS checks in reverse order.
154 JH/17 Support a "tls" option on the ${readsocket } expansion item.
156 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
157 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
158 Previously the "utf8" would be re-prepended for every additional message.
160 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
161 Previously thery were accepted, resulting in issues when attempting to
162 forward messages to a non-supporting MTA.
164 PP/02 Let -n work with printing macros too, not just options.
166 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
167 one parent address was copied, and bogus data was used at delivery-logging
168 time. Either a crash (after delivery) or bogus log data could result.
169 Discovery and analysis by Tim Stewart.
171 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
172 Previously if the string ended mid-character, we did not insert the
173 promised '?' replacement.
175 PP/04 Documentation: current string operators work on bytes, not codepoints.
177 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
178 should pack well giving a smaller memory footprint so better caching and
179 therefore performance. Group the declarations where this can't be done so
180 that the byte-sized flag variables are not interspersed among pointer
181 variables, giving a better chance of good packing by the compiler.
183 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
184 non-null, to avoid issues with sites running BATV. Previously reports were
185 sent with an empty envelope sender so looked like bounces.
187 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
188 The ignore_error flag wasn't being returned from the filter subprocess so
189 was not set for later routers. Investigation and fix by Matthias Kurz.
191 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
192 and a msg:complete for the whole, when a message is manually removed using
193 -Mrm. Developement by Matthias Kurz, hacked on by JH.
195 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
196 a "Gnu special" function, asprintf() in the DB utility binary builds; I
197 hope that is portable enough.
199 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
200 requiring a known-CA anchor certificate; make it now rely entirely on the
201 TLSA as an anchor. Checking the name on the leaf cert against the name
202 on the A-record for the host is still done for TA (but not for EE mode).
204 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
205 included in delivery lines for non-proxied connections, when compiled with
206 SUPPORT_SOCKS and running with proxy logging enabled.
208 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
209 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
210 move the existing event to fire before the normal logging of message
211 failure so that custom logging is bracketed by normal logging.
213 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
214 msg:fail:internal event. Developement by Matthias Kurz.
216 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
217 far too small for todays use of crypto signatures stored there. Go all
218 the way to the max DNS message size of 64kB, even though this might be
219 overmuch for IOT constrained device use.
221 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
222 copy a string over itself. The library routine is documented as not
223 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
225 JH/32 For main options check_spool_space and check_inode_space, where the
226 platform supports 64b integers, support more than the previous 2^31 kB
227 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
228 the previous G, M, k.
230 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
231 $authenticated_fail_id variable on authentication failure. Previously
234 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
235 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
236 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
237 GNUTLS_SEC_PARAM_MEDIUM.
239 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
240 side. Previously we would continue as if no SNI had been received.
242 JH/36 Harden the handling of string-lists. When a list consisted of a sole
243 "<" character, which should be a list-separator specification, we walked
244 off past the nul-terimation.
246 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
247 causes) even when the retry time is not yet met. Previously they were
248 not, meaning that when (say) an account was over-quota and temp-rejecting,
249 and multiple senders' messages were queued, only one sender would get
250 notified on each configured delay_warning cycle.
252 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
254 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
255 cipher-suites, an error can be left on the stack even for a succeeding
256 accept; this results in impossible error messages when a later operation
259 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
260 return error codes indicating retry. Under TLS1.3 this becomes required.
262 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
263 it only wrote the new authenticators, resulting in a lack of tracking of
264 peer changes of ESMTP extensions until the next cache flush.
266 JH/41 Fix the loop reading a message header line to check for integer overflow,
267 and more-often against header_maxsize. Previously a crafted message could
268 induce a crash of the recive process; now the message is cleanly rejected.
270 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
271 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
277 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
278 When redis_servers is set to a list of > 1 element, and the Redis servers
279 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
280 case of MOVED into a DEFER case instead, thus moving the query onto the
281 next server in the list. For a cluster of N elements, all N servers must
282 be defined in redis_servers.
284 GF/02 Catch and remove uninitialized value warning in exiqsumm
285 Check for existence of @ARGV before looking at $ARGV[0]
287 JH/01 Replace the store_release() internal interface with store_newblock(),
288 which internalises the check required to safely use the old one, plus
289 the allocate and data copy operations duplicated in both (!) of the
290 extant use locations.
292 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
293 modifier. This matches the restriction on the commandline.
295 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
296 Previously only the last row was returned.
298 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
299 we assumed that tags in the header were well-formed, and parsed the
300 element content after inspecting only the first char of the tag.
301 Assumptions at that stage could crash the receive process on malformed
304 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
305 While running the DKIM ACL we operate on the Permanent memory pool so that
306 variables created with "set" persist to the DATA ACL. Also (at any time)
307 DNS lookups that fail create cache records using the Permanent pool. But
308 expansions release any allocations made on the current pool - so a dnsdb
309 lookup expansion done in the DKIM ACL releases the memory used for the
310 DNS negative-cache, and bad things result. Solution is to switch to the
311 Main pool for expansions.
312 While we're in that code, add checks on the DNS cache during store_reset,
313 active in the testsuite.
314 Problem spotted, and debugging aided, by Wolfgang Breyha.
316 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
317 When none of the hosts presented to a transport match an already-open
318 connection, close it and proceed with the list. Previously we would
319 queue the message. Spotted by Lena with Yahoo, probably involving
322 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
323 Previously a spurious "250 OK id=" response was appended to the proper
326 JH/08 The "support for" informational output now, which built with Content
327 Scanning support, has a line for the malware scanner interfaces compiled
328 in. Interface can be individually included or not at build time.
330 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
331 by the template makefile "src/EDITME". The "STREAM" support for an older
332 ClamAV interface method is removed.
334 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
335 rows affected is given instead).
337 JH/11 The runtime Berkeley DB library version is now additionally output by
338 "exim -d -bV". Previously only the compile-time version was shown.
340 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
341 SMTP connection. Previously, when one had more receipients than the
342 first, an abortive onward connection was made. Move to full support for
343 multiple onward connections in sequence, handling cutthrough connection
344 for all multi-message initiating connections.
346 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
347 routers. Previously, a multi-recipient message would fail to match the
348 onward-connection opened for the first recipient, and cause its closure.
350 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
351 a timeout on read on a GnuTLS initiating connection, resulting in the
352 initiating connection being dropped. This mattered most when the callout
353 was marked defer_ok. Fix to keep the two timeout-detection methods
356 JH/15 Relax results from ACL control request to enable cutthrough, in
357 unsupported situations, from error to silently (except under debug)
358 ignoring. This covers use with PRDR, frozen messages, queue-only and
361 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
363 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
364 metadata, resulting in a crash in free().
366 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
367 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
368 Broken also in d185889f4, with init system revamp.
370 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
371 Previously we abruptly closed the connection after reading a malware-
372 found indication; now we go on to read the "scan ok" response line,
375 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
376 ACL. Previously, a crash would result.
378 JH/19 Speed up macro lookups during configuration file read, by skipping non-
379 macro text after a replacement (previously it was only once per line) and
380 by skipping builtin macros when searching for an uppercase lead character.
382 JH/20 DANE support moved from Experimental to mainline. The Makefile control
383 for the build is renamed.
385 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
386 was allocated for every new TLS startup, meaning one per message. Fix
387 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
389 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
390 reported the original. Fix to report (as far as possible) the ACL
391 result replacing the original.
393 JH/23 Fix memory leak during multi-message connections using STARTTLS under
394 OpenSSL. Certificate information is loaded for every new TLS startup,
395 and the resources needed to be freed.
397 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
399 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
402 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
403 DATA response info to the (existing) per-recipient response info for
404 the "C=" log element. It can have useful tracking info from the
405 destination system. Patch from Simon Arlott.
407 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
408 length value. Previously this would segfault.
410 HS/02 Support Avast multiline protoocol, this allows passing flags to
411 newer versions of the scanner.
413 JH/28 Ensure that variables possibly set during message acceptance are marked
414 dead before release of memory in the daemon loop. This stops complaints
415 about them when the debug_store option is enabled. Discovered specifically
416 for sender_rate_period, but applies to a whole set of variables.
417 Do the same for the queue-runner and queue-list loops, for variables set
418 from spool message files. Do the same for the SMTP per-message loop, for
419 certain variables indirectly set in ACL operations.
421 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
422 as a multi-recipient message from a mailinglist manager). The coding had
423 an arbitrary cutoff number of characters while checking for more input;
424 enforced by writing a NUL into the buffer. This corrupted long / fast
425 input. The problem was exposed more widely when more pipelineing of SMTP
426 responses was introduced, and one Exim system was feeding another.
427 The symptom is log complaints of SMTP syntax error (NUL chars) on the
428 receiving system, and refused recipients seen by the sending system
429 (propating to people being dropped from mailing lists).
430 Discovered and pinpointed by David Carter.
432 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
433 replaced by the ${authresults } expansion.
435 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
437 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
438 allows proper process termination in container environments.
440 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
441 Previously the "final dot" had a newline after it; ensure it is CR,LF.
443 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
444 and "err_perm", deprecated since 4.83 when the RFC-defined words
445 "temperror" and "permerror" were introduced.
447 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
448 transport-filters or DKIM-signing. The restriction was lost in the
449 consolidation of verify-callout and delivery SMTP handling.
450 Extend the restriction to also cover ARC-signing.
452 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
453 in defer=pass mode supply a 450 to the initiator. Previously the message
456 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
457 tls_require_ciphers is used as before.
459 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
460 "pass_unscanned". Only tmpfails from the scanner are written to
461 the paniclog, as they may require admin intervention (permission
462 denied, license issues). Other scanner errors (like decompression
463 bombs) do not cause a paniclog entry.
465 JH/36 Fix reinitialisation of DKIM logging variable between messages.
466 Previously it was possible to log spurious information in receive log
469 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
470 triggered odd behaviour from Outlook Express clients.
472 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
475 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
476 since the IETF WG has not yet settled on that versus the original
477 "bare" representation.
479 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
480 Previously the millisecond value corrupted the output.
481 Fix also for syslog_pid=no and log_selector +pid, for which the pid
482 corrupted the output.
488 JH/01 Rework error string handling in TLS interface so that the caller in
489 more cases is responsible for logging. This permits library-sourced
490 string to be attached to addresses during delivery, and collapses
491 pairs of long lines into single ones.
493 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
494 during configuration. Wildcards are allowed and expanded.
496 JH/02 Rework error string handling in DKIM to pass more info back to callers.
497 This permits better logging.
499 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
500 do not close it down and have the child transport start it up again on
501 the passed-on TCP connection. Instead, proxy the child (and any
502 subsequent ones) for TLS via a unix-domain socket channel. Logging is
503 affected: the continued delivery log lines do not have any DNSSEC, TLS
504 Certificate or OCSP information. TLS cipher information is still logged.
506 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
507 identical IP addresses on different listening ports. Will also affect
510 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
511 add noisy ifdef guards to special-case this sillyness.
512 Patch from Bernd Kuhls.
514 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
515 than 255 are no longer allowed.
517 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
518 Disable the session-cache too, which might reduce our load. Since we
519 currrectly use a new context for every connection, both as server and
520 client, there is no benefit for these.
521 GnuTLS appears to not support tickets server-side by default (we don't
522 call gnutls_session_ticket_enable_server()) but client side is enabled
523 by default on recent versions (3.1.3 +) unless the PFS priority string
526 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
527 <https://reproducible-builds.org/specs/source-date-epoch/>.
529 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
530 the check for any unsuccessful recipients did not notice the limit, and
531 erroneously found still-pending ones.
533 JH/08 Pipeline CHUNKING command and data together, on kernels that support
534 MSG_MORE. Only in-clear (not on TLS connections).
536 JH/09 Avoid using a temporary file during transport using dkim. Unless a
537 transport-filter is involved we can buffer the headers in memory for
538 creating the signature, and read the spool data file once for the
539 signature and again for transmission.
541 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
542 4.77 as the kernel support then wasn't solid, having issues in 64bit
543 mode. Now, it's been long enough. Add support for FreeBSD also.
545 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
546 case where the routing stage had gathered several addresses to send to
547 a host before calling the transport for the first, we previously failed
548 to close down TLS in the old transport process before passing the TCP
549 connection to the new process. The new one sent a STARTTLS command
550 which naturally failed, giving a failed delivery and bloating the retry
551 database. Investigation and fix prototype from Wolfgang Breyha.
553 JH/12 Fix check on SMTP command input synchronisation. Previously there were
554 false-negatives in the check that the sender had not preempted a response
555 or prompt from Exim (running as a server), due to that code's lack of
556 awareness of the SMTP input buffering.
558 PP/04 Add commandline_checks_require_admin option.
559 Exim drops privileges sanely, various checks such as -be aren't a
560 security problem, as long as you trust local users with access to their
561 own account. When invoked by services which pass untrusted data to
562 Exim, this might be an issue. Set this option in main configuration
563 AND make fixes to the calling application, such as using `--` to stop
566 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
567 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
568 responses to those, into a single TLS record each way (this usually means
569 a single packet). As a side issue, smtp_enforce_sync now works on TLS
572 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
573 affects you only if you're dancing at the edge of the param size limits.
574 If you are, and this message makes sense to you, then: raise the
575 configured limit or use OpenSSL 1.1. Nothing we can do for older
578 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
579 cmdline element to get the documented default one. Previously it was
582 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
583 are unsuable for non-message-specific reasons - eg. connection timeout,
584 banner-time rejection.
586 JH/16 Fix logging of delivery remote port, when specified by router, under
589 PP/06 Repair manualroute's ability to take options in any order, even if one
590 is the name of a transport.
593 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
595 JH/17 Change the list-building routines interface to use the expanding-string
596 triplet model, for better allocation and copying behaviour.
598 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
599 Previously it was constructed the first time a possibly-matching string
600 was met in the configuration file input during startup; now it is done
603 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
604 compatible one, to avoid the (poorly documented) possibility of a config
605 file in the working directory redirecting the DB files, possibly correpting
606 some existing file. CVE-2017-10140 assigned for BDB.
608 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
609 cache-hot. Previously, although the result was properly cached, the
610 initial verify call returned a defer.
612 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
613 the main verify for receipient in uncached-mode.
615 JH/22 Retire historical build files to an "unsupported" subdir. These are
616 defined as "ones for which we have no current evidence of testing".
618 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
619 if present. Previously it was ignored.
621 JH/24 Start using specified-initialisers in C structure init coding. This is
622 a C99 feature (it's 2017, so now considered safe).
624 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
625 if was a fixed-sized field and bitmask ops via macros; it is now more
628 PP/07 GitHub PR 56: Apply MariaDB build fix.
629 Patch provided by Jaroslav Škarvada.
631 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
632 during Coverity cleanups [4.87 JH/47]
633 Diagnosis and fix provided by Michael Fischer v. Mollard.
635 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
636 the right size to place the terminating semicolon on its own folded
637 line, the header hash was calculated to an incorrect value thanks to
638 the (relaxed) space the fold became.
640 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
641 and confused the parent.
643 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
644 which could crash as a result. This could lead to undeliverable messages.
646 JH/28 Logging: "next input sent too soon" now shows where input was truncated
649 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
650 matters on fast-turnover and PID-randomising systems, which were getting
651 out-of-order delivery.
653 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
654 a possibly-overlapping copy. The symptom was that "Remote host closed
655 connection in response to HELO" was logged instead of the actual 4xx
658 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
659 Previously only that bufferd was discarded, resulting in SYMTP command
662 JH/32 DKIM: when a message has multiple signatures matching an identity given
663 in dkim_verify_signers, run the dkim acl once for each. Previously only
664 one run was done. Bug 2189.
666 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
667 "panic the current process" to "deliberately defer". The panic log is
668 still written with the problem list name; the mail and reject logs now
669 get a temp-reject line for the message that was being handled, saying
670 something like "domains check lookup or other defer". The SMTP 451
671 message is still "Temporary local problem".
673 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
674 A crafted sequence of BDAT commands could result in in-use memory beeing
675 freed. CVE-2017-16943.
677 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
678 from SMTP input. Previously it was always done; now only done for DATA
679 and not BDAT commands. CVE-2017-16944.
681 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
682 to the message (such as an overlong header line). Previously this was
683 not done and we did not exit BDAT mode. Followon from the previous item
684 though a different problem.
690 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
691 than -2003 did; needs libidn2 in addition to libidn.
693 JH/02 The path option on a pipe transport is now expanded before use.
695 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
696 Patch provided by "Björn", documentation fix added too.
698 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
699 missing a wire-to-host endian conversion.
701 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
702 close after a BDAT command line could be taken as a following command,
703 giving a synch failure. Fix by only checking for synch immediately
704 before acknowledging the chunk.
706 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
707 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
708 macro. Patches provided by Josh Soref.
710 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
711 Previously we did not; the RFC seems ambiguous and VRFY is not listed
712 by IANA as a service extension. However, John Klensin suggests that we
715 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
716 the dkim code may be unix-mode line endings rather than smtp wire-format
717 CRLF, so prepend a CR to any bare LF.
719 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
720 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
722 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
723 the first were themselves being wrongly included in the feed into dkim
724 processing; with most chunk sizes in use this resulted in an incorrect
725 body hash calculated value.
727 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
728 DKIM signature block, for verification. Although advised against by
729 standards it is specifically not ruled illegal.
731 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
733 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
734 missing a body hash (the bh= tag).
736 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
737 It seems that HAProxy sends the Proxy Protocol information in clear and
738 only then does a TLS startup, so do the same.
740 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
741 TCP connections (such as for Spamd) unless the daemon successfully set
742 Fast Open mode on its listening sockets. This fixes breakage seen on
743 too-old kernels or those not configured for Fast Open, at the cost of
744 requiring both directions being enabled for TFO, and TFO never being used
745 by non-daemon-related Exim processes.
747 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
748 endings, at least on the first header line. Try to canonify any that get
749 past that check, despite the cost.
751 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
752 now limited to an arbitrary five deep, while parsing addresses with the
753 strip_excess_angle_brackets option enabled.
755 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
756 instead leave the unprompted TLS handshake in socket buffer for the
757 TLS library to consume.
759 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
761 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
763 JH/16 Drop variables when they go out of scope. Memory management drops a whole
764 region in one operation, for speed, and this leaves assigned pointers
765 dangling. Add checks run only under the testsuite which checks all
766 variables at a store-reset and panics on a dangling pointer; add code
767 explicitly nulling out all the variables discovered. Fixes one known
768 bug: a transport crash, where a dangling pointer for $sending_ip_address
769 originally assigned in a verify callout, is re-used.
771 PP/06 Drop '.' from @INC in various Perl scripts.
773 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
775 PP/08 Reduce a number of compilation warnings under clang; building with
776 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
777 should be warning-free.
779 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
781 HS/01 Fix portability problems introduced by PP/08 for platforms where
782 realloc(NULL) is not equivalent to malloc() [SunOS et al].
784 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
785 chunk. This allows us to accept broken chunked messages. We need a more
786 general solution here.
788 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
789 already-broken messages in the queue.
791 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
793 JH/19 Fix reference counting bug in routing-generated-address tracking.
799 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
800 supports it and a size is available (ie. the sending peer gave us one).
802 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
803 years of being deprecated). The replacements are the ACLs
804 acl_smtp_mime and acl_not_smtp_mime.
806 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
807 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
808 or even an in-clear connection were permitted. Now, if the host lookup
809 was dnssec and dane was requested then the host is only used if the
810 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
811 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
812 if one fails this test.
813 This means that a poorly-configured remote DNS will make it incommunicado;
814 but it protects against a DNS-interception attack on it.
816 JH/04 Bug 1810: make continued-use of an open smtp transport connection
817 non-noisy when a race steals the message being considered.
819 JH/05 If main configuration option tls_certificate is unset, generate a
820 self-signed certificate for inbound TLS connections.
822 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
823 in rewrites and routers.
825 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
826 and logged a warning sing 4.83; now they are a configuration file error.
828 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
829 (lacking @domain). Apply the same qualification processing as RCPT.
831 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
833 JH/10 Support ${sha256:} applied to a string (as well as the previous
836 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
837 a cutthrough deliver is pending, as we always want to make a connection.
838 This also avoids re-routing the message when later placing the cutthrough
839 connection after a verify cache hit.
840 Do not update it with the verify result either.
842 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
843 when routing results in more than one destination address.
845 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
846 signing (which inhibits the cutthrough capability). Previously only
847 the presence of an option was tested; now an expansion evaluating as
848 empty is permissible (obviously it should depend only on data available
849 when the cutthrough connection is made).
851 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
852 the relevant preceding SMTP command did not note the pipelining mode.
854 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
855 Previously they were not counted.
857 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
858 as one having no matching records. Previously we deferred the message
859 that needed the lookup.
861 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
862 distinguished as "(=".
864 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
865 for missing MX records. Previously it only worked for missing A records.
867 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
869 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
870 after the data-go-ahead and data-ack. Patch from Jason Betts.
872 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
873 even for a "none" policy. Patch from Tony Meyer.
875 JH/22 Fix continued use of a connection for further deliveries. If a port was
876 specified by a router, it must also match for the delivery to be
879 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
880 When one of the recipients of a message was unsuitable for the connection
881 (has no matching addresses), we lost track of needing to mark it
882 deferred. As a result mail would be lost.
884 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
886 JH/25 Decoding ACL controls is now done using a binary search; the source code
887 takes up less space and should be simpler to maintain. Merge the ACL
888 condition decode tables also, with similar effect.
890 JH/26 Fix problem with one_time used on a redirect router which returned the
891 parent address unchanged. A retry would see the parent address marked as
892 delivered, so not attempt the (identical) child. As a result mail would
895 JH/27 Fix a possible security hole, wherein a process operating with the Exim
896 UID can gain a root shell. Credit to http://www.halfdog.net/ for
897 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
900 JH/28 Enable {spool,log} filesystem space and inode checks as default.
901 Main config options check_{log,spool}_{inodes,space} are now
902 100 inodes, 10MB unless set otherwise in the configuration.
904 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
905 Previously it only applied to the main-section connection policy
908 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
910 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
911 by me. Added RFC7919 DH primes as an alternative.
913 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
914 are not in the system include path.
916 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
917 GnuTLS, when a session startup failed (eg because the client disconnected)
918 Exim did stdio operations after fclose. This was exposed by a recent
919 change which nulled out the file handle after the fclose.
921 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
922 signed directly by the cert-signing cert, rather than an intermediate
923 OCSP-signing cert. This is the model used by LetsEncrypt.
925 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
927 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
928 an incoming connection.
930 HS/02 Bug 1802: Do not half-close the connection after sending a request
933 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
934 fallback to "prime256v1".
936 JH/34 SECURITY: Use proper copy of DATA command in error message.
937 Could leak key material. Remotely exploitable. CVE-2016-9963.
943 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
944 and 3.4.4 - once the server is enabled to respond to an OCSP request
945 it does even when not requested, resulting in a stapling non-aware
946 client dropping the TLS connection.
948 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
949 support variable-length bit vectors. No functional change.
951 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
952 The I= interface field on outgoing lines is now after the H= remote
953 host field, same as incoming lines. There is a separate
954 outgoing_interface log selector which allows you to disable the
957 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
958 If not running log_selector +smtp_connection the mainlog would be held
959 open indefinitely after a "too many connections" event, including to a
960 deleted file after a log rotate. Leave the per net connection logging
961 leaving it open for efficiency as that will be quickly detected by the
962 check on the next write.
964 HS/01 Bug 1671: Fix post transport crash.
965 Processing the wait-<transport> messages could crash the delivery
966 process if the message IDs didn't exist for some reason. When
967 using 'split_spool_directory=yes' the construction of the spool
968 file name failed already, exposing the same netto behaviour.
970 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
971 mime_regex ACL conditions.
973 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
974 to DSN fail messages (bounces): remote IP, remote greeting, remote response
975 to HELO, local diagnostic string.
977 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
978 log line to debug. Even when configured with a tls authenticator many
979 client connections are expected to not authenticate in this way, so
980 an authenticate fail is not an error.
982 HS/02 Add the Exim version string to the process info. This way exiwhat
983 gives some more detail about the running daemon.
985 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
986 matter for fast-change records such as DNSBLs.
988 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
989 retry record, even if constant. There may be multiple transports with
990 different interface settings and the retry behaviour needs to be kept
993 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
995 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
997 JH/10 Bug 840: fix log_defer_output option of pipe transport
999 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1000 hosts_randomize. This matters a lot when combined with mua_wrapper.
1002 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1003 ${quote_pgsql:<string>} operator.
1005 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1007 JH/14 Change header file name for freeradius-client. Relevant if compiling
1008 with Radius support; from the Gentoo tree and checked under Fedora.
1010 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1012 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1013 option transport_filter, meaning no filtering.
1015 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1017 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1018 defaults to "*" (all hosts). The variable is now available when not built
1019 with TLS, default unset, mainly to enable keeping the testsuite sane.
1020 If a server certificate is not supplied (via tls_certificate) an error is
1021 logged, and clients will find TLS connections fail on startup. Presumably
1022 they will retry in-clear.
1023 Packagers of Exim are strongly encouraged to create a server certificate
1024 at installation time.
1026 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1027 with the $config_file variable.
1029 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1030 in transport context, after the attempt, and per-recipient. The latter type
1031 is per host attempted. The event data is the error message, and the errno
1032 information encodes the lookup type (A vs. MX) used for the (first) host,
1033 and the trailing two digits of the smtp 4xx response.
1035 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1036 to write to mainlog (or rejectlog, paniclog) in the window between file
1037 creation and permissions/ownership being changed. Particularly affects
1038 installations where exicyclog is run as root, rather than exim user;
1039 result is that the running daemon panics and dies.
1041 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1043 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1044 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1045 "pri" and "weight". Note that the previous implicit priority given by the
1046 list order is no longer honoured.
1048 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1049 for DKIM processing.
1051 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1052 by defining SUPPORT_SOCKS.
1054 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1055 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1056 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1057 variables are renamed to proxy_{local,external}_{address,port}.
1059 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1060 for a build by defining SUPPORT_I18N
1062 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1063 of the query string, and make ${quote_redis:} do that quoting.
1065 JH/29 Move Events support from Experimental to mainline, enabled by default
1066 and removable for a build by defining DISABLE_EVENT.
1068 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1070 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1071 cached by the daemon.
1073 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1074 by defining LOOKUP_REDIS. The libhiredis library is required.
1076 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1077 keys are given for lookup.
1079 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1080 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1081 only supported when built with TLS support. The PolarSSL SHA routines
1082 are still used when the TLS library is too old for convenient support.
1084 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1085 openssl_options), for security. OpenSSL forces this from version 1.1.0
1086 server-side so match that on older versions.
1088 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1089 allocation for $value could be released as the expansion processing
1090 concluded, but leaving the global pointer active for it.
1092 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1093 and to use the domains and local_parts ACL conditions.
1095 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1096 incorrectly not doubled on cutthrough transmission, hence seen as a
1097 body-termination at the receiving system - resulting in truncated mails.
1098 Commonly the sender saw a TCP-level error, and retransmitted the message
1099 via the normal store-and-forward channel. This could result in duplicates
1100 received - but deduplicating mailstores were liable to retain only the
1101 initial truncated version.
1103 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1105 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1107 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1108 we're in there, support oversigning also; bug 1309.
1110 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1112 HS/04 Add support for keep_environment and add_environment options.
1114 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1115 either intentional arithmetic overflow during PRNG, or testing config-
1118 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1119 delivery resulted in actual delivery. Cancel cutthrough before DATA
1122 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1123 rejects a recipient: pass the reject to the originator.
1125 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1126 Many were false-positives and ignorable, but it's worth fixing the
1129 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1130 for the new environment-manipulation done at startup. Move the routines
1131 from being local to tls.c to being global via the os.c file.
1133 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1134 an extract embedded as result-arg for a map, the first arg for extract
1135 is unavailable so we cannot tell if this is a numbered or keyed
1136 extraction. Accept either.
1142 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1145 JH/02 The smtp transport option "multi_domain" is now expanded.
1147 JH/03 The smtp transport now requests PRDR by default, if the server offers
1150 JH/04 Certificate name checking on server certificates, when exim is a client,
1151 is now done by default. The transport option tls_verify_cert_hostnames
1152 can be used to disable this per-host. The build option
1153 EXPERIMENTAL_CERTNAMES is withdrawn.
1155 JH/05 The value of the tls_verify_certificates smtp transport and main options
1156 default to the word "system" to access the system default CA bundle.
1157 For GnuTLS, only version 3.0.20 or later.
1159 JH/06 Verification of the server certificate for a TLS connection is now tried
1160 (but not required) by default. The verification status is now logged by
1161 default, for both outbound TLS and client-certificate supplying inbound
1164 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1167 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1168 Status Notification (bounce) messages are now MIME format per RFC 3464.
1169 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1170 under the control of the dsn_advertise_hosts option, and routers may
1171 have a dsn_lasthop option.
1173 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1174 default, modifiable by a malware= option. The list separator for
1175 the options can now be changed in the usual way. Bug 68.
1177 JH/10 The smtp_receive_timeout main option is now expanded before use.
1179 JH/11 The incoming_interface log option now also enables logging of the
1180 local interface on delivery outgoing connections.
1182 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1183 if the interface and destination host and port all match.
1185 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1188 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1189 Patch from Andrew Lewis.
1191 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1192 now supports optional time-restrictions, weighting, and priority
1193 modifiers per server. Patch originally by <rommer@active.by>.
1195 JH/16 The spamd_address main option now supports a mixed list of local
1196 and remote servers. Remote servers can be IPv6 addresses, and
1197 specify a port-range.
1199 JH/17 Bug 68: The spamd_address main option now supports an optional
1200 timeout value per server.
1202 JH/18 Bug 1581: Router and transport options headers_add/remove can
1203 now have the list separator specified.
1205 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1208 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1211 JH/21 Support for the A6 type of dns record is withdrawn.
1213 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1214 rather than the verbs used.
1216 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1217 from 255 to 1024 chars.
1219 JH/24 Verification callouts now attempt to use TLS by default.
1221 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1222 are generic router options now. The defaults didn't change.
1224 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1225 Original patch from Alexander Shikoff, worked over by JH.
1227 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1228 files if ZCAT_COMMAND is not executable.
1230 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1232 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1234 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1235 Normally benign, it bites when the pair was led to by a CNAME;
1236 modern usage is to not canonicalize the domain to a CNAME target
1237 (and we were inconsistent anyway for A-only vs AAAA+A).
1239 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1241 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1242 when evaluating $sender_host_dnssec.
1244 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1245 $sender_helo_dnssec variable.
1247 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1249 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1251 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1253 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1254 documented as working, but never had. Support all but $spam_report.
1256 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1257 added for tls authenticator.
1259 HS/03 Add perl_taintmode main config option
1265 TL/01 When running the test suite, the README says that variables such as
1266 no_msglog_check are global and can be placed anywhere in a specific
1267 test's script, however it was observed that placement needed to be near
1268 the beginning for it to behave that way. Changed the runtest perl
1269 script to read through the entire script once to detect and set these
1270 variables, reset to the beginning of the script, and then run through
1271 the script parsing/test process like normal.
1273 TL/02 The BSD's have an arc4random API. One of the functions to induce
1274 adding randomness was arc4random_stir(), but it has been removed in
1275 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1276 function when detected.
1278 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1279 cause callback expansion.
1281 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1282 syntax errors in an expansion can be treated as a string instead of
1283 logging or causing an error, due to the internal use of bool_lax
1284 instead of bool when processing it.
1286 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1287 server certificates when making smtp deliveries.
1289 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1291 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1293 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1295 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1296 Merged patch from Sebastian Wiedenroth.
1298 JH/05 Fix results-pipe from transport process. Several recipients, combined
1299 with certificate use, exposed issues where response data items split
1300 over buffer boundaries were not parsed properly. This eventually
1301 resulted in duplicates being sent. This issue only became common enough
1302 to notice due to the introduction of connection certificate information,
1303 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1305 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1306 size buffer was used, resulting in syntax errors when an expansion
1309 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1310 version 3.3.6 or later.
1312 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1313 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1314 both become "event_action", the variables become $event_name, $event_data
1315 and $event_defer_errno. There is a new variable $verify_mode, usable in
1316 routers, transports and related events. The tls:cert event is now also
1317 raised for inbound connections, if the main configuration event_action
1320 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1321 early OCSP support, but no stapling (appears to be less than 1.0.0).
1323 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1324 server certificate names available under the smtp transport option
1325 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1328 JH/10 Time-related extraction expansions from certificates now use the main
1329 option "timezone" setting for output formatting, and are consistent
1330 between OpenSSL and GnuTLS compilations. Bug 1541.
1332 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1333 encoded parameter in the incoming message. Bug 1558.
1335 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1336 include certificate info, eximon was claiming there were spoolfile
1339 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1341 JH/14 Log delivery-related information more consistently, using the sequence
1342 "H=<name> [<ip>]" wherever possible.
1344 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1345 are problematic for Debian distribution, omit them from the release
1348 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1350 JH/16 Fix string representation of time values on 64bit time_t architectures.
1353 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1354 output list separator was used.
1359 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1360 checkers that were complaining about end of non-void function with no
1363 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1364 This was a regression introduced in 4.83 by another bugfix.
1366 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1368 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1369 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1375 TF/01 Correctly close the server side of TLS when forking for delivery.
1377 When a message was received over SMTP with TLS, Exim failed to clear up
1378 the incoming connection properly after forking off the child process to
1379 deliver the message. In some situations the subsequent outgoing
1380 delivery connection happened to have the same fd number as the incoming
1381 connection previously had. Exim would try to use TLS and fail, logging
1382 a "Bad file descriptor" error.
1384 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1385 utilities have not been installed.
1387 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1388 temporary space as the ACL may create new global variables.
1390 TL/01 LDAP support uses per connection or global context settings, depending
1391 upon the detected version of the libraries at build time.
1393 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1394 to extract and use the src ip:port in logging and expansions as if it
1395 were a direct connection from the outside internet. PPv2 support was
1396 updated based on HAProxy spec change in May 2014.
1398 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1400 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1401 Properly escape header and check for NULL return.
1403 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1406 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1408 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1409 characters in header names, implemented as a verify condition.
1410 Contributed by Michael Fischer v. Mollard.
1412 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1413 results permerror and temperror. Previous values are deprecated but
1414 still accepted. In a future release, err_perm and err_temp will be
1415 completely removed, which will be a backward incompatibility if the
1416 ACL tests for either of these two old results. Patch contributed by
1417 user bes-internal on the mailing list.
1419 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1421 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1422 selectors, in both main and reject logs.
1424 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1427 JH/07 Add malware type "sock" for talking to simple daemon.
1429 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1431 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1432 routers/transports under cutthrough routing.
1434 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1435 numbers. Touch up "bool" conditional to keep the same definition.
1437 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1439 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1441 JH/12 Expand items in router/transport headers_add or headers_remove lists
1442 individually rather than the list as a whole. Bug 1452.
1444 Required for reasonable handling of multiple headers_ options when
1445 they may be empty; requires that headers_remove items with embedded
1446 colons must have them doubled (or the list-separator changed).
1448 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1449 view the policy declared in the DMARC record. Currently, $dmarc_status
1450 is a combined value of both the record presence and the result of the
1453 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1455 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1456 dnslookup router and the smtp transport (applying to the forward
1459 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1460 of ldap servers used for a specific lookup. Patch provided by Heiko
1463 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1464 New variable $lookup_dnssec_authenticated for observability.
1466 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1467 Patch submitted by Lars Timman.
1469 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1471 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1472 Requires trusted mode and valid format message id, aborts otherwise.
1473 Patch contributed by Heiko Schlichting.
1475 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1476 certextract with support for various fields. Bug 1358.
1478 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1479 is requested by default, modifiable by smtp transport option
1482 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1483 operate on certificate variables to give certificate fingerprints
1484 Also new ${sha256:cert_variable}.
1486 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1488 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1489 Christian Aistleitner.
1491 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1493 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1494 file. Patch from Wolfgang Breyha.
1496 JH/25 Expand the coverage of the delivery $host and $host_address to
1497 client authenticators run in verify callout. Bug 1476.
1499 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1500 align with daemon_smtp_ports. Bug 72.
1502 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1503 support and error reporting did not work properly.
1505 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1506 and is readable. Patch from Andrew Colin Kissa.
1508 TL/14 Enhance documentation of ${run expansion and how it parses the
1509 commandline after expansion, particularly in the case when an
1510 unquoted variable expansion results in an empty value.
1512 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1514 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1515 by renaming away from it. Addresses GH issue 15, reported by
1518 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1520 TL/15 SECURITY: prevent double expansion in math comparison functions
1521 (can expand unsanitized data). Not remotely exploitable.
1528 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1530 PP/02 Make -n do something, by making it not do something.
1531 When combined with -bP, the name of an option is not output.
1533 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1536 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1537 $sender_host_name and config options to manage this, and basic check
1540 PP/05 DSCP support for outbound connections and control modifier for inbound.
1542 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1543 (Only plugin which currently uses this is kerberos4, which nobody should
1544 be using, but we should make it available and other future plugins might
1545 conceivably use it, even though it would break NAT; stuff *should* be
1546 using channel bindings instead).
1548 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1549 name; added for Sendmail compatibility; requires admin caller.
1550 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1551 just ignore it); requires trusted caller.
1552 Also parse but ignore: -Ac -Am -X<logfile>
1555 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1557 TL/02 Add +smtp_confirmation as a default logging option.
1559 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1560 Patch by Magnus Holmgren from 2007-02-20.
1562 TL/04 Bugzilla 1281 - Spec typo.
1563 Bugzilla 1283 - Spec typo.
1564 Bugzilla 1290 - Spec grammar fixes.
1566 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1568 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1570 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1571 mailing list by by Dmitry Isaikin.
1573 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1575 JH/02 Support "G" suffix to numbers in ${if comparisons.
1577 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1579 NM/01 Bugzilla 1197 - Spec typo
1580 Bugzilla 1196 - Spec examples corrections
1582 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1584 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1585 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1588 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1589 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1591 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1592 "acl {{name}{arg}...}", and optional args on acl condition
1595 JH/05 Permit multiple router/transport headers_add/remove lines.
1597 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1599 JH/07 Avoid using a waiting database for a single-message-only transport.
1600 Performance patch from Paul Fisher. Bugzilla 1262.
1602 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1605 JH/09 Add $headers_added variable, with content from use of ACL modifier
1606 add_header (but not yet added to the message). Bugzilla 199.
1608 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1609 Pulled from Bugzilla 817 by Wolfgang Breyha.
1611 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1613 (nb: this is the same fix as in Exim 4.80.1)
1615 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1618 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1621 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1623 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1624 advertise SMTP AUTH mechanism to us, instead of a generic
1625 protocol violation error. Also, make Exim more robust to bad
1626 data from the Dovecot auth socket.
1628 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1630 When a queue runner is handling a message, Exim first routes the
1631 recipient addresses, during which it prunes them based on the retry
1632 hints database. After that it attempts to deliver the message to
1633 any remaining recipients. It then updates the hints database using
1636 So if a recipient address works intermittently, it can get repeatedly
1637 deferred at routing time. The retry hints record remains fresh so the
1638 address never reaches the final cutoff time.
1640 This is a fairly common occurrence when a user is bumping up against
1641 their storage quota. Exim had some logic in its local delivery code
1642 to deal with this. However it did not apply to per-recipient defers
1643 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1645 This change adds a proper retry rule check during routing so that the
1646 final cutoff time is checked against the message's age. We only do
1647 this check if there is an address retry record and there is not a
1648 domain retry record; this implies that previous attempts to handle
1649 the address had the retry_use_local_parts option turned on. We use
1650 this as an approximation for the destination being like a local
1651 delivery, as in LMTP.
1653 I suspect this new check makes the old local delivery cutoff check
1654 redundant, but I have not verified this so I left the code in place.
1656 TF/02 Correct gecos expansion when From: is a prefix of the username.
1658 Test 0254 submits a message to Exim with the header
1662 When I ran the test suite under the user fanf2, Exim expanded
1663 the header to contain my full name, whereas it should have added
1664 a Resent-Sender: header. It erroneously treats any prefix of the
1665 username as equal to the username.
1667 This change corrects that bug.
1669 GF/01 DCC debug and logging tidyup
1670 Error conditions log to paniclog rather than rejectlog.
1671 Debug lines prefixed by "DCC: " to remove any ambiguity.
1673 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1675 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1676 Bug spotted by Jeremy Harris; was flawed since initial commit.
1677 Would have resulted in OCSP responses post-SNI triggering an Exim
1678 NULL dereference and crash.
1680 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1682 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1683 Bug detection, analysis and fix by Samuel Thibault.
1684 Bugzilla 1331, Debian bug #698092.
1686 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1688 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1689 Server implementation by Todd Lyons, client by JH.
1690 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1691 config variable "prdr_enable" controls whether the server
1692 advertises the facility. If the client requests PRDR a new
1693 acl_data_smtp_prdr ACL is called once for each recipient, after
1694 the body content is received and before the acl_smtp_data ACL.
1695 The client is controlled by both of: a hosts_try_prdr option
1696 on the smtp transport, and the server advertisement.
1697 Default client logging of deliveries and rejections involving
1698 PRDR are flagged with the string "PRDR".
1700 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1701 fclose(). Diagnosis by Todd Lyons.
1703 PP/17 Update configure.default to handle IPv6 localhost better.
1704 Patch by Alain Williams (plus minor tweaks).
1707 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1708 This is now consistent with GnuTLS, and is now documented: the
1709 previous undocumented portable approach to treating the option as
1710 unset was to force an expansion failure. That still works, and
1711 an empty string is now equivalent.
1713 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1714 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1715 not performing validation itself.
1717 PP/20 Added force_command boolean option to pipe transport.
1718 Patch from Nick Koston, of cPanel Inc.
1720 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1723 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1725 PP/21 Fix eximon continuous updating with timestamped log-files.
1726 Broken in a format-string cleanup in 4.80, missed when I repaired the
1727 other false fix of the same issue.
1728 Report and fix from Heiko Schlichting.
1731 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1732 Report from Prashanth Katuri.
1734 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1735 It's SecureTransport, so affects any MacOS clients which use the
1736 system-integrated TLS libraries, including email clients.
1738 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1739 using a MIME ACL for non-SMTP local injection.
1740 Report and assistance in diagnosis by Warren Baker.
1742 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1744 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1746 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1747 last id that failed so it may be referenced in subsequent ACL's.
1749 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1752 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1753 ldap library initialization, allowing self-signed CA's to be
1754 used. Also properly sets require_cert option later in code by
1755 using NULL (global ldap config) instead of ldap handle (per
1756 session). Bug diagnosis and testing by alxgomz.
1758 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1759 the src/util/ subdirectory.
1761 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1762 renamed to Transport Post Delivery Action by Jeremy Harris, as
1765 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1766 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1767 redis_servers = needs to be configured which will be used by the redis
1768 lookup. Patch from Warren Baker, of The Packet Hub.
1770 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1772 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1773 hostname or reverse DNS when processing a host list. Used suggestions
1774 from multiple comments on this bug.
1776 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1778 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1779 Missed a few lines, added it to make the runtest require no keyboard
1782 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1783 contains upper case chars. Make router use caseful_local_part.
1785 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1786 support when GnuTLS has been built with p11-kit.
1792 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1794 This, or similar/improved, will also be change PP/11 of 4.82.
1800 PP/01 Handle short writes when writing local log-files.
1801 In practice, only affects FreeBSD (8 onwards).
1802 Bugzilla 1053, with thanks to Dmitry Isaikin.
1804 NM/01 Bugzilla 949 - Documentation tweak
1806 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1809 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1811 PP/02 Implemented gsasl authenticator.
1813 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1815 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1816 `pkg-config foo` for cflags/libs.
1818 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1819 with rest of GSASL and with heimdal_gssapi.
1821 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1822 `pkg-config foo` for cflags/libs for the TLS implementation.
1824 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1825 properties get this fed in as external SSF. A number of robustness
1826 and debugging improvements to the cyrus_sasl authenticator.
1828 PP/08 cyrus_sasl server now expands the server_realm option.
1830 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1831 Patch by Jeremy Harris.
1833 PP/10 Added dbmjz lookup type.
1835 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1837 PP/12 MAIL args handles TAB as well as SP, for better interop with
1838 non-compliant senders.
1839 Analysis and variant patch by Todd Lyons.
1841 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1842 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1843 Patch from Dirk Mueller <dmueller@suse.com>
1845 PP/13 tls_peerdn now print-escaped for spool files.
1846 Observed some $tls_peerdn in wild which contained \n, which resulted
1847 in spool file corruption.
1849 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1850 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1851 or write after TLS renegotiation, which otherwise led to messages
1854 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1855 as a tracking header (ie: a signed header comes before the signature).
1856 Patch from Wolfgang Breyha.
1858 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1859 comma-sep list; embedded commas doubled.
1861 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1863 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1865 Report and patch from Dmitry Banschikov.
1867 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1868 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1869 failures. We appear to now support TLS1.1+ with Exim.
1871 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1872 lets Exim select keys and certificates based upon TLS SNI from client.
1873 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1874 before an outbound SMTP session. New log_selector, +tls_sni.
1876 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1877 NULL dereference. Report and patch from Alun Jones.
1879 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1880 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1881 Not seeing resolver debug output on NetBSD, but suspect this is a
1882 resolver implementation change.
1884 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1885 Left warnings. Added "eximon gdb" invocation mode.
1887 PP/21 Defaulting "accept_8bitmime" to true, not false.
1889 PP/22 Added -bw for inetd wait mode support.
1891 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1892 locate the relevant includes and libraries. Made this the default.
1894 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1895 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1897 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1898 This may cause build issues on older platforms.
1900 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1901 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1902 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1903 Added SNI support via GnuTLS too.
1904 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1906 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1908 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1909 Applied second patch from Janne, implementing suggestion to default
1910 multiple-strings-in-record handling to match SPF spec.
1912 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1914 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1915 read-only, out of scope).
1916 Patch from Wolfgang Breyha, report from Stuart Northfield.
1918 PP/29 Fix three issues highlighted by clang analyser static analysis.
1919 Only crash-plausible issue would require the Cambridge-specific
1920 iplookup router and a misconfiguration.
1921 Report from Marcin Mirosław.
1923 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1925 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1926 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1927 As part of this, removing so much warning spew let me fix some minor
1928 real issues in debug logging.
1930 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1931 assignment on my part. Fixed.
1933 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1934 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1935 Janne Snabb (who went above and beyond: thank you).
1937 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1938 string otherwise requires a connection and a bunch more work and it's
1939 relatively easy to get wrong. Should also expose TLS library linkage
1942 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1943 64-bit ${eval} (JH/03).
1945 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1946 GNU libc to support some of the 64-bit stuff, should not lead to
1947 conflicts. Defined before os.h is pulled in, so if a given platform
1948 needs to override this, it can.
1950 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1951 protection layer was required, which is not implemented.
1952 Bugzilla 1254, patch from Wolfgang Breyha.
1954 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1955 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1956 tls_dhparam take prime identifiers. Also unbreak combination of
1957 OpenSSL+DH_params+TLSSNI.
1959 PP/39 Disable SSLv2 by default in OpenSSL support.
1965 PP/01 Solaris build fix for Oracle's LDAP libraries.
1966 Bugzilla 1109, patch from Stephen Usher.
1968 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1970 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1973 TF/02 Fix a couple more cases where we did not log the error message
1974 when unlink() failed. See also change 4.74-TF/03.
1976 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1977 lock up or crash if it happened to be inside a call to libc when it
1978 got a SIGUSR1 from exiwhat.
1980 The SIGUSR1 handler appends the current process status to the process
1981 log which is later printed by exiwhat. It used to use the general
1982 purpose logging code to do this, but several functions it calls are
1983 not safe for signals.
1985 The new output code in the SIGUSR1 handler is specific to the process
1986 log, and simple enough that it's easy to inspect for signal safety.
1987 Removing some special cases also simplifies the general logging code.
1988 Removing the spurious timestamps from the process log simplifies
1991 TF/04 Improved ratelimit ACL condition.
1993 The /noupdate option has been deprecated in favour of /readonly which
1994 has clearer semantics. The /leaky, /strict, and /readonly update modes
1995 are mutually exclusive. The update mode is no longer included in the
1996 database key; it just determines when the database is updated. (This
1997 means that when you upgrade Exim will forget old rate measurements.)
1999 Exim now checks that the per_* options are used with an update mode that
2000 makes sense for the current ACL. For example, when Exim is processing a
2001 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2002 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2003 must specify per_mail/readonly. If you omit the update mode it defaults to
2004 /leaky where that makes sense (as before) or /readonly where required.
2006 The /noupdate option is now undocumented but still supported for
2007 backwards compatibility. It is equivalent to /readonly except that in
2008 ACLs where /readonly is required you may specify /leaky/noupdate or
2009 /strict/noupdate which are treated the same as /readonly.
2011 A useful new feature is the /count= option. This is a generalization
2012 of the per_byte option, so that you can measure the throughput of other
2013 aggregate values. For example, the per_byte option is now equivalent
2014 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2016 The per_rcpt option has been generalized using the /count= mechanism
2017 (though it's more complicated than the per_byte equivalence). When it is
2018 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2019 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2020 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2021 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2022 Note that using per_rcpt with a non-readonly update mode in more than
2023 one ACL will cause the recipients to be double-counted. (The per_mail
2024 and per_byte options don't have this problem.)
2026 The handling of very low rates has changed slightly. If the computed rate
2027 is less than the event's count (usually one) then this event is the first
2028 after a long gap. In this case the rate is set to the same as this event's
2029 count, so that the first message of a spam run is counted properly.
2031 The major new feature is a mechanism for counting the rate of unique
2032 events. The new per_addr option counts the number of different
2033 recipients that someone has sent messages to in the last time period. It
2034 behaves like per_rcpt if all the recipient addresses are different, but
2035 duplicate recipient addresses do not increase the measured rate. Like
2036 the /count= option this is a general mechanism, so the per_addr option
2037 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2038 example, measure the rate that a client uses different sender addresses
2039 with the options per_mail/unique=$sender_address. There are further
2040 details in the main documentation.
2042 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2044 TF/06 Removed a few PCRE remnants.
2046 TF/07 Automatically extract Exim's version number from tags in the git
2047 repository when doing development or release builds.
2049 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2050 Bugzilla 879. Patch from Paul Fisher.
2052 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2053 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2056 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2058 PP/05 Variable $av_failed, true if the AV scanner deferred.
2059 Bugzilla 1078. Patch from John Horne.
2061 PP/06 Stop make process more reliably on build failure.
2062 Bugzilla 1087. Patch from Heiko Schlittermann.
2064 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2065 Bugzilla 1089. Patch from Heiko Schlittermann.
2067 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2068 Bugzilla 1131. Patch from Holger Weiß.
2070 PP/09 Handle IPv6 addresses with SPF.
2071 Bugzilla 860. Patch from Wolfgang Breyha.
2073 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2075 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2078 PP/11 match_* no longer expand right-hand-side by default.
2079 New compile-time build option, EXPAND_LISTMATCH_RHS.
2080 New expansion conditions, "inlist", "inlisti".
2082 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2084 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2086 PP/14 fix log_write() format string regression from TF/03.
2087 Bugzilla 1152. Patch from Dmitry Isaikin.
2093 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2095 PP/02 Harmonised TLS library version reporting; only show if debugging.
2096 Layout now matches that introduced for other libraries in 4.74 PP/03.
2098 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2100 PP/04 New "dns_use_edns0" global option.
2102 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2105 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2106 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2108 TK/01 Updated PolarSSL code to 0.14.2.
2109 Bugzilla 1097. Patch from Andreas Metzler.
2111 PP/07 Catch divide-by-zero in ${eval:...}.
2112 Fixes bugzilla 1102.
2114 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2117 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2118 format-string attack -- SECURITY: remote arbitrary code execution.
2120 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2121 time unintentionally subject to list matching rules, letting the header
2122 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2123 arbitrary string expansion). This allowed for information disclosure.
2125 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2126 INT_MIN/-1 -- value coerced to INT_MAX.
2132 NM/01 Workaround for PCRE version dependency in version reporting
2135 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2136 This fixes portability to compilers other than gcc, notably
2137 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2139 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2140 makefiles for portability to HP-UX and POSIX correctness.
2142 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2143 Also via indented variable definition in the Makefile.
2144 (Debugging by Oliver Heesakkers).
2146 PP/02 Restore caching of spamd results with expanded spamd_address.
2147 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2149 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2150 Improves build reliability. Fix from: Frank Elsner
2152 NM/02 Fix wide character breakage in the rfc2047 coding
2153 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2155 NM/03 Allow underscore in dnslist lookups
2156 Fixes bug 1026. Patch from Graeme Fowler
2158 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2159 Code patches from Adam Ciarcinski of NetBSD.
2161 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2164 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2165 is logged, to avoid truncation. Patch from John Horne.
2167 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2168 Patch from Jakob Hirsch.
2170 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2171 SQL string expansion failure details.
2172 Patch from Andrey Oktyabrski.
2174 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2175 Patch from Simon Arlott.
2177 PP/09 New lookups functionality failed to compile on old gcc which rejects
2178 extern declarations in function scope.
2179 Patch from Oliver Fleischmann
2181 PP/10 Use sig_atomic_t for flags set from signal handlers.
2182 Check getgroups() return and improve debugging.
2183 Fixed developed for diagnosis in bug 927 (which turned out to be
2186 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2187 Patch from Mark Zealey.
2189 PP/12 Bugzilla 1056: Improved spamd server selection.
2190 Patch from Mark Zealey.
2192 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2193 Based on patch from Heiko Schlittermann.
2195 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2196 Patch from Uwe Doering, sign-off by Michael Haardt.
2198 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2199 variable declaration deep within a block. Bug and patch from
2202 PP/15 lookups-Makefile IRIX compatibility coercion.
2204 PP/16 Make DISABLE_DKIM build knob functional.
2206 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2207 Patch by Simon Arlott
2209 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2210 variable argument macros. Our copy now differs from upstream.
2216 TF/01 Failure to get a lock on a hints database can have serious
2217 consequences so log it to the panic log.
2219 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2220 controlled using the smtp_confirmation log selector.
2222 TF/03 Include the error message when we fail to unlink a spool file.
2224 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2225 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2226 for maintaining out-of-tree patches for some time.
2228 PP/01 Bugzilla 139: Documentation and portability issues.
2229 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2230 Handle per-OS dynamic-module compilation flags.
2232 PP/02 Let /dev/null have normal permissions.
2233 The 4.73 fixes were a little too stringent and complained about the
2234 permissions on /dev/null. Exempt it from some checks.
2235 Reported by Andreas M. Kirchwitz.
2237 PP/03 Report version information for many libraries, including
2238 Exim version information for dynamically loaded libraries. Created
2239 version.h, now support a version extension string for distributors
2240 who patch heavily. Dynamic module ABI change.
2242 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2243 privilege escalation vulnerability whereby the Exim run-time user
2244 can cause root to append content of the attacker's choosing to
2247 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2250 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2251 If dropping privileges for untrusted macros, we disabled normal logging
2252 on the basis that it would fail; for the Exim run-time user, this is not
2253 the case, and it resulted in successful deliveries going unlogged.
2254 Fixed. Reported by Andreas Metzler.
2260 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2261 only prepend for the Resent-* case. Fixes regression introduced in
2262 Exim 4.70 by NM/22 for Bugzilla 607.
2264 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2265 increasing numbers of administrators be bitten by this.
2267 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2269 PP/03 Bugzilla 994: added openssl_options main configuration option.
2271 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2273 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2275 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2277 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2278 without a peer certificate, leading to a segfault because of an
2279 assumption that peers always have certificates. Be a little more
2280 paranoid. Problem reported by Martin Tscholak.
2282 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2283 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2284 NB: ClamAV planning to remove STREAM in "middle of 2010".
2285 CL also introduces -bmalware, various -d+acl logging additions and
2286 more caution in buffer sizes.
2288 PP/09 Implemented reverse_ip expansion operator.
2290 PP/10 Bugzilla 937: provide a "debug" ACL control.
2292 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2294 PP/12 Bugzilla 973: Implement --version.
2296 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2298 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2300 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2302 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2303 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2304 ignore trailing whitespace.
2306 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2308 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2311 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2312 Notification from Dr Andrew Aitchison.
2314 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2315 ExtendedDetectionInfo response format.
2316 Notification from John Horne.
2318 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2321 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2322 XSL and documented dependency on system catalogs, with examples of how
2325 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2328 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2329 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2330 configuration file which is writeable by the Exim user or group.
2332 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2333 of configuration files to cover files specified with the -C option if
2334 they are going to be used with root privileges, not just the default
2337 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2338 option (effectively making it always true).
2340 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2341 files to be used while preserving root privileges.
2343 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2344 that rogue child processes cannot use them.
2346 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2347 run-time user, instead of root.
2349 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2350 Exim run-time user without dropping privileges.
2352 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2353 result string, instead of calling string_vformat() twice with the same
2356 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2357 for other users. Others should always drop root privileges if they use
2358 -C on the command line, even for a whitelisted configure file.
2360 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2362 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2368 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2369 $data_path, and $header_path variables; fixed documentation bugs and
2372 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2373 exipick to access non-standard spools, including the "frozen" queue
2376 NM/01 Bugzilla 965: Support mysql stored procedures.
2377 Patch from Alain Williams
2379 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2381 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2382 Patch from Andreas Metzler
2384 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2385 Patch from Kirill Miazine
2387 NM/05 Bugzilla 671: Added umask to procmail example.
2389 JJ/03 installed exipick 20100323.0, fixing doc bug
2391 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2392 directory. Notification and patch from Dan Rosenberg.
2394 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2396 TK/02 Improve log output when DKIM signing operation fails.
2398 MH/01 Treat the transport option dkim_domain as a colon separated
2399 list, not as a single string, and sign the message with each element,
2400 omitting multiple occurences of the same signer.
2402 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2403 Bugzilla 985, 986. Patch by Simon Arlott
2405 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2406 Patch by Simon Arlott
2408 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2409 MBX locking. Notification from Dan Rosenberg.
2415 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2417 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2419 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2421 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2423 NM/04 Bugzilla 915: Flag broken perl installation during build.
2429 TK/01 Added patch by Johannes Berg that expands the main option
2430 "spamd_address" if it starts with a dollar sign.
2432 TK/02 Write list of recipients to X-Envelope-Sender header when building
2433 the mbox-format spool file for content scanning (suggested by Jakob
2436 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2437 (http://www.dcc-servers.net/) support via dccifd. Activated by
2438 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2440 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2441 by Mark Daniel Reidel <mr@df.eu>.
2443 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2444 When building exim an external PCRE library is now needed -
2445 PCRE is a system library on the majority of modern systems.
2446 See entry on PCRE_LIBS in EDITME file.
2448 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2449 conversation. Added nologin parameter to request.
2450 Patch contributed by Kirill Miazine.
2452 TF/01 Do not log submission mode rewrites if they do not change the address.
2454 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2456 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2457 log files in place. Contributed by Roberto Lima.
2459 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2461 TF/03 Bugzilla 615: When checking the local_parts router precondition
2462 after a local_part_suffix or local_part_prefix option, Exim now
2463 does not use the address's named list lookup cache, since this
2464 contains cached lookups for the whole local part.
2466 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2467 Robert Millan. Documentation is in experimental-spec.txt.
2469 TF/04 Bugzilla 668: Fix parallel build (make -j).
2471 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2473 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2474 Patch provided by Jan Srzednicki.
2476 TF/05 Leading white space used to be stripped from $spam_report which
2477 wrecked the formatting. Now it is preserved.
2479 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2480 that they are available at delivery time.
2482 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2484 TF/08 TLS error reporting now respects the incoming_interface and
2485 incoming_port log selectors.
2487 TF/09 Produce a more useful error message if an SMTP transport's hosts
2488 setting expands to an empty string.
2490 NM/06 Bugzilla 744: EXPN did not work under TLS.
2491 Patch provided by Phil Pennock.
2493 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2494 Patch provided by Richard Godbee.
2496 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2497 acl_smtp_notquit, added index entry.
2499 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2500 Patch provided by Eugene Bujak.
2502 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2503 accept(). Patch provided by Maxim Dounin.
2505 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2506 Patch provided by Phil Pennock.
2508 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2510 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2511 Patch provided by Brad "anomie" Jorsch.
2513 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2514 Patch provided by Dean Brooks.
2516 TK/05 Add native DKIM support (does not depend on external libraries).
2518 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2519 Patch provided by Graeme Fowler.
2521 NM/16 Bugzilla 851: Documentation example syntax fix.
2523 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2525 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2528 NM/19 Bugzilla 745: TLS version reporting.
2529 Patch provided by Phil Pennock.
2531 NM/20 Bugzilla 167: bool: condition support.
2532 Patch provided by Phil Pennock.
2534 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2535 clients. Patch provided by Phil Pennock.
2537 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2538 Patch provided by Brad "anomie" Jorsch.
2540 NM/23 Bugzilla 687: Fix misparses in eximstats.
2541 Patch provided by Heiko Schlittermann.
2543 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2544 Patch provided by Heiko Schlittermann.
2546 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2547 plus update to original patch.
2549 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2551 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2552 Patch provided by David Brownlee.
2554 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2556 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2558 NM/30 Bugzilla 888: TLS documentation bugfixes.
2560 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2562 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2563 Unlike the original bugzilla I have changed all shell scripts in src tree.
2565 NM/33 Bugzilla 898: Transport filter timeout fix.
2566 Patch by Todd Rinaldo.
2568 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2569 Patch by Serge Demonchaux.
2571 NM/35 Bugzilla 39: Base64 decode bug fixes.
2572 Patch by Jakob Hirsch.
2574 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2576 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2578 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2580 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2586 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2587 ALT-N's libdkim that I have put here:
2588 http://duncanthrax.net/exim-experimental/
2590 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2591 were called 'true' and it seems that C99 defines that as a reserved
2592 keyword to be used with 'bool' variable types. That means you could
2593 not include C99-style headers which use bools without triggering
2594 build errors in sieve.c.
2596 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2597 as mailq or other aliases. Changed the --help handling significantly
2598 to do whats expected. exim_usage() emits usage/help information.
2600 SC/01 Added the -bylocaldomain option to eximstats.
2602 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2604 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2606 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2612 PH/01 Another patch from the Sieve maintainer.
2614 PH/02 When an IPv6 address is converted to a string for single-key lookup
2615 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2616 dots are used instead of colons so that keys in lsearch files need not
2617 contain colons. This was done some time before quoting was made available
2618 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2619 (notated using the quote facility) so as to distinguish them from IPv4
2620 keys. This meant that lookups for IP addresses in host lists did not work
2621 for iplsearch lookups.
2623 This has been fixed by arranging for IPv6 addresses to be expressed with
2624 colons if the lookup type is iplsearch. This is not incompatible, because
2625 previously such lookups could never work.
2627 The situation is now rather anomalous, since one *can* have colons in
2628 ordinary lsearch keys. However, making the change in all cases is
2629 incompatible and would probably break a number of configurations.
2631 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2634 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2635 conversion specification without a maximum field width, thereby enabling
2636 a rogue spamd server to cause a buffer overflow. While nobody in their
2637 right mind would setup Exim to query an untrusted spamd server, an
2638 attacker that gains access to a server running spamd could potentially
2639 exploit this vulnerability to run arbitrary code as the Exim user.
2641 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2642 $primary_hostname instead of what libspf2 thinks the hosts name is.
2644 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2645 a directory entry by the name of the lookup key. Previously, if a
2646 symlink pointed to a non-existing file or a file in a directory that
2647 Exim lacked permissions to read, a lookup for a key matching that
2648 symlink would fail. Now it is enough that a matching directory entry
2649 exists, symlink or not. (Bugzilla 503.)
2651 PH/03 The body_linecount and body_zerocount variables are now exported in the
2654 PH/04 Added the $dnslist_matched variable.
2656 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2657 This means they are set thereafter only if the connection becomes
2660 PH/06 Added the client_condition to authenticators so that some can be skipped
2661 by clients under certain conditions.
2663 PH/07 The error message for a badly-placed control=no_multiline_responses left
2664 "_responses" off the end of the name.
2666 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2668 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2669 (without spaces) instead of just copying the configuration text.
2671 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2673 PH/11 Added $max_received_linelength.
2675 PH/12 Added +ignore_defer and +include_defer to host lists.
2677 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2678 way in which PCRE > 7.0 is built.
2680 PH/14 Implemented queue_only_load_latch.
2682 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2683 MAIL command. The effect was to mangle the value on 64-bit systems.
2685 PH/16 Another patch from the Sieve maintainer.
2687 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2689 PH/18 If a system quota error occurred while trying to create the file for
2690 a maildir delivery, the message "Mailbox is full" was not appended to the
2691 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2692 only to a quota excession during the actual writing of the file.
2694 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2695 characters?) which causes problems in log lines. The DN values are now
2696 passed through string_printing() before being added to log lines.
2698 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2699 and InterBase are left for another time.)
2701 PH/21 Added message_body_newlines option.
2703 PH/22 Guard against possible overflow in moan_check_errorcopy().
2705 PH/23 POSIX allows open() to be a macro; guard against that.
2707 PH/24 If the recipient of an error message contained an @ in the local part
2708 (suitably quoted, of course), incorrect values were put in $domain and
2709 $local_part during the evaluation of errors_copy.
2715 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2716 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2719 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2720 issue a MAIL command.
2722 PH/02 In an ACL statement such as
2724 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2726 if a client was not listed at all, or was listed with a value other than
2727 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2728 the condition was not true (as it should be), so access was not denied.
2729 The bug was that the ! inversion was incorrectly passed on to the second
2730 item. This has been fixed.
2732 PH/03 Added additional dnslists conditions == and =& which are different from
2733 = and & when the dns lookup returns more than one IP address.
2735 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2736 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2738 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2739 FSYNC, which compiles an option called disable_fsync that allows for
2740 bypassing fsync(). The documentation is heavily laced with warnings.
2742 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2744 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2745 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2746 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2747 including adding "make clean"; (3) Added -fPIC when compiling the test
2748 dynamically loaded module, to get rid of a warning.
2750 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2751 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2752 = 0s. The bug is otherwise harmless.
2754 PH/07 There was a bug in the dovecot authenticator such that the value of
2755 $auth1 could be overwritten, and so not correctly preserved, after a
2756 successful authentication. This usually meant that the value preserved by
2757 the server_setid option was incorrect.
2759 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2761 PH/09 Installed PCRE release 7.0.
2763 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2764 run for batched SMTP input. It is now run at the start of every message
2765 in the batch. While fixing this I discovered that the process information
2766 (output by running exiwhat) was not always getting set for -bs and -bS
2767 input. This is fixed, and it now also says "batched" for BSMTP.
2769 PH/11 Added control=no_pipelining.
2771 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2772 patch, slightly modified), and move the expansion of helo_data till after
2773 the connection is made in the smtp transport (so it can use these
2776 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2778 PH/14 Added log_selector = +pid.
2780 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2782 PH/16 Add ${if forany and ${if forall.
2784 PH/17 Added dsn_from option to vary the From: line in DSNs.
2786 PH/18 Flush SMTP output before performing a callout, unless control =
2787 no_callout_flush is set.
2789 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2790 was true (the default) a successful delivery failed to delete the retry
2791 item, thus causing premature timeout of the address. The bug is now
2794 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2796 PH/21 Long custom messages for fakedefer and fakereject are now split up
2797 into multiline responses in the same way that messages for "deny" and
2798 other ACL rejections are.
2800 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2801 with slight modification.
2803 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2804 draft, changing the syntax and factoring some duplicate code".
2806 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2807 for deliveries of the second and subsequent messages over the same SMTP
2810 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2811 ${reduce, with only minor "tidies".
2813 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2815 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2816 expansion side effects.
2818 PH/27 When a message times out after an over-quota error from an Exim-imposed
2819 quota, the bounce message says "mailbox is full". This message was not
2820 being given when it was a system quota that was exceeded. It now should
2823 MH/03 Made $recipients available in local_scan(). local_scan() already has
2824 better access to the recipient list through recipients_list[], but
2825 $recipients can be useful in postmaster-provided expansion strings.
2827 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2828 in the case of a MAIL command with additional options following the
2829 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2830 were accidentally chopped off.
2832 PH/29 SMTP synchronization checks are implemented when a command is read -
2833 there is a check that no more input is waiting when there shouldn't be
2834 any. However, for some commands, a delay in an ACL can mean that it is
2835 some time before the response is written. In this time, more input might
2836 arrive, invalidly. So now there are extra checks after an ACL has run for
2837 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2838 pipelining has not been advertised.
2840 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2842 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2843 returned in the SMTP response when smtp_return_error_details was false.
2844 This has been fixed.
2846 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2847 instead of the C I/O that was originally supplied, because problems were
2848 reported on Solaris.
2850 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2851 Exim which did not show up earlier: it was assuming that a call to
2852 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2853 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2854 was a macro that became an assignment, so it seemed to work. This has
2855 changed to a proper function call with a void return, hence the compile
2856 error. Exim's code has been fixed.
2858 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2861 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2863 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2865 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2866 because it turns out to be too big or there is a timeout) there is no
2867 "Completed" line in the log. When some messages of this type were
2868 selected by exigrep, they were listed as "not completed". Others were
2869 picked up by some special patterns. I have improved the selection
2870 criteria to be more general.
2872 PH/38 The host_find_failed option in the manualroute router can now be set
2873 to "ignore", to completely ignore a host whose IP address cannot be
2874 found. If all hosts are ignored, the behaviour is controlled by the new
2875 host_all_ignored option.
2877 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2878 homing or because of multiple MX records with /mx) generated more than
2879 one IP address, and the following item turned out to be the local host,
2880 all the secondary addresses of the first item were incorrectly removed
2881 from the list, along with the local host and any following hosts (which
2882 is what is supposed to happen).
2884 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2885 whoever called Exim into the -H file. In the case of the daemon it was
2886 behaving confusingly. When first started, it used values for whoever
2887 started the daemon, but after a SIGHUP it used the Exim user (because it
2888 calls itself on a restart). I have changed the code so that it now always
2891 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2892 message are rejected with the same error (e.g. no authentication or bad
2893 sender address), and a DATA command is nevertheless sent (as can happen
2894 with PIPELINING or a stupid MUA), the error message that was given to the
2895 RCPT commands is included in the rejection of the DATA command. This is
2896 intended to be helpful for MUAs that show only the final error to their
2899 PH/42 Another patch from the Sieve maintainer.
2901 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2902 Eximstats - Fixed some broken HTML links and added missing column headers
2904 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2905 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2907 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2909 PH/43 Yet another patch from the Sieve maintainer.
2911 PH/44 I found a way to check for a TCP/IP connection going away before sending
2912 the response to the final '.' that terminates a message, but only in the
2913 case where the client has not sent further data following the '.'
2914 (unfortunately, this is allowed). However, in many cases there won't be
2915 any further data because there won't be any more messages to send. A call
2916 to select() can be used: if it shows that the input is "ready", there is
2917 either input waiting, or the socket has been closed. An attempt to read
2918 the next input character can distinguish the two cases. Previously, Exim
2919 would have sent an OK response which the client would never have see.
2920 This could lead to message repetition. This fix should cure that, at
2921 least in a lot of common cases.
2923 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2924 advertised in response to EHLO.
2930 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2931 fixed by 4.65/MH/01 (is this a record?) are fixed:
2933 (i) An empty string was always treated as zero by the numeric comparison
2934 operators. This behaviour has been restored.
2936 (ii) It is documented that the numeric comparison operators always treat
2937 their arguments as decimal numbers. This was broken in that numbers
2938 starting with 0 were being interpreted as octal.
2940 While fixing these problems I realized that there was another issue that
2941 hadn't been noticed. Values of message_size_limit (both the global option
2942 and the transport option) were treated as octal if they started with 0.
2943 The documentation was vague. These values are now always treated as
2944 decimal, and I will make that clear in the documentation.
2950 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2951 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2954 MH/01 Don't check that the operands of numeric comparison operators are
2955 integers when their expansion is in "skipping" mode (fixes bug
2956 introduced by 4.64-PH/07).
2958 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2959 child addresses, Exim now panics and dies. Previously, because the count
2960 is held in a short int, deliveries were likely to be lost. As such a
2961 large number of recipients for a single message is ridiculous
2962 (performance will be very, very poor), I have chosen to impose a limit
2963 rather than extend the field.
2969 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2970 leftover -K file (the existence of which was triggered by #402).
2971 While we were at it, introduced process PID as part of the -K
2972 filename. This should rule out race conditions when creating
2975 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2976 processing considerably. Previous code took too long for large mails,
2977 triggering a timeout which in turn triggers #401.
2979 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2980 in the DK code in transports.c. sendfile() is not really portable,
2981 hence the _LINUX specificness.
2983 TF/01 In the add_headers option to the mail command in an Exim filter,
2984 there was a bug that Exim would claim a syntax error in any
2985 header after the first one which had an odd number of characters
2988 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2989 callout verification, Exim cached a "reject" for the entire domain. This
2990 is correct for most verifications, but it is not correct for a recipient
2991 verification with use_sender or use_postmaster set, because in that case
2992 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2993 case of MAIL FROM:<> rejection from other early rejections (e.g.
2994 rejection of HELO). When verifying a recipient using a non-null MAIL
2995 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2996 Whatever the result of the callout, the value of the domain cache is
2997 left unchanged (for any other kind of callout, getting as far as trying
2998 RCPT means that the domain itself is ok).
3000 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3003 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3004 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3005 session, the daemon ignores ECONNECT errors and logs others; it now
3006 ignores EPIPE as well.
3008 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3009 (quoted-printable decoding).
3011 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3012 later the small subsequent patch to fix an introduced bug.
3014 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3016 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3018 PH/08 An error is now given if message_size_limit is specified negative.
3020 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3021 to be given (somewhat) arbitrary names.
3023 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3026 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3027 miscellaneous code fixes
3029 PH/10 Added the log_reject_target ACL modifier to specify where to log
3032 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3033 hostname. This is wrong, because it relates to the incoming message (and
3034 probably the interface on which it is arriving) and not to the outgoing
3035 callout (which could be using a different interface). This has been
3036 changed to use the value of the helo_data option from the smtp transport
3037 instead - this is what is used when a message is actually being sent. If
3038 there is no remote transport (possible with a router that sets up host
3039 addresses), $smtp_active_hostname is used.
3041 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3042 tweaks were necessary in order to get it to work (see also 21 below):
3043 (a) The code assumed that strncpy() returns a negative number on buffer
3044 overflow, which isn't the case. Replaced with Exim's string_format()
3046 (b) There were several signed/unsigned issues. I just did the minimum
3047 hacking in of casts. There is scope for a larger refactoring.
3048 (c) The code used strcasecmp() which is not a standard C function.
3049 Replaced with Exim's strcmpic() function.
3050 (d) The code set only $1; it now sets $auth1 as well.
3051 (e) A simple test gave the error "authentication client didn't specify
3052 service in request". It would seem that Dovecot has changed its
3053 interface. Fortunately there's a specification; I followed it and
3054 changed what the client sends and it appears to be working now.
3056 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3059 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3060 address A is aliased to B and C, where B exists and C does not. Without
3061 -v the output is "A verified" because verification stops after a
3062 successful redirection if more than one address is generated. However,
3063 with -v the child addresses are also verified. Exim was outputting "A
3064 failed to verify" and then showing the successful verification for C,
3065 with its parentage. It now outputs "B failed to verify", showing B's
3066 parentage before showing the successful verification of C.
3068 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3069 look up a TXT record in a specific list after matching in a combined
3072 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3073 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3074 they consult the DNS. I had assumed they would set it the way they
3075 wanted; and indeed my experiments on Linux seem to show that in some
3076 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3077 To be on the safe side, however, I have now made the interface to
3078 host_find_byname() similar to host_find_bydns(), with an argument
3079 containing the DNS resolver options. The host_find_byname() function now
3080 sets these options at its start, just as host_find_bydns() does. The smtp
3081 transport options dns_qualify_single and dns_search_parents are passed to
3082 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3083 of host_find_byname() use the default settings of RES_DEFNAMES
3084 (qualify_single) but not RES_DNSRCH (search_parents).
3086 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3087 spool_read_header() do less string testing, by means of a preliminary
3088 switch on the second character of optional "-foo" lines. (This is
3089 overdue, caused by the large number of possibilities that now exist.
3090 Originally there were few.) While I was there, I also converted the
3091 str(n)cmp tests so they don't re-test the leading "-" and the first
3092 character, in the hope this might squeeze out yet more improvement.
3094 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3095 flag allowing group syntax was set by the header_syntax check but not
3096 turned off, possible causing trouble later; (2) The flag was not being
3097 set at all for the header_verify test, causing "group"-style headers to
3098 be rejected. I have now set it in this case, and also caused header_
3099 verify to ignore an empty address taken from a group. While doing this, I
3100 came across some other cases where the code for allowing group syntax
3101 while scanning a header line wasn't quite right (mostly, not resetting
3102 the flag correctly in the right place). These bugs could have caused
3103 trouble for malformed header lines. I hope it is now all correct.
3105 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3106 with the "reply" argument non-NULL. The code, however (which originally
3107 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3108 but it didn't always do it. This confused somebody who was copying the
3109 code for some other use. I have removed all the tests.
3111 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3112 feature that was used to support insecure browsers during the U.S. crypto
3113 embargo. It requires special client support, and Exim is probably the
3114 only MTA that supported it -- and would never use it because real RSA is
3115 always available. This code has been removed, because it had the bad
3116 effect of slowing Exim down by computing (never used) parameters for the
3117 RSA_EXPORT functionality.
3119 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3120 authenticator to fail if there's a tab character in the incoming data
3121 (there should never be unless someone is messing about, as it's supposed
3122 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3123 if the connection is using TLS or if the remote IP is the same as the
3124 local IP, and the "valid-client-cert option" if a client certificate has
3127 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3128 authenticators. This can be used for authorization after authentication
3129 succeeds. (In the case of plaintext, it servers for both authentication
3132 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3133 if any retry times were supplied.
3135 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3136 connection, where there is no client IP address to check. In this
3137 situation, the verify now always succeeds.
3139 PH/25 Applied John Jetmore's -Mset patch.
3141 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3143 PH/27 In a string expansion for a processed (not raw) header when multiple
3144 headers of the same name were present, leading whitespace was being
3145 removed from all of them, but trailing whitespace was being removed only
3146 from the last one. Now trailing whitespace is removed from each header
3147 before concatenation. Completely empty headers in a concatenation (as
3148 before) are ignored.
3150 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3151 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3153 PH/29 [Removed. This was a change that I later backed out, and forgot to
3154 correct the ChangeLog entry (that I had efficiently created) before
3155 committing the later change.]
3157 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3158 address errors (4xx response to RCPT) over the same connection as other
3159 messages routed to the same hosts. Such deliveries are always "forced",
3160 so retry times are not inspected. This resulted in far too many retries
3161 for the affected addresses. The effect occurred only when there were more
3162 hosts than the hosts_max_try setting in the smtp transport when it had
3163 the 4xx errors. Those hosts that it had tried were not added to the list
3164 of hosts for which the message was waiting, so if all were tried, there
3165 was no problem. Two fixes have been applied:
3167 (i) If there are any address or message errors in an SMTP delivery, none
3168 of the hosts (tried or untried) are now added to the list of hosts
3169 for which the message is waiting, so the message should not be a
3170 candidate for sending over the same connection that was used for a
3171 successful delivery of some other message. This seems entirely
3172 reasonable: after all the message is NOT "waiting for some host".
3173 This is so "obvious" that I'm not sure why it wasn't done
3174 previously. Hope I haven't missed anything, but it can't do any
3175 harm, as the worst effect is to miss an optimization.
3177 (ii) If, despite (i), such a delivery is accidentally attempted, the
3178 routing retry time is respected, so at least it doesn't keep
3179 hammering the server.
3181 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3182 in ${readsocket because some servers need this prod.
3184 PH/32 Added some extra debug output when updating a wait-xxx database.
3186 PH/33 The hint "could be header name not terminated by colon", which has been
3187 given for certain expansion errors for a long time, was not being given
3188 for the ${if def:h_colon_omitted{... case.
3190 PH/34 The spec says: "With one important exception, whenever a domain list is
3191 being scanned, $domain contains the subject domain." There was at least
3192 one case where this was not true.
3194 PH/35 The error "getsockname() failed: connection reset by peer" was being
3195 written to the panic log as well as the main log, but it isn't really
3196 panic-worthy as it just means the connection died rather early on. I have
3197 removed the panic log writing for the ECONNRESET error when getsockname()
3200 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3201 runs only) independently of the message's sender address. This meant
3202 that, if the 4xx error was in fact related to the sender, a different
3203 message to the same recipient with a different sender could confuse
3204 things. In particular, this can happen when sending to a greylisting
3205 server, but other circumstances could also provoke similar problems.
3206 I have changed the default so that the retry time for these errors is now
3207 based a combination of the sender and recipient addresses. This change
3208 can be overridden by setting address_retry_include_sender=false in the
3211 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3212 remote server are returned as part of bounce messages. This was not
3213 happening for LMTP over a pipe (the lmtp transport), but now it is the
3214 same for both kinds of LMTP.
3216 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3217 in header lines that were in fact CNAMEs. This is no longer the case.
3219 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3220 and queue runs started by the daemon processed all messages. This has
3221 been fixed so that -R and -S can now usefully be given with -q<time>.
3223 PH/40 Import PCRE release 6.7 (fixes some bugs).
3225 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3227 PH/42 Give an error if -q is specified more than once.
3229 PH/43 Renamed the variables $interface_address and $interface_port as
3230 $received_ip_address and $received_port, to make it clear that these
3231 values apply to message reception, and not to the outgoing interface when
3232 a message is delivered. (The old names remain recognized, of course.)
3234 PH/44 There was no timeout on the connect() call when using a Unix domain
3235 socket in the ${readsocket expansion. There now is.
3237 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3238 be meaningful with "accept".
3240 SC/01 Eximstats V1.43
3241 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3243 SC/02 Eximstats V1.44
3244 Use a glob alias rather than an array ref in the generated
3245 parser. This improves both readability and performance.
3247 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3248 Collect SpamAssassin and rejection statistics.
3249 Don't display local sender or destination tables unless
3250 there is data to show.
3251 Added average volumes into the top table text output.
3253 SC/04 Eximstats V1.46
3254 Collect data on the number of addresses (recipients)
3255 as well as the number of messages.
3257 SC/05 Eximstats V1.47
3258 Added 'Message too big' to the list of mail rejection
3259 reasons (thanks to Marco Gaiarin).
3261 SC/06 Eximstats V1.48
3262 Mainlog lines which have GMT offsets and are too short to
3263 have a flag are now skipped.
3265 SC/07 Eximstats V1.49 (Alain Williams)
3266 Added the -emptyok flag.
3268 SC/08 Eximstats V1.50
3269 Fixes for obtaining the IP address from reject messages.
3271 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3272 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3273 whitespace changes from 4.64-PH/27
3275 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3278 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3279 are found, allow negative numbers in numeric criteria)
3281 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3283 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3284 to match changes made in 4.64-PH/43
3286 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3288 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3289 "domain" is actually an IP address, and force a failure. This locks out
3290 those revolvers/nameservers that support "A-for-A" lookups, in
3291 contravention of the specifications.
3293 PH/48 When a host name was looked up from an IP address, and the subsequent
3294 forward lookup of the name timed out, the host name was left in
3295 $sender_host_name, contrary to the specification.
3297 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3298 restricted to single-key lookups, Exim was not diagnosing an error if
3299 * or *@ was used with a query-style lookup.
3301 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3303 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3304 long ago, but noone interested enough thought of it. Let's just say that
3305 the "1.1" means that there are some new functions that weren't there at
3306 some point in the past.
3308 PH/51 Error processing for expansion failure of helo_data from an smtp
3309 transport during callout processing was broken.
3311 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3312 tested/used via the -bh/-bhc/-bs options.
3314 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3315 bug, fixed in subsequent PCRE releases).
3317 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3318 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3320 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3326 SC/01 Use a glob alias rather than an array ref in eximstats generated
3327 parser. This improves both readability and performance.
3329 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3330 Don't display local sender or destination tables in eximstats unless
3331 there is data to show.
3332 Added average volumes into the eximstats top table text output.
3334 SC/03 Collect data on the number of addresses (recipients) as well
3335 as the number of messages in eximstats.
3337 TF/01 Correct an error in the documentation for the redirect router. Exim
3338 does (usually) call initgroups() when daemonizing.
3340 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3341 with consistent privilege compared to when running as a daemon.
3343 TF/03 Note in the spec that $authenticated_id is not set for local
3344 submissions from trusted users.
3346 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3347 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3349 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3350 by adding some example configuration directives to the default
3351 configuration file. A little bit of work is required to uncomment the
3352 directives and define how usernames and passwords are checked, but
3353 there is now a framework to start from.
3355 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3356 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3357 without this. I don't know how relevant this is to other LDAP libraries.
3359 PH/02 Add the verb name to the "unknown ACL verb" error.
3361 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3363 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3365 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3366 directory not expanded when it should be if an expanded home directory
3367 was set for the address (which is overridden by the transport).
3369 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3372 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3373 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3374 because it is too late at that time, and has no effect.
3376 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3377 security issue with \' (bugzilla #107). I could not use the
3378 PQescapeStringConn() function, because it needs a PGconn value as one of
3381 PH/08 When testing addresses using -bt, indicate those final addresses that
3382 are duplicates that would not cause an additional delivery. At least one
3383 person was confused, thinking that -bt output corresponded to deliveries.
3384 (Suppressing duplicates isn't a good idea as you lose the information
3385 about possibly different redirections that led to the duplicates.)
3387 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3388 systems where poll() doesn't work, in particular OS X.
3390 PH/10 Added more information to debugging output for retry time not reached.
3392 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3393 operations in malware.c.
3395 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3398 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3399 syslog_duplication set false, log lines that would normally be written
3400 both the the main log and to the reject log were not written to syslog at
3403 PH/14 In the default configuration, change the use of "message" in ACL warn
3404 statements to "add_header".
3406 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3407 not followed by a command (e.g. "seen endif").
3409 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3410 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3413 PH/17 Added extra conditions to the default value of delay_warning_condition
3417 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3418 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3419 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3422 The Auto-Submitted: and various List- headers are standardised, whereas I
3423 don't think Precedence: ever was.
3425 PH/18 Refactored debugging code in route_finduser() to show more information,
3426 in particular, the error code if getpwnam() issues one.
3428 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3429 This is apparently needed in addition to the PH/07 change above to avoid
3430 any possible encoding problems.
3432 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3433 but not after initializing Perl.
3435 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3436 output them only if debugging. By default they are written stderr,
3437 apparently, which is not desirable.
3439 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3442 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3445 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3447 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3448 authenticated or an ident call has been made. Suppress the default
3449 values for $authenticated_id and $authenticated_sender (but permit -oMai
3450 and -oMas) when testing with -bh.
3452 PH/24 Re-jigged the order of the tests in the default configuration so that the
3453 tests for valid domains and recipients precede the DNS black list and CSA
3454 tests, on the grounds that those ones are more expensive.
3456 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3457 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3458 This bug exists in every version of Exim that I still have, right back to
3461 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3462 However, an attempt to turn on case-sensitivity in a regex key by
3463 including (?-i) didn't work because the subject string was already
3464 lowercased, and the effects were non-intuitive. It turns out that a
3465 one-line patch can be used to allow (?-i) to work as expected.
3471 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3472 other effects) broke the use of negated acl sub-conditions.
3474 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3477 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3478 "Deny" causes Exim to reject the incoming connection with a 554 error.
3479 Unfortunately, if there is a major crisis, such as a disk failure,
3480 tcp-wrappers gives "deny", whereas what one would like would be some
3481 kind of temporary error. A kludge has been added to help with this.
3482 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3483 554 error is used if errno is still zero or contains ENOENT (which occurs
3484 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3487 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3489 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3490 errors. Otherwise a message that provokes a temporary error (when other
3491 messages do not) can cause a whole host to time out.
3493 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3494 addresses were routed directly to files or pipes from a redirect router.
3495 File deliveries just didn't batch; pipe deliveries might have suffered
3498 PH/06 A failure to get a lock for a hints database would erroneously always say
3499 "Failed to get write lock", even when it was really a read lock.
3501 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3502 of 0600. This has been changed to use the value of the lockfile_mode
3503 option (which defaults to 0600).
3505 PH/08 Applied small patch from the Sieve maintainer.
3507 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3508 folder from quota calculations, a direct delivery into this folder messed
3509 up the contents of the maildirsize file. This was because the regex was
3510 used only to exclude .Trash (or whatever) when the size of the mailbox
3511 was calculated. There was no check that a delivery was happening into an
3512 excluded directory. This bug has been fixed by ignoring all quota
3513 processing for deliveries into excluded directories.
3515 PH/10 Added the maildirfolder_create_regex option to appendfile.
3521 PH/01 The code for finding all the local interface addresses on a FreeBSD
3522 system running IPv6 was broken. This may well have applied to all BSD
3523 systems, as well as to others that have similar system calls. The broken
3524 code found IPv4 interfaces correctly, but gave incorrect values for the
3525 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3526 that it would not match correctly against @[] and not recognize the IPv6
3529 PH/02 The ipliteral router was not recognizing addresses of the form user@
3530 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3532 PH/03 Added disable_ipv6.
3534 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3535 decoded form, because it is most often used to construct To: headers
3536 lines in autoreplies, and the decoded form may well be syntactically
3537 invalid. However, $reply_address has leading white space removed, and all
3538 newlines turned into spaces so that the autoreply transport does not
3541 PH/05 If group was specified without a user on a router, and no group or user
3542 was specified on a transport, the group from the router was ignored.
3544 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3545 for visible compile-time settings that can be used to change these
3546 numbers, for those that want even more. Backwards compatibility with old
3547 spool files has been maintained. However, going back to a previous Exim
3548 release will lost any variables that are in spool files.
3550 PH/07 Two small changes when running in the test harness: increase delay when
3551 passing a TCP/IP connection to a new process, in case the original
3552 process has to generate a bounce, and remove special handling of
3553 127.0.0.2 (sic), which is no longer necessary.
3555 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3556 be the same on different OS.
3558 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3561 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3562 whether --show-vars was specified or not
3564 JJ/02 exipick: Added support for new ACL variable spool format introduced
3567 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3568 syntactically invalid From: or Reply-to: line, and a filter used this to
3569 generate an autoreply, and therefore failed to obtain an address for the
3570 autoreply, Exim could try to deliver to a non-existent relative file
3571 name, causing unrelated and misleading errors. What now happens is that
3572 it logs this as a hard delivery error, but does not attempt to create a
3575 PH/11 The exinext utility has a -C option for testing purposes, but although
3576 the given file was scanned by exinext itself; it wasn't being passed on
3577 when Exim was called.
3579 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3580 an end-of-file indication when reading a command response.
3582 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3583 compiled. In many other places in Exim, IPv6 addresses are always
3584 recognized, so I have changed this. It also means that IPv4 domain
3585 literals of the form [IPV4:n.n.n.n] are now always recognized.
3587 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3588 used if the router is not running as root, for example, when verifying at
3589 ACL time, or when using -bh. The debugging output from this situation was
3590 non-existent - all you got was a failure to exec. I have made two
3593 (a) Failures to set uid/gid, the current directory, or a process leader
3594 in a subprocess such as that created by queryprogram now generate
3595 suitable debugging output when -d is set.
3597 (b) The queryprogram router detects when it is not running as root,
3598 outputs suitable debugging information if -d is set, and then runs
3599 the subprocess without attempting to change uid/gid.
3601 PH/15 Minor change to Makefile for building test_host (undocumented testing
3604 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3605 additional section of a DNS packet that returns MX or SRV records.
3606 Instead, it always explicitly searches for A/AAAA records. This avoids
3607 major problems that occur when a DNS server includes only records of one
3608 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3609 fixed another bug: if SRV records were looked up and the corresponding
3610 address records were *not* found in the additional section, the port
3611 values from the SRV records were lost.
3613 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3614 using the correct key (the original address) when searching the retry
3615 rules in order to find which one to use for generating the retry hint.
3617 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3618 adding the default one. Similarly, if it contains a Reply-To: header, the
3619 errors_reply_to option, if set, is not used.
3621 PH/19 When calculating a retry time, Exim used to measure the "time since
3622 failure" by looking at the "first failed" field in the retry record. Now
3623 it does not use this if it is later than than the arrival time of the
3624 message. Instead it uses the arrival time. This makes for better
3625 behaviour in cases where some deliveries succeed, thus re-setting the
3626 "first failed" field. An example is a quota failure for a huge message
3627 when small messages continue to be delivered. Without this change, the
3628 "time since failure" will always be short, possible causing more frequent
3629 delivery attempts for the huge message than are intended.
3630 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3632 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3633 $1, $2, $3) because the numerical variables can be reset during some
3634 expansion items (e.g. "match"), thereby losing the authentication data.
3636 PH/21 Make -bV show the size of off_t variables so that the test suite can
3637 decide whether to run tests for quotas > 2G.
3639 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3640 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3641 filecount value is greater than 2G or if a quota value is greater than 2G
3642 on a system where the size of off_t is not greater than 4, a panic error
3645 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3646 never match. The debug and -bh output now contains an explicit error
3647 message indicating a malformed IPv4 address or mask.
3649 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3650 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3651 PH/23 above applies.
3653 PH/25 Do not write to syslog when running in the test harness. The only
3654 occasion when this arises is a failure to open the main or panic logs
3655 (for which there is an explicit test).
3657 PH/26 Added the /no_tell option to "control=freeze".
3659 PH/27 If a host name lookup failed very early in a connection, for example, if
3660 the IP address matched host_lookup and the reverse lookup yielded a name
3661 that did not have a forward lookup, an error message of the form "no IP
3662 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3663 could be logged. Now it outputs the IP address instead of "NULL".
3665 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3666 allows the sender and the authenticated sender to be set when
3667 submitting a message from within Exim. Since child_open_exim() is
3668 documented for local_scan(), the new function should be too.
3670 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3671 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3672 results in an empty string is now treated as unset.
3674 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3676 PH/31 Added sender_verify_fail logging option.
3678 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3679 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3680 tidied the source and removed it altogether.
3682 PH/33 When a queue run was abandoned because the load average was too high, a
3683 log line was always written; now it is written only if the queue_run log
3684 selector is set. In addition, the log line for abandonment now contains
3685 information about the queue run such as the pid. This is always present
3686 in "start" and "stop" lines but was omitted from the "abandon" line.
3688 PH/34 Omit spaces between a header name and the colon in the error message that
3689 is given when verify = headers_syntax fails (if there are lots of them,
3690 the message gets confusing).
3692 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3693 names, as there are now some PTR records that contain slashes. This check
3694 is only to protect against broken name servers that fall over on strange
3695 characters, so the fact that it applies to all lookups doesn't matter.
3697 PH/36 Now that the new test suite is complete, we can remove some of the
3698 special code in Exim that was needed for the old test suite. For example,
3699 sorting DNS records because real resolvers return them in an arbitrary
3700 order. The new test suite's fake resolver always returns records in the
3703 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3704 bounces) except when queue_only is set, to avoid logging races between
3705 the different processes.
3707 PH/38 Panic-die if .include specifies a non-absolute path.
3709 PH/39 A tweak to the "H" retry rule from its user.
3711 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3712 a label. They prevented compilation on older perls.
3714 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3715 a warning to be raised on newish perls.
3717 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3718 on queue. Changes to match documented behaviour of showing count of
3719 messages matching specified criteria.
3721 PH/40 Changed the default ident timeout from 30s to 5s.
3723 PH/41 Added support for the use of login_cap features, on those BSD systems
3724 that have them, for controlling the resources used by pipe deliveries.
3726 PH/42 The content-scanning code uses fopen() to create files in which to put
3727 message data. Previously it was not paying any attention to the mode of
3728 the files. Exim runs with umask(0) because the rest of the code creates
3729 files with open(), and sets the required mode explicitly. Thus, these
3730 files were ending up world-writeable. This was not a big issue, because,
3731 being within the spool directory, they were not world-accessible. I have
3732 created a function called modefopen, which takes an additional mode
3733 argument. It sets umask(777), creates the file, chmods it to the required
3734 mode, then resets the umask. All the relevant calls to fopen() in the
3735 content scanning code have been changed to use this function.
3737 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3738 to 24 hours. This avoids potential overflow problems when processing G
3739 and H retry rules. I suspect nobody ever tinkers with this value.
3741 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3743 PH/45 When the plaintext authenticator is running as a client, the server's
3744 challenges are checked to ensure they are valid base64 strings. By
3745 default, the authentication attempt is cancelled if an invalid string is
3746 received. Setting client_ignore_invalid_base64 true ignores these errors.
3747 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3748 they are received. Thus, the responses can be made to depend on the
3749 challenges. If an invalid string is ignored, an empty string is placed in
3752 PH/46 Messages that are created by the autoreply transport now contains a
3753 References: header, in accordance with RFCs 2822 and 3834.
3755 PH/47 Added authenticated_sender_force to the smtp transport.
3757 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3759 PH/49 Installed latest patch from the Sieve maintainer.
3761 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3762 was also set, the appendfile transport was unnecessarily scanning a
3763 directory of message files (e.g. for maildir delivery) to find the count
3764 of files (along with the size), even though it did not need this
3765 information. It now does the scan only if it needs to find either the
3766 size of the count of files.
3768 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3770 PH/52 Two bugs concerned with error handling when the smtp transport is
3773 (i) Exim was not creating retry information for temporary errors given
3774 for individual recipients after the DATA command when the smtp transport
3775 was used in LMTP mode. This meant that they could be retried too
3776 frequently, and not timed out correctly.
3778 (ii) Exim was setting the flag that allows error details to be returned
3779 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3780 recipients that were returned after the DATA command.
3782 PH/53 This is related to PH/52, but is more general: for any failing address,
3783 when detailed error information was permitted to be returned to the
3784 sender, but the error was temporary, then after the final timeout, only
3785 "retry timeout exceeded" was returned. Now it returns the full error as
3786 well as "retry timeout exceeded".
3788 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3789 do this, and (what is worse) MTAs that accept it.
3791 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3792 will now be deprecated.
3794 PH/56 New os.c-cygwin from the Cygwin maintainer.
3796 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3797 formats (previously only available in exim formats via -bpr, -bpru,
3798 and -bpra. Now also available in native and exiqgrep formats)
3800 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3801 with very large, slow to parse queues
3803 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3805 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3807 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3808 responses to authentication challenges, though it was showing the
3809 challenges; (ii) I've removed the CR characters from the debug output for
3812 PH/58 Allow for the insertion of a newline as well as a space when a string
3813 is turned into more than one encoded-word during RFC 2047 encoding. The
3814 Sieve code now uses this.
3816 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3817 data_4xx, lost_connection, tls_required.
3819 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3820 message was being sent as an SMTP response.
3822 PH/61 Add -l and -k options to exicyclog.
3824 PH/62 When verifying, if an address was redirected to one new address, so that
3825 verification continued, and the new address failed or deferred after
3826 having set something in $address_data, the value of $address_data was not
3827 passed back to the ACL. This was different to the case when no
3828 redirection occurred. The value is now passed back in both cases.
3830 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3831 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3832 use login_cap.h, so on its own it isn't the distinguishing feature. The
3833 new name refers directly to the setclassresources() function.
3835 PH/65 Added configuration files for NetBSD3.
3837 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3839 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3840 is preferred over IPv4.
3842 PH/68 The bounce_return_message and bounce_return_body options were not being
3843 honoured for bounces generated during the reception of non-SMTP messages.
3844 In particular, this applied to messages rejected by the ACL. This bug has
3845 been fixed. However, if bounce_return_message is true and bounce_return_
3846 body is false, the headers that are returned for a non-SMTP message
3847 include only those that have been read before the error was detected.
3848 (In the case of an ACL rejection, they have all been read.)
3850 PH/69 The HTML version of the specification is now built in a directory called
3851 spec_html instead of spec.html, because the latter looks like a path with
3852 a MIME-type, and this confuses some software.
3854 PH/70 Catch two compiler warnings in sieve.c.
3856 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3857 function verify_get_ident() calls ip_connect() to connect a socket, but
3858 if the "connect()" function timed out, ip_connect() used to close the
3859 socket. However, verify_get_ident() also closes the socket later, and in
3860 between Exim writes to the log, which may get opened at this point. When
3861 the socket was closed in ip_connect(), the log could get the same file
3862 descriptor number as the socket. This naturally causes chaos. The fix is
3863 not to close the socket in ip_connect(); the socket should be closed by
3864 the function that creates it. There was only one place in the code where
3865 this was missing, in the iplookup router, which I don't think anybody now
3866 uses, but I've fixed it anyway.
3868 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3869 well as to direct DNS lookups. Otherwise the handling of names in host
3870 lists is inconsistent and therefore confusing.
3876 PH/01 Two changes to the default runtime configuration:
3878 (1) Move the checks for relay_from_hosts and authenticated clients from
3879 after to before the (commented out) DNS black list checks.
3881 (2) Add control=submission to the relay_from_hosts and authenticated
3882 clients checks, on the grounds that messages accepted by these
3883 statements are most likely to be submissions.
3885 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3887 (1) Generate an error if the third argument for the ${prvs expansion is
3890 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3893 (3) Reset the variables that are obtained from the first argument of
3894 ${prvscheck and used in the second argument before leaving the code,
3895 because their memory is reclaimed, so using them afterwards may do
3898 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3899 one (it's much easier than Tom thought :-).
3901 (5) Because of (4), we can now allow for the use of $prvscheck_result
3902 inside the third argument.
3904 PH/03 For some reason, the default setting of PATH when running a command from
3905 a pipe transport was just "/usr/bin". I have changed it to
3908 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3909 anything to be listed in the output from -bV.
3911 PH/05 When a filter generated an autoreply, the entire To: header line was
3912 quoted in the delivery log line, like this:
3914 => >A.N.Other <ano@some.domain> <original@ddress> ...
3916 This has been changed so that it extracts the operative address. There
3917 may be more than one such address. If so, they are comma-separated, like
3920 => >ano@some.domain,ona@other.domain <original@ddress> ...
3922 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3923 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3924 not being looked up in the rDNS to get a host name, Exim was showing the
3925 IP address twice in Received: lines, even though the IP addresses were
3926 identical. For example:
3928 Received: from [1.2.3.4] (helo=[1.2.3.4])
3930 However, if the real host name was known, it was omitting the HELO data
3931 if it matched the actual IP address. This has been tidied up so that it
3932 doesn't show the same IP address twice.
3934 PH/07 When both +timestamp and +memory debugging was on, the value given by
3935 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3936 called by the debug printing, thereby overwriting the timestamp buffer.
3937 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3939 PH/08 When the original message was included in an autoreply transport, it
3940 always said "this is a copy of the message, including all the headers",
3941 even if body_only or headers_only was set. It now gives an appropriate
3944 PH/09 Applied a patch from the Sieve maintainer which:
3946 o fixes some comments
3947 o adds the (disabled) notify extension core
3948 o adds some debug output for the result of if/elsif tests
3949 o points to the current vacation draft in the documentation
3950 and documents the missing references header update
3954 o fixes a bug in processing the envelope test (when testing
3955 multiple envelope elements, the last element determined the
3958 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3959 Electronic Mail") by including:
3961 Auto-submitted: auto-generated
3963 in the messages that it generates (bounce messages and others, such as
3964 warnings). In the case of bounce messages for non-SMTP messages, there was
3965 also a typo: it was using "Auto_submitted" (underscore instead of
3966 hyphen). Since every message generated by Exim is necessarily in response
3967 to another message, thes have all been changed to:
3969 Auto-Submitted: auto-replied
3971 in accordance with these statements in the RFC:
3973 The auto-replied keyword:
3975 - SHOULD be used on messages sent in direct response to another
3976 message by an automatic process,
3978 - MUST NOT be used on manually-generated messages,
3980 - MAY be used on Delivery Status Notifications (DSNs) and Message
3981 Disposition Notifications (MDNs),
3983 - MUST NOT be used on messages generated by automatic or periodic
3984 processes, except for messages which are automatic responses to
3987 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3988 to the default Received: header definition.
3990 PH/12 Added log selector acl_warn_skipped (default on).
3992 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3993 variables because (a) they are in the wrong storage pool and (b) even if
3994 they were copied, it wouldn't work properly because of the caching.
3996 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3997 checking when decoding. Apparently there are clients that generate
3998 overlong encoded strings. Why am I not surprised?
4000 PH/15 If the first argument of "${if match_address" was not empty, but did not
4001 contain an "@" character, Exim crashed. Now it writes a panic log message
4002 and treats the condition as false.
4004 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4006 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4007 extension "envelope-auth". The code is finished and in agreement with
4008 other implementations, but there is no documentation so far and in fact,
4009 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4010 not changing the active code.
4012 Print executed "if" and "elsif" statements when debugging is used. This
4013 helps a great deal to understand what a filter does.
4015 Document more things not specified clearly in RFC3028. I had all this
4016 sorted out, when out of a sudden new issues came to my mind. Oops."
4018 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4021 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4022 canonical form (as documented). However, after a host name lookup from
4023 the IP address, check_host() was doing a simple string comparison with
4024 addresses acquired from the DNS when checking that the found name did
4025 have the original IP as one of its addresses. Since any found IPv6
4026 addresses are likely to be in abbreviated form, the comparison could
4027 fail. Luckily, there already exists a function for doing the comparison
4028 by converting both addresses to binary, so now that is used instead of
4029 the text comparison.
4031 PH/20 There was another similar case to PH/19, when a complete host name was
4032 given in a host list; looking up its IP address could give an abbreviated
4033 form, whereas the current host's name might or might not be abbreviated.
4034 The same fix has been applied.
4040 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4041 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4044 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4045 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4047 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4049 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4050 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4051 Also, for "Auto-Submitted", treat anything other than "no" as
4052 non-personal, in accordance with RFC 3834. (Previously it treated
4053 anything starting "auto-" as non-personal.)
4055 TF/01 The control=submission/name=... option had a problem with syntax
4056 errors if the name included a slash character. The /name= option
4057 now slurps the rest of the string, so it can include any characters
4058 but it must come last in the list of options (after /sender_retain
4061 PH/05 Some modifications to the interface to the fake nameserver for the new
4069 TK/01 Added the "success_on_redirect" address verification option. See
4070 NewStuff for rationale and an example.
4072 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4074 PH/02 Patch to exigrep to allow it to work on syslog lines.
4076 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4077 fread() to skip over the body file's header line, because in Cygwin the
4078 header line is locked and is inaccessible.
4080 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4081 co-exist for some time) to make it clear that it is the Exim ID that is
4082 referenced, not the Message-ID: header line.
4084 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4085 string_format() function, because snprintf() does not exist on all
4088 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4089 ${stat: expansion item.
4091 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4092 protocol synchronization error", to keep the pedants happy.
4094 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4095 well as for IRIX systems, when gcc is being used. See the host.c source
4098 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4100 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4103 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4104 transport and to the smtp transport in LMTP mode.
4106 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4108 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4109 run a filter in a subprocess. This could lead to confusion in subsequent
4110 lookups in the parent process. There should also be a search_tidyup() at
4111 the end of the subprocess.
4113 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4114 only if the host matched helo_try_verify_hosts, which caused the
4115 verification to occur when the EHLO/HELO command was issued. The ACL just
4116 tested the remembered result. Now, if a previous verification attempt has
4117 not happened, "verify = helo" does it there and then.
4119 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4121 TK/03 Fix log output including CR from clamd.
4123 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4124 exist provoked a memory error which could cause a segfault.
4126 PH/15 Installed PCRE 6.2
4128 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4130 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4131 of the problem. Specifically, suggested +O2 rather than +O1 for the
4134 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4136 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4137 record, the port was not correctly logged when the outgoing_port log
4138 selector was set (it logged the transort's default port).
4140 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4141 fallback_hosts, and "hosts" in the smtp transport.
4143 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4144 host errors such as "Connection refused".
4146 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4147 authentication with radiusclient 0.4.9:
4149 - Error returned from rc_read_config was caught wrongly
4150 - Username/password not passed on to radius server due to wrong length.
4152 The presumption is that some radiusclient API changes for 4.51/PH/17
4153 were not taken care of correctly. The code is still untested by me (my
4154 Linux distribution still has 0.3.2 of radiusclient), but it was
4155 contributed by a Radius user.
4157 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4158 expanding the "port" option of the smtp transport.
4160 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4161 while reading a MIME header. Thanks to Tom Hughes for a patch.
4163 PH/24 Include config.h inside local_scan.h so that configuration settings are
4166 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4167 that in an ACL for RCPT (for example), you can examine exactly what was
4170 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4171 commands, but it was not correctly comparing the address with the actual
4172 client host address. Thus, it would show the EHLO address in Received:
4173 header lines when this was not necessary.
4175 PH/27 Added the % operator to ${eval:}.
4177 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4178 it should be ignoring failures (because with -C, for example, it has lost
4179 privilege). It wasn't ignoring creation failures other than "already
4182 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4185 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4186 because an input line was too long, either on its own, or by virtue of
4187 too many continuations, the temporary file was not being removed, and the
4188 return code was incorrect.
4190 PH/31 Missing "BOOL" in function definition in filtertest.c.
4192 PH/32 Applied Sieve patches from the maintainer.
4194 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4196 PH/33 Added "verify = not_blind".
4198 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4199 Local/Makefile (with some defaults set). These are used in built scripts
4200 such as exicyclog, but they have never been used in the exim_install
4201 script (though there are many overriding facilities there). I have
4202 arranged that the exim_install script now takes note of these two
4205 PH/35 Installed configuration files for Dragonfly.
4207 PH/36 When a locally submitted message by a trusted user did not contain a
4208 From: header, and the sender address was obtained from -f or from an SMTP
4209 MAIL command, and the trusted user did not use -F to supply a sender
4210 name, $originator_name was incorrectly used when constructing a From:
4211 header. Furthermore, $originator_name was used for submission mode
4212 messages from external hosts without From: headers in a similar way,
4213 which is clearly wrong.
4215 PH/37 Added control=suppress_local_fixups.
4217 PH/38 When log_selector = +received_sender was set, and the addition of the
4218 sender made the log line's construction buffer exactly full, or one byte
4219 less than full, an overflow happened when the terminating "\n" was
4222 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4223 when the result of a list match is failure because a DNS lookup failed.
4225 PH/40 RM_COMMAND is now used in the building process.
4227 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4228 the "build-* directories that it finds.
4230 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4231 address in a domain literal was a prefix of an interface address.
4233 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4234 when verifying a sender address, unless rewrite_headers is false.
4236 PH/44 Wrote a long comment about why errors_to addresses are verified as
4237 recipients, not senders.
4239 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4240 the ratelimit ACL was added.
4242 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4244 PH/46 Added extra information about PostgreSQL errors to the error string.
4246 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4247 avoiding the need to install special zones in a real server. This is
4248 backwards compatible; if it can't find the fake resolver, it drops back.
4249 Thus, both old and new test suites can be run.
4251 TF/02 Added util/ratelimit.pl
4253 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4254 clock is set back in time.
4256 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4257 Candler <B.Candler@pobox.com>.
4259 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4260 OK for addresses that are the result of redirections.
4262 PH/48 A number of further additions for the benefit of the new test suite,
4263 including a fake gethostbyname() that interfaces to the fake DNS resolver
4266 TF/06 The fix for widen_domains has also been applied to qualify_single and
4267 search_parents which are the other dnslookup options that can cause
4270 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4273 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4275 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4276 service" if it encountered an MX record with an empty target hostname.
4277 The message is now "an MX or SRV record indicated no SMTP service".
4279 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4280 if the DNS of the sending site is misconfigured. This is quite a
4281 common situation. This change restores the behaviour of treating a
4282 helo verification defer as a failure.
4284 PH/51 If self=fail was set on a router, the bounce message did not include the
4285 actual error message.
4291 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4293 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4294 command itself ended in error, the underlying message about the transport
4295 filter timeout was being overwritten with the pipe command error. Now the
4296 underlying error message should be appended to the second error message.
4298 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4300 PH/02 Reduce the amount of output that "make" produces by default. Full output
4301 can still be requested.
4303 PH/03 The warning log line about a condition test deferring for a "warn" verb
4304 was being output only once per connection, rather than after each
4305 occurrence (because it was using the same function as for successful
4306 "warn" verbs). This seems wrong, so I have changed it.
4308 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4309 it should not have, which might have caused a crash in the right
4310 circumstances, but probably never did.
4312 PH/04 Installed a modified version of Tony Finch's patch to make submission
4313 mode fix the return path as well as the Sender: header line, and to
4314 add a /name= option so that you can make the user's friendly name appear
4317 TF/03 Added the control = fakedefer ACL modifier.
4319 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4320 Mark Lowes for thorough testing.
4322 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4324 TK/03 Merged latest SRS patch from Miles Wilton.
4326 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4327 with the definition in sysexits.h (which is #included earlier).
4328 Fortunately, Exim does not actually use EX_OK. The code used to try to
4329 preserve the sysexits.h value, by assuming that macro definitions were
4330 scanned for macro replacements. I have been disabused of this notion,
4331 so now the code just undefines EX_OK before #including unistd.h.
4333 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4334 in the smtp transport. When a block could not be written in a single
4335 write() function, the timeout was being re-applied to each part-write.
4336 This seems wrong - if the receiver was accepting one byte at a time it
4337 would take for ever. The timeout is now adjusted when this happens. It
4338 doesn't have to be particularly precise.
4340 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4341 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4343 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4344 without a domain if the check to <postmaster@domain> fails.
4346 SC/01 Eximstats: added -xls and the ability to specify output files
4347 (patch written by Frank Heydlauf).
4349 SC/02 Eximstats: use FileHandles for outputting results.
4351 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4353 SC/04 Eximstats: fixed display of large numbers with -nvr option
4355 SC/05 Eximstats: fixed merging of reports with empty tables.
4357 SC/06 Eximstats: added the -include_original_destination flag
4359 SC/07 Eximstats: removed tabs and trailing whitespace.
4361 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4363 TK/06 MBOX spool code: Add real "From " MBOX separator line
4364 so the .eml file is really in mbox format (even though
4365 most programs do not really care). Patch from Alex Miller.
4367 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4368 The latter is generated from $received_to and is only set if the
4369 message has one envelope recipient. SA can use these headers,
4370 obviously out-of-the-box. Patch from Alex Miller.
4372 PH/08 The ${def test on a variable was returning false if the variable's
4373 value was "0", contrary to what the specification has always said!
4374 The result should be true unless the variable is empty.
4376 PH/09 The syntax error of a character other than { following "${if
4377 def:variable_name" (after optional whitespace) was not being diagnosed.
4378 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4379 accidental colon was present, for example, could give incorrect results.
4381 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4382 result is used (not including appendfile, where other changes are about
4385 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4386 This involved changing a lot of size variables from int to off_t. It
4387 should work with maildirs and everything.
4389 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4390 spamd dying while we are connected to it.
4392 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4395 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4396 function for BDB 4.3.
4398 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4400 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4401 the sources for PCRE have been re-arranged and more files are now
4404 PH/15 The code I had for printing potentially long long variables in PH/11
4405 above was not the best (it lost precision). The length of off_t variables
4406 is now inspected at build time, and an appropriate printing format (%ld
4407 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4408 to be "long long int" or "long int". This is needed for the internal
4409 formatting function string_vformat().
4411 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4412 the configuration file to be ":syslog", then the script "guesses" where
4413 the logs files are, rather than using the compiled in default. In our
4414 case the guess is not the same as the compiled default, so the script
4415 suddenly stopped working when I started to use syslog. The patch checks
4416 to see if log_file_path is "". If so, it attempts to read it from exim
4417 with no configuration file to get the compiled in version, before it
4418 falls back to the previous guessing code."
4420 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4421 implementing BATV in an Exim configuration. See NewStuff for the gory
4424 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4425 Makefile that are specific to HP-UX.
4427 PH/18 If the "use_postmaster" option was set for a recipient callout together
4428 with the "random" option, the postmaster address was used as the MAIL
4429 FROM address for the random test, but not for the subsequent recipient
4430 test. It is now used for both.
4432 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4433 patch removes a few documentation additions to RFC 3028, because the
4434 latest draft now contains them. It adds the new en;ascii-case comparator
4435 and a new error check for 8bit text in MIME parts. Comparator and
4436 require names are now matched exactly. I enabled the subaddress
4437 extension, but it is not well tested yet (read: it works for me)."
4439 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4440 rework some of the code of TK/09 above to avoid the hardwired use of
4441 "%lld" and "long long". Replaced the call to snprintf() with a call to
4444 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4445 records point to non-existent hosts", "retry timeout exceeded", and
4446 "retry time not reached for any host after a long failure period".
4448 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4449 experimental DomainKeys support:
4451 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4452 (2) On an error such as an illegally used "control", the wrong name for
4453 the control was given.
4455 These problems did NOT occur unless DomainKeys support was compiled.
4457 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4459 PH/24 Added ${if match_ip condition.
4461 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4462 databases so that it will be absolutely obvious if a crash occurs in the
4463 DB library. This is a regular occurrence (often caused by mis-matched
4466 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4467 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4468 picked up on a user's system that detects such things. There doesn't seem
4469 to be a gcc warning option for this - only an attribute that has to be
4470 put on the function's prototype. It seems that in Fedora Core 4 they have
4471 set this on a number of new functions. No doubt there will be more in due
4474 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4475 specify a transport. However, if an address that was verified by such a
4476 router was the subject of a callout, Exim crashed because it tried to
4477 read the rcpt_include_affixes from the non-existent transport. Now it
4478 just assumes that the setting of that option is false. This bug was
4479 introduced by 4.51/PH/31.
4481 PH/28 Changed -d+all to exclude +memory, because that information is very
4482 rarely of interest, but it makes the output a lot bigger. People tend to
4483 do -d+all out of habit.
4485 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4486 code in os-type was giving problems when libc.so lives in lib64, like on
4489 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4490 aren't the modern standard, and it seems that some systems' include files
4491 don't always have them. Exim was already checking for some of the newer
4492 ones like T_AAAA, and defining it itself. I've added checks for all the
4493 record types that Exim uses.
4495 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4496 not automatically generating a new one, as it is supposed to. This
4497 prevented TLS from working. If the file did exist, but contained invalid
4498 data, a new version was generated, as expected. It was only the case of a
4499 non-existent file that was broken.
4501 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4502 with a change in libdomainkeys > 0.64.
4504 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4505 from DNS. If the selector record carries the flag, it now has
4506 precedence over the domain-wide flag.
4508 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4510 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4511 the use of a port name that isn't defined in /etc/services) occurred, the
4512 message was deferred as in a normal delivery, and thus remained on the
4513 spool, instead of being failed because of the mua_wrapper setting. This
4514 is now fixed, and I tidied up some of the mua_wrapper messages at the
4517 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4518 the messages in a hash of arrays rather than using individual hashes.
4519 This is a bit cleaner and results in dramatic memory savings, albeit
4520 at a slight CPU cost.
4522 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4523 as requested by Marc Sherman.
4525 SC/10 Eximstats: added histograms for user specified patterns as requested
4528 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4530 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4531 fopen() in the content-scanning modules that did not already have it.
4537 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4538 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4540 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4542 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4544 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4545 to test Sieve filters that use "vacation".
4547 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4548 that changes the way the GnuTLS parameters are stored in the cache file.
4549 The new format can be generated externally. For backward compatibility,
4550 if the data in the cache doesn't make sense, Exim assumes it has read an
4551 old-format file, and it generates new data and writes a new file. This
4552 means that you can't go back to an older release without removing the
4555 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4556 work if there are any delivery delays because "one_time" forces the
4557 parent to be marked "delivered", so its unseen clone is never tried
4558 again. For this reason, Exim now forbids the simultaneous setting of
4561 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4562 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4563 entry if you want to know the details. The fix, however, neglected to
4564 consider the case when local delivery batching is involved. The test for
4565 "previously delivered" was not happening when checking to see if an
4566 address could be batched with a previous (undelivered) one; under
4567 certain circumstances this could lead to multiple deliveries to the same
4570 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4571 in its include files, and this causes problems building Exim.
4573 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4574 header_syntax) but Exim was just ignoring anything given after a slash.
4575 In particular, this caused confusion with an attempt to use "verify =
4576 reverse_host_lookup/defer_ok". An error is now given when options are
4577 supplied for verify items that do not have them. (Maybe reverse_host_
4578 lookup should have a defer_ok option, but that's a different point.)
4580 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4581 defined by RFC 821) to 2048, because there were problems with some AUTH
4582 commands, and RFC 1869 says the size should be increased for extended
4583 SMTP commands that take arguments.
4585 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4588 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4589 "unknown" error; now it says that the functionality isn't in the binary.
4591 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4592 an address' error message when a string expansion fails (syntax or
4593 whatever). Otherwise the password may appear in the log. Following change
4594 PH/42 below, there is no longer a chance of it appearing in a bounce
4597 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4599 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4600 of its addresses was ever tried. (Bugzilla bug #2.)
4602 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4603 the result incorrectly in the debug output. (It correctly added a newline
4604 to what was transported.)
4606 TF/01 Added $received_time.
4608 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4609 commented out examples of how to interface to a virus scanner and to
4610 SpamAssassin. Also added commented examples of av_scanner and
4611 spamd_address settings.
4613 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4614 and controls are allowed in which ACLs. There were a couple of minor
4615 errors. Some of the entries in the conditions table (which is a table of
4616 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4617 negation of where the condition IS allowed.
4619 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4621 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4622 header file does not have a version number, so I've had to invent a new
4623 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4624 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4625 radiusclient), but it was contributed by a Radius user.
4627 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4628 files or directories, for OpenSSL.
4630 PH/19 When an Exim process that is running as root has to create an Exim log
4631 file, it does so in a subprocess that runs as exim:exim so as to get the
4632 ownership right at creation (otherwise, other Exim processes might see
4633 the file with the wrong ownership). There was no test for failure of this
4634 fork() call, which would lead to the process getting stuck as it waited
4635 for a non-existent subprocess. Forks do occasionally fail when resources
4636 run out. I reviewed all the other calls to fork(); they all seem to check
4639 PH/20 When checking for unexpected SMTP input at connect time (before writing
4640 the banner), Exim was not dealing correctly with a non-positive return
4641 from the read() function. If the client had disconnected by this time,
4642 the result was a log entry for a synchronization error with an empty
4643 string after "input=" when read() returned zero. If read() returned -1
4644 (an event I could not check), uninitialized data bytes were printed.
4645 There were reports of junk text (parts of files, etc) appearing after
4648 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4650 PH/22 Added support for macro redefinition, and (re)definition in between
4651 driver and ACL definitions.
4653 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4654 forgetting to use the resulting value; it was using the unexpanded value.
4656 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4657 hadn't been configured. The fix is from Juergen Kreileder, who
4658 understands it better than I do:
4660 "Here's what I see happening with three configured cyrus_sasl
4661 authenticators configured (plain, login, cram-md5):
4663 On startup auth_cyrus_sasl_init() gets called for each of these.
4664 This means three calls to sasl_listmech() without a specified mech_list.
4665 => SASL tests which mechs of all available mechs actually work
4666 => three warnings about OTP not working
4667 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4669 With the patch, sasl_listmech() also gets called three times. But now
4670 SASL's mech_list option is set to the server_mech specified in the the
4671 authenticator. Or in other words, the answer from sasl_listmech()
4672 gets limited to just the mech you're testing for (which is different
4674 => the return list contains just 'plain' or 'login', 'cram-md5' or
4675 nothing depending on the value of ob->server_mech.
4677 I've just tested the patch: Authentication still works fine,
4678 unavailable mechs specified in the exim configuration are still
4679 caught, and the auth.log warnings about OTP are gone."
4681 PH/25 When debugging is enabled, the contents of the command line are added
4682 to the debugging output, even when log_selector=+arguments is not
4685 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4686 answer is "GNU", and only if the return is "GNU/something" is the answer
4689 PH/27 $acl_verify_message is now set immediately after the failure of a
4690 verification in an ACL, and so is available in subsequent modifiers. In
4691 particular, the message can be preserved by coding like this:
4693 warn !verify = sender
4694 set acl_m0 = $acl_verify_message
4696 Previously, $acl_verify_message was set only while expanding "message"
4697 and "log_message" when a very denied access.
4699 PH/28 Modified OS/os.c-Linux with
4701 -#ifndef OS_LOAD_AVERAGE
4702 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4704 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4705 nomenclature these days.)
4707 PH/29 Installed patch from the Sieve maintainer that adds the options
4708 sieve_useraddress and sieve_subaddress to the redirect router.
4710 PH/30 In these circumstances:
4711 . Two addresses routed to the same list of hosts;
4712 . First host does not offer TLS;
4713 . First host accepts first address;
4714 . First host gives temporary error to second address;
4715 . Second host offers TLS and a TLS session is established;
4716 . Second host accepts second address.
4717 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4718 and peerdn, if requested) that were in fact used only for the second
4721 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4722 attention to any local part prefix or suffix that was matched by the
4723 router that accepted the address. It now behaves in the same way as it
4724 does for delivery: the affixes are removed from the local part unless
4725 rcpt_include_affixes is set on the transport.
4727 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4728 timeout during the DATA phase of an incoming message.
4730 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4731 applied a patch sanctioned by the Sieve maintainer.
4733 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4734 the uid or gid is negative. A case of a negative gid caused this to be
4735 noticed. The fix allows for either to be negative.
4737 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4738 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4741 PH/36 The change PH/12 above was broken. Fixed it.
4743 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4744 the grounds that routing the same address twice would always produce the
4745 same answer. This might have been true once, but it is certainly no
4746 longer true now. Routing a child address may depend on the previous
4747 routing that produced that child. Some complicated redirection strategies
4748 went wrong when messages had multiple recipients, and made Exim's
4749 behaviour dependent on the order in which the addresses were given.
4751 I have moved the duplicate checking until after the routing is complete.
4752 Exim scans the addresses that are assigned to local and remote
4753 transports, and removes any duplicates. This means that more work will be
4754 done, as duplicates will always all be routed, but duplicates are
4755 presumably rare, so I don't expect this is of any significance.
4757 For deliveries to pipes, files, and autoreplies, the duplicate checking
4758 still happens during the routing process, since they are not going to be
4761 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4762 It corrects a timeout issue with spamd. This is Ian's comment: "The
4763 background is that sometimes spamd either never reads data from a
4764 connection it has accepted, or it never writes response data. The exiscan
4765 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4766 blindly assumes that writes won't block so it may never time out."
4768 PH/39 Allow G after quota size as well as K and M.
4770 PH/40 The value set for $authenticated_id in an authenticator may not contain
4771 binary zeroes or newlines because the value is written to log lines and
4772 to spool files. There was no check on this. Now the value is run through
4773 the string_printing() function so that such characters are converted to
4774 printable escape sequences.
4776 PH/41 $message_linecount is a new variable that contains the total number of
4777 lines in the message. Compare $body_linecount, which is the count for the
4780 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4781 bounce and delay warning messages, except in certain special cases, which
4784 (a) An SMTP error message from a remote host;
4785 (b) A message specified in a :fail: redirection;
4786 (c) A message specified in a "fail" command in a system filter;
4787 (d) A message specified in a FAIL return from the queryprogram router;
4788 (e) A message specified by the cannot_route_message router option.
4790 In these cases only, Exim does include the error details in bounce and
4791 warning messages. There are also a few cases where bland messages such
4792 as "unrouteable address" or "local delivery error" are given.
4794 PH/43 $value is now also set for the "else" part of a ${run expansion.
4796 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4797 being worked on, but at least Exim now implements the latest version to
4800 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4801 process to complete was treated as a delivery failure, a timeout while
4802 writing the message to the pipe was logged, but erroneously treated as a
4803 successful delivery. Such timeouts include transport filter timeouts. For
4804 consistency with the overall process timeout, these timeouts are now
4805 treated as errors, giving rise to delivery failures by default. However,
4806 there is now a new Boolean option for the pipe transport called
4807 timeout_defer, which, if set TRUE, converts the failures into defers for
4808 both kinds of timeout. A transport filter timeout is now identified in
4811 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4812 systems where "make" and "gmake" are different, calling "gmake" at top
4813 level broke things. I've arranged for the value of $(MAKE) to be passed
4814 from the Makefile to this script so that it can call the same version of
4818 A note about Exim versions 4.44 and 4.50
4819 ----------------------------------------
4821 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4822 changes of various kinds. As a consequence, a big documentation update was
4823 needed. This delayed the release for rather longer than seemed good, especially
4824 in the light of a couple of (minor) security issues. Therefore, the changes
4825 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4826 release. So 4.44 and 4.50 are in effect two different branches that both start
4829 I have left the 4.50 change log unchanged; it contains all the changes since
4830 4.43. The change log for 4.44 is below; many of its items are identical to
4831 those for 4.50. This seems to be the most sensible way to preserve the
4832 historical information.
4838 1. Minor wording change to the doc/README.SIEVE file.
4840 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4841 computation of the current number of files was incorrect.
4843 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4844 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4847 4. Give more explanation in the error message when the command for a transport
4848 filter fails to execute.
4850 5. There are several places where Exim runs a non-Exim command in a
4851 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4852 was being done only for the command run by the queryprogram router. It is
4853 now done for all such subprocesses. The other cases are: ${run, transport
4854 filters, and the commands run by the lmtp and pipe transports.
4856 6. Added CONFIGURE_GROUP build-time option.
4858 7. Some older OS have a limit of 256 on the maximum number of file
4859 descriptors. Exim was using setrlimit() to set 1000 as a large value
4860 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4861 systems. I've change it so that if it can't get 1000, it tries for 256.
4863 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4864 was an oversight, and furthermore, ever since the addition of extra
4865 controls (e.g. 4.43/32), the checks on when to allow different forms of
4866 "control" were broken. There should now be diagnostics for all cases when a
4867 control that does not make sense is encountered.
4869 9. Added the /retain_sender option to "control=submission".
4871 10. $recipients is now available in the predata ACL (oversight).
4873 11. Tidy the search cache before the fork to do a delivery from a message
4874 received from the command line. Otherwise the child will trigger a lookup
4875 failure and thereby defer the delivery if it tries to use (for example) a
4876 cached ldap connection that the parent has called unbind on.
4878 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4879 of $address_data from the recipient verification was clobbered by the
4880 sender verification.
4882 13. The value of address_data from a sender verification is now available in
4883 $sender_address_data in subsequent conditions in the ACL statement.
4885 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4887 15. Added a new option "connect=<time>" to callout options, to set a different
4890 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4891 was its contents. (It was OK if the option was not defined at all.)
4893 17. A "Completed" log line is now written for messages that are removed from
4894 the spool by the -Mrm option.
4896 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4897 information about exactly what failed.
4899 19. Added -dd to debug only the daemon process.
4901 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4902 handles timeouts, both on the server side and network timeouts. Renamed the
4903 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4905 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4906 It is now set to "smtps".
4908 22. $host_address is now set to the target address during the checking of
4909 ignore_target_hosts.
4911 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4912 being passed; this would have caused $sender_host_name to have been used if
4913 matching the list had actually called for a host name (not very likely,
4914 since this list is usually IP addresses). A host name is now passed as
4917 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4918 code that specifies a non-restarting handler (typically sigaction() in
4919 modern systems) in an attempt to fix a rare and obscure crash bug.
4921 25. Narrowed the window for a race in the daemon that could cause it to ignore
4922 SIGCHLD signals. This is not a major problem, because they are used only to
4923 wake it up if nothing else does.
4925 26. A malformed maildirsize file could cause Exim to calculate negative values
4926 for the mailbox size or file count. Odd effects could occur as a result.
4927 The maildirsize information is now recalculated if the size or filecount
4930 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4931 support for a long time. Removed HAVE_SYS_VFS_H.
4933 28. Installed the latest version of exipick from John Jetmore.
4935 29. In an address list, if the pattern was not a regular expression, an empty
4936 subject address (from a bounce message) matched only if the pattern was an
4937 empty string. Non-empty patterns were not even tested. This was the wrong
4938 because it is perfectly reasonable to use an empty address as part of a
4939 database query. An empty address is now tested by patterns that are
4940 lookups. However, all the other forms of pattern expect the subject to
4941 contain a local part and a domain, and therefore, for them, an empty
4942 address still always fails if the pattern is not itself empty.
4944 30. Exim went into a mad DNS loop when attempting to do a callout where the
4945 host was specified on an smtp transport, and looking it up yielded more
4946 than one IP address.
4948 31. Re-factored the code for checking spool and log partition space into a
4949 function that finds that data and another that does the check. The former
4950 is then used to implement four new variables: $spool_space, $log_space,
4951 $spool_inodes, and $log_inodes.
4953 32. The RFC2047 encoding function was originally intended for short strings
4954 such as real names; it was not keeping to the 75-character limit for
4955 encoded words that the RFC imposes. It now respects the limit, and
4956 generates multiple encoded words if necessary. To be on the safe side, I
4957 have increased the buffer size for the ${rfc2047: expansion operator from
4960 33. It is now permitted to omit both strings after an "if" condition; if the
4961 condition is true, the result is "true". As before, when the second string
4962 is omitted, a false condition yields an empty string. This makes it less
4963 cumbersome to write custom ACL and router conditions.
4965 34. Failure to deliver a bounce message always caused it to be frozen, even if
4966 there was an errors_to setting on the router. The errors_to setting is now
4969 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4970 canonical form (fully expanded) before being placed in
4971 $sender_host_address.
4973 36. The table in the code that translates DNS record types into text (T_A to
4974 "A" for instance) was missing entries for NS and CNAME. It is just possible
4975 that this could have caused confusion if both these types were looked up
4976 for the same domain, because the text type is used as part of Exim's
4977 per-process caching. But the chance of anyone hitting this buglet seems
4980 37. The dnsdb lookup has been extended in a number of ways.
4982 (1) There is a new type, "zns", which walks up the domain tree until it
4983 finds some nameserver records. It should be used with care.
4985 (2) There is a new type, "mxh", which is like "mx" except that it returns
4986 just the host names, not the priorities.
4988 (3) It is now possible to give a list of domains (or IP addresses) to be
4989 looked up. The behaviour when one of the lookups defers can be
4990 controlled by a keyword.
4992 (4) It is now possible to specify the separator character for use when
4993 multiple records are returned.
4995 38. The dnslists ACL condition has been extended: it is now possible to supply
4996 a list of IP addresses and/or domains to be looked up in a particular DNS
4999 39. Added log_selector=+queue_time_overall.
5001 40. When running the queue in the test harness, wait just a tad after forking a
5002 delivery process, to get repeatability of debugging output.
5004 41. Include certificate and key file names in error message when GnuTLS fails
5005 to set them up, because the GnuTLS error message doesn't include the name
5006 of the failing file when there is a problem reading it.
5008 42. Allow both -bf and -bF in the same test run.
5010 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5012 44. The "Exiscan patch" is now merged into the mainline Exim source.
5014 45. Sometimes the final signoff response after QUIT could fail to get
5015 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5016 < 0 before doing a fflush(). This bug looks as though it goes back to the
5017 introduction of TLS in release 3.20, but "sometimes" must have been rare
5018 because the tests only now provoked it.
5020 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5021 (this can affect the format of dates).
5023 47. exim_tidydb, when checking for the continued existence of a message for
5024 which it has found a message-specific retry record, was not finding
5025 messages that were in split spool directories. Consequently, it was
5026 deleting retry records that should have stayed in existence.
5028 48. Steve fixed some bugs in eximstats.
5030 49. The SPA authentication driver was not abandoning authentication and moving
5031 on to the next authenticator when an expansion was forced to fail,
5032 contradicting the general specification for all authenticators. Instead it
5033 was generating a temporary error. It now behaves as specified.
5035 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5036 (the order specifies the preference for clients). The order is now AES256,
5037 AES128, 3DES, ARCFOUR128.
5039 51. Small patch to Sieve code - explicitly set From: when generating an
5042 52. Exim crashed if a remote delivery caused a very long error message to be
5043 recorded - for instance if somebody sent an entire SpamAssassin report back
5044 as a large number of 550 error lines. This bug was coincidentally fixed by
5045 increasing the size of one of Exim's internal buffers (big_buffer) that
5046 happened as part of the Exiscan merge. However, to be on the safe side, I
5047 have made the code more robust (and fixed the comments that describe what
5050 53. Now that there can be additional text after "Completed" in log lines (if
5051 the queue_time_overall log selector is set), a one-byte patch to exigrep
5052 was needed to allow it to recognize "Completed" as not the last thing in
5055 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5056 patch that reportedly fixes this has been added. I am not expert enough to
5057 create a test for it. This is what the patch creator wrote:
5059 "I found a little strange behaviour of ldap code when working with
5060 Windows 2003 AD Domain, where users was placed in more than one
5061 Organization Units. When I tried to give exim partial DN, the exit code
5062 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5063 But simultaneously result of request was absolutely normal ldap result,
5064 so I produce this patch..."
5066 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5067 REFERENCE, so I have modified the code to exclude the patch when that macro
5070 55. Some experimental protocols are using DNS PTR records for new purposes. The
5071 keys for these records are domain names, not reversed IP addresses. The
5072 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5073 leaves it alone. Component reversal etc. now happens only for IP addresses.
5076 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5078 57. Double the size of the debug message buffer (to 2048) so that more of very
5079 long debug lines gets shown.
5081 58. The exicyclog utility now does better if the number of log files to keep
5082 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5084 59. Two changes related to the smtp_active_hostname option:
5086 (1) $smtp_active_hostname is now available as a variable.
5087 (2) The default for smtp_banner uses $smtp_active_hostname instead
5088 of $primary_hostname.
5090 60. The host_aton() function is supposed to be passed a string that is known
5091 to be a valid IP address. However, in the case of IPv6 addresses, it was
5092 not checking this. This is a hostage to fortune. Exim now panics and dies
5093 if the condition is not met. A case was found where this could be provoked
5094 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5095 components; fortuitously, this particular loophole had already been fixed
5096 by change 4.50/55 above.
5098 If there are any other similar loopholes, the new check in host_aton()
5099 itself should stop them being exploited. The report I received stated that
5100 data on the command line could provoke the exploit when Exim was running as
5101 exim, but did not say which command line option was involved. All I could
5102 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5103 running as the user.
5106 61. There was a buffer overflow vulnerability in the SPA authentication code
5107 (which came originally from the Samba project). I have added a test to the
5108 spa_base64_to_bits() function which I hope fixes it.
5111 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5112 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5114 63. The daemon start-up calls getloadavg() while still root for those OS that
5115 need the first call to be done as root, but it missed one case: when
5116 deliver_queue_load_max is set with deliver_drop_privilege. This is
5117 necessary for the benefit of the queue runner, because there is no re-exec
5118 when deliver_drop_privilege is set.
5120 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5121 This has been fixed.
5123 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5124 was in use, was not putting the data itself into the right store pool;
5125 consequently, it could be overwritten for a subsequent message in the same
5126 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5129 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5131 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5132 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5133 the function were treating the return as a boolean value, which happened to
5134 work because 0=false and not-0=true, but is not correct code.
5136 68. The host_aton() function was not handling scoped IPv6 addresses (those
5137 with, for example, "%eth0" on the end) correctly.
5139 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5140 negated items (that is, ~something) in unsigned ints. Some compilers
5141 apparently mutter when there is no cast.
5143 70. If an address verification called from an ACL failed, and did not produce a
5144 user-specific message (i.e. there was only a "system" message), nothing was
5145 put in $acl_verify_message. In this situation, it now puts the system
5148 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5149 session; change 4.31/43 added the unwanted input to the log line - except
5150 that it did not do this in the start of session case. It now does.
5152 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5153 This is wrong and can cause the other end to generate a synchronization
5154 error if it is another Exim or anything else that does the synchronization
5155 check. A QUIT command is no longer sent after a timeout.
5157 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5158 during host lookups.
5160 74. The defer_ok option of callout verification was not working if it was used
5161 when verifying addresses in header lines, that is, for this case:
5163 verify = header_sender/callout=defer_ok
5165 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5166 those file descriptors could be used for SMTP connections. If anything
5167 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5168 could be sent to the SMTP client, causing chaos. The daemon now opens
5169 stdin, stdout, and stderr to /dev/null when it puts itself into the
5172 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5173 log by default. The user can override this with suitable Perl magic.
5175 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5176 the log message when discard triggers, was not working for the DATA ACL or
5177 for the non-SMTP ACL.
5179 78. Error message wording change in sieve.c.
5181 79. If smtp_accept_max_per_host was set, the number of connections could be
5182 restricted to fewer than expected, because the daemon was trying to set up
5183 a new connection before checking whether the processes handling previous
5184 connections had finished. The check for completed processes is now done
5185 earlier. On busy systems, this bug wouldn't be noticed because something
5186 else would have woken the daemon, and it would have reaped the completed
5189 80. If a message was submitted locally by a user whose login name contained one
5190 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5191 It caused a spool format error. I have fixed the spool reading code. A
5192 related problem was that the "from" clause in the Received: line became
5193 illegal because of the space(s). It is now covered by ${quote_local_part.
5195 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5198 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5200 83. Patch from Sieve maintainer to fix unterminated string problem in
5201 "vacation" handling.
5203 84. Some minor changes to the Linux configuration files to help with other
5204 OS variants using glibc.
5206 85. One more patch for Sieve to update vacation handling to latest spec.
5209 ----------------------------------------------------
5210 See the note above about the 4.44 and 4.50 releases.
5211 ----------------------------------------------------
5217 1. Change 4.43/35 introduced a bug that caused file counts to be
5218 incorrectly computed when quota_filecount was set in an appendfile
5221 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5222 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5225 3. Give more explanation in the error message when the command for a transport
5226 filter fails to execute.
5228 4. There are several places where Exim runs a non-Exim command in a
5229 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5230 was being done only for the command run by the queryprogram router. It is
5231 now done for all such subprocesses. The other cases are: ${run, transport
5232 filters, and the commands run by the lmtp and pipe transports.
5234 5. Some older OS have a limit of 256 on the maximum number of file
5235 descriptors. Exim was using setrlimit() to set 1000 as a large value
5236 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5237 systems. I've change it so that if it can't get 1000, it tries for 256.
5239 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5240 was an oversight, and furthermore, ever since the addition of extra
5241 controls (e.g. 4.43/32), the checks on when to allow different forms of
5242 "control" were broken. There should now be diagnostics for all cases when a
5243 control that does not make sense is encountered.
5245 7. $recipients is now available in the predata ACL (oversight).
5247 8. Tidy the search cache before the fork to do a delivery from a message
5248 received from the command line. Otherwise the child will trigger a lookup
5249 failure and thereby defer the delivery if it tries to use (for example) a
5250 cached ldap connection that the parent has called unbind on.
5252 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5253 of $address_data from the recipient verification was clobbered by the
5254 sender verification.
5256 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5257 was its contents. (It was OK if the option was not defined at all.)
5259 11. A "Completed" log line is now written for messages that are removed from
5260 the spool by the -Mrm option.
5262 12. $host_address is now set to the target address during the checking of
5263 ignore_target_hosts.
5265 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5266 being passed; this would have caused $sender_host_name to have been used if
5267 matching the list had actually called for a host name (not very likely,
5268 since this list is usually IP addresses). A host name is now passed as
5271 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5272 code that specifies a non-restarting handler (typically sigaction() in
5273 modern systems) in an attempt to fix a rare and obscure crash bug.
5275 15. Narrowed the window for a race in the daemon that could cause it to ignore
5276 SIGCHLD signals. This is not a major problem, because they are used only to
5277 wake it up if nothing else does.
5279 16. A malformed maildirsize file could cause Exim to calculate negative values
5280 for the mailbox size or file count. Odd effects could occur as a result.
5281 The maildirsize information is now recalculated if the size or filecount
5284 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5285 support for a long time. Removed HAVE_SYS_VFS_H.
5287 18. Updated exipick to current release from John Jetmore.
5289 19. Allow an empty sender to be matched against a lookup in an address list.
5290 Previously the only cases considered were a regular expression, or an
5293 20. Exim went into a mad DNS lookup loop when doing a callout where the
5294 host was specified on the transport, if the DNS lookup yielded more than
5297 21. The RFC2047 encoding function was originally intended for short strings
5298 such as real names; it was not keeping to the 75-character limit for
5299 encoded words that the RFC imposes. It now respects the limit, and
5300 generates multiple encoded words if necessary. To be on the safe side, I
5301 have increased the buffer size for the ${rfc2047: expansion operator from
5304 22. Failure to deliver a bounce message always caused it to be frozen, even if
5305 there was an errors_to setting on the router. The errors_to setting is now
5308 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5309 canonical form (fully expanded) before being placed in
5310 $sender_host_address.
5312 24. Updated eximstats to version 1.33
5314 25. Include certificate and key file names in error message when GnuTLS fails
5315 to set them up, because the GnuTLS error message doesn't include the name
5316 of the failing file when there is a problem reading it.
5318 26. Expand error message when OpenSSL has problems setting up cert/key files.
5321 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5322 (this can affect the format of dates).
5324 28. exim_tidydb, when checking for the continued existence of a message for
5325 which it has found a message-specific retry record, was not finding
5326 messages that were in split spool directories. Consequently, it was
5327 deleting retry records that should have stayed in existence.
5329 29. eximstats updated to version 1.35
5330 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5331 1.35 - bugfix such that pie charts by volume are generated correctly
5333 30. The SPA authentication driver was not abandoning authentication and moving
5334 on to the next authenticator when an expansion was forced to fail,
5335 contradicting the general specification for all authenticators. Instead it
5336 was generating a temporary error. It now behaves as specified.
5338 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5339 (the order specifies the preference for clients). The order is now AES256,
5340 AES128, 3DES, ARCFOUR128.
5342 31. Small patch to Sieve code - explicitly set From: when generating an
5345 32. Exim crashed if a remote delivery caused a very long error message to be
5346 recorded - for instance if somebody sent an entire SpamAssassin report back
5347 as a large number of 550 error lines. This bug was coincidentally fixed by
5348 increasing the size of one of Exim's internal buffers (big_buffer) that
5349 happened as part of the Exiscan merge. However, to be on the safe side, I
5350 have made the code more robust (and fixed the comments that describe what
5353 33. Some experimental protocols are using DNS PTR records for new purposes. The
5354 keys for these records are domain names, not reversed IP addresses. The
5355 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5356 leaves it alone. Component reversal etc. now happens only for IP addresses.
5359 34. The host_aton() function is supposed to be passed a string that is known
5360 to be a valid IP address. However, in the case of IPv6 addresses, it was
5361 not checking this. This is a hostage to fortune. Exim now panics and dies
5362 if the condition is not met. A case was found where this could be provoked
5363 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5364 components; fortuitously, this particular loophole had already been fixed
5365 by change 4.50/55 or 4.44/33 above.
5367 If there are any other similar loopholes, the new check in host_aton()
5368 itself should stop them being exploited. The report I received stated that
5369 data on the command line could provoke the exploit when Exim was running as
5370 exim, but did not say which command line option was involved. All I could
5371 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5372 running as the user.
5375 35. There was a buffer overflow vulnerability in the SPA authentication code
5376 (which came originally from the Samba project). I have added a test to the
5377 spa_base64_to_bits() function which I hope fixes it.
5380 36. The daemon start-up calls getloadavg() while still root for those OS that
5381 need the first call to be done as root, but it missed one case: when
5382 deliver_queue_load_max is set with deliver_drop_privilege. This is
5383 necessary for the benefit of the queue runner, because there is no re-exec
5384 when deliver_drop_privilege is set.
5386 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5387 was in use, was not putting the data itself into the right store pool;
5388 consequently, it could be overwritten for a subsequent message in the same
5389 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5392 38. Sometimes the final signoff response after QUIT could fail to get
5393 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5394 < 0 before doing a fflush(). This bug looks as though it goes back to the
5395 introduction of TLS in release 3.20, but "sometimes" must have been rare
5396 because the tests only now provoked it.
5402 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5403 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5404 Now it returns an integer. A number of calls were still expecting a T/F
5405 return. Fortuitously, in all cases, the tests worked in OK situations,
5406 which is the norm. However, things would have gone wrong on any write
5407 failures on the smtp file descriptor. This function is used when sending
5408 messages over SMTP and also when doing verify callouts.
5410 2. When Exim is called to do synchronous delivery of a locally submitted
5411 message (the -odf or -odi options), it no longer closes stderr before doing
5414 3. Implemented the mua_wrapper option.
5416 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5418 5. Implemented the functions header_remove(), header_testname(),
5419 header_add_at_position(), and receive_remove_recipient(), and exported them
5422 6. If an ACL "warn" statement specified the addition of headers, Exim already
5423 inserted X-ACL-Warn: at the start if there was no header name. However, it
5424 was not making this test for the second and subsequent header lines if
5425 there were newlines in the string. This meant that an invalid header could
5426 be inserted if Exim was badly configured.
5428 7. Allow an ACL "warn" statement to add header lines at the start or after all
5429 the Received: headers, as well as at the end.
5431 8. Added the rcpt_4xx retry error code.
5433 9. Added postmaster_mailfrom=xxx to callout verification option.
5435 10. Added mailfrom=xxxx to the callout verification option, for verify=
5438 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5439 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5441 12. Inserted some casts to stop certain compilers warning when using pointer
5442 differences as field lengths or precisions in printf-type calls (mostly
5443 affecting debugging statements).
5445 13. Added optional readline() support for -be (dynamically loaded).
5447 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5448 same clock tick as a message's arrival, so that its received time was the
5449 same as the "first fail" time on the retry record, and that message
5450 remained on the queue past the ultimate address timeout, every queue runner
5451 would try a delivery (because it was past the ultimate address timeout) but
5452 after another failure, the ultimate address timeout, which should have then
5453 bounced the address, did not kick in. This was a "< instead of <=" error;
5454 in most cases the first failure would have been in the next clock tick
5455 after the received time, and all would be well.
5457 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5458 being recognized when the domain list was tested by the match_domain
5459 condition in an expansion string.
5461 16. Added the ${str2b64: operator.
5463 17. Exim was always calling setrlimit() to set a large limit for the number of
5464 processes, without checking whether the existing limit was already
5465 adequate. (It did check for the limit on file descriptors.) Furthermore,
5466 errors from getrlimit() and setrlimit() were being ignored. Now they are
5467 logged to the main and panic logs, but Exim does carry on, to try to do its
5468 job under whatever limits there are.
5470 18. Imported PCRE 5.0.
5472 19. Trivial typo in log message " temporarily refused connection" (the leading
5475 20. If the log selector return_path_on_delivery was set and an address was
5476 redirected to /dev/null, the delivery process crashed because it assumed
5477 that a return path would always be set for a "successful" delivery. In this
5478 case, the whole delivery is bypassed as an optimization, and therefore no
5481 21. Internal re-arrangement: the function for sending a challenge and reading
5482 a response while authentication was assuming a zero-terminated challenge
5483 string. It's now changed to take a pointer and a length, to allow for
5484 binary data in such strings.
5486 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5488 23. Exim was not respecting finduser_retries when seeking the login of the
5489 uid under which it was called; it was always trying 10 times. (The default
5490 setting of finduser_retries is zero.) Also, it was sleeping after the final
5491 failure, which is pointless.
5493 24. Implemented tls_on_connect_ports.
5495 25. Implemented acl_smtp_predata.
5497 26. If the domain in control=submission is set empty, Exim assumes that the
5498 authenticated id is a complete email address when it generates From: or
5499 Sender: header lines.
5501 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5502 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5503 chown and chgrp in /bin and hostname in /usr/bin.
5505 28. Exim was keeping the "process log" file open after each use, just as it
5506 does for the main log. This opens the possibility of it remaining open for
5507 long periods when the USR1 signal hits a daemon. Occasional processlog
5508 errors were reported, that could have been caused by this. Anyway, it seems
5509 much more sensible not to leave this file open at all, so that is what now
5512 29. The long-running daemon process does not normally write to the log once it
5513 has entered its main loop, and it closes the log before doing so. This is
5514 so that log files can straightforwardly be renamed and moved. However,
5515 there are a couple of unusual error situations where the daemon does write
5516 log entries, and I had neglected to close the log afterwards.
5518 30. The text of an SMTP error response that was received during a remote
5519 delivery was being truncated at 512 bytes. This is too short for some of
5520 the long messages that one sometimes sees. I've increased the limit to
5523 31. It is now possible to make retry rules that apply only when a message has a
5524 specific sender, in particular, an empty sender.
5526 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5527 it possible to be selective about when SMTP synchronization is enforced.
5529 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5531 32. Implemented hosts_connection_nolog.
5533 33. Added an ACL for QUIT.
5535 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5538 35. Added mailbox_size and mailbox_filecount to appendfile.
5540 36. Added control = no_multiline_responses to ACLs.
5542 37. There was a bug in the logic of the code that waits for the clock to tick
5543 in the case where the clock went backwards by a substantial amount such
5544 that the microsecond fraction of "now" was more than the microsecond
5545 fraction of "then" (but the whole seconds number was less).
5547 38. Added support for the libradius Radius client library this is found on
5548 FreeBSD (previously only the radiusclient library was supported).
5554 1. When certain lookups returned multiple values in the form name=value, the
5555 quoting of the values was not always being done properly. Specifically:
5556 (a) If the value started with a double quote, but contained no whitespace,
5558 (b) If the value contained whitespace other than a space character (i.e.
5559 tabs or newlines or carriage returns) it was not quoted.
5560 This fix has been applied to the mysql and pgsql lookups by writing a
5561 separate quoting function and calling it from the lookup code. The fix
5562 should probably also be applied to nisplus, ibase and oracle lookups, but
5563 since I cannot test any of those, I have not disturbed their existing code.
5565 2. A hit in the callout cache for a specific address caused a log line with no
5566 reason for rejecting RCPT. Now it says "Previous (cached) callout
5567 verification failure".
5569 3. There was an off-by-one bug in the queryprogram router. An over-long
5570 return line was truncated at 256 instead of 255 characters, thereby
5571 overflowing its buffer with the terminating zero. As well as fixing this, I
5572 have increased the buffer size to 1024 (and made a note to document this).
5574 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5575 when Exim is waiting for an SMTP response from a remote server, Exim
5576 restarts its select() call on the socket, thereby resetting its timeout.
5577 This is not a problem when such interrupts are rare. Somebody set up a cron
5578 job to run exiwhat every 2 minutes, which is less than the normal select()
5579 timeout (5 or 10 minutes). This meant that the select() timeout never
5580 kicked in because it was always reset. I have fixed this by comparing the
5581 time when an interrupt arrives with the time at the start of the first call
5582 to select(). If more time than the timeout has elapsed, the interrupt is
5583 treated as a timeout.
5585 5. Some internal re-factoring in preparation for the addition of Sieve
5586 extensions (by MH). In particular, the "personal" test is moved to a
5587 separate function, and given an option for scanning Cc: and Bcc: (which is
5588 not set for Exim filters).
5590 6. When Exim created an email address using the login of the caller as the
5591 local part (e.g. when creating a From: or Sender: header line), it was not
5592 quoting the local part when it contained special characters such as @.
5594 7. Installed new OpenBSD configuration files.
5596 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5597 try to make them clearer.
5599 9. Callout options, other than the timeout value, were being ignored when
5600 verifying sender addresses in header lines. For example, when using
5602 verify = header_sender/callout=no_cache
5604 the cache was (incorrectly) being used.
5606 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5607 only the Cygwin environment.
5609 11. When return_path_on_delivery was set as a log selector, if different remote
5610 addresses in the same message used different return paths and parallel
5611 remote delivery occurred, the wrong values would sometimes be logged.
5612 (Whenever a remote delivery process finished, the return path value from
5613 the most recently started remote delivery process was logged.)
5615 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5616 lines when AUTH and/or TLS are in use. This is the "received protocol"
5617 field. Exim used to use "asmtp" for authenticated SMTP, without any
5618 indication (in the protocol name) for TLS use. Now it follows the RFC and
5619 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5620 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5621 names appear in log lines as well as in Received: header lines.
5623 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5624 extensions, and comparison tests, and to fix some bugs.
5626 14. Changes to the "personal" filter test:
5628 (1) The test was buggy in that it was just doing the equivalent of
5629 "contains" tests on header lines. For example, if a user's address was
5630 anne@some.where, the "personal" test would incorrectly be true for
5632 To: susanne@some.where
5634 This test is now done by extracting each address from the header in turn,
5635 and checking the entire address. Other tests that are part of "personal"
5636 are now done using regular expressions (for example, to check local parts
5637 of addresses in From: header lines).
5639 (2) The list of non-personal local parts in From: addresses has been
5640 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5641 taken from the Sieve specification recommendations.
5643 (3) If the message contains any header line starting with "List-" it is
5644 treated as non-personal.
5646 (4) The test for "circular" in the Subject: header line has been removed
5647 because it now seems ill-conceived.
5649 15. Minor typos in src/EDITME comments corrected.
5651 16. Installed latest exipick from John Jetmore.
5653 17. If headers_add on a router specified a text string that was too long for
5654 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5655 of string_sprintf() is now avoided.
5657 18. $message_body_size was not set (it was always zero) when running the DATA
5658 ACL and the local_scan() function.
5660 19. For the "mail" command in an Exim filter, no default was being set for
5661 the once_repeat time, causing a random time value to be used if "once" was
5662 specified. (If the value happened to be <= 0, no repeat happened.) The
5663 default is now 0s, meaning "never repeat". The "vacation" command was OK
5664 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5665 (I found it when inspecting the code).
5667 20. There is now an overall timeout for performing a callout verification. It
5668 defaults to 4 times the callout timeout, which applies to individual SMTP
5669 commands during the callout. The overall timeout applies when there is more
5670 than one host that can be tried. The timeout is checked before trying the
5671 next host. This prevents very long delays if there are a large number of
5672 hosts and all are timing out (e.g. when the network connections are timing
5673 out). The value of the overall timeout can be changed by specifying an
5674 additional sub-option for "callout", called "maxwait". For example:
5676 verify = sender/callout=5s,maxwait=20s
5678 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5679 to the end before writing, but this should make it even safer).
5681 22. Exim was forgetting that it had advertised PIPELINING for the second and
5682 subsequent messages on an SMTP connection. It was also not resetting its
5683 memory on STARTTLS and an internal HELO.
5685 23. When Exim logs an SMTP synchronization error within a session, it now
5686 records whether PIPELINING has been advertised or not.
5688 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5689 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5690 rather than long int.
5692 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5694 26. Added the never_mail option to autoreply.
5700 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5701 crash if the getsockname() call failed; this can happen if a connection is
5702 closed very soon after it is established. The problem was simply in the
5703 order in which certain operations were done, causing Exim to try to write
5704 to the SMTP stream before it had set up the file descriptor. The bug has
5705 been fixed by making things happen in the correct order.
5711 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5712 before the connection was closed, thus losing the rejection response.
5714 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5715 some early Solaris releases, but causes trouble in current releases where
5716 socklen_t is defined.
5718 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5721 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5724 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5725 that releasing the top of it at the end releases what was used for sub-
5726 expansions (unless the block got too big). However, discard this block if
5727 the first thing is a variable or header, so that we can use its block when
5728 it is dynamic (useful for very large $message_headers, for example).
5730 6. Lookups now cache *every* query, not just the most recent. A new, separate
5731 store pool is used for this. It can be recovered when all lookup caches are
5732 flushed. Lookups now release memory at the end of their result strings.
5733 This has involved some general refactoring of the lookup sources.
5735 7. Some code has been added to the store_xxx() functions to reduce the amount
5736 of flapping under certain conditions.
5738 8. log_incoming_interface used to affect only the <= reception log lines. Now
5739 it causes the local interface and port to be added to several more SMTP log
5740 lines, for example "SMTP connection from", and rejection lines.
5742 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5744 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5746 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5747 could be overwritten at the end of the current message (or the start of a
5748 new message if it was set in a HELO ACL). The value is now preserved for
5749 the duration of the SMTP connection.
5751 12. If a transport had a headers_rewrite setting, and a matching header line
5752 contained an unqualified address, that address was qualified, even if it
5753 did not match any rewriting rules. The underlying bug was that the values
5754 of the flags that permit the existence of unqualified sender and recipient
5755 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5756 non-local messages, and by -bnq for local messages) were not being
5757 preserved with the message after it was received.
5759 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5760 "next input=" as part of the text comprising the host identity instead of
5761 the correct text. The code was using the same buffer for two different
5762 strings. However, depending on which order the printing function evaluated
5763 its arguments, the bug did not always show up. Under Linux, for example, my
5764 test suite worked just fine.
5766 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5767 doesn't work with some older versions of Perl. It has been changed to "my",
5768 which in any case is probably the better facility to use.
5770 15. A really picky compiler found some instances of statements for creating
5771 error messages that either had too many or two few arguments for the format
5774 16. The size of the buffer for calls to the DNS resolver has been increased
5775 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5776 for addresses that have a lot of PTR records. This alleviates a problem; it
5777 does not fully solve it.
5779 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5780 buffer now truncates the list and logs the incident, which is the same
5781 action as happens when Exim is looking up a host name and its aliases.
5782 Previously in this situation something unpredictable would happen;
5783 sometimes it was "internal error: store_reset failed".
5785 18. If a server dropped the connection unexpectedly when an Exim client was
5786 using GnuTLS and trying to read a response, the client delivery process
5787 crashed while trying to generate an error log message.
5789 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5790 string, for example:
5792 warn message = H1: something\nH2: something
5794 the text was added as a single header line from Exim's point of view
5795 though it ended up OK in the delivered message. However, searching for the
5796 second and subsequent header lines using $h_h2: did not work. This has been
5797 fixed. Similarly, if a system filter added multiple headers in this way,
5798 the routers could not see them.
5800 20. Expanded the error message when iplsearch is called with an invalid key to
5801 suggest using net-iplsearch in a host list.
5803 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5804 ACLs are no longer actually imposed (and a message to that effect is
5807 22. If a "gecos" field in a passwd entry contained escaped characters, in
5808 particular, if it contained a \" sequence, Exim got it wrong when building
5809 a From: or a Sender: header from that name. A second bug also caused
5810 incorrect handling when an unquoted " was present following a character
5811 that needed quoting.
5813 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5814 was not being matched caselessly.
5816 24. Arranged for all hyphens in the exim.8 source to be escaped with
5819 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5820 properly. Recipient callout cache records were still being keyed to include
5821 the sender, even when use_sender was set false. This led to far more
5822 callouts that were necessary. The sender is no longer included in the key
5823 when use_sender is false.
5825 26. Added "control = submission" modifier to ACLs.
5827 27. Added the ${base62d: operator to decode base 62 numbers.
5829 28. dnsdb lookups can now access SRV records.
5831 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5832 the configuration file.
5834 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5835 (-v) mode. This makes the output for a verbose queue run more intelligible.
5837 31. Added a use_postmaster feature to recipient callouts.
5839 32. Added the $body_zerocount variable, containing the number of binary zero
5840 bytes in the message body.
5842 33. The time of last modification of the "new" subdirectory is now used as the
5843 "mailbox time last read" when there is a quota error for a maildir
5846 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5848 35. Added +ignore_unknown as a special item in host lists.
5850 36. Code for decoding IPv6 addresses in host lists is now included, even if
5851 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5852 address was recognized as an IP address, but was then not correctly decoded
5853 into binary, causing unexpected and incorrect effects when compared with
5860 1. Very minor rewording of debugging text in manualroute to say "list of
5861 hosts" instead of "hostlist".
5863 2. If verify=header_syntax was set, and a header line with an unqualified
5864 address (no domain) and a large number of spaces between the end of the
5865 name and the colon was received, the reception process suffered a buffer
5866 overflow, and (when I tested it) crashed. This was caused by some obsolete
5867 code that should have been removed. The fix is to remove it!
5869 3. When running in the test harness, delay a bit after writing a bounce
5870 message to get a bit more predictability in the log output.
5872 4. Added a call to search_tidyup() just before forking a reception process. In
5873 theory, someone could use a lookup in the expansion of smtp_accept_max_
5874 per_host which, without the tidyup, could leave open a database connection.
5876 5. Added the variables $recipient_data and $sender_data which get set from a
5877 lookup success in an ACL "recipients" or "senders" condition, or a router
5878 "senders" option, similar to $domain_data and $local_part_data.
5880 6. Moved the writing of debug_print from before to after the "senders" test
5883 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5884 problems for message scanning, either using a data ACL, or using
5885 local_scan() because the Received: header was not generated till after they
5886 were called (in order to set the time as the time of reception completion).
5887 I have revised the way this works. The header is now generated after the
5888 body is received, but before the ACL or local_scan() are called. After they
5889 are run, the timestamp in the header is updated.
5895 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5896 before starting a queue runner without re-exec. This happened only when
5897 deliver_drop_privilege was set or when the Exim user was set to root. The
5898 effect of the bug was that timeouts during subsequent deliveries caused
5899 crashes instead of being properly handled. The handler is now left at its
5900 default (and expected) setting.
5902 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5903 message, again when deliver_drop_privilege is set or Exim is run as root.
5904 The bug described in (1) was not present in this case, but the tidying up
5905 of the other signals was missing. I have made the two cases consistent.
5907 3. The ignore_target_hosts setting on a manualroute router was being ignored
5908 for hosts that were looked up using the /MX notation.
5910 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5913 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5914 operated on the sender address. After changing the $sender_address to <>
5915 for the sender address verify, Exim was re-instated it as the original
5916 (before rewriting) address, but remembering that it had rewritten it, so it
5917 wasn't rewriting it again. This bug also had the effect of breaking the
5918 sender address verification caching when the sender address was rewritten.
5920 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5921 This has been changed so that if the ip literal address matches
5922 ignore_target_hosts, the router declines.
5924 7. Added expansion conditions match_domain, match_address, and match_local_
5925 part (NOT match_host).
5927 8. The placeholder for the Received: header didn't have a length field set.
5929 9. Added code to Exim itself and to exim_lock to test for a specific race
5930 condition that could lead to file corruption when using MBX delivery. The
5931 issue is with the lockfile that is created in /tmp. If this file is removed
5932 after a process has opened it but before that process has acquired a lock,
5933 there is the potential for a second process to recreate the file and also
5934 acquire a lock. This could lead to two Exim processes writing to the file
5935 at the same time. The added code performs the same test as UW imapd; it
5936 checks after acquiring the lock that its file descriptor still refers to
5937 the same named file.
5939 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5940 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5943 11. Added the smtp_active_hostname option. If used, this will typically be made
5944 to depend on the incoming interface address. Because $interface_address is
5945 not set up until the daemon has forked a reception process, error responses
5946 that can happen earlier (such as "too many connections") no longer contain
5949 12. If an expansion in a condition on a "warn" statement fails because a lookup
5950 defers, the "warn" statement is abandoned, and the next ACL statement is
5951 processed. Previously this caused the whole ACL to be aborted.
5953 13. Added the iplsearch lookup type.
5955 14. Added ident_timeout as a log selector.
5957 15. Added tls_certificate_verified as a log selector.
5959 16. Added a global option tls_require_ciphers (compare the smtp transport
5960 option of the same name). This controls incoming TLS connections.
5962 17. I finally figured out how to make tls_require_ciphers do a similar thing
5963 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5964 before starting the TLS session.
5966 18. Tabs are now shown as \t in -bP output.
5968 19. If the log selector return_path_on_delivery was set, Exim crashed when
5969 bouncing a message because it had too many Received: header lines.
5971 20. If two routers both had headers_remove settings, and the first one included
5972 a superfluous trailing colon, the final name in the first list and the
5973 first name in the second list were incorrectly joined into one item (with a
5974 colon in the middle).
5980 1. Added -C and -D options to the exinext utility, mainly to make it easier
5981 to include in the automated testing, but these could be helpful when
5982 multiple configurations are in use.
5984 2. The exinext utility was not formatting the output nicely when there was
5985 an alternate port involved in the retry record key, nor when there was a
5986 message id as well (for retries that were specific to a specific message
5987 and a specific host). It was also confused by IPv6 addresses, because of
5988 the additional colons they contain. I have fixed the IPv4 problem, and
5989 patched it up to do a reasonable job for IPv6.
5991 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5992 delivery, the log line now contains "pipelined" if PIPELINING was used.
5994 4. An SMTP transport process used to panic and die if the bind() call to set
5995 an explicit outgoing interface failed. This has been changed; it is now
5996 treated in the same way as a connect() failure.
5998 5. A reference to $sender_host_name in the part of a conditional expansion
5999 that was being skipped was still causing a DNS lookup. This no longer
6002 6. The def: expansion condition was not recognizing references to header lines
6003 that used bh_ and bheader_.
6005 7. Added the _cache feature to named lists.
6007 8. The code for checking quota_filecount in the appendfile transport was
6008 allowing one more file than it should have been.
6010 9. For compatibility with Sendmail, the command line option
6018 and sets the incoming protocol and host name (for trusted callers). The
6019 host name and its colon can be omitted when only the protocol is to be set.
6020 Note the Exim already has two private options, -pd and -ps, that refer to
6021 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6022 "s", but I don't think that's a major issue.
6024 10. A number of refactoring changes to the code, none of which should affect
6027 (a) The number of logging options was getting close to filling up the
6028 32-bit word that was used as a bit map. I have split them into two classes:
6029 those that are passed in the argument to log_write(), and those that are
6030 only ever tested independently outside of that function. These are now in
6031 separate 32-bit words, so there is plenty of room for expansion again.
6032 There is no change in the user interface or the logging behaviour.
6034 (b) When building, for example, log lines, the code previously used a
6035 macro that called string_cat() twice, in order to add two strings. This is
6036 not really sufficiently general. Furthermore, there was one instance where
6037 it was actually wrong because one of the argument was used twice, and in
6038 one call a function was used. (As it happened, calling the function twice
6039 did not affect the overall behaviour.) The macro has been replaced by a
6040 function that can join an arbitrary number of extra strings onto a growing
6043 (c) The code for expansion conditions now uses a table and a binary chop
6044 instead of a serial search (which was left over from when there were very
6045 few conditions). Also, it now recognizes conditions like "pam" even when
6046 the relevant support is not compiled in: a suitably worded error message is
6047 given if an attempt is made to use such a condition.
6049 11. Added ${time_interval:xxxxx}.
6051 12. A bug was causing one of the ddress fields not to be passed back correctly
6052 from remote delivery subprocesses. The field in question was not being
6053 subsequently used, so this caused to problems in practice.
6055 13. Added new log selectors queue_time and deliver_time.
6057 14. Might have fixed a bug in maildirsizefile handling that threw up
6058 "unexpected character" debug warnings, and recalculated the data
6059 unnecessarily. In any case, I expanded the warning message to give more
6062 15. Added the message "Restricted characters in address" to the statements in
6063 the default ACL that block characters like @ and % in local parts.
6065 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6066 Three changes have been made:
6068 (a) There was a serious bug; a negative response to MAIL caused the whole
6069 recipient domain to be cached as invalid, thereby blocking all messages
6070 to all local parts at the same domain, from all senders. This bug has
6071 been fixed. The domain is no longer cached after a negative response to
6072 MAIL if the sender used is not empty.
6074 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6077 (c) A new callout option, "use_sender" has been added for people who want
6078 the modified behaviour.
6084 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6087 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6088 indeed breaks things for older releases.
6090 3. Added additional logging to the case where there is a problem reading data
6091 from a filter that is running in a subprocess using a pipe, in order to
6092 try to track down a specific problem.
6094 4. Testing facility fudge: when running in the test harness and attempting
6095 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6096 getting "No route to host". Convert this to a timeout.
6098 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6101 6. Some OS don't have socklen_t but use size_t instead. This affects the
6102 fifth argument of getsockopt() amongst other things. This is now
6103 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6104 can be set for individual OS. I have set it for SunOS5, OSF1, and
6105 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6106 some earlier ones do not.
6108 7. Change 4.30/15 was not doing the test caselessly.
6110 8. The standard form for an IPv6 address literal was being rejected by address
6111 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6112 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6113 this, as well as the form without the "IPv6" on the front (but only when
6114 address literals are enabled, of course).
6116 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6118 10. Exim crashed if a message with an empty sender address specified by -f
6119 encountered a router with an errors_to setting. This could be provoked only
6120 by a command such as
6124 where an empty string was supplied; "<>" did not hit this bug.
6126 11. Installed PCRE release 4.5.
6128 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6129 remained set. It is now erased.
6131 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6132 times from message ids (which are base 36 rather than the normal 62).
6134 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6135 were being counted as actual protocol errors, and logged if the log
6136 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6137 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6138 and DATA following a set of rejected RCPTs do not count as protocol errors.
6139 In other words, Exim assumes they were pipelined, though this may not
6140 actually be the case. Of course, in all cases the client gets an
6141 appropriate error code.
6143 15. If a lookup fails in an ACL condition, a message about the failure may
6144 be available; it is used if testing the ACL cannot continue, because most
6145 such messages specify what the cause of the deferral is. However, some
6146 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6147 that caused an old message to be retained and used if a later statement
6148 caused a defer, replacing the real cause of the deferral.
6150 16. If an IP address had so many PTR records that the DNS lookup buffer
6151 was not large enough to hold them, Exim could crash while trying to process
6152 the truncated data. It now detects and logs this case.
6154 17. Further to 4.21/58, another change has been made: if (and only if) the
6155 first line of a message (the first header line) ends with CRLF, a bare LF
6156 in a subsequent header line has a space inserted after it, so as not to
6157 terminate the header.
6159 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6160 unnecessarily, used atoi() instead of strtol(), and didn't check the
6161 termination when getting file sizes from file names by regex.
6163 19. Completely re-implemented the support for maildirsize files, in the light
6164 of a number of problems with the previous contributed implementation
6165 (4.30/29). In particular:
6167 . If the quota is zero, the maildirsize file is maintained, but no quota is
6170 . If the maildir directory does not exist, it is created before any attempt
6171 to write a maildirsize file.
6173 . The quota value in the file is just a cache; if the quota is changed in
6174 the transport, the new value overrides.
6176 . A regular expression is available for excluding directories from the
6179 20. The autoreply transport checks the characters in options that define the
6180 message's headers; it allows continued headers, but it was checking with
6181 isspace() after an embedded newline instead of explicitly looking for a
6184 21. If all the "regular" hosts to which an address was routed had passed their
6185 expiry times, and had not reached their retry times, the address was
6186 bounced, even if fallback hosts were defined. Now Exim should go on to try
6189 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6190 equivalent code in the SMTP transport. Some hosts send humungous responses
6191 to HELO/EHLO, more than 1024 it seems.
6193 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6194 gives compiler warnings in some environments. I've now done it "properly",
6197 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6198 (because of problems with the built-in one) was declared to return uschar *
6199 instead of char *, causing compiler failure.
6201 25. Fixed a file descriptor leak when processing alias/forward files.
6203 26. Fixed a minor format string issue in dbfn.c.
6205 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6207 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6208 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6211 29. When a sender address is verified, it is cached, to save repeating the test
6212 when there is more than one recipient in a message. However, when the
6213 verification involves a callout, it is possible for different callout
6214 options to be set for different recipients. It is too complicated to keep
6215 track of this in the cache, so now Exim always runs a verification when a
6216 callout is required, relying on the callout cache for the optimization.
6217 The overhead is duplication of the address routing, but this should not be
6220 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6221 to be verified with callout=postmaster, and the main callout worked but the
6222 postmaster check failed, the verification correctly failed. However, if a
6223 subsequent RCPT command asked for sender verification *without* the
6224 postmaster check, incorrect caching caused this verification also to fail,
6227 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6228 it was not caching the DNS options (qualify_single, search_parents) that
6229 were used when the lookup failed. A subsequent lookup with different
6230 options therefore always gave the same answer, though there were cases
6231 where it should not have. (Example: a "domains = !$mx_any" option on a
6232 dnslookup router: the "domains" option is always processed without any
6233 widening, but the router might have qualify_single set.) Now Exim uses the
6234 cached value only when the same options are set.
6236 32. Added John Jetmore's "exipick" utility to the distribution.
6238 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6239 than a timeout (e.g. a certificate is required, and is not provided), an
6240 Exim server now closes the connection immediately. Previously it waited for
6241 the client to close - but if the client is SSL, it seems that they each
6242 wait for each other, leading to a delay before one of them times out.
6244 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6245 maintained 0.8.x compatibility because I don't think many are using it, and
6246 it is clearly obsolete.
6248 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6251 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6252 client certificate was expired. A simple patch fixes this, though I don't
6253 understand the full logic of why the verify callback is called multiple
6256 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6257 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6258 which causes problems with some clients (such as the Certicom SSL Plus
6259 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6260 disables the coutermeasure allowing Eudora to connect."
6262 38. Exim was not checking that a write() to a log file succeeded. This could
6263 lead to Bad Things if a log got too big, in particular if it hit a file
6264 size limit. Exim now panics and dies if it cannot write to a log file, just
6265 as it does if it cannot open a log file.
6267 39. Modified OS/Makefile-Linux so that it now contains
6269 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6271 The two -D definitions ensure that Exim is compiled with large file
6272 support, which makes it possible to handle log files that are bigger than
6275 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6276 instance) a domain was checked against a named list that involved a lookup,
6277 causing $domain_data to be set, then another domain was checked against the
6278 same list, then the first domain was re-checked, the value of $domain_data
6279 after the final check could be wrong. In particular, if the second check
6280 failed, it could be set empty. This bug probably also applied to
6283 41. The strip_trailing_dot option was not being applied to the address given
6284 with the -f command-line option.
6286 42. The code for reading a message's header from the spool was incrementing
6287 $received_count, but never initializing it. This meant that the value was
6288 incorrect (doubled) while delivering a message in the same process in which
6289 it was received. In the most common configuration of Exim, this never
6290 happens - a fresh exec is done - but it can happen when
6291 deliver_drop_privilege is set.
6293 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6294 it now includes up to 150 characters of the unexpected data in the log
6297 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6298 and building data strings. The size of both of these buffers was 10 000
6299 bytes - far larger than anybody would *ever* want, thought I. Needless to
6300 say, somebody hit the limit. I have increased the maximum line length to
6301 20 000 and the maximum data length of concatenated lines to 100 000. I have
6302 also fixed two bugs, because there was no checking on these buffers. Tsk,
6303 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6304 buffer is too small.
6306 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6307 lsearch lookups. Now it does.
6309 46. When parsing a route_list item in a manualroute router, a fixed-length
6310 buffer was used for the list of hosts. I made this 1024 bytes long,
6311 thinking that nobody would ever have a list of hosts that long. Wrong.
6312 Somebody had a whole pile of complicated expansion conditions, and the
6313 string was silently truncated, leading to an expansion error. It turns out
6314 that it is easier to change to an unlimited length (owing to other changes
6315 that have happened since this code was originally written) than to build
6316 structure for giving a limitation error. The length of the item that
6317 expands into the list of hosts is now unlimited.
6319 47. The lsearch lookup could not handle data where the length of text line was
6320 more than 4095 characters. Such lines were truncated, leading to shortened
6321 data being returned. It should now handle lines of any length.
6323 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6324 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6327 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6328 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6329 no longer happens in comment lines. A list of replacements is now placed
6330 at the head of all of the source files, except those whose only change is
6331 to replace PERL_COMMAND in the very first #! line.
6333 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6334 messages on the queue, with a bottom-up merge sort, using code contributed
6335 by Michael Haardt. This should make operations like -bp somewhat faster on
6336 large queues. It won't affect queue runners, except when queue_run_in_order
6339 51. Installed eximstats 1.31 in the distribution.
6341 52. Added support for SRV lookups to the dnslookup router.
6343 53. If an ACL referred to $message_body or $message_body_end, the value was not
6344 reset for any messages that followed in the same SMTP session.
6346 54. The store-handling optimization for building very long strings was not
6347 differentiating between the different store pools. I don't think this
6348 actually made any difference in practice, but I've tidied it.
6350 55. While running the routers to verify a sender address, $sender_address
6351 was still set to the sender address. This is wrong, because when routing to
6352 send a bounce to the sender, it would be empty. Therefore, I have changed
6353 it so that, while verifying a sender address, $sender_address is set to <>.
6354 (There is no change to what happens when verifying a recipient address.)
6356 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6357 target A or AAAA records (if not already returned) without resetting the
6358 qualify_single or search_parents options of the DNS resolver. These are
6359 inappropriate in this case because the targets of MX and SRV records must
6360 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6361 target that, when qualified, matched something in the local domain. These
6362 two options are now turned off when doing these lookups.
6364 57. It seems that at least some releases of Reiserfs (which does not have the
6365 concept of a fixed number of inodes) returns zero and not -1 for the
6366 number of available inodes. This interacted badly with check_spool_inodes,
6367 which assumed that -1 was the "no such thing" setting. What I have done is
6368 to check that the total number of inodes is greater than zero before doing
6369 the test of how many are available.
6371 58. When a "warn" ACL statement has a log_message modifier, the message is
6372 remembered, and not repeated. This is to avoid a lot of repetition when a
6373 message has many recipients that cause the same warning to be written.
6374 However, Exim was preserving the list of already written lines for an
6375 entire SMTP session, which doesn't seem right. The memory is now reset if a
6376 new message is started.
6378 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6379 debugging output unless log_rewrite was also set.
6381 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6382 of (int)(handle) when we know that handle contains (void *)(-1).
6384 61. The Exim daemon panic-logs an error return when it closes the incoming
6385 connection. However "connection reset by peer" seems to be common, and
6386 isn't really an error worthy of noting specially, so that particular error
6389 62. When Exim is trying to find all the local interfaces, it used to panic and
6390 die if the ioctl to get the interface flags failed. However, it seems that
6391 on at least one OS (Solaris 9) it is possible to have an interface that is
6392 included in the list of interfaces, but for which you get a failure error
6393 for this call. This happens when the interface is not "plumbed" into a
6394 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6395 failure of the "get flags" call assumes that the interface is down.
6397 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6398 makes life easier for people who are doing arithmetic on fields extracted
6399 from dates, where you often get leading zeros that should not be
6400 interpreted as octal.
6402 64. Added qualify_domain to the redirect router, to override the global
6405 65. If a pathologically long header line contained very many addresses (the
6406 report of this problem mentioned 10 000) and each of them was rewritten,
6407 Exim could use up a very large amount of memory. (It kept on making new
6408 copies of the header line as it rewrote, and never released the old ones.)
6409 At the expense of a bit more processing, the header rewriting function has
6410 been changed so that it no longer eats memory in this way.
6412 66. The generation of the Received: header has been moved from the time that a
6413 message starts to be received, to the time that it finishes. The timestamp
6414 in the Received: header should now be very close to that of the <= log
6415 line. There are two side-effects of this change:
6417 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6418 logged header lines no longer include the local Received: line, because
6419 it has not yet been created. The same applies to a copy of the message
6420 that is returned to a non-SMTP sender when a message is rejected.
6422 (b) When a filter file is tested using -bf, no additional Received: header
6423 is added to the test message. After some thought, I decided that this
6426 This change does not affect the value of $received_for. It is still set
6427 after address rewriting, but before local_scan() is called.
6429 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6431 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6432 gave an unhelpful panic error message, and a defer error. I have managed to
6433 change this behaviour so that it now rejects any supplied certificate,
6434 which seems right, as the list of acceptable certificates is empty.
6436 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6437 gave an unhelpful defer error. I have not managed to make this reject any
6438 supplied certificates, but the error message it gives is "no certificate
6439 supplied", which is not helpful.
6441 70. exigrep's output now also includes lines that are not associated with any
6442 message, but which match the given pattern. Implemented by a patch from
6443 Martin Sluka, which also tidied up the Perl a bit.
6445 71. Recipient callout verification, like sender verification, was using <> in
6446 the MAIL FROM command. This isn't really the right thing, since the actual
6447 sender may affect whether the remote host accepts the recipient or not. I
6448 have changed it to use the actual sender in the callout; this means that
6449 the cache record is now keyed on a recipient/sender pair, not just the
6450 recipient address. There doesn't seem to be a real danger of callout loops,
6451 since a callout by the remote host to check the sender would use <>.
6452 [SEE ABOVE: changed after hitting problems.]
6454 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6455 temporary errors. However, in the case of such a code being given after
6456 the end of a data transmission (i.e. after ".") Exim was failing to write
6457 a retry record for the message. (Yes, there was some broken host that was
6458 actually sending 8xx at this point.)
6460 73. An unknown lookup type in a host list could cause Exim to panic-die when
6461 the list was checked. (An example that provoked this was putting <; in the
6462 middle of a list instead of at the start.) If this happened during a DATA
6463 ACL check, a -D file could be left lying around. This kind of configuration
6464 error no longer causes Exim to die; instead it causes a defer error. The
6465 incident is still logged to the main and panic logs.
6467 74. Buglet left over from Exim 3 conversion. The message "too many messages
6468 in one connection" was written to the rejectlog but not the mainlog, except
6469 when address rewriting (yes!) was being logged.
6471 75. Added write_rejectlog option.
6473 76. When a system filter was run not as root (that is, when system_filter_user
6474 was set), the values of the $n variables were not being returned to the
6475 main process; thus, they were not subsequently available in the $sn
6478 77. Added +return_path_on_delivery log selector.
6480 78. A connection timeout was being treated differently from recipients deferred
6481 when testing hosts_max_try with a message that was older than the host's
6482 retry timeout. (The host should not be counted, thus allowing all hosts to
6483 be tried at least once before bouncing.) This may have been the cause of an
6484 occasionally reported bug whereby a message would remain on the queue
6485 longer than the retry timeout, but would be bounced if a delivery was
6486 forced. I say "may" because I never totally pinned down the problem;
6487 setting up timeout/retry tests is difficult. See also the next item.
6489 79. The ultimate address timeout was not being applied to errors that involved
6490 a combination of host plus message (for example, a timeout on a MAIL
6491 command). When an address resolved to a number of possible hosts, and they
6492 were not all tried for each delivery (e.g. because of hosts_max_try), a
6493 message could remain on the queue longer than the retry timeout.
6495 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6498 81. Fixed an obscure SMTP outgoing bug which required at least the following
6499 conditions: (a) there was another message waiting for the same server;
6500 (b) the server returned 5xx to all RCPT commands in the first message so
6501 that the message was not completed; (c) the server dropped the connection
6502 or gave a negative response to the RSET that Exim sends to abort the
6503 transaction. The observed case was a dropped connection after DATA that had
6504 been sent in pipelining mode. That is, the server had advertised PIPELINING
6505 but was not implementing it correctly. The effect of the bug was incorrect
6506 behaviour, such as trying another host, and this could lead to a crash.
6512 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6513 and daemon.c were passed as pointers to ints; they should have been
6514 pointers to socklen_t variables (which are typically unsigned ints).
6516 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6519 3. Fixed a really odd bug that affected only the testing scheme; patching a
6520 certain fixed string in the binary changed the value of another string that
6521 happened to be identical to the end of the original first string.
6523 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6524 name", it returns that address as the IP address. On some operating
6525 systems (e.g. Solaris), it also passes back the IP address string as the
6526 "host name". However, on others (e.g. Linux), it passes back an empty
6527 string. Exim wasn't checking for this, and was changing the host name to an
6528 empty string, assuming it had been canonicalized.
6530 5. Although rare, it is permitted to have more than one PTR record for a given
6531 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6532 all the names associated with an address, because they do in Solaris.
6533 However, it seems that they do not in Linux for data that comes from the
6534 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6535 I found this out when I moved to a new Linux workstation and tried to run
6536 the Exim test suite.
6538 To get round this problem I have changed the code so that it now does its
6539 own call to the DNS to look up PTR records when searching for a host name.
6540 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6541 addresses that are only in /etc/hosts are still found.
6543 This behaviour is, however, controlled by an option called host_lookup_
6544 order, which defaults to "bydns:byaddr". If people want to use the other
6545 order, or indeed, just use one or the other means of lookup, they can
6546 specify it in this variable.
6548 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6549 some operating systems, this comes back from gethostbyaddr() as an empty
6550 string, and this is what Exim used to test for. However, it seems that in
6551 other systems, "." is yielded. Exim now tests for this case too.
6553 7. The values of check_spool_space and check_log_space are now held internally
6554 as a number of kilobytes instead of an absolute number of bytes. If a
6555 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6556 kilobyte. This means that much larger values can be stored.
6558 8. Exim monitor: an attempt to get the action menu when not actually pointing
6559 at a message produces an empty menu entitled "No message selected". This
6560 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6561 no entries in it ("Shell widget menu has zero width and/or height"). So I
6562 have added a single, blank menu entry in this case.
6564 9. Added ${quote_local_part.
6566 10. MIME decoding is now applied to the contents of Subject: header lines when
6569 11. Now that a reference to $sender_host_address automatically causes a reverse
6570 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6571 host lookup before query-style lookups in lists that might use this
6572 variable. This has therefore been abolished, and the "net-" prefix is no
6573 longer necessary for query-style lookups.
6575 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6576 have been a typo for LFLAGS, so it has been changed.
6578 13. The install script calls Exim with "-C /dev/null" in order to find the
6579 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6580 to be output. However, since Exim outputs its version number before the
6581 error, it didn't break the script. It just looked ugly. I fixed this by
6582 always allowing "-C /dev/null" if the caller is root.
6584 14. Ignore overlarge ACL variable number when reading spool file - insurance
6585 against a later release with more variables having written the file.
6587 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6588 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6589 this, as well as the form without the "IPv6" on the front.
6591 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6592 OS/Makefile-Darwin file.
6594 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6595 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6597 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6598 variable. However, some OS use ints for inodes. I've added cast to long int
6599 to get rid of the compiler warning.
6601 19. I had forgotten to lock out "/../" in configuration file names when
6602 ALT_CONFIG_PREFIX was set.
6604 20. Routers used for verification do not need to specify transports. However,
6605 if such a router generated a host list, and callout was configured, Exim
6606 crashed, because it could not find a port number from the (non-existent)
6607 transport. It now assumes port 25 in this circumstance.
6609 21. Added the -t option to exigrep.
6611 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6612 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6613 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6614 them is set, LOOKUP_LSEARCH is forced.
6616 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6618 24. Added sender and host information to the "rejected by local_scan()" log
6619 line; previously there was no indication of these.
6621 25. Added .include_if_exists.
6623 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6624 Linux. It turns out that not all file systems support this. Apparently some
6625 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6626 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6627 not supported on the file descriptor, is now ignored when Exim is trying to
6628 sync a directory. This applies only to Linux.
6630 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6632 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6633 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6636 29. Added support for maildirsize files from supplied patch (modified a bit).
6638 30. The use of :fail: followed by an empty string could lead Exim to respond to
6639 sender verification failures with (e.g.):
6641 550 Verification failed for <xxx>
6642 550 Sender verify failed
6644 where the first response line was missing the '-' that indicates it is not
6645 the final line of the response.
6647 31. The loop for finding the name of the user that called Exim had a hardwired
6648 limit of 10; it now uses the value of finduser_retries, which is used for
6649 all other user lookups.
6651 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6654 33. Exim was neglecting to zero errno before one call of strtol() when
6655 expanding a string and expecting an integer value. On some systems this
6656 resulted in spurious "integer overflow" errors. Also, it was casting the
6657 result into an int without checking.
6659 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6660 did not work. The code looks as if it has *never* worked, though it appears
6661 to have been documented since at least release 1.62. I have made it work.
6663 35. The "timeout_DNS" error in retry rules, also documented since at least
6664 1.62, also never worked. As it isn't clear exactly what this means, and
6665 clearly it isn't a major issue, I have abolished the feature by treating it
6666 as "timeout", and writing a warning to the main and panic logs.
6668 36. The display of retry rules for -brt wasn't always showing the error code
6671 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6672 timeout_connect_A, timeout_connect_MX.
6674 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6675 to the empty sender.
6677 39. The daemon was not analysing the content of -oX till after it had closed
6678 stderr and disconnected from the controlling terminal. This meant that any
6679 syntax errors were only noted on the panic log, and the return code from
6680 the command was 0. By re-arranging the code a little, I've made the
6681 decoding happen first, so such errors now appear on stderr, and the return
6682 code is 1. However, the actual setting up of the sockets still happens in
6683 the disconnected process, so errors there are still only recorded on the
6686 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6687 connections (as happens on some IP stacks) was logged at start up time as
6688 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6689 it from "IPv6 and IPv4", which means that two separate sockets are being
6692 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6693 says whether AF_INET or AF_INET6 was passed as an argument.
6695 42. Exiwhat output was messed up when time zones were included in log
6698 43. Exiwhat now gives more information about the daemon's listening ports,
6699 and whether -tls-on-connect was used.
6701 44. The "port" option of the smtp transport is now expanded.
6703 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6704 silently ignored. Now an error message is written to the main and panic
6707 46. There's a new ACL modifier called "logwrite" which writes to a log file
6708 as soon as it is encountered.
6710 47. Added $local_user_uid and $local_user_gid at routing time.
6712 48. Exim crashed when trying to verify a sender address that was being
6715 49. Exim was recognizing only a space character after ".include". It now also
6716 recognizes a tab character.
6718 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6719 extracting the relevant information from the specification. The man page no
6720 longer contains scrambled data for the -d option, and I've added a section
6721 at the front about calling Exim under different names.
6723 51. Added "extra_headers" argument to the "mail" command in filter files.
6725 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6728 53. Installed eximstats 1.29.
6730 54. Added transport_filter_timeout as a generic transport option.
6732 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6733 Cc: header lines. This was required by RFC 822, but it not required by RFC
6736 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6737 incoming messages that did not have them. Now it does so only if the
6738 message originates locally, that is, if there is no associated remote host
6739 address. When Resent- header lines are present, this applies to the Resent-
6740 lines rather than the non-Resent- lines.
6742 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6743 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6745 58. Messages for configuration errors now include the name of the main
6746 configuration file - useful now that there may be more than one file in a
6747 list (.included file names were always shown).
6749 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6750 for those rare installations that do not start the daemon as root or run it
6751 setuid root. I've cut out the call to initgroups() if the daemon is not
6754 60. The Exim user and group can now be bound into the binary as text strings
6755 that are looked up at the start of Exim's processing.
6757 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6759 62. Added $mailstore_basename variable.
6761 63. Installed patch to sieve.c from Michael Haardt.
6763 64. When Exim failed to open the panic log after failing to open the main log,
6764 the original message it was trying to log was written to stderr and debug
6765 output, but if they were not available (the usual case in production), it
6766 was lost. Now it is written to syslog before the two lines that record the
6767 failures to open the logs.
6769 65. Users' Exim filters run in subprocesses under the user's uid. It is
6770 possible for a "deliver" command or an alias in a "personal" command to
6771 provoke an address rewrite. If logging of address rewriting is configured,
6772 this fails because the process is not running as root or exim. There may be
6773 a better way of dealing with this, but for the moment (because 4.30 needs
6774 to be released), I have disabled address rewrite logging when running a
6775 filter in a non-root, non-exim process.
6781 1. The buildconfig auxiliary program wasn't quoting the value set for
6782 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6783 not defined. This bug was masked in 4.22 by the effect that was fixed in
6786 2. Some messages that were rejected after a message id was allocated were
6787 shown as "incomplete" by exigrep. It no longer does this for messages that
6788 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6790 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6791 have allow_domain_literals set, the ID did not get logged in the <= line.
6792 Domain literals are now always recognized in Message-ID: header lines.
6794 4. The first argument for a ${extract expansion item is the key name or field
6795 number. Leading and trailing spaces in this item were not being ignored,
6796 causing some misleading effects.
6798 5. When deliver_drop_privilege was set, single queue runner processes started
6799 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6800 same command in the process it spins off) were not dropping privilege.
6802 6. When the daemon running as "exim" started a queue runner, it always
6803 re-executed Exim in the spun-off process. This is a waste of effort when
6804 deliver_drop_privilege is set. The new process now just calls the
6805 queue-runner function directly.
6811 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6814 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6815 ignored. Though the use of -oP was forcing the writing of a pid file, it
6816 was always written to the default place.
6818 3. If the message "no IP address found for host xxxx" is generated during
6819 incoming verification, it is now followed by identification of the incoming
6820 connection (so you can more easily find what provoked it).
6822 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6824 5. Added some features to "harden" Exim a bit more against certain attacks:
6826 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6827 be put in Local/Makefile. This is like the never_users runtime option,
6828 but it cannot be overridden. The default setting is "root".
6830 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6831 prefix string with which any file named in a -C command line option
6834 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6835 is retained for -C and -D only if the caller of Exim is root. Without
6836 it, the exim user may also use -C and -D and retain privilege.
6838 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6839 command line option is disabled.
6841 6. Macro names set by the -D option must start with an upper case letter, just
6842 like macro names defined in the configuration file.
6844 7. Added "dereference=" facility to LDAP.
6846 8. Two instances of the typo "uknown" in the source files are fixed.
6848 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6849 the Configure-Makefile script screwed up while processing it.
6851 10. Incorporated PCRE 4.4.
6853 11. The SMTP synchronization check was not operating right at the start of an
6854 SMTP session. For example, it could not catch a HELO sent before the client
6855 waited for the greeting. There is now a check for outstanding input at the
6856 point when the greeting is written. Because of the duplex, asynchronous
6857 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6858 way, but not yet received, when the check is performed.
6860 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6861 on TCP/IP sockets, because this apparently causes some broken clients to
6864 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6865 unchanged) from the Cygwin maintainer.
6867 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6868 was supported instead of testing for mbx. Effectively a typo.
6870 15. The spa authenticator server code was not checking that the input it
6871 received was valid base64.
6873 16. The debug output line for the "set" modifier in ACLs was not showing the
6874 name of the variable that was being set.
6876 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6878 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6879 lookup on its own. Something else was needed to trigger the lookup. For
6880 example, a match in host_lookup or the need for a host name in a host list.
6881 Now, if $sender_host_name is referenced and the host name has not yet been
6882 looked up, a lookup is performed. If the lookup fails, the variable remains
6883 empty, and $host_lookup_failed is set to "1".
6885 19. Added "eqi" as a case-independent comparison operator.
6887 20. The saslauthd authentication condition could segfault if neither service
6888 nor realm was specified.
6890 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6891 error message that was logged was misleading, and incoming SMTP
6892 connections were dropped. The message is now more accurate, and temporary
6893 errors are given to SMTP connections.
6895 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6896 (except RSET) with a 421 error, until QUIT is received. However, it was
6897 failing to send a response to QUIT.
6899 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6900 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6901 ACL is now run after the helo_try_verify_hosts code.
6903 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6904 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6905 case-independent, so other case variants are also recognized). Apparently
6906 some systems use these upper case variants.
6908 25. If more than two messages were waiting for the same host, and a transport
6909 filter was specified for the transport, Exim sent two messages over the
6910 same TCP/IP connection, and then failed with "socket operation on non-
6911 socket" when it tried to send the third.
6913 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6915 27. The extern definition of crypt16() in expand.c was not being excluded when
6916 the OS had its own crypt16() function.
6918 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6919 as a preferred synonym for return_size_limit, both as an option and as an
6922 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6924 30. Changed the default configuration ACL to relax the local part checking rule
6925 for addresses that are not in any local domains. For these addresses,
6926 slashes and pipe symbols are allowed within local parts, but the sequence
6927 /../ is explicitly forbidden.
6929 31. SPA server authentication was not clearing the challenge buffer before
6932 32. log_message in a "warn" ACL statement was writing to the reject log as
6933 well as to the main log, which contradicts the documentation and doesn't
6934 seem right (because no rejection is happening). So I have stopped it.
6936 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6937 However, I am unable to do any testing of this.
6939 34. Fixed an infelicity in the appendfile transport. When checking directories
6940 for a mailbox, to see if any needed to be created, it was accidentally
6941 using path names with one or more superfluous leading slashes; tracing
6942 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6944 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6945 contents are added to the log line that is written for every discarded
6946 recipient. (Previously a log_message setting was ignored.)
6948 36. The ${quote: operator now quotes the string if it is empty.
6950 37. The install script runs exim in order to find its version number. If for
6951 some reason other than non-existence or emptiness, which it checks, it
6952 could not run './exim', it was installing it with an empty version number,
6953 i.e. as "exim-". This error state is now caught, and the installation is
6956 38. An argument was missing from the function that creates an error message
6957 when Exim fails to connect to the socket for saslauthd authentication.
6958 This could cause Exim to crash, or give a corrupted message.
6960 39. Added isip, isip4, and isip6 to ${if conditions.
6962 40. The ACL variables $acl_xx are now saved with the message, and can be
6963 accessed later in routers, transports, and filters.
6965 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6966 strings in the file are not string-expanded.
6968 42. If a MAIL command specified a SIZE value that was too large to fit into an
6969 int variable, the check against message_size_limit failed. Such values are
6970 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6971 day this will have to be increased, but I don't think I want to be around
6972 when emails are that large.
6979 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6980 iconv() is not standard in FreeBSD.
6982 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6983 IPv6 enabled. The observed symptom was a segmentation fault on return from
6984 the function os_common_find_running_interfaces() in src/os.c.
6986 3. In the check_special_case() function in daemon.c I had used "errno" as an
6987 argument name, which causes warnings on some systems. This was basically a
6988 typo, since it was named "eno" in the comments!
6990 4. The code that waits for the clock to tick (at a resolution of some fraction
6991 of a second) so as to ensure message-id uniqueness was always waiting for
6992 at least one whole tick, when it could have waited for less. [This is
6993 almost certainly not relevant at current processor speeds, where it is
6994 unlikely to ever wait at all. But we try to future-proof.]
6996 5. The function that sleeps for a time interval that includes fractions of a
6997 second contained a race. It did not block SIGALRM between setting the
6998 timer, and suspending (a couple of lines later). If the interval was short
6999 and the sigsuspend() was delayed until after it had expired, the suspension
7000 never ended. On busy systems this could lead to processes getting stuck for
7003 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7004 process, before it forks any delivery processes. The open lookup caching
7005 mechanism meant that the open file or database connection was passed into
7006 the delivery process. The problem was that delivery processes always tidy
7007 up cached lookup data. This could cause a problem for the next delivery
7008 process started by the queue runner, because the external queue runner
7009 process does not know about the closure. So the next delivery process
7010 still has data in the lookup cache. In the case of a file lookup, there was
7011 no problem because closing a file descriptor in a subprocess doesn't affect
7012 the parent. However, if the lookup was caching a connection to a database,
7013 the connection was closed, and the second delivery process was likely to
7014 see errors such as "PGSQL: query failed: server closed the connection
7015 unexpectedly". The problem has been fixed by closing all cached lookups
7016 in a queue runner before running a delivery process.
7018 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7019 seem to have the "const" qualifier which it has on other OS. I've
7022 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7023 *of the same type* (client or server) with the same public name that an
7024 error should be diagnosed.
7026 9. When Exim looked up a host name for an IP address, but failed to find the
7027 original IP address when looking up the host name (a safety check), it
7028 output the message "<ip address> does not match any IP for NULL", which was
7029 confusing, to say the least. The bug was that the host name should have
7030 appeared instead of "NULL".
7032 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7033 user that is built into the binary, and the -C or -D options is used, root
7034 privilege is dropped before the configuration file is read. In addition,
7035 logging is switched to stderr instead of the normal log files. If the
7036 configuration then re-defines the Exim user, the unprivileged environment
7037 is probably not what is expected, so Exim logs a panic warning message (but
7040 However, if deliver_drop_privilege is set, the unprivileged state may well
7041 be exactly what is intended, so the warning has been cut out in that case,
7042 and Exim is allowed to try to write to its normal log files.
7048 1. smtp_return_error_details was not giving details for temporary sender
7049 or receiver verification errors.
7051 2. Diagnose a configuration error if two authenticators have the same public
7054 3. Exim used not to create the message log file for a message until the first
7055 delivery attempt. This could be confusing when incoming messages were held
7056 for policy or load reasons. The message log file is now created at the time
7057 the message is received, and an initial "Received" line is written to it.
7059 4. The automatically generated man page for command line options had a minor
7060 bug that caused no ill effects; however, a more serious problem was that
7061 the procedure for building the man page automatically didn't always
7062 operate. Consequently, release 4.20 contains an out-of-date version. This
7063 shouldn't happen again.
7065 5. When building Exim with embedded Perl support, the script that builds the
7066 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7067 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7069 6. The freeze_tell option was not being used for messages that were frozen on
7070 arrival, either by an ACL or by local_scan().
7072 7. Added the smtp_incomplete_transaction log selector.
7074 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7075 was accepting AUTH without a new EHLO.
7077 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7078 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7081 10. Logging of TCP/IP connections (when configured) now happens in the main
7082 daemon process instead of the child process, so that the TCP/IP connection
7083 count is more accurate (but it can never be perfect).
7085 11. The use of "drop" in a nested ACL was not being handled correctly in the
7086 outer ACL. Now, if condition failure induced by the nested "drop" causes
7087 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7088 or "require"), the connection is dropped.
7090 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7091 that yield "discard" can now be used with an "accept" or a "discard" verb,
7092 but an error is generated for any others (because I can't see a useful way
7093 to define what should happen).
7095 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7096 are now processed in the same way as lines in the Exim configuration file.
7097 In particular, continuation lines are supported.
7099 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7101 15. Added -ti meaning -t -i.
7103 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7104 domains, and warn by logging if others are found.
7106 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7107 returned from GIFCONF when Exim is trying to find the list of interfaces on
7108 a host. The code in os.c has been modified to copy each ifreq to an aligned
7109 structure in all cases.
7111 Also, in some cases, the returned ifreq's were being copied to a 'struct
7112 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7113 means the last couple of bytes of an IPv6 address could be chopped if the
7114 ifreq contained only a normal sockaddr (14 bytes storage).
7116 18. Named domain lists were not supported in the hosts_treat_as_local option.
7117 An entry such as +xxxx was not recognized, and was treated as a literal
7120 19. Ensure that header lines added by a DATA ACL are included in the reject log
7121 if the ACL subsequently rejects the message.
7123 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7124 MD5 (which is deprecated).
7126 21. When testing a filter file using -bf, Exim was writing a message when it
7127 took the sender from a "From " line in the message, but it was not doing so
7128 when it took $return_path from a Return-Path: header line. It now does.
7130 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7131 with a valid header line field name (a series of printing characters
7132 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7134 23. Changed "disc" in the source to "disk" to conform to the documentation and
7135 the book and for uniformity.
7137 24. Ignore Sendmail's -Ooption=value command line item.
7139 25. When execve() failed while trying to run a command in a pipe transport,
7140 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7141 could be confused with a return value of 69 from the command itself. This
7142 has been changed to 127, the value the shell returns if it is asked to run
7143 a non-existent command. The wording for the related log line suggests a
7144 non-existent command as the problem.
7146 26. If received_header_text expands to an empty string, do not add a Received:
7147 header line to the message. (Well, it adds a token one on the spool, but
7148 marks it "old" so that it doesn't get used or transmitted.)
7150 27. Installed eximstats 1.28 (addition of -nt option).
7152 28. There was no check for failure on the call to getsockname() in the daemon
7153 code. This can fail if there is a shortage of resources on the system, with
7154 ENOMEM, for example. A temporary error is now given on failure.
7156 29. Contrary to the C standard, it seems that in some environments, the
7157 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7158 program. Exim now does this explicitly; it affects the formatting of
7159 timestamps using strftime().
7161 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7162 complaints. I've now initialized all the variables, to avoid this.
7164 32. Header lines added by a system filter were not being "seen" during
7165 transport-time rewrites.
7167 33. The info_callback() function passed to OpenSSL is set up with type void
7168 (*)(SSL *, int, int), as described somewhere. However, when calling the
7169 function (actually a macro) that sets it up, the type void(*)() is
7170 expected. I've put in a cast to prevent warnings from picky compilers.
7172 34. If a DNS black list lookup found a CNAME record, but there were no A
7173 records associated with the domain it pointed at, Exim crashed.
7175 35. If a DNS black list lookup returned more than one A record, Exim ignored
7176 all but the first. It now scans all returned addresses if a particular IP
7177 value is being sought. In this situation, the contents of the
7178 $dnslist_value variable are a list of all the addresses, separated by a
7181 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7182 to accept a host name and all its aliases if the forward lookup for any of
7183 them yielded the IP address of the incoming connection. Now it accepts only
7184 those names whose forward lookup yields the correct IP address. Any other
7185 names are discarded. This closes a loophole whereby a rogue DNS
7186 administrator could create reverse DNS records to break through a
7187 wildcarded host restriction in an ACL.
7189 37. If a user filter or a system filter that ran in a subprocess used any of
7190 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7191 the wrong values were passed to the pipe command ($thisaddress had the
7192 value of $0, $0 had the value of $1, etc). This bug was introduced by
7193 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7195 38. Improved the line breaking for long SMTP error messages from ACLs.
7196 Previously, if there was no break point between 40 and 75 characters, Exim
7197 left the rest of the message alone. Two changes have been made: (a) I've
7198 reduced the minimum length to 35 characters; (b) if it can't find a break
7199 point between 35 and 75 characters, it looks ahead and uses the first one
7200 that it finds. This may give the occasional overlong line, but at least the
7201 remaining text gets split now.
7203 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7204 file descriptors might be low, and that setting 1000 would always raise it.
7205 It turns out that in some environments, the limit is already over 1000 and
7206 that lowering it causes trouble. So now Exim takes care not to decrease it.
7208 40. When delivering a message, the value of $return_path is set to $sender_
7209 address at the start of routing (routers may change the value). By an
7210 oversight, this default was not being set up when an address was tested by
7211 -bt or -bv, which affected the outcome if any router or filter referred to
7214 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7215 log only when "message" or "log_message" are set. However, if one of the
7216 conditions was an address verification, or a call to a nested ACL, the
7217 messages generated by the underlying test were being passed through. This
7218 no longer happens. The underlying message is available in $acl_verify_
7219 message for both "message" and "log_message" expansions, so it can be
7220 passed through if needed.
7222 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7223 new expansion $bh_ to give the encoded byte string without charset
7224 translation. Translation happens only if iconv() is available; HAVE_ICONV
7225 indicates this at build time. HEADERS_CHARSET gives the charset to
7226 translate to; headers_charset can change it in the configuration, and
7227 "headers charset" can change it in an individual filter file.
7229 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7230 that creates RFC 2047 encoded "words" labels them as that charset instead
7231 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7232 expansion operator; (ii) when Exim creates a From: line for a local
7233 message; (iii) when a header line is rewritten to include a "phrase" part.
7235 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7236 buggy, causing it to skip the first lines of messages whose message ID
7237 ended in 'D'. This would not have bitten before Exim release 4.14, because
7238 message IDs were unlikely to end in 'D' before then. The effect was to have
7239 incorrect size information for certain domains.
7241 45. #include "config.h" was missing at the start of the crypt16.c module. This
7242 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7245 46. If there was a timeout during a "random" callout check, Exim treated it as
7246 a failure of the random address, and carried on sending RSET and the real
7247 address. If the delay was just some slowness somewhere, the response to the
7248 original RCPT would be taken as a response to RSET and so on, causing
7249 mayhem of various kinds.
7251 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7252 when I implemented it. It didn't allow for the fact that some option values
7253 may legitimately be negative (e.g. size_addition), and it didn't even do
7254 the right test for positive values.
7256 48. Domain names in DNS records are case-independent. Exim always looks them up
7257 in lower case. Some resolvers return domain names in exactly the case they
7258 appear in the zone file, that is, they may contain uppercase letters. Not
7259 all resolvers do this - some return always lower case. Exim was treating a
7260 change of case by a resolver as a change of domain, similar to a widening
7261 of a domain abbreviation. This triggered its re-routing code and so it was
7262 trying to route what was effectively the same domain again. This normally
7263 caused routing to fail (because the router wouldn't handle the domain
7264 twice). Now Exim checks for this case specially, and just changes the
7265 casing of the domain that it ultimately uses when it transmits the message
7268 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7271 50. If a filter generated a file delivery with a non-absolute name (possible if
7272 no home directory exists for the router), the forbid_file option was not
7275 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7276 the existing equality matching.
7278 52. Exim was using ints instead of ino_t variables in some places where it was
7279 dealing with inode numbers.
7281 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7282 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7283 called TMPDIR, and if it finds it is different, it changes its value.
7285 54. The smtp_printf() function is now made available to local_scan() so
7286 additional output lines can be written before returning. There is also an
7287 smtp_fflush() function to enable the detection of a dropped connection.
7288 The variables smtp_input and smtp_batched_input are exported to
7291 55. Changed the default runtime configuration: the message "Unknown user"
7292 has been removed from the ACL, and instead placed on the localuser router,
7293 using the cannot_route_message feature. This means that any verification
7294 failures that generate their own messages won't get overridden. Similarly,
7295 the "Unrouteable address" message that was in the ACL for unverifiable
7296 relay addresses has also been removed.
7298 56. Added hosts_avoid_esmtp to the smtp transport.
7300 57. The exicyclog script was not checking for the esoteric option
7301 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7302 will work only if exicyclog is run under the appropriate euid.
7304 58. Following a discussion on the list, the rules by which Exim recognises line
7305 endings on incoming messages have been changed. The -dropcr and drop_cr
7306 options are now no-ops, retained only for backwards compatibility. The
7307 following line terminators are recognized: LF CRLF CR. However, special
7308 processing applies to CR:
7310 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7311 nor a local message in the state where . is a terminator.
7313 (ii) If a bare CR is encountered in a header line, an extra space is added
7314 after the line terminator so as not to end the header. The reasoning
7315 behind this is that bare CRs in header lines are most likely either
7316 to be mistakes, or people trying to play silly games.
7318 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7319 was being incorrectly given as 18 bytes larger than it should have been.
7320 This is a VOB (very old bug).
7322 60. This may never have affected anything current, but just in case it has:
7323 When the local host is found other than at the start of a list of hosts,
7324 the local host, those with the same MX, and any that follow, are discarded.
7325 When the list in question was part of a longer list of hosts, the following
7326 hosts (not currently being processed) were also being discarded. This no
7327 longer happens. I'm not sure if this situation could ever has previously
7330 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7333 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7334 header. This is something that is recommended in a new Internet Draft, and
7335 is something that is documented as being done by Sendmail. There are two
7336 possible values. For messages generated by the autoreply transport, Exim
7339 Auto-Submitted: auto-replied
7341 whereas for all other generated messages (e.g. bounces) it adds
7343 Auto-Submitted: auto-generated
7345 63. The "personal" condition in filters now includes a test for the
7346 Auto-Submitted: header. If it contains the string "auto-" the message it
7347 not considered personal.
7349 64. Added rcpt_include_affixes as a generic transport option.
7351 65. Added queue_only_override (default true).
7353 66. Added the syslog_duplication option.
7355 67. If what should have been the first header line of a message consisted of
7356 a space followed by a colon, Exim was mis-interpreting it as a header line.
7357 It isn't of course - it is syntactically invalid and should therefore be
7358 treated as the start of the message body. The misbehaviour could have
7359 caused a number of strange effects, including loss of data in subsequent
7360 header lines, and spool format errors.
7362 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7363 client host had authenticated. This control can now be exercised by an ACL
7364 for more flexibility.
7366 69. By default, callouts do not happen when testing with -bh. There is now a
7367 variant, -bhc, which does actually run the callout code, including
7368 consulting and updating the callout cache.
7370 70. Added support for saslauthd authentication, courtesy of Alexander
7373 71. If statvfs() failed on the spool or log directories while checking their
7374 size for availability, Exim confusingly gave the error "space shortage".
7375 Furthermore, in debugging mode it crashed with a floating point exception.
7376 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7377 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7378 serious problem, Exim now writes to the main and panic logs when this
7379 happens, with details of the failure. It then refuses to accept the
7380 incoming message, giving the message "spool directory problem" or "log
7381 directory problem" with a 421 code for SMTP messages.
7383 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7384 0, 1, and 2 exist, because some OS complain for execs without them (see
7385 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7386 descriptors. However, the code omitted to check that the open succeeded,
7387 causing mysterious errors if for some reason the permissions on /dev/null
7388 got screwed. Now Exim writes a message to the main and panic logs, and
7389 bombs out if it can't open /dev/null.
7391 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7392 interact so that it is all more flexible. It is supposed to remain
7393 backwards compatible. Also added extra_local_interfaces.
7395 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7396 to bomb out with an assertion failure - to the client this appears as a
7397 connection drop. This problem occurs in the part of the code that was taken
7398 from the Samba project. Fortunately, the assertion is in a very simple
7399 function, so I have fixed this by reproducing the function inline in the
7400 one place where it is called, and arranging for authentication to fail
7401 instead of killing the process with assert().
7403 75. The SPA client code was not working when the server requested OEM rather
7404 than Unicode encoding.
7406 76. Added code to make require_files with a specific uid setting more usable in
7407 the case where statting the file as root fails - usually a non-root-mounted
7408 NFS file system. When this happens and the failure is EACCES, Exim now
7409 forks a subprocess and does the per-uid checking as the relevant uid.
7411 77. Added process_log_path.
7413 78. If log_file_path was not explicitly set, a setting of check_log_space or
7414 check_log_inodes was ignored.
7416 79. If a space check for the spool or log partitions fails, the incident is now
7417 logged. Of course, in the latter case the data may get lost...
7419 80. Added the %p formatting code to string_format() so that it can be used to
7420 print addresses in debug_print(). Adjusted all the address printing in the
7421 debugging in store.c to use %p rather than %d.
7423 81. There was a concern that a line of code in smtp_in.c could overflow a
7424 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7425 initially expressed, the concern was not well-founded, because trailing
7426 spaces are removed early. However, if the trailing spaces were followed by
7427 a NULL, they did not get removed, so the overflow was possible. Two fixes
7430 (a) I re-wrote the offending code in a cleaner fashion.
7431 (b) If an incoming SMTP command contains a NULL character, it is rejected
7434 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7435 runs initgroups(), so that if the Exim user is in any additional groups,
7436 they will be used during message reception.
7442 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob