1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
9 JH/01 Avoid costly startup code when not strictly needed. This reduces time
10 for some exim process initialisations. It does mean that the logging
11 of TLS configuration problems is only done for the daemon startup.
13 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
15 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
16 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
19 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
20 DKIM signing is being done. Previously a transport_filter would always
21 disable CHUNKING, falling back to traditional DATA.
23 JH/05 Regard command-line receipients as tainted.
25 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
27 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
28 PAM library frees one of the arguments given to it, despite the
29 documentation. Therefore a plain malloc must be used.
31 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
32 on-stack buffers were used, resulting in a taint trap when DSN information
33 copied from a received message was written into the buffer.
35 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
36 the ordering of its ARC headers. This caused a crash.
38 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
39 a new record was being constructed with information from the peer, a trap
42 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
43 installation would get error messages from DMARC verify, when it hit the
44 nonexistent file indicated by the default. Distros wanting DMARC enabled
45 should both provide the file and set the option.
46 Also enforce no DMARC verification for command-line sourced messages.
48 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
49 could, depending on the platform, hang at the STARTTLS response.
51 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
52 message on a connection. Previously if one message had ARC headers and
53 the following one did not, a crash could result when adding an
54 Authentication-Results: header.
56 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
57 Exim main code and Exim-related utities. The introduction of taint
58 tracking also did many adjustments to string handling. Since then, eximon
59 frequently terminated with an assert failure.
61 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
62 check for 452 responses. This slightly helps the inefficieny of doing
63 a large alias-expansion into a recipient-limited target. The max_rcpt
64 transport option still applies (and at the current default, will override
65 the new feature). The check is done for either cause of synch, and forces
66 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
67 connection. The new facility is not tunable at this time.
69 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
70 library live data was being used, so the results became garbage. Make
71 copies while it is still usable.
73 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
74 on delivery deferred (==) and failed (**) lines (if a delivery was
75 attemtped). Previously it was only on completion (=>) lines.
81 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
82 side, to decode a TLS 1.3 packet capture.
84 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
85 Previously the default library behaviour applied, sending two, each in
88 JH/03 Debug output for ACL now gives the config file name and line number for
91 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
93 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
95 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
96 buffer overrun for (non-chunking) other transports.
98 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
99 TLS1.3, means that a server rejecting a client certificate is not visible
100 to the client until the first read of encrypted data (typically the
101 response to EHLO). Add detection for that case and treat it as a failed
102 TLS connection attempt, so that the normal retry-in-clear can work (if
103 suitably configured).
105 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
106 and/or domain. Found and fixed by Jason Betts.
108 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
109 configuration). If a CNAME target was not a wellformed name pattern, a
112 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
113 the OS reports them interleaved with other addresses.
115 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
116 used both for input and for a verify callout, both encrypted, SMTP
117 responses being sent by the server could be lost. This resulted in
118 dropped connections and sometimes bounces generated by a peer sending
121 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
122 string. Previously it was possible to cause undefined behaviour in a
123 library routine (usually a crash). Found by "zerons".
125 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
128 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
129 API was removed, so update to use the newer ones.
131 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
132 any timeout set, is taking a long time. Previously we would hang on to a
133 rotated logfile "forever" if the input was arriving with long gaps
134 (a previous attempt to fix addressed lack, for a long time, of initial
137 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
138 shared (NFS) environment. The length of the tempfile name is now
139 4 + 16 ("hdr.$message_exim_id") which might break on file
140 systems which restrict the file name length to lower values.
143 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
144 shared (NFS) environment.
146 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
147 did for all versions <4.90). Notably -M, -m, --invert, -I may be
150 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
151 on some platforms for bit 31.
153 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
154 to changes apparently associated with TLS1.3 handling some of the APIs
155 previously used were either nonfunctional or inappropriate. Strings
156 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
157 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
158 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
159 This affects log line X= elements, the $tls_{in,out}_cipher variables,
160 and the use of specific cipher names in the encrypted= ACL condition.
162 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
164 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
165 verification result was not updated unless hosts_require_ocsp applied.
167 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
168 queue_list_requires_admin set to false, non-admin users were denied the
171 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
172 directory-of-certs mode. Previously they were advertised despite the
175 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
176 A single TCP connection by a client will now hold a TLS connection open
177 for multiple message deliveries, by default. Previoud the default was to
180 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
181 default. If built with the facility, DANE will be used. The facility
182 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
184 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
185 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
186 must be defined and you must still, unless you define DISABLE_TLS, manage
187 the the include-dir and library-file requirements that go with that
188 choice. Non-TLS builds are still supported.
190 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
193 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
194 default. If the platform supports and has the facility enabled, it will
195 be requested on all coneections.
197 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
198 controlled by the build-time option SUPPORT_PIPE_CONNECT.
200 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
202 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
203 success-DSN messages. Previously the From: header was always the default
204 one for these; the option was ignored.
206 JH/28 Fix the timeout on smtp response to apply to the whole response.
207 Previously it was reset for every read, so a teergrubing peer sending
208 single bytes within the time limit could extend the connection for a
209 long time. Credit to Qualsys Security Advisory Team for the discovery.
211 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
212 delivery address, which leaked information of the results of local
213 forwarding. Change to the original envelope recipient address, per
216 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
217 requested. Previously not bounce was generated and a log entry of
218 error ignored was made.
220 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
222 JH/32 Introduce a general tainting mechanism for values read from the input
223 channel, and values derived from them. Refuse to expand any tainted
224 values, to catch one form of exploit.
226 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
227 was unused and the unexpanded text used for the test. Found and
228 fixed by Ruben Jenster.
230 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
231 an attempt to use a TLS library read routine dereffed a nul pointer,
234 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
235 them in our smtp response.
237 JH/36 Have the general router option retry_use_local_part default to true when
238 any of the restrictive preconditions are set (to anything). Previously it
239 was only for check_local user. The change removes one item of manual
240 configuration which is required for proper retries when a remote router
241 handles a subset of addresses for a domain.
243 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
244 link count into consideration.
246 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
247 caused the extension of big_buffer, the following lines were ignored.
249 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
250 accordance with RFC 2308. Previously there was no expiry, so a longlived
251 receive process (eg. due to ACL delays) versus a short SOA value could
254 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
256 JH/39 Promote DMARC support to mainline.
258 JH/40 Bug 2452: Add a References: header to DSNs.
260 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
261 parameters. The relevant library call is documented as "Deprecated: This
262 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
263 3.6.0, DH parameters are negotiated following RFC7919."
265 HS/06 Change the default of dnssec_request_domains to "*"
267 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
268 carried on and emitted a BDAT command, even when PIPELINING was not
271 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
272 buffer was used for the filename, resulting in a trap when tainted
273 arguments (eg. $domain) were used.
275 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
276 recommended to avoid a possible server-load attack. The feature can be
277 re-enabled via the openssl_options main cofiguration option.
279 JH/45 local_scan API: documented the current smtp_printf() call. This changed
280 for version 4.90 - adding a "more data" boolean to the arguments.
281 Bumped the ABI version number also, this having been missed previously;
282 release versions 4.90 to 4.92.3 inclusive were effectively broken in
283 respect of usage of smtp_printf() by either local_scan code or libraries
284 accessed via the ${dlfunc } expansion item. Both will need coding
285 adjustment for any calls to smtp_printf() to match the new function
286 signature; a FALSE value for the new argument is always safe.
288 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
289 the file-offset (which the Linux syscall does, and exim expects); this
290 resulted in an indefinite loop.
292 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
293 to do ARC verification. The Authentication-Results: header line added
294 by the configuration then had no ARC item.
300 JH/01 Remove code calling the customisable local_scan function, unless a new
301 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
303 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
304 non-signal-safe functions being used.
306 JH/03 Bug 2269: When presented with a received message having a stupidly large
307 number of DKIM-Signature headers, disable DKIM verification to avoid
308 a resource-consumption attack. The limit is set at twenty.
310 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
311 report of oldest_pass in ${authres } in consequence, and separate out
312 some descriptions of reasons for verification fail.
314 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
315 files in the spool were present and unlocked. A queue-runner could spot
316 them, resulting in a duplicate delivery. Fix that by doing the unlock
317 after the unlink. Investigation by Tim Stewart. Take the opportunity to
318 add more error-checking on spoolfile handling while that code is being
321 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
322 No known attacks, no CVE, this is defensive hardening.
324 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
325 a queue-runner could start a delivery while other operations were ongoing.
326 Cutthrough delivery was a common victim, resulting in duplicate delivery.
327 Found and investigated by Tim Stewart. Fix by using the open message data
328 file handle rather than opening another, and not locally closing it (which
329 releases a lock) for that case, while creating the temporary .eml format
330 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
332 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
333 $sender_verify_failure/$recipient_verify_failure to "random".
335 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
338 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
339 Previously this would segfault.
341 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
344 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
345 like zero, since the resolver should be doing this for us, But we need one
346 as a CNAME but no MX presence gets the CNAME returned; we need to check
347 that doesn't point to an MX to declare it "no MX returned" rather than
348 "error, loop". A new main option is added so the older capability of
349 following some limited number of chain links is maintained.
351 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
353 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
354 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
355 now-deprecated earlier definitions used only specified the range up to TLS
356 1.2 (in the older-version library docs).
358 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
360 JH/15 Rework TLS client-side context management. Stop using a global, and
361 explicitly pass a context around. This enables future use of TLS for
362 connections to service-daemons (eg. malware scanning) while a client smtp
363 connection is using TLS; with cutthrough connections this is quite likely.
365 JH/16 Fix ARC verification to do AS checks in reverse order.
367 JH/17 Support a "tls" option on the ${readsocket } expansion item.
369 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
370 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
371 Previously the "utf8" would be re-prepended for every additional message.
373 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
374 Previously thery were accepted, resulting in issues when attempting to
375 forward messages to a non-supporting MTA.
377 PP/02 Let -n work with printing macros too, not just options.
379 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
380 one parent address was copied, and bogus data was used at delivery-logging
381 time. Either a crash (after delivery) or bogus log data could result.
382 Discovery and analysis by Tim Stewart.
384 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
385 Previously if the string ended mid-character, we did not insert the
386 promised '?' replacement.
388 PP/04 Documentation: current string operators work on bytes, not codepoints.
390 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
391 should pack well giving a smaller memory footprint so better caching and
392 therefore performance. Group the declarations where this can't be done so
393 that the byte-sized flag variables are not interspersed among pointer
394 variables, giving a better chance of good packing by the compiler.
396 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
397 non-null, to avoid issues with sites running BATV. Previously reports were
398 sent with an empty envelope sender so looked like bounces.
400 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
401 The ignore_error flag wasn't being returned from the filter subprocess so
402 was not set for later routers. Investigation and fix by Matthias Kurz.
404 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
405 and a msg:complete for the whole, when a message is manually removed using
406 -Mrm. Developement by Matthias Kurz, hacked on by JH.
408 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
409 a "Gnu special" function, asprintf() in the DB utility binary builds; I
410 hope that is portable enough.
412 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
413 requiring a known-CA anchor certificate; make it now rely entirely on the
414 TLSA as an anchor. Checking the name on the leaf cert against the name
415 on the A-record for the host is still done for TA (but not for EE mode).
417 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
418 included in delivery lines for non-proxied connections, when compiled with
419 SUPPORT_SOCKS and running with proxy logging enabled.
421 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
422 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
423 move the existing event to fire before the normal logging of message
424 failure so that custom logging is bracketed by normal logging.
426 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
427 msg:fail:internal event. Developement by Matthias Kurz.
429 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
430 far too small for todays use of crypto signatures stored there. Go all
431 the way to the max DNS message size of 64kB, even though this might be
432 overmuch for IOT constrained device use.
434 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
435 copy a string over itself. The library routine is documented as not
436 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
438 JH/32 For main options check_spool_space and check_inode_space, where the
439 platform supports 64b integers, support more than the previous 2^31 kB
440 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
441 the previous G, M, k.
443 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
444 $authenticated_fail_id variable on authentication failure. Previously
447 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
448 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
449 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
450 GNUTLS_SEC_PARAM_MEDIUM.
452 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
453 side. Previously we would continue as if no SNI had been received.
455 JH/36 Harden the handling of string-lists. When a list consisted of a sole
456 "<" character, which should be a list-separator specification, we walked
457 off past the nul-terimation.
459 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
460 causes) even when the retry time is not yet met. Previously they were
461 not, meaning that when (say) an account was over-quota and temp-rejecting,
462 and multiple senders' messages were queued, only one sender would get
463 notified on each configured delay_warning cycle.
465 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
467 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
468 cipher-suites, an error can be left on the stack even for a succeeding
469 accept; this results in impossible error messages when a later operation
472 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
473 return error codes indicating retry. Under TLS1.3 this becomes required.
475 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
476 it only wrote the new authenticators, resulting in a lack of tracking of
477 peer changes of ESMTP extensions until the next cache flush.
479 JH/41 Fix the loop reading a message header line to check for integer overflow,
480 and more-often against header_maxsize. Previously a crafted message could
481 induce a crash of the recive process; now the message is cleanly rejected.
483 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
484 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
490 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
491 When redis_servers is set to a list of > 1 element, and the Redis servers
492 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
493 case of MOVED into a DEFER case instead, thus moving the query onto the
494 next server in the list. For a cluster of N elements, all N servers must
495 be defined in redis_servers.
497 GF/02 Catch and remove uninitialized value warning in exiqsumm
498 Check for existence of @ARGV before looking at $ARGV[0]
500 JH/01 Replace the store_release() internal interface with store_newblock(),
501 which internalises the check required to safely use the old one, plus
502 the allocate and data copy operations duplicated in both (!) of the
503 extant use locations.
505 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
506 modifier. This matches the restriction on the commandline.
508 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
509 Previously only the last row was returned.
511 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
512 we assumed that tags in the header were well-formed, and parsed the
513 element content after inspecting only the first char of the tag.
514 Assumptions at that stage could crash the receive process on malformed
517 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
518 While running the DKIM ACL we operate on the Permanent memory pool so that
519 variables created with "set" persist to the DATA ACL. Also (at any time)
520 DNS lookups that fail create cache records using the Permanent pool. But
521 expansions release any allocations made on the current pool - so a dnsdb
522 lookup expansion done in the DKIM ACL releases the memory used for the
523 DNS negative-cache, and bad things result. Solution is to switch to the
524 Main pool for expansions.
525 While we're in that code, add checks on the DNS cache during store_reset,
526 active in the testsuite.
527 Problem spotted, and debugging aided, by Wolfgang Breyha.
529 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
530 When none of the hosts presented to a transport match an already-open
531 connection, close it and proceed with the list. Previously we would
532 queue the message. Spotted by Lena with Yahoo, probably involving
535 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
536 Previously a spurious "250 OK id=" response was appended to the proper
539 JH/08 The "support for" informational output now, which built with Content
540 Scanning support, has a line for the malware scanner interfaces compiled
541 in. Interface can be individually included or not at build time.
543 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
544 by the template makefile "src/EDITME". The "STREAM" support for an older
545 ClamAV interface method is removed.
547 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
548 rows affected is given instead).
550 JH/11 The runtime Berkeley DB library version is now additionally output by
551 "exim -d -bV". Previously only the compile-time version was shown.
553 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
554 SMTP connection. Previously, when one had more receipients than the
555 first, an abortive onward connection was made. Move to full support for
556 multiple onward connections in sequence, handling cutthrough connection
557 for all multi-message initiating connections.
559 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
560 routers. Previously, a multi-recipient message would fail to match the
561 onward-connection opened for the first recipient, and cause its closure.
563 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
564 a timeout on read on a GnuTLS initiating connection, resulting in the
565 initiating connection being dropped. This mattered most when the callout
566 was marked defer_ok. Fix to keep the two timeout-detection methods
569 JH/15 Relax results from ACL control request to enable cutthrough, in
570 unsupported situations, from error to silently (except under debug)
571 ignoring. This covers use with PRDR, frozen messages, queue-only and
574 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
576 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
577 metadata, resulting in a crash in free().
579 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
580 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
581 Broken also in d185889f4, with init system revamp.
583 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
584 Previously we abruptly closed the connection after reading a malware-
585 found indication; now we go on to read the "scan ok" response line,
588 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
589 ACL. Previously, a crash would result.
591 JH/19 Speed up macro lookups during configuration file read, by skipping non-
592 macro text after a replacement (previously it was only once per line) and
593 by skipping builtin macros when searching for an uppercase lead character.
595 JH/20 DANE support moved from Experimental to mainline. The Makefile control
596 for the build is renamed.
598 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
599 was allocated for every new TLS startup, meaning one per message. Fix
600 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
602 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
603 reported the original. Fix to report (as far as possible) the ACL
604 result replacing the original.
606 JH/23 Fix memory leak during multi-message connections using STARTTLS under
607 OpenSSL. Certificate information is loaded for every new TLS startup,
608 and the resources needed to be freed.
610 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
612 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
615 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
616 DATA response info to the (existing) per-recipient response info for
617 the "C=" log element. It can have useful tracking info from the
618 destination system. Patch from Simon Arlott.
620 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
621 length value. Previously this would segfault.
623 HS/02 Support Avast multiline protoocol, this allows passing flags to
624 newer versions of the scanner.
626 JH/28 Ensure that variables possibly set during message acceptance are marked
627 dead before release of memory in the daemon loop. This stops complaints
628 about them when the debug_store option is enabled. Discovered specifically
629 for sender_rate_period, but applies to a whole set of variables.
630 Do the same for the queue-runner and queue-list loops, for variables set
631 from spool message files. Do the same for the SMTP per-message loop, for
632 certain variables indirectly set in ACL operations.
634 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
635 as a multi-recipient message from a mailinglist manager). The coding had
636 an arbitrary cutoff number of characters while checking for more input;
637 enforced by writing a NUL into the buffer. This corrupted long / fast
638 input. The problem was exposed more widely when more pipelineing of SMTP
639 responses was introduced, and one Exim system was feeding another.
640 The symptom is log complaints of SMTP syntax error (NUL chars) on the
641 receiving system, and refused recipients seen by the sending system
642 (propating to people being dropped from mailing lists).
643 Discovered and pinpointed by David Carter.
645 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
646 replaced by the ${authresults } expansion.
648 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
650 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
651 allows proper process termination in container environments.
653 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
654 Previously the "final dot" had a newline after it; ensure it is CR,LF.
656 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
657 and "err_perm", deprecated since 4.83 when the RFC-defined words
658 "temperror" and "permerror" were introduced.
660 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
661 transport-filters or DKIM-signing. The restriction was lost in the
662 consolidation of verify-callout and delivery SMTP handling.
663 Extend the restriction to also cover ARC-signing.
665 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
666 in defer=pass mode supply a 450 to the initiator. Previously the message
669 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
670 tls_require_ciphers is used as before.
672 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
673 "pass_unscanned". Only tmpfails from the scanner are written to
674 the paniclog, as they may require admin intervention (permission
675 denied, license issues). Other scanner errors (like decompression
676 bombs) do not cause a paniclog entry.
678 JH/36 Fix reinitialisation of DKIM logging variable between messages.
679 Previously it was possible to log spurious information in receive log
682 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
683 triggered odd behaviour from Outlook Express clients.
685 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
688 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
689 since the IETF WG has not yet settled on that versus the original
690 "bare" representation.
692 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
693 Previously the millisecond value corrupted the output.
694 Fix also for syslog_pid=no and log_selector +pid, for which the pid
695 corrupted the output.
701 JH/01 Rework error string handling in TLS interface so that the caller in
702 more cases is responsible for logging. This permits library-sourced
703 string to be attached to addresses during delivery, and collapses
704 pairs of long lines into single ones.
706 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
707 during configuration. Wildcards are allowed and expanded.
709 JH/02 Rework error string handling in DKIM to pass more info back to callers.
710 This permits better logging.
712 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
713 do not close it down and have the child transport start it up again on
714 the passed-on TCP connection. Instead, proxy the child (and any
715 subsequent ones) for TLS via a unix-domain socket channel. Logging is
716 affected: the continued delivery log lines do not have any DNSSEC, TLS
717 Certificate or OCSP information. TLS cipher information is still logged.
719 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
720 identical IP addresses on different listening ports. Will also affect
723 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
724 add noisy ifdef guards to special-case this sillyness.
725 Patch from Bernd Kuhls.
727 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
728 than 255 are no longer allowed.
730 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
731 Disable the session-cache too, which might reduce our load. Since we
732 currrectly use a new context for every connection, both as server and
733 client, there is no benefit for these.
734 GnuTLS appears to not support tickets server-side by default (we don't
735 call gnutls_session_ticket_enable_server()) but client side is enabled
736 by default on recent versions (3.1.3 +) unless the PFS priority string
739 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
740 <https://reproducible-builds.org/specs/source-date-epoch/>.
742 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
743 the check for any unsuccessful recipients did not notice the limit, and
744 erroneously found still-pending ones.
746 JH/08 Pipeline CHUNKING command and data together, on kernels that support
747 MSG_MORE. Only in-clear (not on TLS connections).
749 JH/09 Avoid using a temporary file during transport using dkim. Unless a
750 transport-filter is involved we can buffer the headers in memory for
751 creating the signature, and read the spool data file once for the
752 signature and again for transmission.
754 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
755 4.77 as the kernel support then wasn't solid, having issues in 64bit
756 mode. Now, it's been long enough. Add support for FreeBSD also.
758 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
759 case where the routing stage had gathered several addresses to send to
760 a host before calling the transport for the first, we previously failed
761 to close down TLS in the old transport process before passing the TCP
762 connection to the new process. The new one sent a STARTTLS command
763 which naturally failed, giving a failed delivery and bloating the retry
764 database. Investigation and fix prototype from Wolfgang Breyha.
766 JH/12 Fix check on SMTP command input synchronisation. Previously there were
767 false-negatives in the check that the sender had not preempted a response
768 or prompt from Exim (running as a server), due to that code's lack of
769 awareness of the SMTP input buffering.
771 PP/04 Add commandline_checks_require_admin option.
772 Exim drops privileges sanely, various checks such as -be aren't a
773 security problem, as long as you trust local users with access to their
774 own account. When invoked by services which pass untrusted data to
775 Exim, this might be an issue. Set this option in main configuration
776 AND make fixes to the calling application, such as using `--` to stop
779 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
780 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
781 responses to those, into a single TLS record each way (this usually means
782 a single packet). As a side issue, smtp_enforce_sync now works on TLS
785 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
786 affects you only if you're dancing at the edge of the param size limits.
787 If you are, and this message makes sense to you, then: raise the
788 configured limit or use OpenSSL 1.1. Nothing we can do for older
791 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
792 cmdline element to get the documented default one. Previously it was
795 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
796 are unsuable for non-message-specific reasons - eg. connection timeout,
797 banner-time rejection.
799 JH/16 Fix logging of delivery remote port, when specified by router, under
802 PP/06 Repair manualroute's ability to take options in any order, even if one
803 is the name of a transport.
806 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
808 JH/17 Change the list-building routines interface to use the expanding-string
809 triplet model, for better allocation and copying behaviour.
811 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
812 Previously it was constructed the first time a possibly-matching string
813 was met in the configuration file input during startup; now it is done
816 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
817 compatible one, to avoid the (poorly documented) possibility of a config
818 file in the working directory redirecting the DB files, possibly correpting
819 some existing file. CVE-2017-10140 assigned for BDB.
821 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
822 cache-hot. Previously, although the result was properly cached, the
823 initial verify call returned a defer.
825 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
826 the main verify for receipient in uncached-mode.
828 JH/22 Retire historical build files to an "unsupported" subdir. These are
829 defined as "ones for which we have no current evidence of testing".
831 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
832 if present. Previously it was ignored.
834 JH/24 Start using specified-initialisers in C structure init coding. This is
835 a C99 feature (it's 2017, so now considered safe).
837 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
838 if was a fixed-sized field and bitmask ops via macros; it is now more
841 PP/07 GitHub PR 56: Apply MariaDB build fix.
842 Patch provided by Jaroslav Škarvada.
844 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
845 during Coverity cleanups [4.87 JH/47]
846 Diagnosis and fix provided by Michael Fischer v. Mollard.
848 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
849 the right size to place the terminating semicolon on its own folded
850 line, the header hash was calculated to an incorrect value thanks to
851 the (relaxed) space the fold became.
853 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
854 and confused the parent.
856 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
857 which could crash as a result. This could lead to undeliverable messages.
859 JH/28 Logging: "next input sent too soon" now shows where input was truncated
862 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
863 matters on fast-turnover and PID-randomising systems, which were getting
864 out-of-order delivery.
866 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
867 a possibly-overlapping copy. The symptom was that "Remote host closed
868 connection in response to HELO" was logged instead of the actual 4xx
871 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
872 Previously only that bufferd was discarded, resulting in SYMTP command
875 JH/32 DKIM: when a message has multiple signatures matching an identity given
876 in dkim_verify_signers, run the dkim acl once for each. Previously only
877 one run was done. Bug 2189.
879 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
880 "panic the current process" to "deliberately defer". The panic log is
881 still written with the problem list name; the mail and reject logs now
882 get a temp-reject line for the message that was being handled, saying
883 something like "domains check lookup or other defer". The SMTP 451
884 message is still "Temporary local problem".
886 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
887 A crafted sequence of BDAT commands could result in in-use memory beeing
888 freed. CVE-2017-16943.
890 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
891 from SMTP input. Previously it was always done; now only done for DATA
892 and not BDAT commands. CVE-2017-16944.
894 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
895 to the message (such as an overlong header line). Previously this was
896 not done and we did not exit BDAT mode. Followon from the previous item
897 though a different problem.
903 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
904 than -2003 did; needs libidn2 in addition to libidn.
906 JH/02 The path option on a pipe transport is now expanded before use.
908 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
909 Patch provided by "Björn", documentation fix added too.
911 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
912 missing a wire-to-host endian conversion.
914 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
915 close after a BDAT command line could be taken as a following command,
916 giving a synch failure. Fix by only checking for synch immediately
917 before acknowledging the chunk.
919 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
920 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
921 macro. Patches provided by Josh Soref.
923 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
924 Previously we did not; the RFC seems ambiguous and VRFY is not listed
925 by IANA as a service extension. However, John Klensin suggests that we
928 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
929 the dkim code may be unix-mode line endings rather than smtp wire-format
930 CRLF, so prepend a CR to any bare LF.
932 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
933 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
935 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
936 the first were themselves being wrongly included in the feed into dkim
937 processing; with most chunk sizes in use this resulted in an incorrect
938 body hash calculated value.
940 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
941 DKIM signature block, for verification. Although advised against by
942 standards it is specifically not ruled illegal.
944 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
946 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
947 missing a body hash (the bh= tag).
949 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
950 It seems that HAProxy sends the Proxy Protocol information in clear and
951 only then does a TLS startup, so do the same.
953 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
954 TCP connections (such as for Spamd) unless the daemon successfully set
955 Fast Open mode on its listening sockets. This fixes breakage seen on
956 too-old kernels or those not configured for Fast Open, at the cost of
957 requiring both directions being enabled for TFO, and TFO never being used
958 by non-daemon-related Exim processes.
960 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
961 endings, at least on the first header line. Try to canonify any that get
962 past that check, despite the cost.
964 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
965 now limited to an arbitrary five deep, while parsing addresses with the
966 strip_excess_angle_brackets option enabled.
968 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
969 instead leave the unprompted TLS handshake in socket buffer for the
970 TLS library to consume.
972 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
974 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
976 JH/16 Drop variables when they go out of scope. Memory management drops a whole
977 region in one operation, for speed, and this leaves assigned pointers
978 dangling. Add checks run only under the testsuite which checks all
979 variables at a store-reset and panics on a dangling pointer; add code
980 explicitly nulling out all the variables discovered. Fixes one known
981 bug: a transport crash, where a dangling pointer for $sending_ip_address
982 originally assigned in a verify callout, is re-used.
984 PP/06 Drop '.' from @INC in various Perl scripts.
986 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
988 PP/08 Reduce a number of compilation warnings under clang; building with
989 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
990 should be warning-free.
992 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
994 HS/01 Fix portability problems introduced by PP/08 for platforms where
995 realloc(NULL) is not equivalent to malloc() [SunOS et al].
997 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
998 chunk. This allows us to accept broken chunked messages. We need a more
999 general solution here.
1001 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1002 already-broken messages in the queue.
1004 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1006 JH/19 Fix reference counting bug in routing-generated-address tracking.
1012 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1013 supports it and a size is available (ie. the sending peer gave us one).
1015 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1016 years of being deprecated). The replacements are the ACLs
1017 acl_smtp_mime and acl_not_smtp_mime.
1019 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1020 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1021 or even an in-clear connection were permitted. Now, if the host lookup
1022 was dnssec and dane was requested then the host is only used if the
1023 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1024 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1025 if one fails this test.
1026 This means that a poorly-configured remote DNS will make it incommunicado;
1027 but it protects against a DNS-interception attack on it.
1029 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1030 non-noisy when a race steals the message being considered.
1032 JH/05 If main configuration option tls_certificate is unset, generate a
1033 self-signed certificate for inbound TLS connections.
1035 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1036 in rewrites and routers.
1038 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1039 and logged a warning sing 4.83; now they are a configuration file error.
1041 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1042 (lacking @domain). Apply the same qualification processing as RCPT.
1044 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1046 JH/10 Support ${sha256:} applied to a string (as well as the previous
1049 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1050 a cutthrough deliver is pending, as we always want to make a connection.
1051 This also avoids re-routing the message when later placing the cutthrough
1052 connection after a verify cache hit.
1053 Do not update it with the verify result either.
1055 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1056 when routing results in more than one destination address.
1058 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1059 signing (which inhibits the cutthrough capability). Previously only
1060 the presence of an option was tested; now an expansion evaluating as
1061 empty is permissible (obviously it should depend only on data available
1062 when the cutthrough connection is made).
1064 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1065 the relevant preceding SMTP command did not note the pipelining mode.
1067 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1068 Previously they were not counted.
1070 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1071 as one having no matching records. Previously we deferred the message
1072 that needed the lookup.
1074 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1075 distinguished as "(=".
1077 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1078 for missing MX records. Previously it only worked for missing A records.
1080 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1082 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1083 after the data-go-ahead and data-ack. Patch from Jason Betts.
1085 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1086 even for a "none" policy. Patch from Tony Meyer.
1088 JH/22 Fix continued use of a connection for further deliveries. If a port was
1089 specified by a router, it must also match for the delivery to be
1092 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1093 When one of the recipients of a message was unsuitable for the connection
1094 (has no matching addresses), we lost track of needing to mark it
1095 deferred. As a result mail would be lost.
1097 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1099 JH/25 Decoding ACL controls is now done using a binary search; the source code
1100 takes up less space and should be simpler to maintain. Merge the ACL
1101 condition decode tables also, with similar effect.
1103 JH/26 Fix problem with one_time used on a redirect router which returned the
1104 parent address unchanged. A retry would see the parent address marked as
1105 delivered, so not attempt the (identical) child. As a result mail would
1108 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1109 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1110 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1113 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1114 Main config options check_{log,spool}_{inodes,space} are now
1115 100 inodes, 10MB unless set otherwise in the configuration.
1117 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1118 Previously it only applied to the main-section connection policy
1121 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1123 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1124 by me. Added RFC7919 DH primes as an alternative.
1126 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1127 are not in the system include path.
1129 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1130 GnuTLS, when a session startup failed (eg because the client disconnected)
1131 Exim did stdio operations after fclose. This was exposed by a recent
1132 change which nulled out the file handle after the fclose.
1134 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1135 signed directly by the cert-signing cert, rather than an intermediate
1136 OCSP-signing cert. This is the model used by LetsEncrypt.
1138 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1140 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1141 an incoming connection.
1143 HS/02 Bug 1802: Do not half-close the connection after sending a request
1146 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1147 fallback to "prime256v1".
1149 JH/34 SECURITY: Use proper copy of DATA command in error message.
1150 Could leak key material. Remotely exploitable. CVE-2016-9963.
1156 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1157 and 3.4.4 - once the server is enabled to respond to an OCSP request
1158 it does even when not requested, resulting in a stapling non-aware
1159 client dropping the TLS connection.
1161 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1162 support variable-length bit vectors. No functional change.
1164 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1165 The I= interface field on outgoing lines is now after the H= remote
1166 host field, same as incoming lines. There is a separate
1167 outgoing_interface log selector which allows you to disable the
1170 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1171 If not running log_selector +smtp_connection the mainlog would be held
1172 open indefinitely after a "too many connections" event, including to a
1173 deleted file after a log rotate. Leave the per net connection logging
1174 leaving it open for efficiency as that will be quickly detected by the
1175 check on the next write.
1177 HS/01 Bug 1671: Fix post transport crash.
1178 Processing the wait-<transport> messages could crash the delivery
1179 process if the message IDs didn't exist for some reason. When
1180 using 'split_spool_directory=yes' the construction of the spool
1181 file name failed already, exposing the same netto behaviour.
1183 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1184 mime_regex ACL conditions.
1186 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1187 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1188 to HELO, local diagnostic string.
1190 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1191 log line to debug. Even when configured with a tls authenticator many
1192 client connections are expected to not authenticate in this way, so
1193 an authenticate fail is not an error.
1195 HS/02 Add the Exim version string to the process info. This way exiwhat
1196 gives some more detail about the running daemon.
1198 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1199 matter for fast-change records such as DNSBLs.
1201 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1202 retry record, even if constant. There may be multiple transports with
1203 different interface settings and the retry behaviour needs to be kept
1206 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1208 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1210 JH/10 Bug 840: fix log_defer_output option of pipe transport
1212 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1213 hosts_randomize. This matters a lot when combined with mua_wrapper.
1215 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1216 ${quote_pgsql:<string>} operator.
1218 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1220 JH/14 Change header file name for freeradius-client. Relevant if compiling
1221 with Radius support; from the Gentoo tree and checked under Fedora.
1223 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1225 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1226 option transport_filter, meaning no filtering.
1228 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1230 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1231 defaults to "*" (all hosts). The variable is now available when not built
1232 with TLS, default unset, mainly to enable keeping the testsuite sane.
1233 If a server certificate is not supplied (via tls_certificate) an error is
1234 logged, and clients will find TLS connections fail on startup. Presumably
1235 they will retry in-clear.
1236 Packagers of Exim are strongly encouraged to create a server certificate
1237 at installation time.
1239 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1240 with the $config_file variable.
1242 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1243 in transport context, after the attempt, and per-recipient. The latter type
1244 is per host attempted. The event data is the error message, and the errno
1245 information encodes the lookup type (A vs. MX) used for the (first) host,
1246 and the trailing two digits of the smtp 4xx response.
1248 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1249 to write to mainlog (or rejectlog, paniclog) in the window between file
1250 creation and permissions/ownership being changed. Particularly affects
1251 installations where exicyclog is run as root, rather than exim user;
1252 result is that the running daemon panics and dies.
1254 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1256 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1257 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1258 "pri" and "weight". Note that the previous implicit priority given by the
1259 list order is no longer honoured.
1261 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1262 for DKIM processing.
1264 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1265 by defining SUPPORT_SOCKS.
1267 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1268 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1269 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1270 variables are renamed to proxy_{local,external}_{address,port}.
1272 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1273 for a build by defining SUPPORT_I18N
1275 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1276 of the query string, and make ${quote_redis:} do that quoting.
1278 JH/29 Move Events support from Experimental to mainline, enabled by default
1279 and removable for a build by defining DISABLE_EVENT.
1281 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1283 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1284 cached by the daemon.
1286 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1287 by defining LOOKUP_REDIS. The libhiredis library is required.
1289 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1290 keys are given for lookup.
1292 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1293 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1294 only supported when built with TLS support. The PolarSSL SHA routines
1295 are still used when the TLS library is too old for convenient support.
1297 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1298 openssl_options), for security. OpenSSL forces this from version 1.1.0
1299 server-side so match that on older versions.
1301 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1302 allocation for $value could be released as the expansion processing
1303 concluded, but leaving the global pointer active for it.
1305 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1306 and to use the domains and local_parts ACL conditions.
1308 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1309 incorrectly not doubled on cutthrough transmission, hence seen as a
1310 body-termination at the receiving system - resulting in truncated mails.
1311 Commonly the sender saw a TCP-level error, and retransmitted the message
1312 via the normal store-and-forward channel. This could result in duplicates
1313 received - but deduplicating mailstores were liable to retain only the
1314 initial truncated version.
1316 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1318 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1320 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1321 we're in there, support oversigning also; bug 1309.
1323 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1325 HS/04 Add support for keep_environment and add_environment options.
1327 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1328 either intentional arithmetic overflow during PRNG, or testing config-
1331 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1332 delivery resulted in actual delivery. Cancel cutthrough before DATA
1335 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1336 rejects a recipient: pass the reject to the originator.
1338 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1339 Many were false-positives and ignorable, but it's worth fixing the
1342 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1343 for the new environment-manipulation done at startup. Move the routines
1344 from being local to tls.c to being global via the os.c file.
1346 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1347 an extract embedded as result-arg for a map, the first arg for extract
1348 is unavailable so we cannot tell if this is a numbered or keyed
1349 extraction. Accept either.
1355 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1358 JH/02 The smtp transport option "multi_domain" is now expanded.
1360 JH/03 The smtp transport now requests PRDR by default, if the server offers
1363 JH/04 Certificate name checking on server certificates, when exim is a client,
1364 is now done by default. The transport option tls_verify_cert_hostnames
1365 can be used to disable this per-host. The build option
1366 EXPERIMENTAL_CERTNAMES is withdrawn.
1368 JH/05 The value of the tls_verify_certificates smtp transport and main options
1369 default to the word "system" to access the system default CA bundle.
1370 For GnuTLS, only version 3.0.20 or later.
1372 JH/06 Verification of the server certificate for a TLS connection is now tried
1373 (but not required) by default. The verification status is now logged by
1374 default, for both outbound TLS and client-certificate supplying inbound
1377 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1380 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1381 Status Notification (bounce) messages are now MIME format per RFC 3464.
1382 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1383 under the control of the dsn_advertise_hosts option, and routers may
1384 have a dsn_lasthop option.
1386 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1387 default, modifiable by a malware= option. The list separator for
1388 the options can now be changed in the usual way. Bug 68.
1390 JH/10 The smtp_receive_timeout main option is now expanded before use.
1392 JH/11 The incoming_interface log option now also enables logging of the
1393 local interface on delivery outgoing connections.
1395 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1396 if the interface and destination host and port all match.
1398 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1401 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1402 Patch from Andrew Lewis.
1404 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1405 now supports optional time-restrictions, weighting, and priority
1406 modifiers per server. Patch originally by <rommer@active.by>.
1408 JH/16 The spamd_address main option now supports a mixed list of local
1409 and remote servers. Remote servers can be IPv6 addresses, and
1410 specify a port-range.
1412 JH/17 Bug 68: The spamd_address main option now supports an optional
1413 timeout value per server.
1415 JH/18 Bug 1581: Router and transport options headers_add/remove can
1416 now have the list separator specified.
1418 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1421 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1424 JH/21 Support for the A6 type of dns record is withdrawn.
1426 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1427 rather than the verbs used.
1429 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1430 from 255 to 1024 chars.
1432 JH/24 Verification callouts now attempt to use TLS by default.
1434 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1435 are generic router options now. The defaults didn't change.
1437 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1438 Original patch from Alexander Shikoff, worked over by JH.
1440 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1441 files if ZCAT_COMMAND is not executable.
1443 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1445 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1447 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1448 Normally benign, it bites when the pair was led to by a CNAME;
1449 modern usage is to not canonicalize the domain to a CNAME target
1450 (and we were inconsistent anyway for A-only vs AAAA+A).
1452 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1454 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1455 when evaluating $sender_host_dnssec.
1457 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1458 $sender_helo_dnssec variable.
1460 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1462 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1464 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1466 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1467 documented as working, but never had. Support all but $spam_report.
1469 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1470 added for tls authenticator.
1472 HS/03 Add perl_taintmode main config option
1478 TL/01 When running the test suite, the README says that variables such as
1479 no_msglog_check are global and can be placed anywhere in a specific
1480 test's script, however it was observed that placement needed to be near
1481 the beginning for it to behave that way. Changed the runtest perl
1482 script to read through the entire script once to detect and set these
1483 variables, reset to the beginning of the script, and then run through
1484 the script parsing/test process like normal.
1486 TL/02 The BSD's have an arc4random API. One of the functions to induce
1487 adding randomness was arc4random_stir(), but it has been removed in
1488 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1489 function when detected.
1491 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1492 cause callback expansion.
1494 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1495 syntax errors in an expansion can be treated as a string instead of
1496 logging or causing an error, due to the internal use of bool_lax
1497 instead of bool when processing it.
1499 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1500 server certificates when making smtp deliveries.
1502 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1504 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1506 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1508 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1509 Merged patch from Sebastian Wiedenroth.
1511 JH/05 Fix results-pipe from transport process. Several recipients, combined
1512 with certificate use, exposed issues where response data items split
1513 over buffer boundaries were not parsed properly. This eventually
1514 resulted in duplicates being sent. This issue only became common enough
1515 to notice due to the introduction of connection certificate information,
1516 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1518 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1519 size buffer was used, resulting in syntax errors when an expansion
1522 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1523 version 3.3.6 or later.
1525 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1526 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1527 both become "event_action", the variables become $event_name, $event_data
1528 and $event_defer_errno. There is a new variable $verify_mode, usable in
1529 routers, transports and related events. The tls:cert event is now also
1530 raised for inbound connections, if the main configuration event_action
1533 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1534 early OCSP support, but no stapling (appears to be less than 1.0.0).
1536 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1537 server certificate names available under the smtp transport option
1538 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1541 JH/10 Time-related extraction expansions from certificates now use the main
1542 option "timezone" setting for output formatting, and are consistent
1543 between OpenSSL and GnuTLS compilations. Bug 1541.
1545 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1546 encoded parameter in the incoming message. Bug 1558.
1548 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1549 include certificate info, eximon was claiming there were spoolfile
1552 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1554 JH/14 Log delivery-related information more consistently, using the sequence
1555 "H=<name> [<ip>]" wherever possible.
1557 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1558 are problematic for Debian distribution, omit them from the release
1561 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1563 JH/16 Fix string representation of time values on 64bit time_t architectures.
1566 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1567 output list separator was used.
1572 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1573 checkers that were complaining about end of non-void function with no
1576 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1577 This was a regression introduced in 4.83 by another bugfix.
1579 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1581 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1582 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1588 TF/01 Correctly close the server side of TLS when forking for delivery.
1590 When a message was received over SMTP with TLS, Exim failed to clear up
1591 the incoming connection properly after forking off the child process to
1592 deliver the message. In some situations the subsequent outgoing
1593 delivery connection happened to have the same fd number as the incoming
1594 connection previously had. Exim would try to use TLS and fail, logging
1595 a "Bad file descriptor" error.
1597 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1598 utilities have not been installed.
1600 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1601 temporary space as the ACL may create new global variables.
1603 TL/01 LDAP support uses per connection or global context settings, depending
1604 upon the detected version of the libraries at build time.
1606 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1607 to extract and use the src ip:port in logging and expansions as if it
1608 were a direct connection from the outside internet. PPv2 support was
1609 updated based on HAProxy spec change in May 2014.
1611 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1613 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1614 Properly escape header and check for NULL return.
1616 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1619 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1621 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1622 characters in header names, implemented as a verify condition.
1623 Contributed by Michael Fischer v. Mollard.
1625 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1626 results permerror and temperror. Previous values are deprecated but
1627 still accepted. In a future release, err_perm and err_temp will be
1628 completely removed, which will be a backward incompatibility if the
1629 ACL tests for either of these two old results. Patch contributed by
1630 user bes-internal on the mailing list.
1632 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1634 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1635 selectors, in both main and reject logs.
1637 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1640 JH/07 Add malware type "sock" for talking to simple daemon.
1642 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1644 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1645 routers/transports under cutthrough routing.
1647 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1648 numbers. Touch up "bool" conditional to keep the same definition.
1650 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1652 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1654 JH/12 Expand items in router/transport headers_add or headers_remove lists
1655 individually rather than the list as a whole. Bug 1452.
1657 Required for reasonable handling of multiple headers_ options when
1658 they may be empty; requires that headers_remove items with embedded
1659 colons must have them doubled (or the list-separator changed).
1661 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1662 view the policy declared in the DMARC record. Currently, $dmarc_status
1663 is a combined value of both the record presence and the result of the
1666 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1668 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1669 dnslookup router and the smtp transport (applying to the forward
1672 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1673 of ldap servers used for a specific lookup. Patch provided by Heiko
1676 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1677 New variable $lookup_dnssec_authenticated for observability.
1679 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1680 Patch submitted by Lars Timman.
1682 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1684 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1685 Requires trusted mode and valid format message id, aborts otherwise.
1686 Patch contributed by Heiko Schlichting.
1688 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1689 certextract with support for various fields. Bug 1358.
1691 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1692 is requested by default, modifiable by smtp transport option
1695 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1696 operate on certificate variables to give certificate fingerprints
1697 Also new ${sha256:cert_variable}.
1699 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1701 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1702 Christian Aistleitner.
1704 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1706 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1707 file. Patch from Wolfgang Breyha.
1709 JH/25 Expand the coverage of the delivery $host and $host_address to
1710 client authenticators run in verify callout. Bug 1476.
1712 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1713 align with daemon_smtp_ports. Bug 72.
1715 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1716 support and error reporting did not work properly.
1718 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1719 and is readable. Patch from Andrew Colin Kissa.
1721 TL/14 Enhance documentation of ${run expansion and how it parses the
1722 commandline after expansion, particularly in the case when an
1723 unquoted variable expansion results in an empty value.
1725 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1727 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1728 by renaming away from it. Addresses GH issue 15, reported by
1731 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1733 TL/15 SECURITY: prevent double expansion in math comparison functions
1734 (can expand unsanitized data). Not remotely exploitable.
1741 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1743 PP/02 Make -n do something, by making it not do something.
1744 When combined with -bP, the name of an option is not output.
1746 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1749 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1750 $sender_host_name and config options to manage this, and basic check
1753 PP/05 DSCP support for outbound connections and control modifier for inbound.
1755 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1756 (Only plugin which currently uses this is kerberos4, which nobody should
1757 be using, but we should make it available and other future plugins might
1758 conceivably use it, even though it would break NAT; stuff *should* be
1759 using channel bindings instead).
1761 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1762 name; added for Sendmail compatibility; requires admin caller.
1763 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1764 just ignore it); requires trusted caller.
1765 Also parse but ignore: -Ac -Am -X<logfile>
1768 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1770 TL/02 Add +smtp_confirmation as a default logging option.
1772 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1773 Patch by Magnus Holmgren from 2007-02-20.
1775 TL/04 Bugzilla 1281 - Spec typo.
1776 Bugzilla 1283 - Spec typo.
1777 Bugzilla 1290 - Spec grammar fixes.
1779 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1781 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1783 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1784 mailing list by by Dmitry Isaikin.
1786 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1788 JH/02 Support "G" suffix to numbers in ${if comparisons.
1790 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1792 NM/01 Bugzilla 1197 - Spec typo
1793 Bugzilla 1196 - Spec examples corrections
1795 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1797 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1798 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1801 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1802 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1804 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1805 "acl {{name}{arg}...}", and optional args on acl condition
1808 JH/05 Permit multiple router/transport headers_add/remove lines.
1810 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1812 JH/07 Avoid using a waiting database for a single-message-only transport.
1813 Performance patch from Paul Fisher. Bugzilla 1262.
1815 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1818 JH/09 Add $headers_added variable, with content from use of ACL modifier
1819 add_header (but not yet added to the message). Bugzilla 199.
1821 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1822 Pulled from Bugzilla 817 by Wolfgang Breyha.
1824 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1826 (nb: this is the same fix as in Exim 4.80.1)
1828 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1831 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1834 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1836 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1837 advertise SMTP AUTH mechanism to us, instead of a generic
1838 protocol violation error. Also, make Exim more robust to bad
1839 data from the Dovecot auth socket.
1841 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1843 When a queue runner is handling a message, Exim first routes the
1844 recipient addresses, during which it prunes them based on the retry
1845 hints database. After that it attempts to deliver the message to
1846 any remaining recipients. It then updates the hints database using
1849 So if a recipient address works intermittently, it can get repeatedly
1850 deferred at routing time. The retry hints record remains fresh so the
1851 address never reaches the final cutoff time.
1853 This is a fairly common occurrence when a user is bumping up against
1854 their storage quota. Exim had some logic in its local delivery code
1855 to deal with this. However it did not apply to per-recipient defers
1856 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1858 This change adds a proper retry rule check during routing so that the
1859 final cutoff time is checked against the message's age. We only do
1860 this check if there is an address retry record and there is not a
1861 domain retry record; this implies that previous attempts to handle
1862 the address had the retry_use_local_parts option turned on. We use
1863 this as an approximation for the destination being like a local
1864 delivery, as in LMTP.
1866 I suspect this new check makes the old local delivery cutoff check
1867 redundant, but I have not verified this so I left the code in place.
1869 TF/02 Correct gecos expansion when From: is a prefix of the username.
1871 Test 0254 submits a message to Exim with the header
1875 When I ran the test suite under the user fanf2, Exim expanded
1876 the header to contain my full name, whereas it should have added
1877 a Resent-Sender: header. It erroneously treats any prefix of the
1878 username as equal to the username.
1880 This change corrects that bug.
1882 GF/01 DCC debug and logging tidyup
1883 Error conditions log to paniclog rather than rejectlog.
1884 Debug lines prefixed by "DCC: " to remove any ambiguity.
1886 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1888 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1889 Bug spotted by Jeremy Harris; was flawed since initial commit.
1890 Would have resulted in OCSP responses post-SNI triggering an Exim
1891 NULL dereference and crash.
1893 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1895 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1896 Bug detection, analysis and fix by Samuel Thibault.
1897 Bugzilla 1331, Debian bug #698092.
1899 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1901 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1902 Server implementation by Todd Lyons, client by JH.
1903 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1904 config variable "prdr_enable" controls whether the server
1905 advertises the facility. If the client requests PRDR a new
1906 acl_data_smtp_prdr ACL is called once for each recipient, after
1907 the body content is received and before the acl_smtp_data ACL.
1908 The client is controlled by both of: a hosts_try_prdr option
1909 on the smtp transport, and the server advertisement.
1910 Default client logging of deliveries and rejections involving
1911 PRDR are flagged with the string "PRDR".
1913 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1914 fclose(). Diagnosis by Todd Lyons.
1916 PP/17 Update configure.default to handle IPv6 localhost better.
1917 Patch by Alain Williams (plus minor tweaks).
1920 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1921 This is now consistent with GnuTLS, and is now documented: the
1922 previous undocumented portable approach to treating the option as
1923 unset was to force an expansion failure. That still works, and
1924 an empty string is now equivalent.
1926 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1927 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1928 not performing validation itself.
1930 PP/20 Added force_command boolean option to pipe transport.
1931 Patch from Nick Koston, of cPanel Inc.
1933 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1936 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1938 PP/21 Fix eximon continuous updating with timestamped log-files.
1939 Broken in a format-string cleanup in 4.80, missed when I repaired the
1940 other false fix of the same issue.
1941 Report and fix from Heiko Schlichting.
1944 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1945 Report from Prashanth Katuri.
1947 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1948 It's SecureTransport, so affects any MacOS clients which use the
1949 system-integrated TLS libraries, including email clients.
1951 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1952 using a MIME ACL for non-SMTP local injection.
1953 Report and assistance in diagnosis by Warren Baker.
1955 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1957 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1959 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1960 last id that failed so it may be referenced in subsequent ACL's.
1962 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1965 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1966 ldap library initialization, allowing self-signed CA's to be
1967 used. Also properly sets require_cert option later in code by
1968 using NULL (global ldap config) instead of ldap handle (per
1969 session). Bug diagnosis and testing by alxgomz.
1971 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1972 the src/util/ subdirectory.
1974 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1975 renamed to Transport Post Delivery Action by Jeremy Harris, as
1978 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1979 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1980 redis_servers = needs to be configured which will be used by the redis
1981 lookup. Patch from Warren Baker, of The Packet Hub.
1983 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1985 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1986 hostname or reverse DNS when processing a host list. Used suggestions
1987 from multiple comments on this bug.
1989 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1991 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1992 Missed a few lines, added it to make the runtest require no keyboard
1995 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1996 contains upper case chars. Make router use caseful_local_part.
1998 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1999 support when GnuTLS has been built with p11-kit.
2005 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2007 This, or similar/improved, will also be change PP/11 of 4.82.
2013 PP/01 Handle short writes when writing local log-files.
2014 In practice, only affects FreeBSD (8 onwards).
2015 Bugzilla 1053, with thanks to Dmitry Isaikin.
2017 NM/01 Bugzilla 949 - Documentation tweak
2019 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2022 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2024 PP/02 Implemented gsasl authenticator.
2026 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2028 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2029 `pkg-config foo` for cflags/libs.
2031 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2032 with rest of GSASL and with heimdal_gssapi.
2034 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2035 `pkg-config foo` for cflags/libs for the TLS implementation.
2037 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2038 properties get this fed in as external SSF. A number of robustness
2039 and debugging improvements to the cyrus_sasl authenticator.
2041 PP/08 cyrus_sasl server now expands the server_realm option.
2043 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2044 Patch by Jeremy Harris.
2046 PP/10 Added dbmjz lookup type.
2048 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2050 PP/12 MAIL args handles TAB as well as SP, for better interop with
2051 non-compliant senders.
2052 Analysis and variant patch by Todd Lyons.
2054 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2055 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2056 Patch from Dirk Mueller <dmueller@suse.com>
2058 PP/13 tls_peerdn now print-escaped for spool files.
2059 Observed some $tls_peerdn in wild which contained \n, which resulted
2060 in spool file corruption.
2062 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2063 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2064 or write after TLS renegotiation, which otherwise led to messages
2067 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2068 as a tracking header (ie: a signed header comes before the signature).
2069 Patch from Wolfgang Breyha.
2071 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2072 comma-sep list; embedded commas doubled.
2074 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2076 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2078 Report and patch from Dmitry Banschikov.
2080 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2081 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2082 failures. We appear to now support TLS1.1+ with Exim.
2084 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2085 lets Exim select keys and certificates based upon TLS SNI from client.
2086 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2087 before an outbound SMTP session. New log_selector, +tls_sni.
2089 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2090 NULL dereference. Report and patch from Alun Jones.
2092 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2093 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2094 Not seeing resolver debug output on NetBSD, but suspect this is a
2095 resolver implementation change.
2097 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2098 Left warnings. Added "eximon gdb" invocation mode.
2100 PP/21 Defaulting "accept_8bitmime" to true, not false.
2102 PP/22 Added -bw for inetd wait mode support.
2104 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2105 locate the relevant includes and libraries. Made this the default.
2107 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2108 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2110 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2111 This may cause build issues on older platforms.
2113 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2114 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2115 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2116 Added SNI support via GnuTLS too.
2117 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2119 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2121 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2122 Applied second patch from Janne, implementing suggestion to default
2123 multiple-strings-in-record handling to match SPF spec.
2125 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2127 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2128 read-only, out of scope).
2129 Patch from Wolfgang Breyha, report from Stuart Northfield.
2131 PP/29 Fix three issues highlighted by clang analyser static analysis.
2132 Only crash-plausible issue would require the Cambridge-specific
2133 iplookup router and a misconfiguration.
2134 Report from Marcin Mirosław.
2136 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2138 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2139 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2140 As part of this, removing so much warning spew let me fix some minor
2141 real issues in debug logging.
2143 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2144 assignment on my part. Fixed.
2146 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2147 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2148 Janne Snabb (who went above and beyond: thank you).
2150 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2151 string otherwise requires a connection and a bunch more work and it's
2152 relatively easy to get wrong. Should also expose TLS library linkage
2155 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2156 64-bit ${eval} (JH/03).
2158 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2159 GNU libc to support some of the 64-bit stuff, should not lead to
2160 conflicts. Defined before os.h is pulled in, so if a given platform
2161 needs to override this, it can.
2163 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2164 protection layer was required, which is not implemented.
2165 Bugzilla 1254, patch from Wolfgang Breyha.
2167 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2168 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2169 tls_dhparam take prime identifiers. Also unbreak combination of
2170 OpenSSL+DH_params+TLSSNI.
2172 PP/39 Disable SSLv2 by default in OpenSSL support.
2178 PP/01 Solaris build fix for Oracle's LDAP libraries.
2179 Bugzilla 1109, patch from Stephen Usher.
2181 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2183 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2186 TF/02 Fix a couple more cases where we did not log the error message
2187 when unlink() failed. See also change 4.74-TF/03.
2189 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2190 lock up or crash if it happened to be inside a call to libc when it
2191 got a SIGUSR1 from exiwhat.
2193 The SIGUSR1 handler appends the current process status to the process
2194 log which is later printed by exiwhat. It used to use the general
2195 purpose logging code to do this, but several functions it calls are
2196 not safe for signals.
2198 The new output code in the SIGUSR1 handler is specific to the process
2199 log, and simple enough that it's easy to inspect for signal safety.
2200 Removing some special cases also simplifies the general logging code.
2201 Removing the spurious timestamps from the process log simplifies
2204 TF/04 Improved ratelimit ACL condition.
2206 The /noupdate option has been deprecated in favour of /readonly which
2207 has clearer semantics. The /leaky, /strict, and /readonly update modes
2208 are mutually exclusive. The update mode is no longer included in the
2209 database key; it just determines when the database is updated. (This
2210 means that when you upgrade Exim will forget old rate measurements.)
2212 Exim now checks that the per_* options are used with an update mode that
2213 makes sense for the current ACL. For example, when Exim is processing a
2214 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2215 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2216 must specify per_mail/readonly. If you omit the update mode it defaults to
2217 /leaky where that makes sense (as before) or /readonly where required.
2219 The /noupdate option is now undocumented but still supported for
2220 backwards compatibility. It is equivalent to /readonly except that in
2221 ACLs where /readonly is required you may specify /leaky/noupdate or
2222 /strict/noupdate which are treated the same as /readonly.
2224 A useful new feature is the /count= option. This is a generalization
2225 of the per_byte option, so that you can measure the throughput of other
2226 aggregate values. For example, the per_byte option is now equivalent
2227 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2229 The per_rcpt option has been generalized using the /count= mechanism
2230 (though it's more complicated than the per_byte equivalence). When it is
2231 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2232 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2233 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2234 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2235 Note that using per_rcpt with a non-readonly update mode in more than
2236 one ACL will cause the recipients to be double-counted. (The per_mail
2237 and per_byte options don't have this problem.)
2239 The handling of very low rates has changed slightly. If the computed rate
2240 is less than the event's count (usually one) then this event is the first
2241 after a long gap. In this case the rate is set to the same as this event's
2242 count, so that the first message of a spam run is counted properly.
2244 The major new feature is a mechanism for counting the rate of unique
2245 events. The new per_addr option counts the number of different
2246 recipients that someone has sent messages to in the last time period. It
2247 behaves like per_rcpt if all the recipient addresses are different, but
2248 duplicate recipient addresses do not increase the measured rate. Like
2249 the /count= option this is a general mechanism, so the per_addr option
2250 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2251 example, measure the rate that a client uses different sender addresses
2252 with the options per_mail/unique=$sender_address. There are further
2253 details in the main documentation.
2255 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2257 TF/06 Removed a few PCRE remnants.
2259 TF/07 Automatically extract Exim's version number from tags in the git
2260 repository when doing development or release builds.
2262 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2263 Bugzilla 879. Patch from Paul Fisher.
2265 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2266 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2269 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2271 PP/05 Variable $av_failed, true if the AV scanner deferred.
2272 Bugzilla 1078. Patch from John Horne.
2274 PP/06 Stop make process more reliably on build failure.
2275 Bugzilla 1087. Patch from Heiko Schlittermann.
2277 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2278 Bugzilla 1089. Patch from Heiko Schlittermann.
2280 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2281 Bugzilla 1131. Patch from Holger Weiß.
2283 PP/09 Handle IPv6 addresses with SPF.
2284 Bugzilla 860. Patch from Wolfgang Breyha.
2286 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2288 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2291 PP/11 match_* no longer expand right-hand-side by default.
2292 New compile-time build option, EXPAND_LISTMATCH_RHS.
2293 New expansion conditions, "inlist", "inlisti".
2295 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2297 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2299 PP/14 fix log_write() format string regression from TF/03.
2300 Bugzilla 1152. Patch from Dmitry Isaikin.
2306 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2308 PP/02 Harmonised TLS library version reporting; only show if debugging.
2309 Layout now matches that introduced for other libraries in 4.74 PP/03.
2311 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2313 PP/04 New "dns_use_edns0" global option.
2315 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2318 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2319 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2321 TK/01 Updated PolarSSL code to 0.14.2.
2322 Bugzilla 1097. Patch from Andreas Metzler.
2324 PP/07 Catch divide-by-zero in ${eval:...}.
2325 Fixes bugzilla 1102.
2327 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2330 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2331 format-string attack -- SECURITY: remote arbitrary code execution.
2333 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2334 time unintentionally subject to list matching rules, letting the header
2335 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2336 arbitrary string expansion). This allowed for information disclosure.
2338 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2339 INT_MIN/-1 -- value coerced to INT_MAX.
2345 NM/01 Workaround for PCRE version dependency in version reporting
2348 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2349 This fixes portability to compilers other than gcc, notably
2350 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2352 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2353 makefiles for portability to HP-UX and POSIX correctness.
2355 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2356 Also via indented variable definition in the Makefile.
2357 (Debugging by Oliver Heesakkers).
2359 PP/02 Restore caching of spamd results with expanded spamd_address.
2360 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2362 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2363 Improves build reliability. Fix from: Frank Elsner
2365 NM/02 Fix wide character breakage in the rfc2047 coding
2366 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2368 NM/03 Allow underscore in dnslist lookups
2369 Fixes bug 1026. Patch from Graeme Fowler
2371 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2372 Code patches from Adam Ciarcinski of NetBSD.
2374 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2377 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2378 is logged, to avoid truncation. Patch from John Horne.
2380 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2381 Patch from Jakob Hirsch.
2383 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2384 SQL string expansion failure details.
2385 Patch from Andrey Oktyabrski.
2387 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2388 Patch from Simon Arlott.
2390 PP/09 New lookups functionality failed to compile on old gcc which rejects
2391 extern declarations in function scope.
2392 Patch from Oliver Fleischmann
2394 PP/10 Use sig_atomic_t for flags set from signal handlers.
2395 Check getgroups() return and improve debugging.
2396 Fixed developed for diagnosis in bug 927 (which turned out to be
2399 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2400 Patch from Mark Zealey.
2402 PP/12 Bugzilla 1056: Improved spamd server selection.
2403 Patch from Mark Zealey.
2405 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2406 Based on patch from Heiko Schlittermann.
2408 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2409 Patch from Uwe Doering, sign-off by Michael Haardt.
2411 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2412 variable declaration deep within a block. Bug and patch from
2415 PP/15 lookups-Makefile IRIX compatibility coercion.
2417 PP/16 Make DISABLE_DKIM build knob functional.
2419 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2420 Patch by Simon Arlott
2422 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2423 variable argument macros. Our copy now differs from upstream.
2429 TF/01 Failure to get a lock on a hints database can have serious
2430 consequences so log it to the panic log.
2432 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2433 controlled using the smtp_confirmation log selector.
2435 TF/03 Include the error message when we fail to unlink a spool file.
2437 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2438 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2439 for maintaining out-of-tree patches for some time.
2441 PP/01 Bugzilla 139: Documentation and portability issues.
2442 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2443 Handle per-OS dynamic-module compilation flags.
2445 PP/02 Let /dev/null have normal permissions.
2446 The 4.73 fixes were a little too stringent and complained about the
2447 permissions on /dev/null. Exempt it from some checks.
2448 Reported by Andreas M. Kirchwitz.
2450 PP/03 Report version information for many libraries, including
2451 Exim version information for dynamically loaded libraries. Created
2452 version.h, now support a version extension string for distributors
2453 who patch heavily. Dynamic module ABI change.
2455 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2456 privilege escalation vulnerability whereby the Exim run-time user
2457 can cause root to append content of the attacker's choosing to
2460 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2463 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2464 If dropping privileges for untrusted macros, we disabled normal logging
2465 on the basis that it would fail; for the Exim run-time user, this is not
2466 the case, and it resulted in successful deliveries going unlogged.
2467 Fixed. Reported by Andreas Metzler.
2473 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2474 only prepend for the Resent-* case. Fixes regression introduced in
2475 Exim 4.70 by NM/22 for Bugzilla 607.
2477 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2478 increasing numbers of administrators be bitten by this.
2480 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2482 PP/03 Bugzilla 994: added openssl_options main configuration option.
2484 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2486 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2488 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2490 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2491 without a peer certificate, leading to a segfault because of an
2492 assumption that peers always have certificates. Be a little more
2493 paranoid. Problem reported by Martin Tscholak.
2495 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2496 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2497 NB: ClamAV planning to remove STREAM in "middle of 2010".
2498 CL also introduces -bmalware, various -d+acl logging additions and
2499 more caution in buffer sizes.
2501 PP/09 Implemented reverse_ip expansion operator.
2503 PP/10 Bugzilla 937: provide a "debug" ACL control.
2505 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2507 PP/12 Bugzilla 973: Implement --version.
2509 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2511 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2513 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2515 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2516 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2517 ignore trailing whitespace.
2519 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2521 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2524 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2525 Notification from Dr Andrew Aitchison.
2527 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2528 ExtendedDetectionInfo response format.
2529 Notification from John Horne.
2531 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2534 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2535 XSL and documented dependency on system catalogs, with examples of how
2538 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2541 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2542 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2543 configuration file which is writeable by the Exim user or group.
2545 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2546 of configuration files to cover files specified with the -C option if
2547 they are going to be used with root privileges, not just the default
2550 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2551 option (effectively making it always true).
2553 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2554 files to be used while preserving root privileges.
2556 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2557 that rogue child processes cannot use them.
2559 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2560 run-time user, instead of root.
2562 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2563 Exim run-time user without dropping privileges.
2565 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2566 result string, instead of calling string_vformat() twice with the same
2569 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2570 for other users. Others should always drop root privileges if they use
2571 -C on the command line, even for a whitelisted configure file.
2573 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2575 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2581 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2582 $data_path, and $header_path variables; fixed documentation bugs and
2585 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2586 exipick to access non-standard spools, including the "frozen" queue
2589 NM/01 Bugzilla 965: Support mysql stored procedures.
2590 Patch from Alain Williams
2592 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2594 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2595 Patch from Andreas Metzler
2597 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2598 Patch from Kirill Miazine
2600 NM/05 Bugzilla 671: Added umask to procmail example.
2602 JJ/03 installed exipick 20100323.0, fixing doc bug
2604 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2605 directory. Notification and patch from Dan Rosenberg.
2607 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2609 TK/02 Improve log output when DKIM signing operation fails.
2611 MH/01 Treat the transport option dkim_domain as a colon separated
2612 list, not as a single string, and sign the message with each element,
2613 omitting multiple occurences of the same signer.
2615 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2616 Bugzilla 985, 986. Patch by Simon Arlott
2618 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2619 Patch by Simon Arlott
2621 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2622 MBX locking. Notification from Dan Rosenberg.
2628 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2630 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2632 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2634 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2636 NM/04 Bugzilla 915: Flag broken perl installation during build.
2642 TK/01 Added patch by Johannes Berg that expands the main option
2643 "spamd_address" if it starts with a dollar sign.
2645 TK/02 Write list of recipients to X-Envelope-Sender header when building
2646 the mbox-format spool file for content scanning (suggested by Jakob
2649 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2650 (http://www.dcc-servers.net/) support via dccifd. Activated by
2651 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2653 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2654 by Mark Daniel Reidel <mr@df.eu>.
2656 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2657 When building exim an external PCRE library is now needed -
2658 PCRE is a system library on the majority of modern systems.
2659 See entry on PCRE_LIBS in EDITME file.
2661 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2662 conversation. Added nologin parameter to request.
2663 Patch contributed by Kirill Miazine.
2665 TF/01 Do not log submission mode rewrites if they do not change the address.
2667 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2669 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2670 log files in place. Contributed by Roberto Lima.
2672 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2674 TF/03 Bugzilla 615: When checking the local_parts router precondition
2675 after a local_part_suffix or local_part_prefix option, Exim now
2676 does not use the address's named list lookup cache, since this
2677 contains cached lookups for the whole local part.
2679 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2680 Robert Millan. Documentation is in experimental-spec.txt.
2682 TF/04 Bugzilla 668: Fix parallel build (make -j).
2684 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2686 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2687 Patch provided by Jan Srzednicki.
2689 TF/05 Leading white space used to be stripped from $spam_report which
2690 wrecked the formatting. Now it is preserved.
2692 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2693 that they are available at delivery time.
2695 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2697 TF/08 TLS error reporting now respects the incoming_interface and
2698 incoming_port log selectors.
2700 TF/09 Produce a more useful error message if an SMTP transport's hosts
2701 setting expands to an empty string.
2703 NM/06 Bugzilla 744: EXPN did not work under TLS.
2704 Patch provided by Phil Pennock.
2706 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2707 Patch provided by Richard Godbee.
2709 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2710 acl_smtp_notquit, added index entry.
2712 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2713 Patch provided by Eugene Bujak.
2715 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2716 accept(). Patch provided by Maxim Dounin.
2718 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2719 Patch provided by Phil Pennock.
2721 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2723 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2724 Patch provided by Brad "anomie" Jorsch.
2726 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2727 Patch provided by Dean Brooks.
2729 TK/05 Add native DKIM support (does not depend on external libraries).
2731 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2732 Patch provided by Graeme Fowler.
2734 NM/16 Bugzilla 851: Documentation example syntax fix.
2736 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2738 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2741 NM/19 Bugzilla 745: TLS version reporting.
2742 Patch provided by Phil Pennock.
2744 NM/20 Bugzilla 167: bool: condition support.
2745 Patch provided by Phil Pennock.
2747 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2748 clients. Patch provided by Phil Pennock.
2750 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2751 Patch provided by Brad "anomie" Jorsch.
2753 NM/23 Bugzilla 687: Fix misparses in eximstats.
2754 Patch provided by Heiko Schlittermann.
2756 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2757 Patch provided by Heiko Schlittermann.
2759 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2760 plus update to original patch.
2762 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2764 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2765 Patch provided by David Brownlee.
2767 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2769 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2771 NM/30 Bugzilla 888: TLS documentation bugfixes.
2773 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2775 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2776 Unlike the original bugzilla I have changed all shell scripts in src tree.
2778 NM/33 Bugzilla 898: Transport filter timeout fix.
2779 Patch by Todd Rinaldo.
2781 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2782 Patch by Serge Demonchaux.
2784 NM/35 Bugzilla 39: Base64 decode bug fixes.
2785 Patch by Jakob Hirsch.
2787 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2789 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2791 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2793 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2799 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2800 ALT-N's libdkim that I have put here:
2801 http://duncanthrax.net/exim-experimental/
2803 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2804 were called 'true' and it seems that C99 defines that as a reserved
2805 keyword to be used with 'bool' variable types. That means you could
2806 not include C99-style headers which use bools without triggering
2807 build errors in sieve.c.
2809 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2810 as mailq or other aliases. Changed the --help handling significantly
2811 to do whats expected. exim_usage() emits usage/help information.
2813 SC/01 Added the -bylocaldomain option to eximstats.
2815 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2817 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2819 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2825 PH/01 Another patch from the Sieve maintainer.
2827 PH/02 When an IPv6 address is converted to a string for single-key lookup
2828 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2829 dots are used instead of colons so that keys in lsearch files need not
2830 contain colons. This was done some time before quoting was made available
2831 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2832 (notated using the quote facility) so as to distinguish them from IPv4
2833 keys. This meant that lookups for IP addresses in host lists did not work
2834 for iplsearch lookups.
2836 This has been fixed by arranging for IPv6 addresses to be expressed with
2837 colons if the lookup type is iplsearch. This is not incompatible, because
2838 previously such lookups could never work.
2840 The situation is now rather anomalous, since one *can* have colons in
2841 ordinary lsearch keys. However, making the change in all cases is
2842 incompatible and would probably break a number of configurations.
2844 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2847 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2848 conversion specification without a maximum field width, thereby enabling
2849 a rogue spamd server to cause a buffer overflow. While nobody in their
2850 right mind would setup Exim to query an untrusted spamd server, an
2851 attacker that gains access to a server running spamd could potentially
2852 exploit this vulnerability to run arbitrary code as the Exim user.
2854 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2855 $primary_hostname instead of what libspf2 thinks the hosts name is.
2857 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2858 a directory entry by the name of the lookup key. Previously, if a
2859 symlink pointed to a non-existing file or a file in a directory that
2860 Exim lacked permissions to read, a lookup for a key matching that
2861 symlink would fail. Now it is enough that a matching directory entry
2862 exists, symlink or not. (Bugzilla 503.)
2864 PH/03 The body_linecount and body_zerocount variables are now exported in the
2867 PH/04 Added the $dnslist_matched variable.
2869 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2870 This means they are set thereafter only if the connection becomes
2873 PH/06 Added the client_condition to authenticators so that some can be skipped
2874 by clients under certain conditions.
2876 PH/07 The error message for a badly-placed control=no_multiline_responses left
2877 "_responses" off the end of the name.
2879 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2881 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2882 (without spaces) instead of just copying the configuration text.
2884 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2886 PH/11 Added $max_received_linelength.
2888 PH/12 Added +ignore_defer and +include_defer to host lists.
2890 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2891 way in which PCRE > 7.0 is built.
2893 PH/14 Implemented queue_only_load_latch.
2895 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2896 MAIL command. The effect was to mangle the value on 64-bit systems.
2898 PH/16 Another patch from the Sieve maintainer.
2900 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2902 PH/18 If a system quota error occurred while trying to create the file for
2903 a maildir delivery, the message "Mailbox is full" was not appended to the
2904 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2905 only to a quota excession during the actual writing of the file.
2907 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2908 characters?) which causes problems in log lines. The DN values are now
2909 passed through string_printing() before being added to log lines.
2911 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2912 and InterBase are left for another time.)
2914 PH/21 Added message_body_newlines option.
2916 PH/22 Guard against possible overflow in moan_check_errorcopy().
2918 PH/23 POSIX allows open() to be a macro; guard against that.
2920 PH/24 If the recipient of an error message contained an @ in the local part
2921 (suitably quoted, of course), incorrect values were put in $domain and
2922 $local_part during the evaluation of errors_copy.
2928 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2929 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2932 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2933 issue a MAIL command.
2935 PH/02 In an ACL statement such as
2937 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2939 if a client was not listed at all, or was listed with a value other than
2940 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2941 the condition was not true (as it should be), so access was not denied.
2942 The bug was that the ! inversion was incorrectly passed on to the second
2943 item. This has been fixed.
2945 PH/03 Added additional dnslists conditions == and =& which are different from
2946 = and & when the dns lookup returns more than one IP address.
2948 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2949 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2951 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2952 FSYNC, which compiles an option called disable_fsync that allows for
2953 bypassing fsync(). The documentation is heavily laced with warnings.
2955 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2957 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2958 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2959 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2960 including adding "make clean"; (3) Added -fPIC when compiling the test
2961 dynamically loaded module, to get rid of a warning.
2963 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2964 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2965 = 0s. The bug is otherwise harmless.
2967 PH/07 There was a bug in the dovecot authenticator such that the value of
2968 $auth1 could be overwritten, and so not correctly preserved, after a
2969 successful authentication. This usually meant that the value preserved by
2970 the server_setid option was incorrect.
2972 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2974 PH/09 Installed PCRE release 7.0.
2976 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2977 run for batched SMTP input. It is now run at the start of every message
2978 in the batch. While fixing this I discovered that the process information
2979 (output by running exiwhat) was not always getting set for -bs and -bS
2980 input. This is fixed, and it now also says "batched" for BSMTP.
2982 PH/11 Added control=no_pipelining.
2984 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2985 patch, slightly modified), and move the expansion of helo_data till after
2986 the connection is made in the smtp transport (so it can use these
2989 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2991 PH/14 Added log_selector = +pid.
2993 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2995 PH/16 Add ${if forany and ${if forall.
2997 PH/17 Added dsn_from option to vary the From: line in DSNs.
2999 PH/18 Flush SMTP output before performing a callout, unless control =
3000 no_callout_flush is set.
3002 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3003 was true (the default) a successful delivery failed to delete the retry
3004 item, thus causing premature timeout of the address. The bug is now
3007 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3009 PH/21 Long custom messages for fakedefer and fakereject are now split up
3010 into multiline responses in the same way that messages for "deny" and
3011 other ACL rejections are.
3013 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3014 with slight modification.
3016 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3017 draft, changing the syntax and factoring some duplicate code".
3019 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3020 for deliveries of the second and subsequent messages over the same SMTP
3023 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3024 ${reduce, with only minor "tidies".
3026 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3028 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3029 expansion side effects.
3031 PH/27 When a message times out after an over-quota error from an Exim-imposed
3032 quota, the bounce message says "mailbox is full". This message was not
3033 being given when it was a system quota that was exceeded. It now should
3036 MH/03 Made $recipients available in local_scan(). local_scan() already has
3037 better access to the recipient list through recipients_list[], but
3038 $recipients can be useful in postmaster-provided expansion strings.
3040 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3041 in the case of a MAIL command with additional options following the
3042 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3043 were accidentally chopped off.
3045 PH/29 SMTP synchronization checks are implemented when a command is read -
3046 there is a check that no more input is waiting when there shouldn't be
3047 any. However, for some commands, a delay in an ACL can mean that it is
3048 some time before the response is written. In this time, more input might
3049 arrive, invalidly. So now there are extra checks after an ACL has run for
3050 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3051 pipelining has not been advertised.
3053 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3055 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3056 returned in the SMTP response when smtp_return_error_details was false.
3057 This has been fixed.
3059 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3060 instead of the C I/O that was originally supplied, because problems were
3061 reported on Solaris.
3063 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3064 Exim which did not show up earlier: it was assuming that a call to
3065 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3066 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3067 was a macro that became an assignment, so it seemed to work. This has
3068 changed to a proper function call with a void return, hence the compile
3069 error. Exim's code has been fixed.
3071 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3074 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3076 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3078 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3079 because it turns out to be too big or there is a timeout) there is no
3080 "Completed" line in the log. When some messages of this type were
3081 selected by exigrep, they were listed as "not completed". Others were
3082 picked up by some special patterns. I have improved the selection
3083 criteria to be more general.
3085 PH/38 The host_find_failed option in the manualroute router can now be set
3086 to "ignore", to completely ignore a host whose IP address cannot be
3087 found. If all hosts are ignored, the behaviour is controlled by the new
3088 host_all_ignored option.
3090 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3091 homing or because of multiple MX records with /mx) generated more than
3092 one IP address, and the following item turned out to be the local host,
3093 all the secondary addresses of the first item were incorrectly removed
3094 from the list, along with the local host and any following hosts (which
3095 is what is supposed to happen).
3097 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3098 whoever called Exim into the -H file. In the case of the daemon it was
3099 behaving confusingly. When first started, it used values for whoever
3100 started the daemon, but after a SIGHUP it used the Exim user (because it
3101 calls itself on a restart). I have changed the code so that it now always
3104 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3105 message are rejected with the same error (e.g. no authentication or bad
3106 sender address), and a DATA command is nevertheless sent (as can happen
3107 with PIPELINING or a stupid MUA), the error message that was given to the
3108 RCPT commands is included in the rejection of the DATA command. This is
3109 intended to be helpful for MUAs that show only the final error to their
3112 PH/42 Another patch from the Sieve maintainer.
3114 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3115 Eximstats - Fixed some broken HTML links and added missing column headers
3117 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3118 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3120 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3122 PH/43 Yet another patch from the Sieve maintainer.
3124 PH/44 I found a way to check for a TCP/IP connection going away before sending
3125 the response to the final '.' that terminates a message, but only in the
3126 case where the client has not sent further data following the '.'
3127 (unfortunately, this is allowed). However, in many cases there won't be
3128 any further data because there won't be any more messages to send. A call
3129 to select() can be used: if it shows that the input is "ready", there is
3130 either input waiting, or the socket has been closed. An attempt to read
3131 the next input character can distinguish the two cases. Previously, Exim
3132 would have sent an OK response which the client would never have see.
3133 This could lead to message repetition. This fix should cure that, at
3134 least in a lot of common cases.
3136 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3137 advertised in response to EHLO.
3143 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3144 fixed by 4.65/MH/01 (is this a record?) are fixed:
3146 (i) An empty string was always treated as zero by the numeric comparison
3147 operators. This behaviour has been restored.
3149 (ii) It is documented that the numeric comparison operators always treat
3150 their arguments as decimal numbers. This was broken in that numbers
3151 starting with 0 were being interpreted as octal.
3153 While fixing these problems I realized that there was another issue that
3154 hadn't been noticed. Values of message_size_limit (both the global option
3155 and the transport option) were treated as octal if they started with 0.
3156 The documentation was vague. These values are now always treated as
3157 decimal, and I will make that clear in the documentation.
3163 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3164 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3167 MH/01 Don't check that the operands of numeric comparison operators are
3168 integers when their expansion is in "skipping" mode (fixes bug
3169 introduced by 4.64-PH/07).
3171 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3172 child addresses, Exim now panics and dies. Previously, because the count
3173 is held in a short int, deliveries were likely to be lost. As such a
3174 large number of recipients for a single message is ridiculous
3175 (performance will be very, very poor), I have chosen to impose a limit
3176 rather than extend the field.
3182 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3183 leftover -K file (the existence of which was triggered by #402).
3184 While we were at it, introduced process PID as part of the -K
3185 filename. This should rule out race conditions when creating
3188 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3189 processing considerably. Previous code took too long for large mails,
3190 triggering a timeout which in turn triggers #401.
3192 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3193 in the DK code in transports.c. sendfile() is not really portable,
3194 hence the _LINUX specificness.
3196 TF/01 In the add_headers option to the mail command in an Exim filter,
3197 there was a bug that Exim would claim a syntax error in any
3198 header after the first one which had an odd number of characters
3201 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3202 callout verification, Exim cached a "reject" for the entire domain. This
3203 is correct for most verifications, but it is not correct for a recipient
3204 verification with use_sender or use_postmaster set, because in that case
3205 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3206 case of MAIL FROM:<> rejection from other early rejections (e.g.
3207 rejection of HELO). When verifying a recipient using a non-null MAIL
3208 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3209 Whatever the result of the callout, the value of the domain cache is
3210 left unchanged (for any other kind of callout, getting as far as trying
3211 RCPT means that the domain itself is ok).
3213 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3216 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3217 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3218 session, the daemon ignores ECONNECT errors and logs others; it now
3219 ignores EPIPE as well.
3221 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3222 (quoted-printable decoding).
3224 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3225 later the small subsequent patch to fix an introduced bug.
3227 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3229 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3231 PH/08 An error is now given if message_size_limit is specified negative.
3233 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3234 to be given (somewhat) arbitrary names.
3236 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3239 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3240 miscellaneous code fixes
3242 PH/10 Added the log_reject_target ACL modifier to specify where to log
3245 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3246 hostname. This is wrong, because it relates to the incoming message (and
3247 probably the interface on which it is arriving) and not to the outgoing
3248 callout (which could be using a different interface). This has been
3249 changed to use the value of the helo_data option from the smtp transport
3250 instead - this is what is used when a message is actually being sent. If
3251 there is no remote transport (possible with a router that sets up host
3252 addresses), $smtp_active_hostname is used.
3254 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3255 tweaks were necessary in order to get it to work (see also 21 below):
3256 (a) The code assumed that strncpy() returns a negative number on buffer
3257 overflow, which isn't the case. Replaced with Exim's string_format()
3259 (b) There were several signed/unsigned issues. I just did the minimum
3260 hacking in of casts. There is scope for a larger refactoring.
3261 (c) The code used strcasecmp() which is not a standard C function.
3262 Replaced with Exim's strcmpic() function.
3263 (d) The code set only $1; it now sets $auth1 as well.
3264 (e) A simple test gave the error "authentication client didn't specify
3265 service in request". It would seem that Dovecot has changed its
3266 interface. Fortunately there's a specification; I followed it and
3267 changed what the client sends and it appears to be working now.
3269 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3272 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3273 address A is aliased to B and C, where B exists and C does not. Without
3274 -v the output is "A verified" because verification stops after a
3275 successful redirection if more than one address is generated. However,
3276 with -v the child addresses are also verified. Exim was outputting "A
3277 failed to verify" and then showing the successful verification for C,
3278 with its parentage. It now outputs "B failed to verify", showing B's
3279 parentage before showing the successful verification of C.
3281 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3282 look up a TXT record in a specific list after matching in a combined
3285 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3286 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3287 they consult the DNS. I had assumed they would set it the way they
3288 wanted; and indeed my experiments on Linux seem to show that in some
3289 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3290 To be on the safe side, however, I have now made the interface to
3291 host_find_byname() similar to host_find_bydns(), with an argument
3292 containing the DNS resolver options. The host_find_byname() function now
3293 sets these options at its start, just as host_find_bydns() does. The smtp
3294 transport options dns_qualify_single and dns_search_parents are passed to
3295 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3296 of host_find_byname() use the default settings of RES_DEFNAMES
3297 (qualify_single) but not RES_DNSRCH (search_parents).
3299 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3300 spool_read_header() do less string testing, by means of a preliminary
3301 switch on the second character of optional "-foo" lines. (This is
3302 overdue, caused by the large number of possibilities that now exist.
3303 Originally there were few.) While I was there, I also converted the
3304 str(n)cmp tests so they don't re-test the leading "-" and the first
3305 character, in the hope this might squeeze out yet more improvement.
3307 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3308 flag allowing group syntax was set by the header_syntax check but not
3309 turned off, possible causing trouble later; (2) The flag was not being
3310 set at all for the header_verify test, causing "group"-style headers to
3311 be rejected. I have now set it in this case, and also caused header_
3312 verify to ignore an empty address taken from a group. While doing this, I
3313 came across some other cases where the code for allowing group syntax
3314 while scanning a header line wasn't quite right (mostly, not resetting
3315 the flag correctly in the right place). These bugs could have caused
3316 trouble for malformed header lines. I hope it is now all correct.
3318 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3319 with the "reply" argument non-NULL. The code, however (which originally
3320 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3321 but it didn't always do it. This confused somebody who was copying the
3322 code for some other use. I have removed all the tests.
3324 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3325 feature that was used to support insecure browsers during the U.S. crypto
3326 embargo. It requires special client support, and Exim is probably the
3327 only MTA that supported it -- and would never use it because real RSA is
3328 always available. This code has been removed, because it had the bad
3329 effect of slowing Exim down by computing (never used) parameters for the
3330 RSA_EXPORT functionality.
3332 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3333 authenticator to fail if there's a tab character in the incoming data
3334 (there should never be unless someone is messing about, as it's supposed
3335 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3336 if the connection is using TLS or if the remote IP is the same as the
3337 local IP, and the "valid-client-cert option" if a client certificate has
3340 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3341 authenticators. This can be used for authorization after authentication
3342 succeeds. (In the case of plaintext, it servers for both authentication
3345 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3346 if any retry times were supplied.
3348 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3349 connection, where there is no client IP address to check. In this
3350 situation, the verify now always succeeds.
3352 PH/25 Applied John Jetmore's -Mset patch.
3354 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3356 PH/27 In a string expansion for a processed (not raw) header when multiple
3357 headers of the same name were present, leading whitespace was being
3358 removed from all of them, but trailing whitespace was being removed only
3359 from the last one. Now trailing whitespace is removed from each header
3360 before concatenation. Completely empty headers in a concatenation (as
3361 before) are ignored.
3363 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3364 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3366 PH/29 [Removed. This was a change that I later backed out, and forgot to
3367 correct the ChangeLog entry (that I had efficiently created) before
3368 committing the later change.]
3370 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3371 address errors (4xx response to RCPT) over the same connection as other
3372 messages routed to the same hosts. Such deliveries are always "forced",
3373 so retry times are not inspected. This resulted in far too many retries
3374 for the affected addresses. The effect occurred only when there were more
3375 hosts than the hosts_max_try setting in the smtp transport when it had
3376 the 4xx errors. Those hosts that it had tried were not added to the list
3377 of hosts for which the message was waiting, so if all were tried, there
3378 was no problem. Two fixes have been applied:
3380 (i) If there are any address or message errors in an SMTP delivery, none
3381 of the hosts (tried or untried) are now added to the list of hosts
3382 for which the message is waiting, so the message should not be a
3383 candidate for sending over the same connection that was used for a
3384 successful delivery of some other message. This seems entirely
3385 reasonable: after all the message is NOT "waiting for some host".
3386 This is so "obvious" that I'm not sure why it wasn't done
3387 previously. Hope I haven't missed anything, but it can't do any
3388 harm, as the worst effect is to miss an optimization.
3390 (ii) If, despite (i), such a delivery is accidentally attempted, the
3391 routing retry time is respected, so at least it doesn't keep
3392 hammering the server.
3394 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3395 in ${readsocket because some servers need this prod.
3397 PH/32 Added some extra debug output when updating a wait-xxx database.
3399 PH/33 The hint "could be header name not terminated by colon", which has been
3400 given for certain expansion errors for a long time, was not being given
3401 for the ${if def:h_colon_omitted{... case.
3403 PH/34 The spec says: "With one important exception, whenever a domain list is
3404 being scanned, $domain contains the subject domain." There was at least
3405 one case where this was not true.
3407 PH/35 The error "getsockname() failed: connection reset by peer" was being
3408 written to the panic log as well as the main log, but it isn't really
3409 panic-worthy as it just means the connection died rather early on. I have
3410 removed the panic log writing for the ECONNRESET error when getsockname()
3413 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3414 runs only) independently of the message's sender address. This meant
3415 that, if the 4xx error was in fact related to the sender, a different
3416 message to the same recipient with a different sender could confuse
3417 things. In particular, this can happen when sending to a greylisting
3418 server, but other circumstances could also provoke similar problems.
3419 I have changed the default so that the retry time for these errors is now
3420 based a combination of the sender and recipient addresses. This change
3421 can be overridden by setting address_retry_include_sender=false in the
3424 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3425 remote server are returned as part of bounce messages. This was not
3426 happening for LMTP over a pipe (the lmtp transport), but now it is the
3427 same for both kinds of LMTP.
3429 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3430 in header lines that were in fact CNAMEs. This is no longer the case.
3432 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3433 and queue runs started by the daemon processed all messages. This has
3434 been fixed so that -R and -S can now usefully be given with -q<time>.
3436 PH/40 Import PCRE release 6.7 (fixes some bugs).
3438 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3440 PH/42 Give an error if -q is specified more than once.
3442 PH/43 Renamed the variables $interface_address and $interface_port as
3443 $received_ip_address and $received_port, to make it clear that these
3444 values apply to message reception, and not to the outgoing interface when
3445 a message is delivered. (The old names remain recognized, of course.)
3447 PH/44 There was no timeout on the connect() call when using a Unix domain
3448 socket in the ${readsocket expansion. There now is.
3450 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3451 be meaningful with "accept".
3453 SC/01 Eximstats V1.43
3454 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3456 SC/02 Eximstats V1.44
3457 Use a glob alias rather than an array ref in the generated
3458 parser. This improves both readability and performance.
3460 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3461 Collect SpamAssassin and rejection statistics.
3462 Don't display local sender or destination tables unless
3463 there is data to show.
3464 Added average volumes into the top table text output.
3466 SC/04 Eximstats V1.46
3467 Collect data on the number of addresses (recipients)
3468 as well as the number of messages.
3470 SC/05 Eximstats V1.47
3471 Added 'Message too big' to the list of mail rejection
3472 reasons (thanks to Marco Gaiarin).
3474 SC/06 Eximstats V1.48
3475 Mainlog lines which have GMT offsets and are too short to
3476 have a flag are now skipped.
3478 SC/07 Eximstats V1.49 (Alain Williams)
3479 Added the -emptyok flag.
3481 SC/08 Eximstats V1.50
3482 Fixes for obtaining the IP address from reject messages.
3484 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3485 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3486 whitespace changes from 4.64-PH/27
3488 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3491 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3492 are found, allow negative numbers in numeric criteria)
3494 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3496 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3497 to match changes made in 4.64-PH/43
3499 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3501 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3502 "domain" is actually an IP address, and force a failure. This locks out
3503 those revolvers/nameservers that support "A-for-A" lookups, in
3504 contravention of the specifications.
3506 PH/48 When a host name was looked up from an IP address, and the subsequent
3507 forward lookup of the name timed out, the host name was left in
3508 $sender_host_name, contrary to the specification.
3510 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3511 restricted to single-key lookups, Exim was not diagnosing an error if
3512 * or *@ was used with a query-style lookup.
3514 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3516 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3517 long ago, but noone interested enough thought of it. Let's just say that
3518 the "1.1" means that there are some new functions that weren't there at
3519 some point in the past.
3521 PH/51 Error processing for expansion failure of helo_data from an smtp
3522 transport during callout processing was broken.
3524 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3525 tested/used via the -bh/-bhc/-bs options.
3527 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3528 bug, fixed in subsequent PCRE releases).
3530 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3531 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3533 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3539 SC/01 Use a glob alias rather than an array ref in eximstats generated
3540 parser. This improves both readability and performance.
3542 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3543 Don't display local sender or destination tables in eximstats unless
3544 there is data to show.
3545 Added average volumes into the eximstats top table text output.
3547 SC/03 Collect data on the number of addresses (recipients) as well
3548 as the number of messages in eximstats.
3550 TF/01 Correct an error in the documentation for the redirect router. Exim
3551 does (usually) call initgroups() when daemonizing.
3553 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3554 with consistent privilege compared to when running as a daemon.
3556 TF/03 Note in the spec that $authenticated_id is not set for local
3557 submissions from trusted users.
3559 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3560 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3562 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3563 by adding some example configuration directives to the default
3564 configuration file. A little bit of work is required to uncomment the
3565 directives and define how usernames and passwords are checked, but
3566 there is now a framework to start from.
3568 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3569 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3570 without this. I don't know how relevant this is to other LDAP libraries.
3572 PH/02 Add the verb name to the "unknown ACL verb" error.
3574 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3576 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3578 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3579 directory not expanded when it should be if an expanded home directory
3580 was set for the address (which is overridden by the transport).
3582 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3585 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3586 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3587 because it is too late at that time, and has no effect.
3589 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3590 security issue with \' (bugzilla #107). I could not use the
3591 PQescapeStringConn() function, because it needs a PGconn value as one of
3594 PH/08 When testing addresses using -bt, indicate those final addresses that
3595 are duplicates that would not cause an additional delivery. At least one
3596 person was confused, thinking that -bt output corresponded to deliveries.
3597 (Suppressing duplicates isn't a good idea as you lose the information
3598 about possibly different redirections that led to the duplicates.)
3600 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3601 systems where poll() doesn't work, in particular OS X.
3603 PH/10 Added more information to debugging output for retry time not reached.
3605 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3606 operations in malware.c.
3608 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3611 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3612 syslog_duplication set false, log lines that would normally be written
3613 both the the main log and to the reject log were not written to syslog at
3616 PH/14 In the default configuration, change the use of "message" in ACL warn
3617 statements to "add_header".
3619 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3620 not followed by a command (e.g. "seen endif").
3622 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3623 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3626 PH/17 Added extra conditions to the default value of delay_warning_condition
3630 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3631 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3632 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3635 The Auto-Submitted: and various List- headers are standardised, whereas I
3636 don't think Precedence: ever was.
3638 PH/18 Refactored debugging code in route_finduser() to show more information,
3639 in particular, the error code if getpwnam() issues one.
3641 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3642 This is apparently needed in addition to the PH/07 change above to avoid
3643 any possible encoding problems.
3645 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3646 but not after initializing Perl.
3648 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3649 output them only if debugging. By default they are written stderr,
3650 apparently, which is not desirable.
3652 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3655 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3658 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3660 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3661 authenticated or an ident call has been made. Suppress the default
3662 values for $authenticated_id and $authenticated_sender (but permit -oMai
3663 and -oMas) when testing with -bh.
3665 PH/24 Re-jigged the order of the tests in the default configuration so that the
3666 tests for valid domains and recipients precede the DNS black list and CSA
3667 tests, on the grounds that those ones are more expensive.
3669 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3670 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3671 This bug exists in every version of Exim that I still have, right back to
3674 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3675 However, an attempt to turn on case-sensitivity in a regex key by
3676 including (?-i) didn't work because the subject string was already
3677 lowercased, and the effects were non-intuitive. It turns out that a
3678 one-line patch can be used to allow (?-i) to work as expected.
3684 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3685 other effects) broke the use of negated acl sub-conditions.
3687 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3690 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3691 "Deny" causes Exim to reject the incoming connection with a 554 error.
3692 Unfortunately, if there is a major crisis, such as a disk failure,
3693 tcp-wrappers gives "deny", whereas what one would like would be some
3694 kind of temporary error. A kludge has been added to help with this.
3695 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3696 554 error is used if errno is still zero or contains ENOENT (which occurs
3697 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3700 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3702 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3703 errors. Otherwise a message that provokes a temporary error (when other
3704 messages do not) can cause a whole host to time out.
3706 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3707 addresses were routed directly to files or pipes from a redirect router.
3708 File deliveries just didn't batch; pipe deliveries might have suffered
3711 PH/06 A failure to get a lock for a hints database would erroneously always say
3712 "Failed to get write lock", even when it was really a read lock.
3714 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3715 of 0600. This has been changed to use the value of the lockfile_mode
3716 option (which defaults to 0600).
3718 PH/08 Applied small patch from the Sieve maintainer.
3720 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3721 folder from quota calculations, a direct delivery into this folder messed
3722 up the contents of the maildirsize file. This was because the regex was
3723 used only to exclude .Trash (or whatever) when the size of the mailbox
3724 was calculated. There was no check that a delivery was happening into an
3725 excluded directory. This bug has been fixed by ignoring all quota
3726 processing for deliveries into excluded directories.
3728 PH/10 Added the maildirfolder_create_regex option to appendfile.
3734 PH/01 The code for finding all the local interface addresses on a FreeBSD
3735 system running IPv6 was broken. This may well have applied to all BSD
3736 systems, as well as to others that have similar system calls. The broken
3737 code found IPv4 interfaces correctly, but gave incorrect values for the
3738 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3739 that it would not match correctly against @[] and not recognize the IPv6
3742 PH/02 The ipliteral router was not recognizing addresses of the form user@
3743 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3745 PH/03 Added disable_ipv6.
3747 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3748 decoded form, because it is most often used to construct To: headers
3749 lines in autoreplies, and the decoded form may well be syntactically
3750 invalid. However, $reply_address has leading white space removed, and all
3751 newlines turned into spaces so that the autoreply transport does not
3754 PH/05 If group was specified without a user on a router, and no group or user
3755 was specified on a transport, the group from the router was ignored.
3757 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3758 for visible compile-time settings that can be used to change these
3759 numbers, for those that want even more. Backwards compatibility with old
3760 spool files has been maintained. However, going back to a previous Exim
3761 release will lost any variables that are in spool files.
3763 PH/07 Two small changes when running in the test harness: increase delay when
3764 passing a TCP/IP connection to a new process, in case the original
3765 process has to generate a bounce, and remove special handling of
3766 127.0.0.2 (sic), which is no longer necessary.
3768 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3769 be the same on different OS.
3771 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3774 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3775 whether --show-vars was specified or not
3777 JJ/02 exipick: Added support for new ACL variable spool format introduced
3780 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3781 syntactically invalid From: or Reply-to: line, and a filter used this to
3782 generate an autoreply, and therefore failed to obtain an address for the
3783 autoreply, Exim could try to deliver to a non-existent relative file
3784 name, causing unrelated and misleading errors. What now happens is that
3785 it logs this as a hard delivery error, but does not attempt to create a
3788 PH/11 The exinext utility has a -C option for testing purposes, but although
3789 the given file was scanned by exinext itself; it wasn't being passed on
3790 when Exim was called.
3792 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3793 an end-of-file indication when reading a command response.
3795 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3796 compiled. In many other places in Exim, IPv6 addresses are always
3797 recognized, so I have changed this. It also means that IPv4 domain
3798 literals of the form [IPV4:n.n.n.n] are now always recognized.
3800 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3801 used if the router is not running as root, for example, when verifying at
3802 ACL time, or when using -bh. The debugging output from this situation was
3803 non-existent - all you got was a failure to exec. I have made two
3806 (a) Failures to set uid/gid, the current directory, or a process leader
3807 in a subprocess such as that created by queryprogram now generate
3808 suitable debugging output when -d is set.
3810 (b) The queryprogram router detects when it is not running as root,
3811 outputs suitable debugging information if -d is set, and then runs
3812 the subprocess without attempting to change uid/gid.
3814 PH/15 Minor change to Makefile for building test_host (undocumented testing
3817 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3818 additional section of a DNS packet that returns MX or SRV records.
3819 Instead, it always explicitly searches for A/AAAA records. This avoids
3820 major problems that occur when a DNS server includes only records of one
3821 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3822 fixed another bug: if SRV records were looked up and the corresponding
3823 address records were *not* found in the additional section, the port
3824 values from the SRV records were lost.
3826 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3827 using the correct key (the original address) when searching the retry
3828 rules in order to find which one to use for generating the retry hint.
3830 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3831 adding the default one. Similarly, if it contains a Reply-To: header, the
3832 errors_reply_to option, if set, is not used.
3834 PH/19 When calculating a retry time, Exim used to measure the "time since
3835 failure" by looking at the "first failed" field in the retry record. Now
3836 it does not use this if it is later than than the arrival time of the
3837 message. Instead it uses the arrival time. This makes for better
3838 behaviour in cases where some deliveries succeed, thus re-setting the
3839 "first failed" field. An example is a quota failure for a huge message
3840 when small messages continue to be delivered. Without this change, the
3841 "time since failure" will always be short, possible causing more frequent
3842 delivery attempts for the huge message than are intended.
3843 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3845 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3846 $1, $2, $3) because the numerical variables can be reset during some
3847 expansion items (e.g. "match"), thereby losing the authentication data.
3849 PH/21 Make -bV show the size of off_t variables so that the test suite can
3850 decide whether to run tests for quotas > 2G.
3852 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3853 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3854 filecount value is greater than 2G or if a quota value is greater than 2G
3855 on a system where the size of off_t is not greater than 4, a panic error
3858 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3859 never match. The debug and -bh output now contains an explicit error
3860 message indicating a malformed IPv4 address or mask.
3862 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3863 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3864 PH/23 above applies.
3866 PH/25 Do not write to syslog when running in the test harness. The only
3867 occasion when this arises is a failure to open the main or panic logs
3868 (for which there is an explicit test).
3870 PH/26 Added the /no_tell option to "control=freeze".
3872 PH/27 If a host name lookup failed very early in a connection, for example, if
3873 the IP address matched host_lookup and the reverse lookup yielded a name
3874 that did not have a forward lookup, an error message of the form "no IP
3875 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3876 could be logged. Now it outputs the IP address instead of "NULL".
3878 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3879 allows the sender and the authenticated sender to be set when
3880 submitting a message from within Exim. Since child_open_exim() is
3881 documented for local_scan(), the new function should be too.
3883 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3884 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3885 results in an empty string is now treated as unset.
3887 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3889 PH/31 Added sender_verify_fail logging option.
3891 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3892 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3893 tidied the source and removed it altogether.
3895 PH/33 When a queue run was abandoned because the load average was too high, a
3896 log line was always written; now it is written only if the queue_run log
3897 selector is set. In addition, the log line for abandonment now contains
3898 information about the queue run such as the pid. This is always present
3899 in "start" and "stop" lines but was omitted from the "abandon" line.
3901 PH/34 Omit spaces between a header name and the colon in the error message that
3902 is given when verify = headers_syntax fails (if there are lots of them,
3903 the message gets confusing).
3905 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3906 names, as there are now some PTR records that contain slashes. This check
3907 is only to protect against broken name servers that fall over on strange
3908 characters, so the fact that it applies to all lookups doesn't matter.
3910 PH/36 Now that the new test suite is complete, we can remove some of the
3911 special code in Exim that was needed for the old test suite. For example,
3912 sorting DNS records because real resolvers return them in an arbitrary
3913 order. The new test suite's fake resolver always returns records in the
3916 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3917 bounces) except when queue_only is set, to avoid logging races between
3918 the different processes.
3920 PH/38 Panic-die if .include specifies a non-absolute path.
3922 PH/39 A tweak to the "H" retry rule from its user.
3924 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3925 a label. They prevented compilation on older perls.
3927 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3928 a warning to be raised on newish perls.
3930 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3931 on queue. Changes to match documented behaviour of showing count of
3932 messages matching specified criteria.
3934 PH/40 Changed the default ident timeout from 30s to 5s.
3936 PH/41 Added support for the use of login_cap features, on those BSD systems
3937 that have them, for controlling the resources used by pipe deliveries.
3939 PH/42 The content-scanning code uses fopen() to create files in which to put
3940 message data. Previously it was not paying any attention to the mode of
3941 the files. Exim runs with umask(0) because the rest of the code creates
3942 files with open(), and sets the required mode explicitly. Thus, these
3943 files were ending up world-writeable. This was not a big issue, because,
3944 being within the spool directory, they were not world-accessible. I have
3945 created a function called modefopen, which takes an additional mode
3946 argument. It sets umask(777), creates the file, chmods it to the required
3947 mode, then resets the umask. All the relevant calls to fopen() in the
3948 content scanning code have been changed to use this function.
3950 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3951 to 24 hours. This avoids potential overflow problems when processing G
3952 and H retry rules. I suspect nobody ever tinkers with this value.
3954 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3956 PH/45 When the plaintext authenticator is running as a client, the server's
3957 challenges are checked to ensure they are valid base64 strings. By
3958 default, the authentication attempt is cancelled if an invalid string is
3959 received. Setting client_ignore_invalid_base64 true ignores these errors.
3960 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3961 they are received. Thus, the responses can be made to depend on the
3962 challenges. If an invalid string is ignored, an empty string is placed in
3965 PH/46 Messages that are created by the autoreply transport now contains a
3966 References: header, in accordance with RFCs 2822 and 3834.
3968 PH/47 Added authenticated_sender_force to the smtp transport.
3970 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3972 PH/49 Installed latest patch from the Sieve maintainer.
3974 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3975 was also set, the appendfile transport was unnecessarily scanning a
3976 directory of message files (e.g. for maildir delivery) to find the count
3977 of files (along with the size), even though it did not need this
3978 information. It now does the scan only if it needs to find either the
3979 size of the count of files.
3981 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3983 PH/52 Two bugs concerned with error handling when the smtp transport is
3986 (i) Exim was not creating retry information for temporary errors given
3987 for individual recipients after the DATA command when the smtp transport
3988 was used in LMTP mode. This meant that they could be retried too
3989 frequently, and not timed out correctly.
3991 (ii) Exim was setting the flag that allows error details to be returned
3992 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3993 recipients that were returned after the DATA command.
3995 PH/53 This is related to PH/52, but is more general: for any failing address,
3996 when detailed error information was permitted to be returned to the
3997 sender, but the error was temporary, then after the final timeout, only
3998 "retry timeout exceeded" was returned. Now it returns the full error as
3999 well as "retry timeout exceeded".
4001 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4002 do this, and (what is worse) MTAs that accept it.
4004 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4005 will now be deprecated.
4007 PH/56 New os.c-cygwin from the Cygwin maintainer.
4009 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4010 formats (previously only available in exim formats via -bpr, -bpru,
4011 and -bpra. Now also available in native and exiqgrep formats)
4013 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4014 with very large, slow to parse queues
4016 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4018 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4020 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4021 responses to authentication challenges, though it was showing the
4022 challenges; (ii) I've removed the CR characters from the debug output for
4025 PH/58 Allow for the insertion of a newline as well as a space when a string
4026 is turned into more than one encoded-word during RFC 2047 encoding. The
4027 Sieve code now uses this.
4029 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4030 data_4xx, lost_connection, tls_required.
4032 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4033 message was being sent as an SMTP response.
4035 PH/61 Add -l and -k options to exicyclog.
4037 PH/62 When verifying, if an address was redirected to one new address, so that
4038 verification continued, and the new address failed or deferred after
4039 having set something in $address_data, the value of $address_data was not
4040 passed back to the ACL. This was different to the case when no
4041 redirection occurred. The value is now passed back in both cases.
4043 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4044 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4045 use login_cap.h, so on its own it isn't the distinguishing feature. The
4046 new name refers directly to the setclassresources() function.
4048 PH/65 Added configuration files for NetBSD3.
4050 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4052 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4053 is preferred over IPv4.
4055 PH/68 The bounce_return_message and bounce_return_body options were not being
4056 honoured for bounces generated during the reception of non-SMTP messages.
4057 In particular, this applied to messages rejected by the ACL. This bug has
4058 been fixed. However, if bounce_return_message is true and bounce_return_
4059 body is false, the headers that are returned for a non-SMTP message
4060 include only those that have been read before the error was detected.
4061 (In the case of an ACL rejection, they have all been read.)
4063 PH/69 The HTML version of the specification is now built in a directory called
4064 spec_html instead of spec.html, because the latter looks like a path with
4065 a MIME-type, and this confuses some software.
4067 PH/70 Catch two compiler warnings in sieve.c.
4069 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4070 function verify_get_ident() calls ip_connect() to connect a socket, but
4071 if the "connect()" function timed out, ip_connect() used to close the
4072 socket. However, verify_get_ident() also closes the socket later, and in
4073 between Exim writes to the log, which may get opened at this point. When
4074 the socket was closed in ip_connect(), the log could get the same file
4075 descriptor number as the socket. This naturally causes chaos. The fix is
4076 not to close the socket in ip_connect(); the socket should be closed by
4077 the function that creates it. There was only one place in the code where
4078 this was missing, in the iplookup router, which I don't think anybody now
4079 uses, but I've fixed it anyway.
4081 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4082 well as to direct DNS lookups. Otherwise the handling of names in host
4083 lists is inconsistent and therefore confusing.
4089 PH/01 Two changes to the default runtime configuration:
4091 (1) Move the checks for relay_from_hosts and authenticated clients from
4092 after to before the (commented out) DNS black list checks.
4094 (2) Add control=submission to the relay_from_hosts and authenticated
4095 clients checks, on the grounds that messages accepted by these
4096 statements are most likely to be submissions.
4098 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4100 (1) Generate an error if the third argument for the ${prvs expansion is
4103 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4106 (3) Reset the variables that are obtained from the first argument of
4107 ${prvscheck and used in the second argument before leaving the code,
4108 because their memory is reclaimed, so using them afterwards may do
4111 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4112 one (it's much easier than Tom thought :-).
4114 (5) Because of (4), we can now allow for the use of $prvscheck_result
4115 inside the third argument.
4117 PH/03 For some reason, the default setting of PATH when running a command from
4118 a pipe transport was just "/usr/bin". I have changed it to
4121 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4122 anything to be listed in the output from -bV.
4124 PH/05 When a filter generated an autoreply, the entire To: header line was
4125 quoted in the delivery log line, like this:
4127 => >A.N.Other <ano@some.domain> <original@ddress> ...
4129 This has been changed so that it extracts the operative address. There
4130 may be more than one such address. If so, they are comma-separated, like
4133 => >ano@some.domain,ona@other.domain <original@ddress> ...
4135 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4136 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4137 not being looked up in the rDNS to get a host name, Exim was showing the
4138 IP address twice in Received: lines, even though the IP addresses were
4139 identical. For example:
4141 Received: from [1.2.3.4] (helo=[1.2.3.4])
4143 However, if the real host name was known, it was omitting the HELO data
4144 if it matched the actual IP address. This has been tidied up so that it
4145 doesn't show the same IP address twice.
4147 PH/07 When both +timestamp and +memory debugging was on, the value given by
4148 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4149 called by the debug printing, thereby overwriting the timestamp buffer.
4150 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4152 PH/08 When the original message was included in an autoreply transport, it
4153 always said "this is a copy of the message, including all the headers",
4154 even if body_only or headers_only was set. It now gives an appropriate
4157 PH/09 Applied a patch from the Sieve maintainer which:
4159 o fixes some comments
4160 o adds the (disabled) notify extension core
4161 o adds some debug output for the result of if/elsif tests
4162 o points to the current vacation draft in the documentation
4163 and documents the missing references header update
4167 o fixes a bug in processing the envelope test (when testing
4168 multiple envelope elements, the last element determined the
4171 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4172 Electronic Mail") by including:
4174 Auto-submitted: auto-generated
4176 in the messages that it generates (bounce messages and others, such as
4177 warnings). In the case of bounce messages for non-SMTP messages, there was
4178 also a typo: it was using "Auto_submitted" (underscore instead of
4179 hyphen). Since every message generated by Exim is necessarily in response
4180 to another message, thes have all been changed to:
4182 Auto-Submitted: auto-replied
4184 in accordance with these statements in the RFC:
4186 The auto-replied keyword:
4188 - SHOULD be used on messages sent in direct response to another
4189 message by an automatic process,
4191 - MUST NOT be used on manually-generated messages,
4193 - MAY be used on Delivery Status Notifications (DSNs) and Message
4194 Disposition Notifications (MDNs),
4196 - MUST NOT be used on messages generated by automatic or periodic
4197 processes, except for messages which are automatic responses to
4200 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4201 to the default Received: header definition.
4203 PH/12 Added log selector acl_warn_skipped (default on).
4205 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4206 variables because (a) they are in the wrong storage pool and (b) even if
4207 they were copied, it wouldn't work properly because of the caching.
4209 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4210 checking when decoding. Apparently there are clients that generate
4211 overlong encoded strings. Why am I not surprised?
4213 PH/15 If the first argument of "${if match_address" was not empty, but did not
4214 contain an "@" character, Exim crashed. Now it writes a panic log message
4215 and treats the condition as false.
4217 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4219 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4220 extension "envelope-auth". The code is finished and in agreement with
4221 other implementations, but there is no documentation so far and in fact,
4222 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4223 not changing the active code.
4225 Print executed "if" and "elsif" statements when debugging is used. This
4226 helps a great deal to understand what a filter does.
4228 Document more things not specified clearly in RFC3028. I had all this
4229 sorted out, when out of a sudden new issues came to my mind. Oops."
4231 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4234 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4235 canonical form (as documented). However, after a host name lookup from
4236 the IP address, check_host() was doing a simple string comparison with
4237 addresses acquired from the DNS when checking that the found name did
4238 have the original IP as one of its addresses. Since any found IPv6
4239 addresses are likely to be in abbreviated form, the comparison could
4240 fail. Luckily, there already exists a function for doing the comparison
4241 by converting both addresses to binary, so now that is used instead of
4242 the text comparison.
4244 PH/20 There was another similar case to PH/19, when a complete host name was
4245 given in a host list; looking up its IP address could give an abbreviated
4246 form, whereas the current host's name might or might not be abbreviated.
4247 The same fix has been applied.
4253 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4254 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4257 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4258 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4260 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4262 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4263 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4264 Also, for "Auto-Submitted", treat anything other than "no" as
4265 non-personal, in accordance with RFC 3834. (Previously it treated
4266 anything starting "auto-" as non-personal.)
4268 TF/01 The control=submission/name=... option had a problem with syntax
4269 errors if the name included a slash character. The /name= option
4270 now slurps the rest of the string, so it can include any characters
4271 but it must come last in the list of options (after /sender_retain
4274 PH/05 Some modifications to the interface to the fake nameserver for the new
4282 TK/01 Added the "success_on_redirect" address verification option. See
4283 NewStuff for rationale and an example.
4285 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4287 PH/02 Patch to exigrep to allow it to work on syslog lines.
4289 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4290 fread() to skip over the body file's header line, because in Cygwin the
4291 header line is locked and is inaccessible.
4293 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4294 co-exist for some time) to make it clear that it is the Exim ID that is
4295 referenced, not the Message-ID: header line.
4297 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4298 string_format() function, because snprintf() does not exist on all
4301 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4302 ${stat: expansion item.
4304 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4305 protocol synchronization error", to keep the pedants happy.
4307 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4308 well as for IRIX systems, when gcc is being used. See the host.c source
4311 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4313 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4316 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4317 transport and to the smtp transport in LMTP mode.
4319 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4321 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4322 run a filter in a subprocess. This could lead to confusion in subsequent
4323 lookups in the parent process. There should also be a search_tidyup() at
4324 the end of the subprocess.
4326 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4327 only if the host matched helo_try_verify_hosts, which caused the
4328 verification to occur when the EHLO/HELO command was issued. The ACL just
4329 tested the remembered result. Now, if a previous verification attempt has
4330 not happened, "verify = helo" does it there and then.
4332 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4334 TK/03 Fix log output including CR from clamd.
4336 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4337 exist provoked a memory error which could cause a segfault.
4339 PH/15 Installed PCRE 6.2
4341 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4343 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4344 of the problem. Specifically, suggested +O2 rather than +O1 for the
4347 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4349 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4350 record, the port was not correctly logged when the outgoing_port log
4351 selector was set (it logged the transort's default port).
4353 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4354 fallback_hosts, and "hosts" in the smtp transport.
4356 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4357 host errors such as "Connection refused".
4359 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4360 authentication with radiusclient 0.4.9:
4362 - Error returned from rc_read_config was caught wrongly
4363 - Username/password not passed on to radius server due to wrong length.
4365 The presumption is that some radiusclient API changes for 4.51/PH/17
4366 were not taken care of correctly. The code is still untested by me (my
4367 Linux distribution still has 0.3.2 of radiusclient), but it was
4368 contributed by a Radius user.
4370 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4371 expanding the "port" option of the smtp transport.
4373 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4374 while reading a MIME header. Thanks to Tom Hughes for a patch.
4376 PH/24 Include config.h inside local_scan.h so that configuration settings are
4379 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4380 that in an ACL for RCPT (for example), you can examine exactly what was
4383 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4384 commands, but it was not correctly comparing the address with the actual
4385 client host address. Thus, it would show the EHLO address in Received:
4386 header lines when this was not necessary.
4388 PH/27 Added the % operator to ${eval:}.
4390 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4391 it should be ignoring failures (because with -C, for example, it has lost
4392 privilege). It wasn't ignoring creation failures other than "already
4395 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4398 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4399 because an input line was too long, either on its own, or by virtue of
4400 too many continuations, the temporary file was not being removed, and the
4401 return code was incorrect.
4403 PH/31 Missing "BOOL" in function definition in filtertest.c.
4405 PH/32 Applied Sieve patches from the maintainer.
4407 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4409 PH/33 Added "verify = not_blind".
4411 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4412 Local/Makefile (with some defaults set). These are used in built scripts
4413 such as exicyclog, but they have never been used in the exim_install
4414 script (though there are many overriding facilities there). I have
4415 arranged that the exim_install script now takes note of these two
4418 PH/35 Installed configuration files for Dragonfly.
4420 PH/36 When a locally submitted message by a trusted user did not contain a
4421 From: header, and the sender address was obtained from -f or from an SMTP
4422 MAIL command, and the trusted user did not use -F to supply a sender
4423 name, $originator_name was incorrectly used when constructing a From:
4424 header. Furthermore, $originator_name was used for submission mode
4425 messages from external hosts without From: headers in a similar way,
4426 which is clearly wrong.
4428 PH/37 Added control=suppress_local_fixups.
4430 PH/38 When log_selector = +received_sender was set, and the addition of the
4431 sender made the log line's construction buffer exactly full, or one byte
4432 less than full, an overflow happened when the terminating "\n" was
4435 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4436 when the result of a list match is failure because a DNS lookup failed.
4438 PH/40 RM_COMMAND is now used in the building process.
4440 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4441 the "build-* directories that it finds.
4443 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4444 address in a domain literal was a prefix of an interface address.
4446 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4447 when verifying a sender address, unless rewrite_headers is false.
4449 PH/44 Wrote a long comment about why errors_to addresses are verified as
4450 recipients, not senders.
4452 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4453 the ratelimit ACL was added.
4455 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4457 PH/46 Added extra information about PostgreSQL errors to the error string.
4459 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4460 avoiding the need to install special zones in a real server. This is
4461 backwards compatible; if it can't find the fake resolver, it drops back.
4462 Thus, both old and new test suites can be run.
4464 TF/02 Added util/ratelimit.pl
4466 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4467 clock is set back in time.
4469 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4470 Candler <B.Candler@pobox.com>.
4472 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4473 OK for addresses that are the result of redirections.
4475 PH/48 A number of further additions for the benefit of the new test suite,
4476 including a fake gethostbyname() that interfaces to the fake DNS resolver
4479 TF/06 The fix for widen_domains has also been applied to qualify_single and
4480 search_parents which are the other dnslookup options that can cause
4483 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4486 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4488 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4489 service" if it encountered an MX record with an empty target hostname.
4490 The message is now "an MX or SRV record indicated no SMTP service".
4492 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4493 if the DNS of the sending site is misconfigured. This is quite a
4494 common situation. This change restores the behaviour of treating a
4495 helo verification defer as a failure.
4497 PH/51 If self=fail was set on a router, the bounce message did not include the
4498 actual error message.
4504 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4506 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4507 command itself ended in error, the underlying message about the transport
4508 filter timeout was being overwritten with the pipe command error. Now the
4509 underlying error message should be appended to the second error message.
4511 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4513 PH/02 Reduce the amount of output that "make" produces by default. Full output
4514 can still be requested.
4516 PH/03 The warning log line about a condition test deferring for a "warn" verb
4517 was being output only once per connection, rather than after each
4518 occurrence (because it was using the same function as for successful
4519 "warn" verbs). This seems wrong, so I have changed it.
4521 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4522 it should not have, which might have caused a crash in the right
4523 circumstances, but probably never did.
4525 PH/04 Installed a modified version of Tony Finch's patch to make submission
4526 mode fix the return path as well as the Sender: header line, and to
4527 add a /name= option so that you can make the user's friendly name appear
4530 TF/03 Added the control = fakedefer ACL modifier.
4532 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4533 Mark Lowes for thorough testing.
4535 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4537 TK/03 Merged latest SRS patch from Miles Wilton.
4539 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4540 with the definition in sysexits.h (which is #included earlier).
4541 Fortunately, Exim does not actually use EX_OK. The code used to try to
4542 preserve the sysexits.h value, by assuming that macro definitions were
4543 scanned for macro replacements. I have been disabused of this notion,
4544 so now the code just undefines EX_OK before #including unistd.h.
4546 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4547 in the smtp transport. When a block could not be written in a single
4548 write() function, the timeout was being re-applied to each part-write.
4549 This seems wrong - if the receiver was accepting one byte at a time it
4550 would take for ever. The timeout is now adjusted when this happens. It
4551 doesn't have to be particularly precise.
4553 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4554 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4556 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4557 without a domain if the check to <postmaster@domain> fails.
4559 SC/01 Eximstats: added -xls and the ability to specify output files
4560 (patch written by Frank Heydlauf).
4562 SC/02 Eximstats: use FileHandles for outputting results.
4564 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4566 SC/04 Eximstats: fixed display of large numbers with -nvr option
4568 SC/05 Eximstats: fixed merging of reports with empty tables.
4570 SC/06 Eximstats: added the -include_original_destination flag
4572 SC/07 Eximstats: removed tabs and trailing whitespace.
4574 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4576 TK/06 MBOX spool code: Add real "From " MBOX separator line
4577 so the .eml file is really in mbox format (even though
4578 most programs do not really care). Patch from Alex Miller.
4580 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4581 The latter is generated from $received_to and is only set if the
4582 message has one envelope recipient. SA can use these headers,
4583 obviously out-of-the-box. Patch from Alex Miller.
4585 PH/08 The ${def test on a variable was returning false if the variable's
4586 value was "0", contrary to what the specification has always said!
4587 The result should be true unless the variable is empty.
4589 PH/09 The syntax error of a character other than { following "${if
4590 def:variable_name" (after optional whitespace) was not being diagnosed.
4591 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4592 accidental colon was present, for example, could give incorrect results.
4594 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4595 result is used (not including appendfile, where other changes are about
4598 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4599 This involved changing a lot of size variables from int to off_t. It
4600 should work with maildirs and everything.
4602 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4603 spamd dying while we are connected to it.
4605 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4608 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4609 function for BDB 4.3.
4611 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4613 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4614 the sources for PCRE have been re-arranged and more files are now
4617 PH/15 The code I had for printing potentially long long variables in PH/11
4618 above was not the best (it lost precision). The length of off_t variables
4619 is now inspected at build time, and an appropriate printing format (%ld
4620 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4621 to be "long long int" or "long int". This is needed for the internal
4622 formatting function string_vformat().
4624 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4625 the configuration file to be ":syslog", then the script "guesses" where
4626 the logs files are, rather than using the compiled in default. In our
4627 case the guess is not the same as the compiled default, so the script
4628 suddenly stopped working when I started to use syslog. The patch checks
4629 to see if log_file_path is "". If so, it attempts to read it from exim
4630 with no configuration file to get the compiled in version, before it
4631 falls back to the previous guessing code."
4633 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4634 implementing BATV in an Exim configuration. See NewStuff for the gory
4637 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4638 Makefile that are specific to HP-UX.
4640 PH/18 If the "use_postmaster" option was set for a recipient callout together
4641 with the "random" option, the postmaster address was used as the MAIL
4642 FROM address for the random test, but not for the subsequent recipient
4643 test. It is now used for both.
4645 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4646 patch removes a few documentation additions to RFC 3028, because the
4647 latest draft now contains them. It adds the new en;ascii-case comparator
4648 and a new error check for 8bit text in MIME parts. Comparator and
4649 require names are now matched exactly. I enabled the subaddress
4650 extension, but it is not well tested yet (read: it works for me)."
4652 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4653 rework some of the code of TK/09 above to avoid the hardwired use of
4654 "%lld" and "long long". Replaced the call to snprintf() with a call to
4657 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4658 records point to non-existent hosts", "retry timeout exceeded", and
4659 "retry time not reached for any host after a long failure period".
4661 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4662 experimental DomainKeys support:
4664 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4665 (2) On an error such as an illegally used "control", the wrong name for
4666 the control was given.
4668 These problems did NOT occur unless DomainKeys support was compiled.
4670 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4672 PH/24 Added ${if match_ip condition.
4674 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4675 databases so that it will be absolutely obvious if a crash occurs in the
4676 DB library. This is a regular occurrence (often caused by mis-matched
4679 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4680 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4681 picked up on a user's system that detects such things. There doesn't seem
4682 to be a gcc warning option for this - only an attribute that has to be
4683 put on the function's prototype. It seems that in Fedora Core 4 they have
4684 set this on a number of new functions. No doubt there will be more in due
4687 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4688 specify a transport. However, if an address that was verified by such a
4689 router was the subject of a callout, Exim crashed because it tried to
4690 read the rcpt_include_affixes from the non-existent transport. Now it
4691 just assumes that the setting of that option is false. This bug was
4692 introduced by 4.51/PH/31.
4694 PH/28 Changed -d+all to exclude +memory, because that information is very
4695 rarely of interest, but it makes the output a lot bigger. People tend to
4696 do -d+all out of habit.
4698 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4699 code in os-type was giving problems when libc.so lives in lib64, like on
4702 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4703 aren't the modern standard, and it seems that some systems' include files
4704 don't always have them. Exim was already checking for some of the newer
4705 ones like T_AAAA, and defining it itself. I've added checks for all the
4706 record types that Exim uses.
4708 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4709 not automatically generating a new one, as it is supposed to. This
4710 prevented TLS from working. If the file did exist, but contained invalid
4711 data, a new version was generated, as expected. It was only the case of a
4712 non-existent file that was broken.
4714 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4715 with a change in libdomainkeys > 0.64.
4717 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4718 from DNS. If the selector record carries the flag, it now has
4719 precedence over the domain-wide flag.
4721 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4723 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4724 the use of a port name that isn't defined in /etc/services) occurred, the
4725 message was deferred as in a normal delivery, and thus remained on the
4726 spool, instead of being failed because of the mua_wrapper setting. This
4727 is now fixed, and I tidied up some of the mua_wrapper messages at the
4730 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4731 the messages in a hash of arrays rather than using individual hashes.
4732 This is a bit cleaner and results in dramatic memory savings, albeit
4733 at a slight CPU cost.
4735 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4736 as requested by Marc Sherman.
4738 SC/10 Eximstats: added histograms for user specified patterns as requested
4741 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4743 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4744 fopen() in the content-scanning modules that did not already have it.
4750 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4751 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4753 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4755 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4757 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4758 to test Sieve filters that use "vacation".
4760 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4761 that changes the way the GnuTLS parameters are stored in the cache file.
4762 The new format can be generated externally. For backward compatibility,
4763 if the data in the cache doesn't make sense, Exim assumes it has read an
4764 old-format file, and it generates new data and writes a new file. This
4765 means that you can't go back to an older release without removing the
4768 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4769 work if there are any delivery delays because "one_time" forces the
4770 parent to be marked "delivered", so its unseen clone is never tried
4771 again. For this reason, Exim now forbids the simultaneous setting of
4774 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4775 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4776 entry if you want to know the details. The fix, however, neglected to
4777 consider the case when local delivery batching is involved. The test for
4778 "previously delivered" was not happening when checking to see if an
4779 address could be batched with a previous (undelivered) one; under
4780 certain circumstances this could lead to multiple deliveries to the same
4783 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4784 in its include files, and this causes problems building Exim.
4786 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4787 header_syntax) but Exim was just ignoring anything given after a slash.
4788 In particular, this caused confusion with an attempt to use "verify =
4789 reverse_host_lookup/defer_ok". An error is now given when options are
4790 supplied for verify items that do not have them. (Maybe reverse_host_
4791 lookup should have a defer_ok option, but that's a different point.)
4793 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4794 defined by RFC 821) to 2048, because there were problems with some AUTH
4795 commands, and RFC 1869 says the size should be increased for extended
4796 SMTP commands that take arguments.
4798 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4801 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4802 "unknown" error; now it says that the functionality isn't in the binary.
4804 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4805 an address' error message when a string expansion fails (syntax or
4806 whatever). Otherwise the password may appear in the log. Following change
4807 PH/42 below, there is no longer a chance of it appearing in a bounce
4810 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4812 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4813 of its addresses was ever tried. (Bugzilla bug #2.)
4815 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4816 the result incorrectly in the debug output. (It correctly added a newline
4817 to what was transported.)
4819 TF/01 Added $received_time.
4821 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4822 commented out examples of how to interface to a virus scanner and to
4823 SpamAssassin. Also added commented examples of av_scanner and
4824 spamd_address settings.
4826 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4827 and controls are allowed in which ACLs. There were a couple of minor
4828 errors. Some of the entries in the conditions table (which is a table of
4829 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4830 negation of where the condition IS allowed.
4832 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4834 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4835 header file does not have a version number, so I've had to invent a new
4836 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4837 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4838 radiusclient), but it was contributed by a Radius user.
4840 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4841 files or directories, for OpenSSL.
4843 PH/19 When an Exim process that is running as root has to create an Exim log
4844 file, it does so in a subprocess that runs as exim:exim so as to get the
4845 ownership right at creation (otherwise, other Exim processes might see
4846 the file with the wrong ownership). There was no test for failure of this
4847 fork() call, which would lead to the process getting stuck as it waited
4848 for a non-existent subprocess. Forks do occasionally fail when resources
4849 run out. I reviewed all the other calls to fork(); they all seem to check
4852 PH/20 When checking for unexpected SMTP input at connect time (before writing
4853 the banner), Exim was not dealing correctly with a non-positive return
4854 from the read() function. If the client had disconnected by this time,
4855 the result was a log entry for a synchronization error with an empty
4856 string after "input=" when read() returned zero. If read() returned -1
4857 (an event I could not check), uninitialized data bytes were printed.
4858 There were reports of junk text (parts of files, etc) appearing after
4861 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4863 PH/22 Added support for macro redefinition, and (re)definition in between
4864 driver and ACL definitions.
4866 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4867 forgetting to use the resulting value; it was using the unexpanded value.
4869 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4870 hadn't been configured. The fix is from Juergen Kreileder, who
4871 understands it better than I do:
4873 "Here's what I see happening with three configured cyrus_sasl
4874 authenticators configured (plain, login, cram-md5):
4876 On startup auth_cyrus_sasl_init() gets called for each of these.
4877 This means three calls to sasl_listmech() without a specified mech_list.
4878 => SASL tests which mechs of all available mechs actually work
4879 => three warnings about OTP not working
4880 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4882 With the patch, sasl_listmech() also gets called three times. But now
4883 SASL's mech_list option is set to the server_mech specified in the the
4884 authenticator. Or in other words, the answer from sasl_listmech()
4885 gets limited to just the mech you're testing for (which is different
4887 => the return list contains just 'plain' or 'login', 'cram-md5' or
4888 nothing depending on the value of ob->server_mech.
4890 I've just tested the patch: Authentication still works fine,
4891 unavailable mechs specified in the exim configuration are still
4892 caught, and the auth.log warnings about OTP are gone."
4894 PH/25 When debugging is enabled, the contents of the command line are added
4895 to the debugging output, even when log_selector=+arguments is not
4898 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4899 answer is "GNU", and only if the return is "GNU/something" is the answer
4902 PH/27 $acl_verify_message is now set immediately after the failure of a
4903 verification in an ACL, and so is available in subsequent modifiers. In
4904 particular, the message can be preserved by coding like this:
4906 warn !verify = sender
4907 set acl_m0 = $acl_verify_message
4909 Previously, $acl_verify_message was set only while expanding "message"
4910 and "log_message" when a very denied access.
4912 PH/28 Modified OS/os.c-Linux with
4914 -#ifndef OS_LOAD_AVERAGE
4915 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4917 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4918 nomenclature these days.)
4920 PH/29 Installed patch from the Sieve maintainer that adds the options
4921 sieve_useraddress and sieve_subaddress to the redirect router.
4923 PH/30 In these circumstances:
4924 . Two addresses routed to the same list of hosts;
4925 . First host does not offer TLS;
4926 . First host accepts first address;
4927 . First host gives temporary error to second address;
4928 . Second host offers TLS and a TLS session is established;
4929 . Second host accepts second address.
4930 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4931 and peerdn, if requested) that were in fact used only for the second
4934 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4935 attention to any local part prefix or suffix that was matched by the
4936 router that accepted the address. It now behaves in the same way as it
4937 does for delivery: the affixes are removed from the local part unless
4938 rcpt_include_affixes is set on the transport.
4940 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4941 timeout during the DATA phase of an incoming message.
4943 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4944 applied a patch sanctioned by the Sieve maintainer.
4946 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4947 the uid or gid is negative. A case of a negative gid caused this to be
4948 noticed. The fix allows for either to be negative.
4950 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4951 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4954 PH/36 The change PH/12 above was broken. Fixed it.
4956 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4957 the grounds that routing the same address twice would always produce the
4958 same answer. This might have been true once, but it is certainly no
4959 longer true now. Routing a child address may depend on the previous
4960 routing that produced that child. Some complicated redirection strategies
4961 went wrong when messages had multiple recipients, and made Exim's
4962 behaviour dependent on the order in which the addresses were given.
4964 I have moved the duplicate checking until after the routing is complete.
4965 Exim scans the addresses that are assigned to local and remote
4966 transports, and removes any duplicates. This means that more work will be
4967 done, as duplicates will always all be routed, but duplicates are
4968 presumably rare, so I don't expect this is of any significance.
4970 For deliveries to pipes, files, and autoreplies, the duplicate checking
4971 still happens during the routing process, since they are not going to be
4974 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4975 It corrects a timeout issue with spamd. This is Ian's comment: "The
4976 background is that sometimes spamd either never reads data from a
4977 connection it has accepted, or it never writes response data. The exiscan
4978 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4979 blindly assumes that writes won't block so it may never time out."
4981 PH/39 Allow G after quota size as well as K and M.
4983 PH/40 The value set for $authenticated_id in an authenticator may not contain
4984 binary zeroes or newlines because the value is written to log lines and
4985 to spool files. There was no check on this. Now the value is run through
4986 the string_printing() function so that such characters are converted to
4987 printable escape sequences.
4989 PH/41 $message_linecount is a new variable that contains the total number of
4990 lines in the message. Compare $body_linecount, which is the count for the
4993 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4994 bounce and delay warning messages, except in certain special cases, which
4997 (a) An SMTP error message from a remote host;
4998 (b) A message specified in a :fail: redirection;
4999 (c) A message specified in a "fail" command in a system filter;
5000 (d) A message specified in a FAIL return from the queryprogram router;
5001 (e) A message specified by the cannot_route_message router option.
5003 In these cases only, Exim does include the error details in bounce and
5004 warning messages. There are also a few cases where bland messages such
5005 as "unrouteable address" or "local delivery error" are given.
5007 PH/43 $value is now also set for the "else" part of a ${run expansion.
5009 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5010 being worked on, but at least Exim now implements the latest version to
5013 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5014 process to complete was treated as a delivery failure, a timeout while
5015 writing the message to the pipe was logged, but erroneously treated as a
5016 successful delivery. Such timeouts include transport filter timeouts. For
5017 consistency with the overall process timeout, these timeouts are now
5018 treated as errors, giving rise to delivery failures by default. However,
5019 there is now a new Boolean option for the pipe transport called
5020 timeout_defer, which, if set TRUE, converts the failures into defers for
5021 both kinds of timeout. A transport filter timeout is now identified in
5024 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5025 systems where "make" and "gmake" are different, calling "gmake" at top
5026 level broke things. I've arranged for the value of $(MAKE) to be passed
5027 from the Makefile to this script so that it can call the same version of
5031 A note about Exim versions 4.44 and 4.50
5032 ----------------------------------------
5034 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5035 changes of various kinds. As a consequence, a big documentation update was
5036 needed. This delayed the release for rather longer than seemed good, especially
5037 in the light of a couple of (minor) security issues. Therefore, the changes
5038 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5039 release. So 4.44 and 4.50 are in effect two different branches that both start
5042 I have left the 4.50 change log unchanged; it contains all the changes since
5043 4.43. The change log for 4.44 is below; many of its items are identical to
5044 those for 4.50. This seems to be the most sensible way to preserve the
5045 historical information.
5051 1. Minor wording change to the doc/README.SIEVE file.
5053 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5054 computation of the current number of files was incorrect.
5056 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5057 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5060 4. Give more explanation in the error message when the command for a transport
5061 filter fails to execute.
5063 5. There are several places where Exim runs a non-Exim command in a
5064 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5065 was being done only for the command run by the queryprogram router. It is
5066 now done for all such subprocesses. The other cases are: ${run, transport
5067 filters, and the commands run by the lmtp and pipe transports.
5069 6. Added CONFIGURE_GROUP build-time option.
5071 7. Some older OS have a limit of 256 on the maximum number of file
5072 descriptors. Exim was using setrlimit() to set 1000 as a large value
5073 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5074 systems. I've change it so that if it can't get 1000, it tries for 256.
5076 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5077 was an oversight, and furthermore, ever since the addition of extra
5078 controls (e.g. 4.43/32), the checks on when to allow different forms of
5079 "control" were broken. There should now be diagnostics for all cases when a
5080 control that does not make sense is encountered.
5082 9. Added the /retain_sender option to "control=submission".
5084 10. $recipients is now available in the predata ACL (oversight).
5086 11. Tidy the search cache before the fork to do a delivery from a message
5087 received from the command line. Otherwise the child will trigger a lookup
5088 failure and thereby defer the delivery if it tries to use (for example) a
5089 cached ldap connection that the parent has called unbind on.
5091 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5092 of $address_data from the recipient verification was clobbered by the
5093 sender verification.
5095 13. The value of address_data from a sender verification is now available in
5096 $sender_address_data in subsequent conditions in the ACL statement.
5098 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5100 15. Added a new option "connect=<time>" to callout options, to set a different
5103 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5104 was its contents. (It was OK if the option was not defined at all.)
5106 17. A "Completed" log line is now written for messages that are removed from
5107 the spool by the -Mrm option.
5109 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5110 information about exactly what failed.
5112 19. Added -dd to debug only the daemon process.
5114 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5115 handles timeouts, both on the server side and network timeouts. Renamed the
5116 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5118 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5119 It is now set to "smtps".
5121 22. $host_address is now set to the target address during the checking of
5122 ignore_target_hosts.
5124 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5125 being passed; this would have caused $sender_host_name to have been used if
5126 matching the list had actually called for a host name (not very likely,
5127 since this list is usually IP addresses). A host name is now passed as
5130 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5131 code that specifies a non-restarting handler (typically sigaction() in
5132 modern systems) in an attempt to fix a rare and obscure crash bug.
5134 25. Narrowed the window for a race in the daemon that could cause it to ignore
5135 SIGCHLD signals. This is not a major problem, because they are used only to
5136 wake it up if nothing else does.
5138 26. A malformed maildirsize file could cause Exim to calculate negative values
5139 for the mailbox size or file count. Odd effects could occur as a result.
5140 The maildirsize information is now recalculated if the size or filecount
5143 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5144 support for a long time. Removed HAVE_SYS_VFS_H.
5146 28. Installed the latest version of exipick from John Jetmore.
5148 29. In an address list, if the pattern was not a regular expression, an empty
5149 subject address (from a bounce message) matched only if the pattern was an
5150 empty string. Non-empty patterns were not even tested. This was the wrong
5151 because it is perfectly reasonable to use an empty address as part of a
5152 database query. An empty address is now tested by patterns that are
5153 lookups. However, all the other forms of pattern expect the subject to
5154 contain a local part and a domain, and therefore, for them, an empty
5155 address still always fails if the pattern is not itself empty.
5157 30. Exim went into a mad DNS loop when attempting to do a callout where the
5158 host was specified on an smtp transport, and looking it up yielded more
5159 than one IP address.
5161 31. Re-factored the code for checking spool and log partition space into a
5162 function that finds that data and another that does the check. The former
5163 is then used to implement four new variables: $spool_space, $log_space,
5164 $spool_inodes, and $log_inodes.
5166 32. The RFC2047 encoding function was originally intended for short strings
5167 such as real names; it was not keeping to the 75-character limit for
5168 encoded words that the RFC imposes. It now respects the limit, and
5169 generates multiple encoded words if necessary. To be on the safe side, I
5170 have increased the buffer size for the ${rfc2047: expansion operator from
5173 33. It is now permitted to omit both strings after an "if" condition; if the
5174 condition is true, the result is "true". As before, when the second string
5175 is omitted, a false condition yields an empty string. This makes it less
5176 cumbersome to write custom ACL and router conditions.
5178 34. Failure to deliver a bounce message always caused it to be frozen, even if
5179 there was an errors_to setting on the router. The errors_to setting is now
5182 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5183 canonical form (fully expanded) before being placed in
5184 $sender_host_address.
5186 36. The table in the code that translates DNS record types into text (T_A to
5187 "A" for instance) was missing entries for NS and CNAME. It is just possible
5188 that this could have caused confusion if both these types were looked up
5189 for the same domain, because the text type is used as part of Exim's
5190 per-process caching. But the chance of anyone hitting this buglet seems
5193 37. The dnsdb lookup has been extended in a number of ways.
5195 (1) There is a new type, "zns", which walks up the domain tree until it
5196 finds some nameserver records. It should be used with care.
5198 (2) There is a new type, "mxh", which is like "mx" except that it returns
5199 just the host names, not the priorities.
5201 (3) It is now possible to give a list of domains (or IP addresses) to be
5202 looked up. The behaviour when one of the lookups defers can be
5203 controlled by a keyword.
5205 (4) It is now possible to specify the separator character for use when
5206 multiple records are returned.
5208 38. The dnslists ACL condition has been extended: it is now possible to supply
5209 a list of IP addresses and/or domains to be looked up in a particular DNS
5212 39. Added log_selector=+queue_time_overall.
5214 40. When running the queue in the test harness, wait just a tad after forking a
5215 delivery process, to get repeatability of debugging output.
5217 41. Include certificate and key file names in error message when GnuTLS fails
5218 to set them up, because the GnuTLS error message doesn't include the name
5219 of the failing file when there is a problem reading it.
5221 42. Allow both -bf and -bF in the same test run.
5223 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5225 44. The "Exiscan patch" is now merged into the mainline Exim source.
5227 45. Sometimes the final signoff response after QUIT could fail to get
5228 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5229 < 0 before doing a fflush(). This bug looks as though it goes back to the
5230 introduction of TLS in release 3.20, but "sometimes" must have been rare
5231 because the tests only now provoked it.
5233 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5234 (this can affect the format of dates).
5236 47. exim_tidydb, when checking for the continued existence of a message for
5237 which it has found a message-specific retry record, was not finding
5238 messages that were in split spool directories. Consequently, it was
5239 deleting retry records that should have stayed in existence.
5241 48. Steve fixed some bugs in eximstats.
5243 49. The SPA authentication driver was not abandoning authentication and moving
5244 on to the next authenticator when an expansion was forced to fail,
5245 contradicting the general specification for all authenticators. Instead it
5246 was generating a temporary error. It now behaves as specified.
5248 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5249 (the order specifies the preference for clients). The order is now AES256,
5250 AES128, 3DES, ARCFOUR128.
5252 51. Small patch to Sieve code - explicitly set From: when generating an
5255 52. Exim crashed if a remote delivery caused a very long error message to be
5256 recorded - for instance if somebody sent an entire SpamAssassin report back
5257 as a large number of 550 error lines. This bug was coincidentally fixed by
5258 increasing the size of one of Exim's internal buffers (big_buffer) that
5259 happened as part of the Exiscan merge. However, to be on the safe side, I
5260 have made the code more robust (and fixed the comments that describe what
5263 53. Now that there can be additional text after "Completed" in log lines (if
5264 the queue_time_overall log selector is set), a one-byte patch to exigrep
5265 was needed to allow it to recognize "Completed" as not the last thing in
5268 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5269 patch that reportedly fixes this has been added. I am not expert enough to
5270 create a test for it. This is what the patch creator wrote:
5272 "I found a little strange behaviour of ldap code when working with
5273 Windows 2003 AD Domain, where users was placed in more than one
5274 Organization Units. When I tried to give exim partial DN, the exit code
5275 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5276 But simultaneously result of request was absolutely normal ldap result,
5277 so I produce this patch..."
5279 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5280 REFERENCE, so I have modified the code to exclude the patch when that macro
5283 55. Some experimental protocols are using DNS PTR records for new purposes. The
5284 keys for these records are domain names, not reversed IP addresses. The
5285 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5286 leaves it alone. Component reversal etc. now happens only for IP addresses.
5289 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5291 57. Double the size of the debug message buffer (to 2048) so that more of very
5292 long debug lines gets shown.
5294 58. The exicyclog utility now does better if the number of log files to keep
5295 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5297 59. Two changes related to the smtp_active_hostname option:
5299 (1) $smtp_active_hostname is now available as a variable.
5300 (2) The default for smtp_banner uses $smtp_active_hostname instead
5301 of $primary_hostname.
5303 60. The host_aton() function is supposed to be passed a string that is known
5304 to be a valid IP address. However, in the case of IPv6 addresses, it was
5305 not checking this. This is a hostage to fortune. Exim now panics and dies
5306 if the condition is not met. A case was found where this could be provoked
5307 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5308 components; fortuitously, this particular loophole had already been fixed
5309 by change 4.50/55 above.
5311 If there are any other similar loopholes, the new check in host_aton()
5312 itself should stop them being exploited. The report I received stated that
5313 data on the command line could provoke the exploit when Exim was running as
5314 exim, but did not say which command line option was involved. All I could
5315 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5316 running as the user.
5319 61. There was a buffer overflow vulnerability in the SPA authentication code
5320 (which came originally from the Samba project). I have added a test to the
5321 spa_base64_to_bits() function which I hope fixes it.
5324 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5325 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5327 63. The daemon start-up calls getloadavg() while still root for those OS that
5328 need the first call to be done as root, but it missed one case: when
5329 deliver_queue_load_max is set with deliver_drop_privilege. This is
5330 necessary for the benefit of the queue runner, because there is no re-exec
5331 when deliver_drop_privilege is set.
5333 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5334 This has been fixed.
5336 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5337 was in use, was not putting the data itself into the right store pool;
5338 consequently, it could be overwritten for a subsequent message in the same
5339 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5342 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5344 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5345 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5346 the function were treating the return as a boolean value, which happened to
5347 work because 0=false and not-0=true, but is not correct code.
5349 68. The host_aton() function was not handling scoped IPv6 addresses (those
5350 with, for example, "%eth0" on the end) correctly.
5352 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5353 negated items (that is, ~something) in unsigned ints. Some compilers
5354 apparently mutter when there is no cast.
5356 70. If an address verification called from an ACL failed, and did not produce a
5357 user-specific message (i.e. there was only a "system" message), nothing was
5358 put in $acl_verify_message. In this situation, it now puts the system
5361 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5362 session; change 4.31/43 added the unwanted input to the log line - except
5363 that it did not do this in the start of session case. It now does.
5365 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5366 This is wrong and can cause the other end to generate a synchronization
5367 error if it is another Exim or anything else that does the synchronization
5368 check. A QUIT command is no longer sent after a timeout.
5370 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5371 during host lookups.
5373 74. The defer_ok option of callout verification was not working if it was used
5374 when verifying addresses in header lines, that is, for this case:
5376 verify = header_sender/callout=defer_ok
5378 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5379 those file descriptors could be used for SMTP connections. If anything
5380 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5381 could be sent to the SMTP client, causing chaos. The daemon now opens
5382 stdin, stdout, and stderr to /dev/null when it puts itself into the
5385 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5386 log by default. The user can override this with suitable Perl magic.
5388 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5389 the log message when discard triggers, was not working for the DATA ACL or
5390 for the non-SMTP ACL.
5392 78. Error message wording change in sieve.c.
5394 79. If smtp_accept_max_per_host was set, the number of connections could be
5395 restricted to fewer than expected, because the daemon was trying to set up
5396 a new connection before checking whether the processes handling previous
5397 connections had finished. The check for completed processes is now done
5398 earlier. On busy systems, this bug wouldn't be noticed because something
5399 else would have woken the daemon, and it would have reaped the completed
5402 80. If a message was submitted locally by a user whose login name contained one
5403 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5404 It caused a spool format error. I have fixed the spool reading code. A
5405 related problem was that the "from" clause in the Received: line became
5406 illegal because of the space(s). It is now covered by ${quote_local_part.
5408 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5411 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5413 83. Patch from Sieve maintainer to fix unterminated string problem in
5414 "vacation" handling.
5416 84. Some minor changes to the Linux configuration files to help with other
5417 OS variants using glibc.
5419 85. One more patch for Sieve to update vacation handling to latest spec.
5422 ----------------------------------------------------
5423 See the note above about the 4.44 and 4.50 releases.
5424 ----------------------------------------------------
5430 1. Change 4.43/35 introduced a bug that caused file counts to be
5431 incorrectly computed when quota_filecount was set in an appendfile
5434 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5435 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5438 3. Give more explanation in the error message when the command for a transport
5439 filter fails to execute.
5441 4. There are several places where Exim runs a non-Exim command in a
5442 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5443 was being done only for the command run by the queryprogram router. It is
5444 now done for all such subprocesses. The other cases are: ${run, transport
5445 filters, and the commands run by the lmtp and pipe transports.
5447 5. Some older OS have a limit of 256 on the maximum number of file
5448 descriptors. Exim was using setrlimit() to set 1000 as a large value
5449 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5450 systems. I've change it so that if it can't get 1000, it tries for 256.
5452 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5453 was an oversight, and furthermore, ever since the addition of extra
5454 controls (e.g. 4.43/32), the checks on when to allow different forms of
5455 "control" were broken. There should now be diagnostics for all cases when a
5456 control that does not make sense is encountered.
5458 7. $recipients is now available in the predata ACL (oversight).
5460 8. Tidy the search cache before the fork to do a delivery from a message
5461 received from the command line. Otherwise the child will trigger a lookup
5462 failure and thereby defer the delivery if it tries to use (for example) a
5463 cached ldap connection that the parent has called unbind on.
5465 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5466 of $address_data from the recipient verification was clobbered by the
5467 sender verification.
5469 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5470 was its contents. (It was OK if the option was not defined at all.)
5472 11. A "Completed" log line is now written for messages that are removed from
5473 the spool by the -Mrm option.
5475 12. $host_address is now set to the target address during the checking of
5476 ignore_target_hosts.
5478 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5479 being passed; this would have caused $sender_host_name to have been used if
5480 matching the list had actually called for a host name (not very likely,
5481 since this list is usually IP addresses). A host name is now passed as
5484 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5485 code that specifies a non-restarting handler (typically sigaction() in
5486 modern systems) in an attempt to fix a rare and obscure crash bug.
5488 15. Narrowed the window for a race in the daemon that could cause it to ignore
5489 SIGCHLD signals. This is not a major problem, because they are used only to
5490 wake it up if nothing else does.
5492 16. A malformed maildirsize file could cause Exim to calculate negative values
5493 for the mailbox size or file count. Odd effects could occur as a result.
5494 The maildirsize information is now recalculated if the size or filecount
5497 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5498 support for a long time. Removed HAVE_SYS_VFS_H.
5500 18. Updated exipick to current release from John Jetmore.
5502 19. Allow an empty sender to be matched against a lookup in an address list.
5503 Previously the only cases considered were a regular expression, or an
5506 20. Exim went into a mad DNS lookup loop when doing a callout where the
5507 host was specified on the transport, if the DNS lookup yielded more than
5510 21. The RFC2047 encoding function was originally intended for short strings
5511 such as real names; it was not keeping to the 75-character limit for
5512 encoded words that the RFC imposes. It now respects the limit, and
5513 generates multiple encoded words if necessary. To be on the safe side, I
5514 have increased the buffer size for the ${rfc2047: expansion operator from
5517 22. Failure to deliver a bounce message always caused it to be frozen, even if
5518 there was an errors_to setting on the router. The errors_to setting is now
5521 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5522 canonical form (fully expanded) before being placed in
5523 $sender_host_address.
5525 24. Updated eximstats to version 1.33
5527 25. Include certificate and key file names in error message when GnuTLS fails
5528 to set them up, because the GnuTLS error message doesn't include the name
5529 of the failing file when there is a problem reading it.
5531 26. Expand error message when OpenSSL has problems setting up cert/key files.
5534 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5535 (this can affect the format of dates).
5537 28. exim_tidydb, when checking for the continued existence of a message for
5538 which it has found a message-specific retry record, was not finding
5539 messages that were in split spool directories. Consequently, it was
5540 deleting retry records that should have stayed in existence.
5542 29. eximstats updated to version 1.35
5543 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5544 1.35 - bugfix such that pie charts by volume are generated correctly
5546 30. The SPA authentication driver was not abandoning authentication and moving
5547 on to the next authenticator when an expansion was forced to fail,
5548 contradicting the general specification for all authenticators. Instead it
5549 was generating a temporary error. It now behaves as specified.
5551 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5552 (the order specifies the preference for clients). The order is now AES256,
5553 AES128, 3DES, ARCFOUR128.
5555 31. Small patch to Sieve code - explicitly set From: when generating an
5558 32. Exim crashed if a remote delivery caused a very long error message to be
5559 recorded - for instance if somebody sent an entire SpamAssassin report back
5560 as a large number of 550 error lines. This bug was coincidentally fixed by
5561 increasing the size of one of Exim's internal buffers (big_buffer) that
5562 happened as part of the Exiscan merge. However, to be on the safe side, I
5563 have made the code more robust (and fixed the comments that describe what
5566 33. Some experimental protocols are using DNS PTR records for new purposes. The
5567 keys for these records are domain names, not reversed IP addresses. The
5568 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5569 leaves it alone. Component reversal etc. now happens only for IP addresses.
5572 34. The host_aton() function is supposed to be passed a string that is known
5573 to be a valid IP address. However, in the case of IPv6 addresses, it was
5574 not checking this. This is a hostage to fortune. Exim now panics and dies
5575 if the condition is not met. A case was found where this could be provoked
5576 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5577 components; fortuitously, this particular loophole had already been fixed
5578 by change 4.50/55 or 4.44/33 above.
5580 If there are any other similar loopholes, the new check in host_aton()
5581 itself should stop them being exploited. The report I received stated that
5582 data on the command line could provoke the exploit when Exim was running as
5583 exim, but did not say which command line option was involved. All I could
5584 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5585 running as the user.
5588 35. There was a buffer overflow vulnerability in the SPA authentication code
5589 (which came originally from the Samba project). I have added a test to the
5590 spa_base64_to_bits() function which I hope fixes it.
5593 36. The daemon start-up calls getloadavg() while still root for those OS that
5594 need the first call to be done as root, but it missed one case: when
5595 deliver_queue_load_max is set with deliver_drop_privilege. This is
5596 necessary for the benefit of the queue runner, because there is no re-exec
5597 when deliver_drop_privilege is set.
5599 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5600 was in use, was not putting the data itself into the right store pool;
5601 consequently, it could be overwritten for a subsequent message in the same
5602 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5605 38. Sometimes the final signoff response after QUIT could fail to get
5606 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5607 < 0 before doing a fflush(). This bug looks as though it goes back to the
5608 introduction of TLS in release 3.20, but "sometimes" must have been rare
5609 because the tests only now provoked it.
5615 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5616 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5617 Now it returns an integer. A number of calls were still expecting a T/F
5618 return. Fortuitously, in all cases, the tests worked in OK situations,
5619 which is the norm. However, things would have gone wrong on any write
5620 failures on the smtp file descriptor. This function is used when sending
5621 messages over SMTP and also when doing verify callouts.
5623 2. When Exim is called to do synchronous delivery of a locally submitted
5624 message (the -odf or -odi options), it no longer closes stderr before doing
5627 3. Implemented the mua_wrapper option.
5629 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5631 5. Implemented the functions header_remove(), header_testname(),
5632 header_add_at_position(), and receive_remove_recipient(), and exported them
5635 6. If an ACL "warn" statement specified the addition of headers, Exim already
5636 inserted X-ACL-Warn: at the start if there was no header name. However, it
5637 was not making this test for the second and subsequent header lines if
5638 there were newlines in the string. This meant that an invalid header could
5639 be inserted if Exim was badly configured.
5641 7. Allow an ACL "warn" statement to add header lines at the start or after all
5642 the Received: headers, as well as at the end.
5644 8. Added the rcpt_4xx retry error code.
5646 9. Added postmaster_mailfrom=xxx to callout verification option.
5648 10. Added mailfrom=xxxx to the callout verification option, for verify=
5651 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5652 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5654 12. Inserted some casts to stop certain compilers warning when using pointer
5655 differences as field lengths or precisions in printf-type calls (mostly
5656 affecting debugging statements).
5658 13. Added optional readline() support for -be (dynamically loaded).
5660 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5661 same clock tick as a message's arrival, so that its received time was the
5662 same as the "first fail" time on the retry record, and that message
5663 remained on the queue past the ultimate address timeout, every queue runner
5664 would try a delivery (because it was past the ultimate address timeout) but
5665 after another failure, the ultimate address timeout, which should have then
5666 bounced the address, did not kick in. This was a "< instead of <=" error;
5667 in most cases the first failure would have been in the next clock tick
5668 after the received time, and all would be well.
5670 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5671 being recognized when the domain list was tested by the match_domain
5672 condition in an expansion string.
5674 16. Added the ${str2b64: operator.
5676 17. Exim was always calling setrlimit() to set a large limit for the number of
5677 processes, without checking whether the existing limit was already
5678 adequate. (It did check for the limit on file descriptors.) Furthermore,
5679 errors from getrlimit() and setrlimit() were being ignored. Now they are
5680 logged to the main and panic logs, but Exim does carry on, to try to do its
5681 job under whatever limits there are.
5683 18. Imported PCRE 5.0.
5685 19. Trivial typo in log message " temporarily refused connection" (the leading
5688 20. If the log selector return_path_on_delivery was set and an address was
5689 redirected to /dev/null, the delivery process crashed because it assumed
5690 that a return path would always be set for a "successful" delivery. In this
5691 case, the whole delivery is bypassed as an optimization, and therefore no
5694 21. Internal re-arrangement: the function for sending a challenge and reading
5695 a response while authentication was assuming a zero-terminated challenge
5696 string. It's now changed to take a pointer and a length, to allow for
5697 binary data in such strings.
5699 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5701 23. Exim was not respecting finduser_retries when seeking the login of the
5702 uid under which it was called; it was always trying 10 times. (The default
5703 setting of finduser_retries is zero.) Also, it was sleeping after the final
5704 failure, which is pointless.
5706 24. Implemented tls_on_connect_ports.
5708 25. Implemented acl_smtp_predata.
5710 26. If the domain in control=submission is set empty, Exim assumes that the
5711 authenticated id is a complete email address when it generates From: or
5712 Sender: header lines.
5714 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5715 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5716 chown and chgrp in /bin and hostname in /usr/bin.
5718 28. Exim was keeping the "process log" file open after each use, just as it
5719 does for the main log. This opens the possibility of it remaining open for
5720 long periods when the USR1 signal hits a daemon. Occasional processlog
5721 errors were reported, that could have been caused by this. Anyway, it seems
5722 much more sensible not to leave this file open at all, so that is what now
5725 29. The long-running daemon process does not normally write to the log once it
5726 has entered its main loop, and it closes the log before doing so. This is
5727 so that log files can straightforwardly be renamed and moved. However,
5728 there are a couple of unusual error situations where the daemon does write
5729 log entries, and I had neglected to close the log afterwards.
5731 30. The text of an SMTP error response that was received during a remote
5732 delivery was being truncated at 512 bytes. This is too short for some of
5733 the long messages that one sometimes sees. I've increased the limit to
5736 31. It is now possible to make retry rules that apply only when a message has a
5737 specific sender, in particular, an empty sender.
5739 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5740 it possible to be selective about when SMTP synchronization is enforced.
5742 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5744 32. Implemented hosts_connection_nolog.
5746 33. Added an ACL for QUIT.
5748 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5751 35. Added mailbox_size and mailbox_filecount to appendfile.
5753 36. Added control = no_multiline_responses to ACLs.
5755 37. There was a bug in the logic of the code that waits for the clock to tick
5756 in the case where the clock went backwards by a substantial amount such
5757 that the microsecond fraction of "now" was more than the microsecond
5758 fraction of "then" (but the whole seconds number was less).
5760 38. Added support for the libradius Radius client library this is found on
5761 FreeBSD (previously only the radiusclient library was supported).
5767 1. When certain lookups returned multiple values in the form name=value, the
5768 quoting of the values was not always being done properly. Specifically:
5769 (a) If the value started with a double quote, but contained no whitespace,
5771 (b) If the value contained whitespace other than a space character (i.e.
5772 tabs or newlines or carriage returns) it was not quoted.
5773 This fix has been applied to the mysql and pgsql lookups by writing a
5774 separate quoting function and calling it from the lookup code. The fix
5775 should probably also be applied to nisplus, ibase and oracle lookups, but
5776 since I cannot test any of those, I have not disturbed their existing code.
5778 2. A hit in the callout cache for a specific address caused a log line with no
5779 reason for rejecting RCPT. Now it says "Previous (cached) callout
5780 verification failure".
5782 3. There was an off-by-one bug in the queryprogram router. An over-long
5783 return line was truncated at 256 instead of 255 characters, thereby
5784 overflowing its buffer with the terminating zero. As well as fixing this, I
5785 have increased the buffer size to 1024 (and made a note to document this).
5787 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5788 when Exim is waiting for an SMTP response from a remote server, Exim
5789 restarts its select() call on the socket, thereby resetting its timeout.
5790 This is not a problem when such interrupts are rare. Somebody set up a cron
5791 job to run exiwhat every 2 minutes, which is less than the normal select()
5792 timeout (5 or 10 minutes). This meant that the select() timeout never
5793 kicked in because it was always reset. I have fixed this by comparing the
5794 time when an interrupt arrives with the time at the start of the first call
5795 to select(). If more time than the timeout has elapsed, the interrupt is
5796 treated as a timeout.
5798 5. Some internal re-factoring in preparation for the addition of Sieve
5799 extensions (by MH). In particular, the "personal" test is moved to a
5800 separate function, and given an option for scanning Cc: and Bcc: (which is
5801 not set for Exim filters).
5803 6. When Exim created an email address using the login of the caller as the
5804 local part (e.g. when creating a From: or Sender: header line), it was not
5805 quoting the local part when it contained special characters such as @.
5807 7. Installed new OpenBSD configuration files.
5809 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5810 try to make them clearer.
5812 9. Callout options, other than the timeout value, were being ignored when
5813 verifying sender addresses in header lines. For example, when using
5815 verify = header_sender/callout=no_cache
5817 the cache was (incorrectly) being used.
5819 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5820 only the Cygwin environment.
5822 11. When return_path_on_delivery was set as a log selector, if different remote
5823 addresses in the same message used different return paths and parallel
5824 remote delivery occurred, the wrong values would sometimes be logged.
5825 (Whenever a remote delivery process finished, the return path value from
5826 the most recently started remote delivery process was logged.)
5828 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5829 lines when AUTH and/or TLS are in use. This is the "received protocol"
5830 field. Exim used to use "asmtp" for authenticated SMTP, without any
5831 indication (in the protocol name) for TLS use. Now it follows the RFC and
5832 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5833 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5834 names appear in log lines as well as in Received: header lines.
5836 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5837 extensions, and comparison tests, and to fix some bugs.
5839 14. Changes to the "personal" filter test:
5841 (1) The test was buggy in that it was just doing the equivalent of
5842 "contains" tests on header lines. For example, if a user's address was
5843 anne@some.where, the "personal" test would incorrectly be true for
5845 To: susanne@some.where
5847 This test is now done by extracting each address from the header in turn,
5848 and checking the entire address. Other tests that are part of "personal"
5849 are now done using regular expressions (for example, to check local parts
5850 of addresses in From: header lines).
5852 (2) The list of non-personal local parts in From: addresses has been
5853 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5854 taken from the Sieve specification recommendations.
5856 (3) If the message contains any header line starting with "List-" it is
5857 treated as non-personal.
5859 (4) The test for "circular" in the Subject: header line has been removed
5860 because it now seems ill-conceived.
5862 15. Minor typos in src/EDITME comments corrected.
5864 16. Installed latest exipick from John Jetmore.
5866 17. If headers_add on a router specified a text string that was too long for
5867 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5868 of string_sprintf() is now avoided.
5870 18. $message_body_size was not set (it was always zero) when running the DATA
5871 ACL and the local_scan() function.
5873 19. For the "mail" command in an Exim filter, no default was being set for
5874 the once_repeat time, causing a random time value to be used if "once" was
5875 specified. (If the value happened to be <= 0, no repeat happened.) The
5876 default is now 0s, meaning "never repeat". The "vacation" command was OK
5877 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5878 (I found it when inspecting the code).
5880 20. There is now an overall timeout for performing a callout verification. It
5881 defaults to 4 times the callout timeout, which applies to individual SMTP
5882 commands during the callout. The overall timeout applies when there is more
5883 than one host that can be tried. The timeout is checked before trying the
5884 next host. This prevents very long delays if there are a large number of
5885 hosts and all are timing out (e.g. when the network connections are timing
5886 out). The value of the overall timeout can be changed by specifying an
5887 additional sub-option for "callout", called "maxwait". For example:
5889 verify = sender/callout=5s,maxwait=20s
5891 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5892 to the end before writing, but this should make it even safer).
5894 22. Exim was forgetting that it had advertised PIPELINING for the second and
5895 subsequent messages on an SMTP connection. It was also not resetting its
5896 memory on STARTTLS and an internal HELO.
5898 23. When Exim logs an SMTP synchronization error within a session, it now
5899 records whether PIPELINING has been advertised or not.
5901 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5902 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5903 rather than long int.
5905 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5907 26. Added the never_mail option to autoreply.
5913 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5914 crash if the getsockname() call failed; this can happen if a connection is
5915 closed very soon after it is established. The problem was simply in the
5916 order in which certain operations were done, causing Exim to try to write
5917 to the SMTP stream before it had set up the file descriptor. The bug has
5918 been fixed by making things happen in the correct order.
5924 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5925 before the connection was closed, thus losing the rejection response.
5927 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5928 some early Solaris releases, but causes trouble in current releases where
5929 socklen_t is defined.
5931 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5934 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5937 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5938 that releasing the top of it at the end releases what was used for sub-
5939 expansions (unless the block got too big). However, discard this block if
5940 the first thing is a variable or header, so that we can use its block when
5941 it is dynamic (useful for very large $message_headers, for example).
5943 6. Lookups now cache *every* query, not just the most recent. A new, separate
5944 store pool is used for this. It can be recovered when all lookup caches are
5945 flushed. Lookups now release memory at the end of their result strings.
5946 This has involved some general refactoring of the lookup sources.
5948 7. Some code has been added to the store_xxx() functions to reduce the amount
5949 of flapping under certain conditions.
5951 8. log_incoming_interface used to affect only the <= reception log lines. Now
5952 it causes the local interface and port to be added to several more SMTP log
5953 lines, for example "SMTP connection from", and rejection lines.
5955 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5957 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5959 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5960 could be overwritten at the end of the current message (or the start of a
5961 new message if it was set in a HELO ACL). The value is now preserved for
5962 the duration of the SMTP connection.
5964 12. If a transport had a headers_rewrite setting, and a matching header line
5965 contained an unqualified address, that address was qualified, even if it
5966 did not match any rewriting rules. The underlying bug was that the values
5967 of the flags that permit the existence of unqualified sender and recipient
5968 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5969 non-local messages, and by -bnq for local messages) were not being
5970 preserved with the message after it was received.
5972 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5973 "next input=" as part of the text comprising the host identity instead of
5974 the correct text. The code was using the same buffer for two different
5975 strings. However, depending on which order the printing function evaluated
5976 its arguments, the bug did not always show up. Under Linux, for example, my
5977 test suite worked just fine.
5979 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5980 doesn't work with some older versions of Perl. It has been changed to "my",
5981 which in any case is probably the better facility to use.
5983 15. A really picky compiler found some instances of statements for creating
5984 error messages that either had too many or two few arguments for the format
5987 16. The size of the buffer for calls to the DNS resolver has been increased
5988 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5989 for addresses that have a lot of PTR records. This alleviates a problem; it
5990 does not fully solve it.
5992 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5993 buffer now truncates the list and logs the incident, which is the same
5994 action as happens when Exim is looking up a host name and its aliases.
5995 Previously in this situation something unpredictable would happen;
5996 sometimes it was "internal error: store_reset failed".
5998 18. If a server dropped the connection unexpectedly when an Exim client was
5999 using GnuTLS and trying to read a response, the client delivery process
6000 crashed while trying to generate an error log message.
6002 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6003 string, for example:
6005 warn message = H1: something\nH2: something
6007 the text was added as a single header line from Exim's point of view
6008 though it ended up OK in the delivered message. However, searching for the
6009 second and subsequent header lines using $h_h2: did not work. This has been
6010 fixed. Similarly, if a system filter added multiple headers in this way,
6011 the routers could not see them.
6013 20. Expanded the error message when iplsearch is called with an invalid key to
6014 suggest using net-iplsearch in a host list.
6016 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6017 ACLs are no longer actually imposed (and a message to that effect is
6020 22. If a "gecos" field in a passwd entry contained escaped characters, in
6021 particular, if it contained a \" sequence, Exim got it wrong when building
6022 a From: or a Sender: header from that name. A second bug also caused
6023 incorrect handling when an unquoted " was present following a character
6024 that needed quoting.
6026 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6027 was not being matched caselessly.
6029 24. Arranged for all hyphens in the exim.8 source to be escaped with
6032 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6033 properly. Recipient callout cache records were still being keyed to include
6034 the sender, even when use_sender was set false. This led to far more
6035 callouts that were necessary. The sender is no longer included in the key
6036 when use_sender is false.
6038 26. Added "control = submission" modifier to ACLs.
6040 27. Added the ${base62d: operator to decode base 62 numbers.
6042 28. dnsdb lookups can now access SRV records.
6044 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6045 the configuration file.
6047 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6048 (-v) mode. This makes the output for a verbose queue run more intelligible.
6050 31. Added a use_postmaster feature to recipient callouts.
6052 32. Added the $body_zerocount variable, containing the number of binary zero
6053 bytes in the message body.
6055 33. The time of last modification of the "new" subdirectory is now used as the
6056 "mailbox time last read" when there is a quota error for a maildir
6059 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6061 35. Added +ignore_unknown as a special item in host lists.
6063 36. Code for decoding IPv6 addresses in host lists is now included, even if
6064 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6065 address was recognized as an IP address, but was then not correctly decoded
6066 into binary, causing unexpected and incorrect effects when compared with
6073 1. Very minor rewording of debugging text in manualroute to say "list of
6074 hosts" instead of "hostlist".
6076 2. If verify=header_syntax was set, and a header line with an unqualified
6077 address (no domain) and a large number of spaces between the end of the
6078 name and the colon was received, the reception process suffered a buffer
6079 overflow, and (when I tested it) crashed. This was caused by some obsolete
6080 code that should have been removed. The fix is to remove it!
6082 3. When running in the test harness, delay a bit after writing a bounce
6083 message to get a bit more predictability in the log output.
6085 4. Added a call to search_tidyup() just before forking a reception process. In
6086 theory, someone could use a lookup in the expansion of smtp_accept_max_
6087 per_host which, without the tidyup, could leave open a database connection.
6089 5. Added the variables $recipient_data and $sender_data which get set from a
6090 lookup success in an ACL "recipients" or "senders" condition, or a router
6091 "senders" option, similar to $domain_data and $local_part_data.
6093 6. Moved the writing of debug_print from before to after the "senders" test
6096 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6097 problems for message scanning, either using a data ACL, or using
6098 local_scan() because the Received: header was not generated till after they
6099 were called (in order to set the time as the time of reception completion).
6100 I have revised the way this works. The header is now generated after the
6101 body is received, but before the ACL or local_scan() are called. After they
6102 are run, the timestamp in the header is updated.
6108 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6109 before starting a queue runner without re-exec. This happened only when
6110 deliver_drop_privilege was set or when the Exim user was set to root. The
6111 effect of the bug was that timeouts during subsequent deliveries caused
6112 crashes instead of being properly handled. The handler is now left at its
6113 default (and expected) setting.
6115 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6116 message, again when deliver_drop_privilege is set or Exim is run as root.
6117 The bug described in (1) was not present in this case, but the tidying up
6118 of the other signals was missing. I have made the two cases consistent.
6120 3. The ignore_target_hosts setting on a manualroute router was being ignored
6121 for hosts that were looked up using the /MX notation.
6123 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6126 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6127 operated on the sender address. After changing the $sender_address to <>
6128 for the sender address verify, Exim was re-instated it as the original
6129 (before rewriting) address, but remembering that it had rewritten it, so it
6130 wasn't rewriting it again. This bug also had the effect of breaking the
6131 sender address verification caching when the sender address was rewritten.
6133 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6134 This has been changed so that if the ip literal address matches
6135 ignore_target_hosts, the router declines.
6137 7. Added expansion conditions match_domain, match_address, and match_local_
6138 part (NOT match_host).
6140 8. The placeholder for the Received: header didn't have a length field set.
6142 9. Added code to Exim itself and to exim_lock to test for a specific race
6143 condition that could lead to file corruption when using MBX delivery. The
6144 issue is with the lockfile that is created in /tmp. If this file is removed
6145 after a process has opened it but before that process has acquired a lock,
6146 there is the potential for a second process to recreate the file and also
6147 acquire a lock. This could lead to two Exim processes writing to the file
6148 at the same time. The added code performs the same test as UW imapd; it
6149 checks after acquiring the lock that its file descriptor still refers to
6150 the same named file.
6152 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6153 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6156 11. Added the smtp_active_hostname option. If used, this will typically be made
6157 to depend on the incoming interface address. Because $interface_address is
6158 not set up until the daemon has forked a reception process, error responses
6159 that can happen earlier (such as "too many connections") no longer contain
6162 12. If an expansion in a condition on a "warn" statement fails because a lookup
6163 defers, the "warn" statement is abandoned, and the next ACL statement is
6164 processed. Previously this caused the whole ACL to be aborted.
6166 13. Added the iplsearch lookup type.
6168 14. Added ident_timeout as a log selector.
6170 15. Added tls_certificate_verified as a log selector.
6172 16. Added a global option tls_require_ciphers (compare the smtp transport
6173 option of the same name). This controls incoming TLS connections.
6175 17. I finally figured out how to make tls_require_ciphers do a similar thing
6176 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6177 before starting the TLS session.
6179 18. Tabs are now shown as \t in -bP output.
6181 19. If the log selector return_path_on_delivery was set, Exim crashed when
6182 bouncing a message because it had too many Received: header lines.
6184 20. If two routers both had headers_remove settings, and the first one included
6185 a superfluous trailing colon, the final name in the first list and the
6186 first name in the second list were incorrectly joined into one item (with a
6187 colon in the middle).
6193 1. Added -C and -D options to the exinext utility, mainly to make it easier
6194 to include in the automated testing, but these could be helpful when
6195 multiple configurations are in use.
6197 2. The exinext utility was not formatting the output nicely when there was
6198 an alternate port involved in the retry record key, nor when there was a
6199 message id as well (for retries that were specific to a specific message
6200 and a specific host). It was also confused by IPv6 addresses, because of
6201 the additional colons they contain. I have fixed the IPv4 problem, and
6202 patched it up to do a reasonable job for IPv6.
6204 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6205 delivery, the log line now contains "pipelined" if PIPELINING was used.
6207 4. An SMTP transport process used to panic and die if the bind() call to set
6208 an explicit outgoing interface failed. This has been changed; it is now
6209 treated in the same way as a connect() failure.
6211 5. A reference to $sender_host_name in the part of a conditional expansion
6212 that was being skipped was still causing a DNS lookup. This no longer
6215 6. The def: expansion condition was not recognizing references to header lines
6216 that used bh_ and bheader_.
6218 7. Added the _cache feature to named lists.
6220 8. The code for checking quota_filecount in the appendfile transport was
6221 allowing one more file than it should have been.
6223 9. For compatibility with Sendmail, the command line option
6231 and sets the incoming protocol and host name (for trusted callers). The
6232 host name and its colon can be omitted when only the protocol is to be set.
6233 Note the Exim already has two private options, -pd and -ps, that refer to
6234 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6235 "s", but I don't think that's a major issue.
6237 10. A number of refactoring changes to the code, none of which should affect
6240 (a) The number of logging options was getting close to filling up the
6241 32-bit word that was used as a bit map. I have split them into two classes:
6242 those that are passed in the argument to log_write(), and those that are
6243 only ever tested independently outside of that function. These are now in
6244 separate 32-bit words, so there is plenty of room for expansion again.
6245 There is no change in the user interface or the logging behaviour.
6247 (b) When building, for example, log lines, the code previously used a
6248 macro that called string_cat() twice, in order to add two strings. This is
6249 not really sufficiently general. Furthermore, there was one instance where
6250 it was actually wrong because one of the argument was used twice, and in
6251 one call a function was used. (As it happened, calling the function twice
6252 did not affect the overall behaviour.) The macro has been replaced by a
6253 function that can join an arbitrary number of extra strings onto a growing
6256 (c) The code for expansion conditions now uses a table and a binary chop
6257 instead of a serial search (which was left over from when there were very
6258 few conditions). Also, it now recognizes conditions like "pam" even when
6259 the relevant support is not compiled in: a suitably worded error message is
6260 given if an attempt is made to use such a condition.
6262 11. Added ${time_interval:xxxxx}.
6264 12. A bug was causing one of the ddress fields not to be passed back correctly
6265 from remote delivery subprocesses. The field in question was not being
6266 subsequently used, so this caused to problems in practice.
6268 13. Added new log selectors queue_time and deliver_time.
6270 14. Might have fixed a bug in maildirsizefile handling that threw up
6271 "unexpected character" debug warnings, and recalculated the data
6272 unnecessarily. In any case, I expanded the warning message to give more
6275 15. Added the message "Restricted characters in address" to the statements in
6276 the default ACL that block characters like @ and % in local parts.
6278 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6279 Three changes have been made:
6281 (a) There was a serious bug; a negative response to MAIL caused the whole
6282 recipient domain to be cached as invalid, thereby blocking all messages
6283 to all local parts at the same domain, from all senders. This bug has
6284 been fixed. The domain is no longer cached after a negative response to
6285 MAIL if the sender used is not empty.
6287 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6290 (c) A new callout option, "use_sender" has been added for people who want
6291 the modified behaviour.
6297 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6300 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6301 indeed breaks things for older releases.
6303 3. Added additional logging to the case where there is a problem reading data
6304 from a filter that is running in a subprocess using a pipe, in order to
6305 try to track down a specific problem.
6307 4. Testing facility fudge: when running in the test harness and attempting
6308 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6309 getting "No route to host". Convert this to a timeout.
6311 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6314 6. Some OS don't have socklen_t but use size_t instead. This affects the
6315 fifth argument of getsockopt() amongst other things. This is now
6316 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6317 can be set for individual OS. I have set it for SunOS5, OSF1, and
6318 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6319 some earlier ones do not.
6321 7. Change 4.30/15 was not doing the test caselessly.
6323 8. The standard form for an IPv6 address literal was being rejected by address
6324 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6325 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6326 this, as well as the form without the "IPv6" on the front (but only when
6327 address literals are enabled, of course).
6329 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6331 10. Exim crashed if a message with an empty sender address specified by -f
6332 encountered a router with an errors_to setting. This could be provoked only
6333 by a command such as
6337 where an empty string was supplied; "<>" did not hit this bug.
6339 11. Installed PCRE release 4.5.
6341 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6342 remained set. It is now erased.
6344 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6345 times from message ids (which are base 36 rather than the normal 62).
6347 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6348 were being counted as actual protocol errors, and logged if the log
6349 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6350 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6351 and DATA following a set of rejected RCPTs do not count as protocol errors.
6352 In other words, Exim assumes they were pipelined, though this may not
6353 actually be the case. Of course, in all cases the client gets an
6354 appropriate error code.
6356 15. If a lookup fails in an ACL condition, a message about the failure may
6357 be available; it is used if testing the ACL cannot continue, because most
6358 such messages specify what the cause of the deferral is. However, some
6359 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6360 that caused an old message to be retained and used if a later statement
6361 caused a defer, replacing the real cause of the deferral.
6363 16. If an IP address had so many PTR records that the DNS lookup buffer
6364 was not large enough to hold them, Exim could crash while trying to process
6365 the truncated data. It now detects and logs this case.
6367 17. Further to 4.21/58, another change has been made: if (and only if) the
6368 first line of a message (the first header line) ends with CRLF, a bare LF
6369 in a subsequent header line has a space inserted after it, so as not to
6370 terminate the header.
6372 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6373 unnecessarily, used atoi() instead of strtol(), and didn't check the
6374 termination when getting file sizes from file names by regex.
6376 19. Completely re-implemented the support for maildirsize files, in the light
6377 of a number of problems with the previous contributed implementation
6378 (4.30/29). In particular:
6380 . If the quota is zero, the maildirsize file is maintained, but no quota is
6383 . If the maildir directory does not exist, it is created before any attempt
6384 to write a maildirsize file.
6386 . The quota value in the file is just a cache; if the quota is changed in
6387 the transport, the new value overrides.
6389 . A regular expression is available for excluding directories from the
6392 20. The autoreply transport checks the characters in options that define the
6393 message's headers; it allows continued headers, but it was checking with
6394 isspace() after an embedded newline instead of explicitly looking for a
6397 21. If all the "regular" hosts to which an address was routed had passed their
6398 expiry times, and had not reached their retry times, the address was
6399 bounced, even if fallback hosts were defined. Now Exim should go on to try
6402 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6403 equivalent code in the SMTP transport. Some hosts send humungous responses
6404 to HELO/EHLO, more than 1024 it seems.
6406 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6407 gives compiler warnings in some environments. I've now done it "properly",
6410 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6411 (because of problems with the built-in one) was declared to return uschar *
6412 instead of char *, causing compiler failure.
6414 25. Fixed a file descriptor leak when processing alias/forward files.
6416 26. Fixed a minor format string issue in dbfn.c.
6418 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6420 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6421 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6424 29. When a sender address is verified, it is cached, to save repeating the test
6425 when there is more than one recipient in a message. However, when the
6426 verification involves a callout, it is possible for different callout
6427 options to be set for different recipients. It is too complicated to keep
6428 track of this in the cache, so now Exim always runs a verification when a
6429 callout is required, relying on the callout cache for the optimization.
6430 The overhead is duplication of the address routing, but this should not be
6433 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6434 to be verified with callout=postmaster, and the main callout worked but the
6435 postmaster check failed, the verification correctly failed. However, if a
6436 subsequent RCPT command asked for sender verification *without* the
6437 postmaster check, incorrect caching caused this verification also to fail,
6440 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6441 it was not caching the DNS options (qualify_single, search_parents) that
6442 were used when the lookup failed. A subsequent lookup with different
6443 options therefore always gave the same answer, though there were cases
6444 where it should not have. (Example: a "domains = !$mx_any" option on a
6445 dnslookup router: the "domains" option is always processed without any
6446 widening, but the router might have qualify_single set.) Now Exim uses the
6447 cached value only when the same options are set.
6449 32. Added John Jetmore's "exipick" utility to the distribution.
6451 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6452 than a timeout (e.g. a certificate is required, and is not provided), an
6453 Exim server now closes the connection immediately. Previously it waited for
6454 the client to close - but if the client is SSL, it seems that they each
6455 wait for each other, leading to a delay before one of them times out.
6457 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6458 maintained 0.8.x compatibility because I don't think many are using it, and
6459 it is clearly obsolete.
6461 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6464 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6465 client certificate was expired. A simple patch fixes this, though I don't
6466 understand the full logic of why the verify callback is called multiple
6469 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6470 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6471 which causes problems with some clients (such as the Certicom SSL Plus
6472 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6473 disables the coutermeasure allowing Eudora to connect."
6475 38. Exim was not checking that a write() to a log file succeeded. This could
6476 lead to Bad Things if a log got too big, in particular if it hit a file
6477 size limit. Exim now panics and dies if it cannot write to a log file, just
6478 as it does if it cannot open a log file.
6480 39. Modified OS/Makefile-Linux so that it now contains
6482 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6484 The two -D definitions ensure that Exim is compiled with large file
6485 support, which makes it possible to handle log files that are bigger than
6488 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6489 instance) a domain was checked against a named list that involved a lookup,
6490 causing $domain_data to be set, then another domain was checked against the
6491 same list, then the first domain was re-checked, the value of $domain_data
6492 after the final check could be wrong. In particular, if the second check
6493 failed, it could be set empty. This bug probably also applied to
6496 41. The strip_trailing_dot option was not being applied to the address given
6497 with the -f command-line option.
6499 42. The code for reading a message's header from the spool was incrementing
6500 $received_count, but never initializing it. This meant that the value was
6501 incorrect (doubled) while delivering a message in the same process in which
6502 it was received. In the most common configuration of Exim, this never
6503 happens - a fresh exec is done - but it can happen when
6504 deliver_drop_privilege is set.
6506 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6507 it now includes up to 150 characters of the unexpected data in the log
6510 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6511 and building data strings. The size of both of these buffers was 10 000
6512 bytes - far larger than anybody would *ever* want, thought I. Needless to
6513 say, somebody hit the limit. I have increased the maximum line length to
6514 20 000 and the maximum data length of concatenated lines to 100 000. I have
6515 also fixed two bugs, because there was no checking on these buffers. Tsk,
6516 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6517 buffer is too small.
6519 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6520 lsearch lookups. Now it does.
6522 46. When parsing a route_list item in a manualroute router, a fixed-length
6523 buffer was used for the list of hosts. I made this 1024 bytes long,
6524 thinking that nobody would ever have a list of hosts that long. Wrong.
6525 Somebody had a whole pile of complicated expansion conditions, and the
6526 string was silently truncated, leading to an expansion error. It turns out
6527 that it is easier to change to an unlimited length (owing to other changes
6528 that have happened since this code was originally written) than to build
6529 structure for giving a limitation error. The length of the item that
6530 expands into the list of hosts is now unlimited.
6532 47. The lsearch lookup could not handle data where the length of text line was
6533 more than 4095 characters. Such lines were truncated, leading to shortened
6534 data being returned. It should now handle lines of any length.
6536 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6537 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6540 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6541 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6542 no longer happens in comment lines. A list of replacements is now placed
6543 at the head of all of the source files, except those whose only change is
6544 to replace PERL_COMMAND in the very first #! line.
6546 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6547 messages on the queue, with a bottom-up merge sort, using code contributed
6548 by Michael Haardt. This should make operations like -bp somewhat faster on
6549 large queues. It won't affect queue runners, except when queue_run_in_order
6552 51. Installed eximstats 1.31 in the distribution.
6554 52. Added support for SRV lookups to the dnslookup router.
6556 53. If an ACL referred to $message_body or $message_body_end, the value was not
6557 reset for any messages that followed in the same SMTP session.
6559 54. The store-handling optimization for building very long strings was not
6560 differentiating between the different store pools. I don't think this
6561 actually made any difference in practice, but I've tidied it.
6563 55. While running the routers to verify a sender address, $sender_address
6564 was still set to the sender address. This is wrong, because when routing to
6565 send a bounce to the sender, it would be empty. Therefore, I have changed
6566 it so that, while verifying a sender address, $sender_address is set to <>.
6567 (There is no change to what happens when verifying a recipient address.)
6569 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6570 target A or AAAA records (if not already returned) without resetting the
6571 qualify_single or search_parents options of the DNS resolver. These are
6572 inappropriate in this case because the targets of MX and SRV records must
6573 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6574 target that, when qualified, matched something in the local domain. These
6575 two options are now turned off when doing these lookups.
6577 57. It seems that at least some releases of Reiserfs (which does not have the
6578 concept of a fixed number of inodes) returns zero and not -1 for the
6579 number of available inodes. This interacted badly with check_spool_inodes,
6580 which assumed that -1 was the "no such thing" setting. What I have done is
6581 to check that the total number of inodes is greater than zero before doing
6582 the test of how many are available.
6584 58. When a "warn" ACL statement has a log_message modifier, the message is
6585 remembered, and not repeated. This is to avoid a lot of repetition when a
6586 message has many recipients that cause the same warning to be written.
6587 However, Exim was preserving the list of already written lines for an
6588 entire SMTP session, which doesn't seem right. The memory is now reset if a
6589 new message is started.
6591 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6592 debugging output unless log_rewrite was also set.
6594 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6595 of (int)(handle) when we know that handle contains (void *)(-1).
6597 61. The Exim daemon panic-logs an error return when it closes the incoming
6598 connection. However "connection reset by peer" seems to be common, and
6599 isn't really an error worthy of noting specially, so that particular error
6602 62. When Exim is trying to find all the local interfaces, it used to panic and
6603 die if the ioctl to get the interface flags failed. However, it seems that
6604 on at least one OS (Solaris 9) it is possible to have an interface that is
6605 included in the list of interfaces, but for which you get a failure error
6606 for this call. This happens when the interface is not "plumbed" into a
6607 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6608 failure of the "get flags" call assumes that the interface is down.
6610 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6611 makes life easier for people who are doing arithmetic on fields extracted
6612 from dates, where you often get leading zeros that should not be
6613 interpreted as octal.
6615 64. Added qualify_domain to the redirect router, to override the global
6618 65. If a pathologically long header line contained very many addresses (the
6619 report of this problem mentioned 10 000) and each of them was rewritten,
6620 Exim could use up a very large amount of memory. (It kept on making new
6621 copies of the header line as it rewrote, and never released the old ones.)
6622 At the expense of a bit more processing, the header rewriting function has
6623 been changed so that it no longer eats memory in this way.
6625 66. The generation of the Received: header has been moved from the time that a
6626 message starts to be received, to the time that it finishes. The timestamp
6627 in the Received: header should now be very close to that of the <= log
6628 line. There are two side-effects of this change:
6630 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6631 logged header lines no longer include the local Received: line, because
6632 it has not yet been created. The same applies to a copy of the message
6633 that is returned to a non-SMTP sender when a message is rejected.
6635 (b) When a filter file is tested using -bf, no additional Received: header
6636 is added to the test message. After some thought, I decided that this
6639 This change does not affect the value of $received_for. It is still set
6640 after address rewriting, but before local_scan() is called.
6642 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6644 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6645 gave an unhelpful panic error message, and a defer error. I have managed to
6646 change this behaviour so that it now rejects any supplied certificate,
6647 which seems right, as the list of acceptable certificates is empty.
6649 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6650 gave an unhelpful defer error. I have not managed to make this reject any
6651 supplied certificates, but the error message it gives is "no certificate
6652 supplied", which is not helpful.
6654 70. exigrep's output now also includes lines that are not associated with any
6655 message, but which match the given pattern. Implemented by a patch from
6656 Martin Sluka, which also tidied up the Perl a bit.
6658 71. Recipient callout verification, like sender verification, was using <> in
6659 the MAIL FROM command. This isn't really the right thing, since the actual
6660 sender may affect whether the remote host accepts the recipient or not. I
6661 have changed it to use the actual sender in the callout; this means that
6662 the cache record is now keyed on a recipient/sender pair, not just the
6663 recipient address. There doesn't seem to be a real danger of callout loops,
6664 since a callout by the remote host to check the sender would use <>.
6665 [SEE ABOVE: changed after hitting problems.]
6667 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6668 temporary errors. However, in the case of such a code being given after
6669 the end of a data transmission (i.e. after ".") Exim was failing to write
6670 a retry record for the message. (Yes, there was some broken host that was
6671 actually sending 8xx at this point.)
6673 73. An unknown lookup type in a host list could cause Exim to panic-die when
6674 the list was checked. (An example that provoked this was putting <; in the
6675 middle of a list instead of at the start.) If this happened during a DATA
6676 ACL check, a -D file could be left lying around. This kind of configuration
6677 error no longer causes Exim to die; instead it causes a defer error. The
6678 incident is still logged to the main and panic logs.
6680 74. Buglet left over from Exim 3 conversion. The message "too many messages
6681 in one connection" was written to the rejectlog but not the mainlog, except
6682 when address rewriting (yes!) was being logged.
6684 75. Added write_rejectlog option.
6686 76. When a system filter was run not as root (that is, when system_filter_user
6687 was set), the values of the $n variables were not being returned to the
6688 main process; thus, they were not subsequently available in the $sn
6691 77. Added +return_path_on_delivery log selector.
6693 78. A connection timeout was being treated differently from recipients deferred
6694 when testing hosts_max_try with a message that was older than the host's
6695 retry timeout. (The host should not be counted, thus allowing all hosts to
6696 be tried at least once before bouncing.) This may have been the cause of an
6697 occasionally reported bug whereby a message would remain on the queue
6698 longer than the retry timeout, but would be bounced if a delivery was
6699 forced. I say "may" because I never totally pinned down the problem;
6700 setting up timeout/retry tests is difficult. See also the next item.
6702 79. The ultimate address timeout was not being applied to errors that involved
6703 a combination of host plus message (for example, a timeout on a MAIL
6704 command). When an address resolved to a number of possible hosts, and they
6705 were not all tried for each delivery (e.g. because of hosts_max_try), a
6706 message could remain on the queue longer than the retry timeout.
6708 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6711 81. Fixed an obscure SMTP outgoing bug which required at least the following
6712 conditions: (a) there was another message waiting for the same server;
6713 (b) the server returned 5xx to all RCPT commands in the first message so
6714 that the message was not completed; (c) the server dropped the connection
6715 or gave a negative response to the RSET that Exim sends to abort the
6716 transaction. The observed case was a dropped connection after DATA that had
6717 been sent in pipelining mode. That is, the server had advertised PIPELINING
6718 but was not implementing it correctly. The effect of the bug was incorrect
6719 behaviour, such as trying another host, and this could lead to a crash.
6725 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6726 and daemon.c were passed as pointers to ints; they should have been
6727 pointers to socklen_t variables (which are typically unsigned ints).
6729 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6732 3. Fixed a really odd bug that affected only the testing scheme; patching a
6733 certain fixed string in the binary changed the value of another string that
6734 happened to be identical to the end of the original first string.
6736 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6737 name", it returns that address as the IP address. On some operating
6738 systems (e.g. Solaris), it also passes back the IP address string as the
6739 "host name". However, on others (e.g. Linux), it passes back an empty
6740 string. Exim wasn't checking for this, and was changing the host name to an
6741 empty string, assuming it had been canonicalized.
6743 5. Although rare, it is permitted to have more than one PTR record for a given
6744 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6745 all the names associated with an address, because they do in Solaris.
6746 However, it seems that they do not in Linux for data that comes from the
6747 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6748 I found this out when I moved to a new Linux workstation and tried to run
6749 the Exim test suite.
6751 To get round this problem I have changed the code so that it now does its
6752 own call to the DNS to look up PTR records when searching for a host name.
6753 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6754 addresses that are only in /etc/hosts are still found.
6756 This behaviour is, however, controlled by an option called host_lookup_
6757 order, which defaults to "bydns:byaddr". If people want to use the other
6758 order, or indeed, just use one or the other means of lookup, they can
6759 specify it in this variable.
6761 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6762 some operating systems, this comes back from gethostbyaddr() as an empty
6763 string, and this is what Exim used to test for. However, it seems that in
6764 other systems, "." is yielded. Exim now tests for this case too.
6766 7. The values of check_spool_space and check_log_space are now held internally
6767 as a number of kilobytes instead of an absolute number of bytes. If a
6768 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6769 kilobyte. This means that much larger values can be stored.
6771 8. Exim monitor: an attempt to get the action menu when not actually pointing
6772 at a message produces an empty menu entitled "No message selected". This
6773 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6774 no entries in it ("Shell widget menu has zero width and/or height"). So I
6775 have added a single, blank menu entry in this case.
6777 9. Added ${quote_local_part.
6779 10. MIME decoding is now applied to the contents of Subject: header lines when
6782 11. Now that a reference to $sender_host_address automatically causes a reverse
6783 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6784 host lookup before query-style lookups in lists that might use this
6785 variable. This has therefore been abolished, and the "net-" prefix is no
6786 longer necessary for query-style lookups.
6788 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6789 have been a typo for LFLAGS, so it has been changed.
6791 13. The install script calls Exim with "-C /dev/null" in order to find the
6792 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6793 to be output. However, since Exim outputs its version number before the
6794 error, it didn't break the script. It just looked ugly. I fixed this by
6795 always allowing "-C /dev/null" if the caller is root.
6797 14. Ignore overlarge ACL variable number when reading spool file - insurance
6798 against a later release with more variables having written the file.
6800 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6801 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6802 this, as well as the form without the "IPv6" on the front.
6804 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6805 OS/Makefile-Darwin file.
6807 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6808 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6810 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6811 variable. However, some OS use ints for inodes. I've added cast to long int
6812 to get rid of the compiler warning.
6814 19. I had forgotten to lock out "/../" in configuration file names when
6815 ALT_CONFIG_PREFIX was set.
6817 20. Routers used for verification do not need to specify transports. However,
6818 if such a router generated a host list, and callout was configured, Exim
6819 crashed, because it could not find a port number from the (non-existent)
6820 transport. It now assumes port 25 in this circumstance.
6822 21. Added the -t option to exigrep.
6824 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6825 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6826 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6827 them is set, LOOKUP_LSEARCH is forced.
6829 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6831 24. Added sender and host information to the "rejected by local_scan()" log
6832 line; previously there was no indication of these.
6834 25. Added .include_if_exists.
6836 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6837 Linux. It turns out that not all file systems support this. Apparently some
6838 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6839 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6840 not supported on the file descriptor, is now ignored when Exim is trying to
6841 sync a directory. This applies only to Linux.
6843 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6845 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6846 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6849 29. Added support for maildirsize files from supplied patch (modified a bit).
6851 30. The use of :fail: followed by an empty string could lead Exim to respond to
6852 sender verification failures with (e.g.):
6854 550 Verification failed for <xxx>
6855 550 Sender verify failed
6857 where the first response line was missing the '-' that indicates it is not
6858 the final line of the response.
6860 31. The loop for finding the name of the user that called Exim had a hardwired
6861 limit of 10; it now uses the value of finduser_retries, which is used for
6862 all other user lookups.
6864 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6867 33. Exim was neglecting to zero errno before one call of strtol() when
6868 expanding a string and expecting an integer value. On some systems this
6869 resulted in spurious "integer overflow" errors. Also, it was casting the
6870 result into an int without checking.
6872 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6873 did not work. The code looks as if it has *never* worked, though it appears
6874 to have been documented since at least release 1.62. I have made it work.
6876 35. The "timeout_DNS" error in retry rules, also documented since at least
6877 1.62, also never worked. As it isn't clear exactly what this means, and
6878 clearly it isn't a major issue, I have abolished the feature by treating it
6879 as "timeout", and writing a warning to the main and panic logs.
6881 36. The display of retry rules for -brt wasn't always showing the error code
6884 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6885 timeout_connect_A, timeout_connect_MX.
6887 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6888 to the empty sender.
6890 39. The daemon was not analysing the content of -oX till after it had closed
6891 stderr and disconnected from the controlling terminal. This meant that any
6892 syntax errors were only noted on the panic log, and the return code from
6893 the command was 0. By re-arranging the code a little, I've made the
6894 decoding happen first, so such errors now appear on stderr, and the return
6895 code is 1. However, the actual setting up of the sockets still happens in
6896 the disconnected process, so errors there are still only recorded on the
6899 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6900 connections (as happens on some IP stacks) was logged at start up time as
6901 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6902 it from "IPv6 and IPv4", which means that two separate sockets are being
6905 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6906 says whether AF_INET or AF_INET6 was passed as an argument.
6908 42. Exiwhat output was messed up when time zones were included in log
6911 43. Exiwhat now gives more information about the daemon's listening ports,
6912 and whether -tls-on-connect was used.
6914 44. The "port" option of the smtp transport is now expanded.
6916 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6917 silently ignored. Now an error message is written to the main and panic
6920 46. There's a new ACL modifier called "logwrite" which writes to a log file
6921 as soon as it is encountered.
6923 47. Added $local_user_uid and $local_user_gid at routing time.
6925 48. Exim crashed when trying to verify a sender address that was being
6928 49. Exim was recognizing only a space character after ".include". It now also
6929 recognizes a tab character.
6931 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6932 extracting the relevant information from the specification. The man page no
6933 longer contains scrambled data for the -d option, and I've added a section
6934 at the front about calling Exim under different names.
6936 51. Added "extra_headers" argument to the "mail" command in filter files.
6938 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6941 53. Installed eximstats 1.29.
6943 54. Added transport_filter_timeout as a generic transport option.
6945 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6946 Cc: header lines. This was required by RFC 822, but it not required by RFC
6949 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6950 incoming messages that did not have them. Now it does so only if the
6951 message originates locally, that is, if there is no associated remote host
6952 address. When Resent- header lines are present, this applies to the Resent-
6953 lines rather than the non-Resent- lines.
6955 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6956 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6958 58. Messages for configuration errors now include the name of the main
6959 configuration file - useful now that there may be more than one file in a
6960 list (.included file names were always shown).
6962 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6963 for those rare installations that do not start the daemon as root or run it
6964 setuid root. I've cut out the call to initgroups() if the daemon is not
6967 60. The Exim user and group can now be bound into the binary as text strings
6968 that are looked up at the start of Exim's processing.
6970 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6972 62. Added $mailstore_basename variable.
6974 63. Installed patch to sieve.c from Michael Haardt.
6976 64. When Exim failed to open the panic log after failing to open the main log,
6977 the original message it was trying to log was written to stderr and debug
6978 output, but if they were not available (the usual case in production), it
6979 was lost. Now it is written to syslog before the two lines that record the
6980 failures to open the logs.
6982 65. Users' Exim filters run in subprocesses under the user's uid. It is
6983 possible for a "deliver" command or an alias in a "personal" command to
6984 provoke an address rewrite. If logging of address rewriting is configured,
6985 this fails because the process is not running as root or exim. There may be
6986 a better way of dealing with this, but for the moment (because 4.30 needs
6987 to be released), I have disabled address rewrite logging when running a
6988 filter in a non-root, non-exim process.
6994 1. The buildconfig auxiliary program wasn't quoting the value set for
6995 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6996 not defined. This bug was masked in 4.22 by the effect that was fixed in
6999 2. Some messages that were rejected after a message id was allocated were
7000 shown as "incomplete" by exigrep. It no longer does this for messages that
7001 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7003 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7004 have allow_domain_literals set, the ID did not get logged in the <= line.
7005 Domain literals are now always recognized in Message-ID: header lines.
7007 4. The first argument for a ${extract expansion item is the key name or field
7008 number. Leading and trailing spaces in this item were not being ignored,
7009 causing some misleading effects.
7011 5. When deliver_drop_privilege was set, single queue runner processes started
7012 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7013 same command in the process it spins off) were not dropping privilege.
7015 6. When the daemon running as "exim" started a queue runner, it always
7016 re-executed Exim in the spun-off process. This is a waste of effort when
7017 deliver_drop_privilege is set. The new process now just calls the
7018 queue-runner function directly.
7024 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7027 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7028 ignored. Though the use of -oP was forcing the writing of a pid file, it
7029 was always written to the default place.
7031 3. If the message "no IP address found for host xxxx" is generated during
7032 incoming verification, it is now followed by identification of the incoming
7033 connection (so you can more easily find what provoked it).
7035 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7037 5. Added some features to "harden" Exim a bit more against certain attacks:
7039 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7040 be put in Local/Makefile. This is like the never_users runtime option,
7041 but it cannot be overridden. The default setting is "root".
7043 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7044 prefix string with which any file named in a -C command line option
7047 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7048 is retained for -C and -D only if the caller of Exim is root. Without
7049 it, the exim user may also use -C and -D and retain privilege.
7051 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7052 command line option is disabled.
7054 6. Macro names set by the -D option must start with an upper case letter, just
7055 like macro names defined in the configuration file.
7057 7. Added "dereference=" facility to LDAP.
7059 8. Two instances of the typo "uknown" in the source files are fixed.
7061 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7062 the Configure-Makefile script screwed up while processing it.
7064 10. Incorporated PCRE 4.4.
7066 11. The SMTP synchronization check was not operating right at the start of an
7067 SMTP session. For example, it could not catch a HELO sent before the client
7068 waited for the greeting. There is now a check for outstanding input at the
7069 point when the greeting is written. Because of the duplex, asynchronous
7070 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7071 way, but not yet received, when the check is performed.
7073 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7074 on TCP/IP sockets, because this apparently causes some broken clients to
7077 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7078 unchanged) from the Cygwin maintainer.
7080 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7081 was supported instead of testing for mbx. Effectively a typo.
7083 15. The spa authenticator server code was not checking that the input it
7084 received was valid base64.
7086 16. The debug output line for the "set" modifier in ACLs was not showing the
7087 name of the variable that was being set.
7089 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7091 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7092 lookup on its own. Something else was needed to trigger the lookup. For
7093 example, a match in host_lookup or the need for a host name in a host list.
7094 Now, if $sender_host_name is referenced and the host name has not yet been
7095 looked up, a lookup is performed. If the lookup fails, the variable remains
7096 empty, and $host_lookup_failed is set to "1".
7098 19. Added "eqi" as a case-independent comparison operator.
7100 20. The saslauthd authentication condition could segfault if neither service
7101 nor realm was specified.
7103 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7104 error message that was logged was misleading, and incoming SMTP
7105 connections were dropped. The message is now more accurate, and temporary
7106 errors are given to SMTP connections.
7108 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7109 (except RSET) with a 421 error, until QUIT is received. However, it was
7110 failing to send a response to QUIT.
7112 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7113 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7114 ACL is now run after the helo_try_verify_hosts code.
7116 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7117 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7118 case-independent, so other case variants are also recognized). Apparently
7119 some systems use these upper case variants.
7121 25. If more than two messages were waiting for the same host, and a transport
7122 filter was specified for the transport, Exim sent two messages over the
7123 same TCP/IP connection, and then failed with "socket operation on non-
7124 socket" when it tried to send the third.
7126 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7128 27. The extern definition of crypt16() in expand.c was not being excluded when
7129 the OS had its own crypt16() function.
7131 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7132 as a preferred synonym for return_size_limit, both as an option and as an
7135 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7137 30. Changed the default configuration ACL to relax the local part checking rule
7138 for addresses that are not in any local domains. For these addresses,
7139 slashes and pipe symbols are allowed within local parts, but the sequence
7140 /../ is explicitly forbidden.
7142 31. SPA server authentication was not clearing the challenge buffer before
7145 32. log_message in a "warn" ACL statement was writing to the reject log as
7146 well as to the main log, which contradicts the documentation and doesn't
7147 seem right (because no rejection is happening). So I have stopped it.
7149 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7150 However, I am unable to do any testing of this.
7152 34. Fixed an infelicity in the appendfile transport. When checking directories
7153 for a mailbox, to see if any needed to be created, it was accidentally
7154 using path names with one or more superfluous leading slashes; tracing
7155 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7157 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7158 contents are added to the log line that is written for every discarded
7159 recipient. (Previously a log_message setting was ignored.)
7161 36. The ${quote: operator now quotes the string if it is empty.
7163 37. The install script runs exim in order to find its version number. If for
7164 some reason other than non-existence or emptiness, which it checks, it
7165 could not run './exim', it was installing it with an empty version number,
7166 i.e. as "exim-". This error state is now caught, and the installation is
7169 38. An argument was missing from the function that creates an error message
7170 when Exim fails to connect to the socket for saslauthd authentication.
7171 This could cause Exim to crash, or give a corrupted message.
7173 39. Added isip, isip4, and isip6 to ${if conditions.
7175 40. The ACL variables $acl_xx are now saved with the message, and can be
7176 accessed later in routers, transports, and filters.
7178 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7179 strings in the file are not string-expanded.
7181 42. If a MAIL command specified a SIZE value that was too large to fit into an
7182 int variable, the check against message_size_limit failed. Such values are
7183 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7184 day this will have to be increased, but I don't think I want to be around
7185 when emails are that large.
7192 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7193 iconv() is not standard in FreeBSD.
7195 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7196 IPv6 enabled. The observed symptom was a segmentation fault on return from
7197 the function os_common_find_running_interfaces() in src/os.c.
7199 3. In the check_special_case() function in daemon.c I had used "errno" as an
7200 argument name, which causes warnings on some systems. This was basically a
7201 typo, since it was named "eno" in the comments!
7203 4. The code that waits for the clock to tick (at a resolution of some fraction
7204 of a second) so as to ensure message-id uniqueness was always waiting for
7205 at least one whole tick, when it could have waited for less. [This is
7206 almost certainly not relevant at current processor speeds, where it is
7207 unlikely to ever wait at all. But we try to future-proof.]
7209 5. The function that sleeps for a time interval that includes fractions of a
7210 second contained a race. It did not block SIGALRM between setting the
7211 timer, and suspending (a couple of lines later). If the interval was short
7212 and the sigsuspend() was delayed until after it had expired, the suspension
7213 never ended. On busy systems this could lead to processes getting stuck for
7216 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7217 process, before it forks any delivery processes. The open lookup caching
7218 mechanism meant that the open file or database connection was passed into
7219 the delivery process. The problem was that delivery processes always tidy
7220 up cached lookup data. This could cause a problem for the next delivery
7221 process started by the queue runner, because the external queue runner
7222 process does not know about the closure. So the next delivery process
7223 still has data in the lookup cache. In the case of a file lookup, there was
7224 no problem because closing a file descriptor in a subprocess doesn't affect
7225 the parent. However, if the lookup was caching a connection to a database,
7226 the connection was closed, and the second delivery process was likely to
7227 see errors such as "PGSQL: query failed: server closed the connection
7228 unexpectedly". The problem has been fixed by closing all cached lookups
7229 in a queue runner before running a delivery process.
7231 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7232 seem to have the "const" qualifier which it has on other OS. I've
7235 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7236 *of the same type* (client or server) with the same public name that an
7237 error should be diagnosed.
7239 9. When Exim looked up a host name for an IP address, but failed to find the
7240 original IP address when looking up the host name (a safety check), it
7241 output the message "<ip address> does not match any IP for NULL", which was
7242 confusing, to say the least. The bug was that the host name should have
7243 appeared instead of "NULL".
7245 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7246 user that is built into the binary, and the -C or -D options is used, root
7247 privilege is dropped before the configuration file is read. In addition,
7248 logging is switched to stderr instead of the normal log files. If the
7249 configuration then re-defines the Exim user, the unprivileged environment
7250 is probably not what is expected, so Exim logs a panic warning message (but
7253 However, if deliver_drop_privilege is set, the unprivileged state may well
7254 be exactly what is intended, so the warning has been cut out in that case,
7255 and Exim is allowed to try to write to its normal log files.
7261 1. smtp_return_error_details was not giving details for temporary sender
7262 or receiver verification errors.
7264 2. Diagnose a configuration error if two authenticators have the same public
7267 3. Exim used not to create the message log file for a message until the first
7268 delivery attempt. This could be confusing when incoming messages were held
7269 for policy or load reasons. The message log file is now created at the time
7270 the message is received, and an initial "Received" line is written to it.
7272 4. The automatically generated man page for command line options had a minor
7273 bug that caused no ill effects; however, a more serious problem was that
7274 the procedure for building the man page automatically didn't always
7275 operate. Consequently, release 4.20 contains an out-of-date version. This
7276 shouldn't happen again.
7278 5. When building Exim with embedded Perl support, the script that builds the
7279 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7280 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7282 6. The freeze_tell option was not being used for messages that were frozen on
7283 arrival, either by an ACL or by local_scan().
7285 7. Added the smtp_incomplete_transaction log selector.
7287 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7288 was accepting AUTH without a new EHLO.
7290 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7291 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7294 10. Logging of TCP/IP connections (when configured) now happens in the main
7295 daemon process instead of the child process, so that the TCP/IP connection
7296 count is more accurate (but it can never be perfect).
7298 11. The use of "drop" in a nested ACL was not being handled correctly in the
7299 outer ACL. Now, if condition failure induced by the nested "drop" causes
7300 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7301 or "require"), the connection is dropped.
7303 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7304 that yield "discard" can now be used with an "accept" or a "discard" verb,
7305 but an error is generated for any others (because I can't see a useful way
7306 to define what should happen).
7308 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7309 are now processed in the same way as lines in the Exim configuration file.
7310 In particular, continuation lines are supported.
7312 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7314 15. Added -ti meaning -t -i.
7316 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7317 domains, and warn by logging if others are found.
7319 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7320 returned from GIFCONF when Exim is trying to find the list of interfaces on
7321 a host. The code in os.c has been modified to copy each ifreq to an aligned
7322 structure in all cases.
7324 Also, in some cases, the returned ifreq's were being copied to a 'struct
7325 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7326 means the last couple of bytes of an IPv6 address could be chopped if the
7327 ifreq contained only a normal sockaddr (14 bytes storage).
7329 18. Named domain lists were not supported in the hosts_treat_as_local option.
7330 An entry such as +xxxx was not recognized, and was treated as a literal
7333 19. Ensure that header lines added by a DATA ACL are included in the reject log
7334 if the ACL subsequently rejects the message.
7336 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7337 MD5 (which is deprecated).
7339 21. When testing a filter file using -bf, Exim was writing a message when it
7340 took the sender from a "From " line in the message, but it was not doing so
7341 when it took $return_path from a Return-Path: header line. It now does.
7343 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7344 with a valid header line field name (a series of printing characters
7345 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7347 23. Changed "disc" in the source to "disk" to conform to the documentation and
7348 the book and for uniformity.
7350 24. Ignore Sendmail's -Ooption=value command line item.
7352 25. When execve() failed while trying to run a command in a pipe transport,
7353 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7354 could be confused with a return value of 69 from the command itself. This
7355 has been changed to 127, the value the shell returns if it is asked to run
7356 a non-existent command. The wording for the related log line suggests a
7357 non-existent command as the problem.
7359 26. If received_header_text expands to an empty string, do not add a Received:
7360 header line to the message. (Well, it adds a token one on the spool, but
7361 marks it "old" so that it doesn't get used or transmitted.)
7363 27. Installed eximstats 1.28 (addition of -nt option).
7365 28. There was no check for failure on the call to getsockname() in the daemon
7366 code. This can fail if there is a shortage of resources on the system, with
7367 ENOMEM, for example. A temporary error is now given on failure.
7369 29. Contrary to the C standard, it seems that in some environments, the
7370 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7371 program. Exim now does this explicitly; it affects the formatting of
7372 timestamps using strftime().
7374 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7375 complaints. I've now initialized all the variables, to avoid this.
7377 32. Header lines added by a system filter were not being "seen" during
7378 transport-time rewrites.
7380 33. The info_callback() function passed to OpenSSL is set up with type void
7381 (*)(SSL *, int, int), as described somewhere. However, when calling the
7382 function (actually a macro) that sets it up, the type void(*)() is
7383 expected. I've put in a cast to prevent warnings from picky compilers.
7385 34. If a DNS black list lookup found a CNAME record, but there were no A
7386 records associated with the domain it pointed at, Exim crashed.
7388 35. If a DNS black list lookup returned more than one A record, Exim ignored
7389 all but the first. It now scans all returned addresses if a particular IP
7390 value is being sought. In this situation, the contents of the
7391 $dnslist_value variable are a list of all the addresses, separated by a
7394 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7395 to accept a host name and all its aliases if the forward lookup for any of
7396 them yielded the IP address of the incoming connection. Now it accepts only
7397 those names whose forward lookup yields the correct IP address. Any other
7398 names are discarded. This closes a loophole whereby a rogue DNS
7399 administrator could create reverse DNS records to break through a
7400 wildcarded host restriction in an ACL.
7402 37. If a user filter or a system filter that ran in a subprocess used any of
7403 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7404 the wrong values were passed to the pipe command ($thisaddress had the
7405 value of $0, $0 had the value of $1, etc). This bug was introduced by
7406 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7408 38. Improved the line breaking for long SMTP error messages from ACLs.
7409 Previously, if there was no break point between 40 and 75 characters, Exim
7410 left the rest of the message alone. Two changes have been made: (a) I've
7411 reduced the minimum length to 35 characters; (b) if it can't find a break
7412 point between 35 and 75 characters, it looks ahead and uses the first one
7413 that it finds. This may give the occasional overlong line, but at least the
7414 remaining text gets split now.
7416 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7417 file descriptors might be low, and that setting 1000 would always raise it.
7418 It turns out that in some environments, the limit is already over 1000 and
7419 that lowering it causes trouble. So now Exim takes care not to decrease it.
7421 40. When delivering a message, the value of $return_path is set to $sender_
7422 address at the start of routing (routers may change the value). By an
7423 oversight, this default was not being set up when an address was tested by
7424 -bt or -bv, which affected the outcome if any router or filter referred to
7427 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7428 log only when "message" or "log_message" are set. However, if one of the
7429 conditions was an address verification, or a call to a nested ACL, the
7430 messages generated by the underlying test were being passed through. This
7431 no longer happens. The underlying message is available in $acl_verify_
7432 message for both "message" and "log_message" expansions, so it can be
7433 passed through if needed.
7435 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7436 new expansion $bh_ to give the encoded byte string without charset
7437 translation. Translation happens only if iconv() is available; HAVE_ICONV
7438 indicates this at build time. HEADERS_CHARSET gives the charset to
7439 translate to; headers_charset can change it in the configuration, and
7440 "headers charset" can change it in an individual filter file.
7442 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7443 that creates RFC 2047 encoded "words" labels them as that charset instead
7444 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7445 expansion operator; (ii) when Exim creates a From: line for a local
7446 message; (iii) when a header line is rewritten to include a "phrase" part.
7448 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7449 buggy, causing it to skip the first lines of messages whose message ID
7450 ended in 'D'. This would not have bitten before Exim release 4.14, because
7451 message IDs were unlikely to end in 'D' before then. The effect was to have
7452 incorrect size information for certain domains.
7454 45. #include "config.h" was missing at the start of the crypt16.c module. This
7455 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7458 46. If there was a timeout during a "random" callout check, Exim treated it as
7459 a failure of the random address, and carried on sending RSET and the real
7460 address. If the delay was just some slowness somewhere, the response to the
7461 original RCPT would be taken as a response to RSET and so on, causing
7462 mayhem of various kinds.
7464 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7465 when I implemented it. It didn't allow for the fact that some option values
7466 may legitimately be negative (e.g. size_addition), and it didn't even do
7467 the right test for positive values.
7469 48. Domain names in DNS records are case-independent. Exim always looks them up
7470 in lower case. Some resolvers return domain names in exactly the case they
7471 appear in the zone file, that is, they may contain uppercase letters. Not
7472 all resolvers do this - some return always lower case. Exim was treating a
7473 change of case by a resolver as a change of domain, similar to a widening
7474 of a domain abbreviation. This triggered its re-routing code and so it was
7475 trying to route what was effectively the same domain again. This normally
7476 caused routing to fail (because the router wouldn't handle the domain
7477 twice). Now Exim checks for this case specially, and just changes the
7478 casing of the domain that it ultimately uses when it transmits the message
7481 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7484 50. If a filter generated a file delivery with a non-absolute name (possible if
7485 no home directory exists for the router), the forbid_file option was not
7488 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7489 the existing equality matching.
7491 52. Exim was using ints instead of ino_t variables in some places where it was
7492 dealing with inode numbers.
7494 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7495 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7496 called TMPDIR, and if it finds it is different, it changes its value.
7498 54. The smtp_printf() function is now made available to local_scan() so
7499 additional output lines can be written before returning. There is also an
7500 smtp_fflush() function to enable the detection of a dropped connection.
7501 The variables smtp_input and smtp_batched_input are exported to
7504 55. Changed the default runtime configuration: the message "Unknown user"
7505 has been removed from the ACL, and instead placed on the localuser router,
7506 using the cannot_route_message feature. This means that any verification
7507 failures that generate their own messages won't get overridden. Similarly,
7508 the "Unrouteable address" message that was in the ACL for unverifiable
7509 relay addresses has also been removed.
7511 56. Added hosts_avoid_esmtp to the smtp transport.
7513 57. The exicyclog script was not checking for the esoteric option
7514 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7515 will work only if exicyclog is run under the appropriate euid.
7517 58. Following a discussion on the list, the rules by which Exim recognises line
7518 endings on incoming messages have been changed. The -dropcr and drop_cr
7519 options are now no-ops, retained only for backwards compatibility. The
7520 following line terminators are recognized: LF CRLF CR. However, special
7521 processing applies to CR:
7523 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7524 nor a local message in the state where . is a terminator.
7526 (ii) If a bare CR is encountered in a header line, an extra space is added
7527 after the line terminator so as not to end the header. The reasoning
7528 behind this is that bare CRs in header lines are most likely either
7529 to be mistakes, or people trying to play silly games.
7531 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7532 was being incorrectly given as 18 bytes larger than it should have been.
7533 This is a VOB (very old bug).
7535 60. This may never have affected anything current, but just in case it has:
7536 When the local host is found other than at the start of a list of hosts,
7537 the local host, those with the same MX, and any that follow, are discarded.
7538 When the list in question was part of a longer list of hosts, the following
7539 hosts (not currently being processed) were also being discarded. This no
7540 longer happens. I'm not sure if this situation could ever has previously
7543 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7546 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7547 header. This is something that is recommended in a new Internet Draft, and
7548 is something that is documented as being done by Sendmail. There are two
7549 possible values. For messages generated by the autoreply transport, Exim
7552 Auto-Submitted: auto-replied
7554 whereas for all other generated messages (e.g. bounces) it adds
7556 Auto-Submitted: auto-generated
7558 63. The "personal" condition in filters now includes a test for the
7559 Auto-Submitted: header. If it contains the string "auto-" the message it
7560 not considered personal.
7562 64. Added rcpt_include_affixes as a generic transport option.
7564 65. Added queue_only_override (default true).
7566 66. Added the syslog_duplication option.
7568 67. If what should have been the first header line of a message consisted of
7569 a space followed by a colon, Exim was mis-interpreting it as a header line.
7570 It isn't of course - it is syntactically invalid and should therefore be
7571 treated as the start of the message body. The misbehaviour could have
7572 caused a number of strange effects, including loss of data in subsequent
7573 header lines, and spool format errors.
7575 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7576 client host had authenticated. This control can now be exercised by an ACL
7577 for more flexibility.
7579 69. By default, callouts do not happen when testing with -bh. There is now a
7580 variant, -bhc, which does actually run the callout code, including
7581 consulting and updating the callout cache.
7583 70. Added support for saslauthd authentication, courtesy of Alexander
7586 71. If statvfs() failed on the spool or log directories while checking their
7587 size for availability, Exim confusingly gave the error "space shortage".
7588 Furthermore, in debugging mode it crashed with a floating point exception.
7589 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7590 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7591 serious problem, Exim now writes to the main and panic logs when this
7592 happens, with details of the failure. It then refuses to accept the
7593 incoming message, giving the message "spool directory problem" or "log
7594 directory problem" with a 421 code for SMTP messages.
7596 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7597 0, 1, and 2 exist, because some OS complain for execs without them (see
7598 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7599 descriptors. However, the code omitted to check that the open succeeded,
7600 causing mysterious errors if for some reason the permissions on /dev/null
7601 got screwed. Now Exim writes a message to the main and panic logs, and
7602 bombs out if it can't open /dev/null.
7604 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7605 interact so that it is all more flexible. It is supposed to remain
7606 backwards compatible. Also added extra_local_interfaces.
7608 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7609 to bomb out with an assertion failure - to the client this appears as a
7610 connection drop. This problem occurs in the part of the code that was taken
7611 from the Samba project. Fortunately, the assertion is in a very simple
7612 function, so I have fixed this by reproducing the function inline in the
7613 one place where it is called, and arranging for authentication to fail
7614 instead of killing the process with assert().
7616 75. The SPA client code was not working when the server requested OEM rather
7617 than Unicode encoding.
7619 76. Added code to make require_files with a specific uid setting more usable in
7620 the case where statting the file as root fails - usually a non-root-mounted
7621 NFS file system. When this happens and the failure is EACCES, Exim now
7622 forks a subprocess and does the per-uid checking as the relevant uid.
7624 77. Added process_log_path.
7626 78. If log_file_path was not explicitly set, a setting of check_log_space or
7627 check_log_inodes was ignored.
7629 79. If a space check for the spool or log partitions fails, the incident is now
7630 logged. Of course, in the latter case the data may get lost...
7632 80. Added the %p formatting code to string_format() so that it can be used to
7633 print addresses in debug_print(). Adjusted all the address printing in the
7634 debugging in store.c to use %p rather than %d.
7636 81. There was a concern that a line of code in smtp_in.c could overflow a
7637 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7638 initially expressed, the concern was not well-founded, because trailing
7639 spaces are removed early. However, if the trailing spaces were followed by
7640 a NULL, they did not get removed, so the overflow was possible. Two fixes
7643 (a) I re-wrote the offending code in a cleaner fashion.
7644 (b) If an incoming SMTP command contains a NULL character, it is rejected
7647 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7648 runs initgroups(), so that if the Exim user is in any additional groups,
7649 they will be used during message reception.
7655 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob