1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Avoid costly startup code when not strictly needed. This reduces time
12 for some exim process initialisations. It does mean that the logging
13 of TLS configuration problems is only done for the daemon startup.
15 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
17 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
18 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
21 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
22 DKIM signing is being done. Previously a transport_filter would always
23 disable CHUNKING, falling back to traditional DATA.
25 JH/05 Regard command-line receipients as tainted.
31 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
32 side, to decode a TLS 1.3 packet capture.
34 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
35 Previously the default library behaviour applied, sending two, each in
38 JH/03 Debug output for ACL now gives the config file name and line number for
41 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
43 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
45 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
46 buffer overrun for (non-chunking) other transports.
48 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
49 TLS1.3, means that a server rejecting a client certificate is not visible
50 to the client until the first read of encrypted data (typically the
51 response to EHLO). Add detection for that case and treat it as a failed
52 TLS connection attempt, so that the normal retry-in-clear can work (if
55 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
56 and/or domain. Found and fixed by Jason Betts.
58 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
59 configuration). If a CNAME target was not a wellformed name pattern, a
62 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
63 the OS reports them interleaved with other addresses.
65 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
66 used both for input and for a verify callout, both encrypted, SMTP
67 responses being sent by the server could be lost. This resulted in
68 dropped connections and sometimes bounces generated by a peer sending
71 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
72 string. Previously it was possible to cause undefined behaviour in a
73 library routine (usually a crash). Found by "zerons".
75 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
78 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
79 API was removed, so update to use the newer ones.
81 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
82 any timeout set, is taking a long time. Previously we would hang on to a
83 rotated logfile "forever" if the input was arriving with long gaps
84 (a previous attempt to fix addressed lack, for a long time, of initial
87 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
88 shared (NFS) environment. The length of the tempfile name is now
89 4 + 16 ("hdr.$message_exim_id") which might break on file
90 systems which restrict the file name length to lower values.
93 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
94 shared (NFS) environment.
96 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
97 did for all versions <4.90). Notably -M, -m, --invert, -I may be
100 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
101 on some platforms for bit 31.
103 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
104 to changes apparently associated with TLS1.3 handling some of the APIs
105 previously used were either nonfunctional or inappropriate. Strings
106 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
107 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
108 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
109 This affects log line X= elements, the $tls_{in,out}_cipher variables,
110 and the use of specific cipher names in the encrypted= ACL condition.
112 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
114 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
115 verification result was not updated unless hosts_require_ocsp applied.
117 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
118 queue_list_requires_admin set to false, non-admin users were denied the
121 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
122 directory-of-certs mode. Previously they were advertised despite the
125 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
126 A single TCP connection by a client will now hold a TLS connection open
127 for multiple message deliveries, by default. Previoud the default was to
130 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
131 default. If built with the facility, DANE will be used. The facility
132 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
134 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
135 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
136 must be defined and you must still, unless you define DISABLE_TLS, manage
137 the the include-dir and library-file requirements that go with that
138 choice. Non-TLS builds are still supported.
140 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
143 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
144 default. If the platform supports and has the facility enabled, it will
145 be requested on all coneections.
147 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
148 controlled by the build-time option SUPPORT_PIPE_CONNECT.
150 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
152 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
153 success-DSN messages. Previously the From: header was always the default
154 one for these; the option was ignored.
156 JH/28 Fix the timeout on smtp response to apply to the whole response.
157 Previously it was reset for every read, so a teergrubing peer sending
158 single bytes within the time limit could extend the connection for a
159 long time. Credit to Qualsys Security Advisory Team for the discovery.
161 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
162 delivery address, which leaked information of the results of local
163 forwarding. Change to the original envelope recipient address, per
166 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
167 requested. Previously not bounce was generated and a log entry of
168 error ignored was made.
170 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
172 JH/32 Introduce a general tainting mechanism for values read from the input
173 channel, and values derived from them. Refuse to expand any tainted
174 values, to catch one form of exploit.
176 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
177 was unused and the unexpanded text used for the test. Found and
178 fixed by Ruben Jenster.
180 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
181 an attempt to use a TLS library read routine dereffed a nul pointer,
184 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
185 them in our smtp response.
187 JH/36 Have the general router option retry_use_local_part default to true when
188 any of the restrictive preconditions are set (to anything). Previously it
189 was only for check_local user. The change removes one item of manual
190 configuration which is required for proper retries when a remote router
191 handles a subset of addresses for a domain.
193 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
194 link count into consideration.
196 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
197 caused the extension of big_buffer, the following lines were ignored.
199 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
200 accordance with RFC 2308. Previously there was no expiry, so a longlived
201 receive process (eg. due to ACL delays) versus a short SOA value could
204 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
206 JH/39 Promote DMARC support to mainline.
208 JH/40 Bug 2452: Add a References: header to DSNs.
210 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
211 parameters. The relevant library call is documented as "Deprecated: This
212 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
213 3.6.0, DH parameters are negotiated following RFC7919."
215 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
216 buffer was used for the filename, resulting in a trap when tainted
217 arguments (eg. $domain) were used.
223 JH/01 Remove code calling the customisable local_scan function, unless a new
224 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
226 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
227 non-signal-safe functions being used.
229 JH/03 Bug 2269: When presented with a received message having a stupidly large
230 number of DKIM-Signature headers, disable DKIM verification to avoid
231 a resource-consumption attack. The limit is set at twenty.
233 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
234 report of oldest_pass in ${authres } in consequence, and separate out
235 some descriptions of reasons for verification fail.
237 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
238 files in the spool were present and unlocked. A queue-runner could spot
239 them, resulting in a duplicate delivery. Fix that by doing the unlock
240 after the unlink. Investigation by Tim Stewart. Take the opportunity to
241 add more error-checking on spoolfile handling while that code is being
244 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
245 No known attacks, no CVE, this is defensive hardening.
247 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
248 a queue-runner could start a delivery while other operations were ongoing.
249 Cutthrough delivery was a common victim, resulting in duplicate delivery.
250 Found and investigated by Tim Stewart. Fix by using the open message data
251 file handle rather than opening another, and not locally closing it (which
252 releases a lock) for that case, while creating the temporary .eml format
253 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
255 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
256 $sender_verify_failure/$recipient_verify_failure to "random".
258 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
261 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
262 Previously this would segfault.
264 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
267 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
268 like zero, since the resolver should be doing this for us, But we need one
269 as a CNAME but no MX presence gets the CNAME returned; we need to check
270 that doesn't point to an MX to declare it "no MX returned" rather than
271 "error, loop". A new main option is added so the older capability of
272 following some limited number of chain links is maintained.
274 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
276 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
277 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
278 now-deprecated earlier definitions used only specified the range up to TLS
279 1.2 (in the older-version library docs).
281 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
283 JH/15 Rework TLS client-side context management. Stop using a global, and
284 explicitly pass a context around. This enables future use of TLS for
285 connections to service-daemons (eg. malware scanning) while a client smtp
286 connection is using TLS; with cutthrough connections this is quite likely.
288 JH/16 Fix ARC verification to do AS checks in reverse order.
290 JH/17 Support a "tls" option on the ${readsocket } expansion item.
292 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
293 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
294 Previously the "utf8" would be re-prepended for every additional message.
296 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
297 Previously thery were accepted, resulting in issues when attempting to
298 forward messages to a non-supporting MTA.
300 PP/02 Let -n work with printing macros too, not just options.
302 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
303 one parent address was copied, and bogus data was used at delivery-logging
304 time. Either a crash (after delivery) or bogus log data could result.
305 Discovery and analysis by Tim Stewart.
307 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
308 Previously if the string ended mid-character, we did not insert the
309 promised '?' replacement.
311 PP/04 Documentation: current string operators work on bytes, not codepoints.
313 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
314 should pack well giving a smaller memory footprint so better caching and
315 therefore performance. Group the declarations where this can't be done so
316 that the byte-sized flag variables are not interspersed among pointer
317 variables, giving a better chance of good packing by the compiler.
319 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
320 non-null, to avoid issues with sites running BATV. Previously reports were
321 sent with an empty envelope sender so looked like bounces.
323 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
324 The ignore_error flag wasn't being returned from the filter subprocess so
325 was not set for later routers. Investigation and fix by Matthias Kurz.
327 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
328 and a msg:complete for the whole, when a message is manually removed using
329 -Mrm. Developement by Matthias Kurz, hacked on by JH.
331 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
332 a "Gnu special" function, asprintf() in the DB utility binary builds; I
333 hope that is portable enough.
335 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
336 requiring a known-CA anchor certificate; make it now rely entirely on the
337 TLSA as an anchor. Checking the name on the leaf cert against the name
338 on the A-record for the host is still done for TA (but not for EE mode).
340 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
341 included in delivery lines for non-proxied connections, when compiled with
342 SUPPORT_SOCKS and running with proxy logging enabled.
344 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
345 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
346 move the existing event to fire before the normal logging of message
347 failure so that custom logging is bracketed by normal logging.
349 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
350 msg:fail:internal event. Developement by Matthias Kurz.
352 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
353 far too small for todays use of crypto signatures stored there. Go all
354 the way to the max DNS message size of 64kB, even though this might be
355 overmuch for IOT constrained device use.
357 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
358 copy a string over itself. The library routine is documented as not
359 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
361 JH/32 For main options check_spool_space and check_inode_space, where the
362 platform supports 64b integers, support more than the previous 2^31 kB
363 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
364 the previous G, M, k.
366 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
367 $authenticated_fail_id variable on authentication failure. Previously
370 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
371 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
372 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
373 GNUTLS_SEC_PARAM_MEDIUM.
375 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
376 side. Previously we would continue as if no SNI had been received.
378 JH/36 Harden the handling of string-lists. When a list consisted of a sole
379 "<" character, which should be a list-separator specification, we walked
380 off past the nul-terimation.
382 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
383 causes) even when the retry time is not yet met. Previously they were
384 not, meaning that when (say) an account was over-quota and temp-rejecting,
385 and multiple senders' messages were queued, only one sender would get
386 notified on each configured delay_warning cycle.
388 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
390 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
391 cipher-suites, an error can be left on the stack even for a succeeding
392 accept; this results in impossible error messages when a later operation
395 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
396 return error codes indicating retry. Under TLS1.3 this becomes required.
398 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
399 it only wrote the new authenticators, resulting in a lack of tracking of
400 peer changes of ESMTP extensions until the next cache flush.
402 JH/41 Fix the loop reading a message header line to check for integer overflow,
403 and more-often against header_maxsize. Previously a crafted message could
404 induce a crash of the recive process; now the message is cleanly rejected.
406 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
407 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
413 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
414 When redis_servers is set to a list of > 1 element, and the Redis servers
415 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
416 case of MOVED into a DEFER case instead, thus moving the query onto the
417 next server in the list. For a cluster of N elements, all N servers must
418 be defined in redis_servers.
420 GF/02 Catch and remove uninitialized value warning in exiqsumm
421 Check for existence of @ARGV before looking at $ARGV[0]
423 JH/01 Replace the store_release() internal interface with store_newblock(),
424 which internalises the check required to safely use the old one, plus
425 the allocate and data copy operations duplicated in both (!) of the
426 extant use locations.
428 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
429 modifier. This matches the restriction on the commandline.
431 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
432 Previously only the last row was returned.
434 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
435 we assumed that tags in the header were well-formed, and parsed the
436 element content after inspecting only the first char of the tag.
437 Assumptions at that stage could crash the receive process on malformed
440 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
441 While running the DKIM ACL we operate on the Permanent memory pool so that
442 variables created with "set" persist to the DATA ACL. Also (at any time)
443 DNS lookups that fail create cache records using the Permanent pool. But
444 expansions release any allocations made on the current pool - so a dnsdb
445 lookup expansion done in the DKIM ACL releases the memory used for the
446 DNS negative-cache, and bad things result. Solution is to switch to the
447 Main pool for expansions.
448 While we're in that code, add checks on the DNS cache during store_reset,
449 active in the testsuite.
450 Problem spotted, and debugging aided, by Wolfgang Breyha.
452 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
453 When none of the hosts presented to a transport match an already-open
454 connection, close it and proceed with the list. Previously we would
455 queue the message. Spotted by Lena with Yahoo, probably involving
458 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
459 Previously a spurious "250 OK id=" response was appended to the proper
462 JH/08 The "support for" informational output now, which built with Content
463 Scanning support, has a line for the malware scanner interfaces compiled
464 in. Interface can be individually included or not at build time.
466 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
467 by the template makefile "src/EDITME". The "STREAM" support for an older
468 ClamAV interface method is removed.
470 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
471 rows affected is given instead).
473 JH/11 The runtime Berkeley DB library version is now additionally output by
474 "exim -d -bV". Previously only the compile-time version was shown.
476 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
477 SMTP connection. Previously, when one had more receipients than the
478 first, an abortive onward connection was made. Move to full support for
479 multiple onward connections in sequence, handling cutthrough connection
480 for all multi-message initiating connections.
482 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
483 routers. Previously, a multi-recipient message would fail to match the
484 onward-connection opened for the first recipient, and cause its closure.
486 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
487 a timeout on read on a GnuTLS initiating connection, resulting in the
488 initiating connection being dropped. This mattered most when the callout
489 was marked defer_ok. Fix to keep the two timeout-detection methods
492 JH/15 Relax results from ACL control request to enable cutthrough, in
493 unsupported situations, from error to silently (except under debug)
494 ignoring. This covers use with PRDR, frozen messages, queue-only and
497 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
499 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
500 metadata, resulting in a crash in free().
502 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
503 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
504 Broken also in d185889f4, with init system revamp.
506 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
507 Previously we abruptly closed the connection after reading a malware-
508 found indication; now we go on to read the "scan ok" response line,
511 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
512 ACL. Previously, a crash would result.
514 JH/19 Speed up macro lookups during configuration file read, by skipping non-
515 macro text after a replacement (previously it was only once per line) and
516 by skipping builtin macros when searching for an uppercase lead character.
518 JH/20 DANE support moved from Experimental to mainline. The Makefile control
519 for the build is renamed.
521 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
522 was allocated for every new TLS startup, meaning one per message. Fix
523 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
525 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
526 reported the original. Fix to report (as far as possible) the ACL
527 result replacing the original.
529 JH/23 Fix memory leak during multi-message connections using STARTTLS under
530 OpenSSL. Certificate information is loaded for every new TLS startup,
531 and the resources needed to be freed.
533 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
535 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
538 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
539 DATA response info to the (existing) per-recipient response info for
540 the "C=" log element. It can have useful tracking info from the
541 destination system. Patch from Simon Arlott.
543 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
544 length value. Previously this would segfault.
546 HS/02 Support Avast multiline protoocol, this allows passing flags to
547 newer versions of the scanner.
549 JH/28 Ensure that variables possibly set during message acceptance are marked
550 dead before release of memory in the daemon loop. This stops complaints
551 about them when the debug_store option is enabled. Discovered specifically
552 for sender_rate_period, but applies to a whole set of variables.
553 Do the same for the queue-runner and queue-list loops, for variables set
554 from spool message files. Do the same for the SMTP per-message loop, for
555 certain variables indirectly set in ACL operations.
557 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
558 as a multi-recipient message from a mailinglist manager). The coding had
559 an arbitrary cutoff number of characters while checking for more input;
560 enforced by writing a NUL into the buffer. This corrupted long / fast
561 input. The problem was exposed more widely when more pipelineing of SMTP
562 responses was introduced, and one Exim system was feeding another.
563 The symptom is log complaints of SMTP syntax error (NUL chars) on the
564 receiving system, and refused recipients seen by the sending system
565 (propating to people being dropped from mailing lists).
566 Discovered and pinpointed by David Carter.
568 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
569 replaced by the ${authresults } expansion.
571 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
573 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
574 allows proper process termination in container environments.
576 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
577 Previously the "final dot" had a newline after it; ensure it is CR,LF.
579 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
580 and "err_perm", deprecated since 4.83 when the RFC-defined words
581 "temperror" and "permerror" were introduced.
583 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
584 transport-filters or DKIM-signing. The restriction was lost in the
585 consolidation of verify-callout and delivery SMTP handling.
586 Extend the restriction to also cover ARC-signing.
588 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
589 in defer=pass mode supply a 450 to the initiator. Previously the message
592 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
593 tls_require_ciphers is used as before.
595 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
596 "pass_unscanned". Only tmpfails from the scanner are written to
597 the paniclog, as they may require admin intervention (permission
598 denied, license issues). Other scanner errors (like decompression
599 bombs) do not cause a paniclog entry.
601 JH/36 Fix reinitialisation of DKIM logging variable between messages.
602 Previously it was possible to log spurious information in receive log
605 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
606 triggered odd behaviour from Outlook Express clients.
608 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
611 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
612 since the IETF WG has not yet settled on that versus the original
613 "bare" representation.
615 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
616 Previously the millisecond value corrupted the output.
617 Fix also for syslog_pid=no and log_selector +pid, for which the pid
618 corrupted the output.
624 JH/01 Rework error string handling in TLS interface so that the caller in
625 more cases is responsible for logging. This permits library-sourced
626 string to be attached to addresses during delivery, and collapses
627 pairs of long lines into single ones.
629 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
630 during configuration. Wildcards are allowed and expanded.
632 JH/02 Rework error string handling in DKIM to pass more info back to callers.
633 This permits better logging.
635 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
636 do not close it down and have the child transport start it up again on
637 the passed-on TCP connection. Instead, proxy the child (and any
638 subsequent ones) for TLS via a unix-domain socket channel. Logging is
639 affected: the continued delivery log lines do not have any DNSSEC, TLS
640 Certificate or OCSP information. TLS cipher information is still logged.
642 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
643 identical IP addresses on different listening ports. Will also affect
646 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
647 add noisy ifdef guards to special-case this sillyness.
648 Patch from Bernd Kuhls.
650 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
651 than 255 are no longer allowed.
653 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
654 Disable the session-cache too, which might reduce our load. Since we
655 currrectly use a new context for every connection, both as server and
656 client, there is no benefit for these.
657 GnuTLS appears to not support tickets server-side by default (we don't
658 call gnutls_session_ticket_enable_server()) but client side is enabled
659 by default on recent versions (3.1.3 +) unless the PFS priority string
662 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
663 <https://reproducible-builds.org/specs/source-date-epoch/>.
665 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
666 the check for any unsuccessful recipients did not notice the limit, and
667 erroneously found still-pending ones.
669 JH/08 Pipeline CHUNKING command and data together, on kernels that support
670 MSG_MORE. Only in-clear (not on TLS connections).
672 JH/09 Avoid using a temporary file during transport using dkim. Unless a
673 transport-filter is involved we can buffer the headers in memory for
674 creating the signature, and read the spool data file once for the
675 signature and again for transmission.
677 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
678 4.77 as the kernel support then wasn't solid, having issues in 64bit
679 mode. Now, it's been long enough. Add support for FreeBSD also.
681 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
682 case where the routing stage had gathered several addresses to send to
683 a host before calling the transport for the first, we previously failed
684 to close down TLS in the old transport process before passing the TCP
685 connection to the new process. The new one sent a STARTTLS command
686 which naturally failed, giving a failed delivery and bloating the retry
687 database. Investigation and fix prototype from Wolfgang Breyha.
689 JH/12 Fix check on SMTP command input synchronisation. Previously there were
690 false-negatives in the check that the sender had not preempted a response
691 or prompt from Exim (running as a server), due to that code's lack of
692 awareness of the SMTP input buffering.
694 PP/04 Add commandline_checks_require_admin option.
695 Exim drops privileges sanely, various checks such as -be aren't a
696 security problem, as long as you trust local users with access to their
697 own account. When invoked by services which pass untrusted data to
698 Exim, this might be an issue. Set this option in main configuration
699 AND make fixes to the calling application, such as using `--` to stop
702 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
703 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
704 responses to those, into a single TLS record each way (this usually means
705 a single packet). As a side issue, smtp_enforce_sync now works on TLS
708 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
709 affects you only if you're dancing at the edge of the param size limits.
710 If you are, and this message makes sense to you, then: raise the
711 configured limit or use OpenSSL 1.1. Nothing we can do for older
714 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
715 cmdline element to get the documented default one. Previously it was
718 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
719 are unsuable for non-message-specific reasons - eg. connection timeout,
720 banner-time rejection.
722 JH/16 Fix logging of delivery remote port, when specified by router, under
725 PP/06 Repair manualroute's ability to take options in any order, even if one
726 is the name of a transport.
729 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
731 JH/17 Change the list-building routines interface to use the expanding-string
732 triplet model, for better allocation and copying behaviour.
734 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
735 Previously it was constructed the first time a possibly-matching string
736 was met in the configuration file input during startup; now it is done
739 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
740 compatible one, to avoid the (poorly documented) possibility of a config
741 file in the working directory redirecting the DB files, possibly correpting
742 some existing file. CVE-2017-10140 assigned for BDB.
744 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
745 cache-hot. Previously, although the result was properly cached, the
746 initial verify call returned a defer.
748 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
749 the main verify for receipient in uncached-mode.
751 JH/22 Retire historical build files to an "unsupported" subdir. These are
752 defined as "ones for which we have no current evidence of testing".
754 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
755 if present. Previously it was ignored.
757 JH/24 Start using specified-initialisers in C structure init coding. This is
758 a C99 feature (it's 2017, so now considered safe).
760 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
761 if was a fixed-sized field and bitmask ops via macros; it is now more
764 PP/07 GitHub PR 56: Apply MariaDB build fix.
765 Patch provided by Jaroslav Škarvada.
767 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
768 during Coverity cleanups [4.87 JH/47]
769 Diagnosis and fix provided by Michael Fischer v. Mollard.
771 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
772 the right size to place the terminating semicolon on its own folded
773 line, the header hash was calculated to an incorrect value thanks to
774 the (relaxed) space the fold became.
776 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
777 and confused the parent.
779 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
780 which could crash as a result. This could lead to undeliverable messages.
782 JH/28 Logging: "next input sent too soon" now shows where input was truncated
785 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
786 matters on fast-turnover and PID-randomising systems, which were getting
787 out-of-order delivery.
789 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
790 a possibly-overlapping copy. The symptom was that "Remote host closed
791 connection in response to HELO" was logged instead of the actual 4xx
794 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
795 Previously only that bufferd was discarded, resulting in SYMTP command
798 JH/32 DKIM: when a message has multiple signatures matching an identity given
799 in dkim_verify_signers, run the dkim acl once for each. Previously only
800 one run was done. Bug 2189.
802 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
803 "panic the current process" to "deliberately defer". The panic log is
804 still written with the problem list name; the mail and reject logs now
805 get a temp-reject line for the message that was being handled, saying
806 something like "domains check lookup or other defer". The SMTP 451
807 message is still "Temporary local problem".
809 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
810 A crafted sequence of BDAT commands could result in in-use memory beeing
811 freed. CVE-2017-16943.
813 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
814 from SMTP input. Previously it was always done; now only done for DATA
815 and not BDAT commands. CVE-2017-16944.
817 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
818 to the message (such as an overlong header line). Previously this was
819 not done and we did not exit BDAT mode. Followon from the previous item
820 though a different problem.
826 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
827 than -2003 did; needs libidn2 in addition to libidn.
829 JH/02 The path option on a pipe transport is now expanded before use.
831 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
832 Patch provided by "Björn", documentation fix added too.
834 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
835 missing a wire-to-host endian conversion.
837 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
838 close after a BDAT command line could be taken as a following command,
839 giving a synch failure. Fix by only checking for synch immediately
840 before acknowledging the chunk.
842 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
843 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
844 macro. Patches provided by Josh Soref.
846 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
847 Previously we did not; the RFC seems ambiguous and VRFY is not listed
848 by IANA as a service extension. However, John Klensin suggests that we
851 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
852 the dkim code may be unix-mode line endings rather than smtp wire-format
853 CRLF, so prepend a CR to any bare LF.
855 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
856 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
858 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
859 the first were themselves being wrongly included in the feed into dkim
860 processing; with most chunk sizes in use this resulted in an incorrect
861 body hash calculated value.
863 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
864 DKIM signature block, for verification. Although advised against by
865 standards it is specifically not ruled illegal.
867 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
869 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
870 missing a body hash (the bh= tag).
872 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
873 It seems that HAProxy sends the Proxy Protocol information in clear and
874 only then does a TLS startup, so do the same.
876 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
877 TCP connections (such as for Spamd) unless the daemon successfully set
878 Fast Open mode on its listening sockets. This fixes breakage seen on
879 too-old kernels or those not configured for Fast Open, at the cost of
880 requiring both directions being enabled for TFO, and TFO never being used
881 by non-daemon-related Exim processes.
883 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
884 endings, at least on the first header line. Try to canonify any that get
885 past that check, despite the cost.
887 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
888 now limited to an arbitrary five deep, while parsing addresses with the
889 strip_excess_angle_brackets option enabled.
891 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
892 instead leave the unprompted TLS handshake in socket buffer for the
893 TLS library to consume.
895 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
897 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
899 JH/16 Drop variables when they go out of scope. Memory management drops a whole
900 region in one operation, for speed, and this leaves assigned pointers
901 dangling. Add checks run only under the testsuite which checks all
902 variables at a store-reset and panics on a dangling pointer; add code
903 explicitly nulling out all the variables discovered. Fixes one known
904 bug: a transport crash, where a dangling pointer for $sending_ip_address
905 originally assigned in a verify callout, is re-used.
907 PP/06 Drop '.' from @INC in various Perl scripts.
909 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
911 PP/08 Reduce a number of compilation warnings under clang; building with
912 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
913 should be warning-free.
915 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
917 HS/01 Fix portability problems introduced by PP/08 for platforms where
918 realloc(NULL) is not equivalent to malloc() [SunOS et al].
920 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
921 chunk. This allows us to accept broken chunked messages. We need a more
922 general solution here.
924 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
925 already-broken messages in the queue.
927 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
929 JH/19 Fix reference counting bug in routing-generated-address tracking.
935 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
936 supports it and a size is available (ie. the sending peer gave us one).
938 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
939 years of being deprecated). The replacements are the ACLs
940 acl_smtp_mime and acl_not_smtp_mime.
942 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
943 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
944 or even an in-clear connection were permitted. Now, if the host lookup
945 was dnssec and dane was requested then the host is only used if the
946 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
947 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
948 if one fails this test.
949 This means that a poorly-configured remote DNS will make it incommunicado;
950 but it protects against a DNS-interception attack on it.
952 JH/04 Bug 1810: make continued-use of an open smtp transport connection
953 non-noisy when a race steals the message being considered.
955 JH/05 If main configuration option tls_certificate is unset, generate a
956 self-signed certificate for inbound TLS connections.
958 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
959 in rewrites and routers.
961 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
962 and logged a warning sing 4.83; now they are a configuration file error.
964 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
965 (lacking @domain). Apply the same qualification processing as RCPT.
967 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
969 JH/10 Support ${sha256:} applied to a string (as well as the previous
972 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
973 a cutthrough deliver is pending, as we always want to make a connection.
974 This also avoids re-routing the message when later placing the cutthrough
975 connection after a verify cache hit.
976 Do not update it with the verify result either.
978 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
979 when routing results in more than one destination address.
981 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
982 signing (which inhibits the cutthrough capability). Previously only
983 the presence of an option was tested; now an expansion evaluating as
984 empty is permissible (obviously it should depend only on data available
985 when the cutthrough connection is made).
987 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
988 the relevant preceding SMTP command did not note the pipelining mode.
990 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
991 Previously they were not counted.
993 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
994 as one having no matching records. Previously we deferred the message
995 that needed the lookup.
997 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
998 distinguished as "(=".
1000 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1001 for missing MX records. Previously it only worked for missing A records.
1003 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1005 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1006 after the data-go-ahead and data-ack. Patch from Jason Betts.
1008 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1009 even for a "none" policy. Patch from Tony Meyer.
1011 JH/22 Fix continued use of a connection for further deliveries. If a port was
1012 specified by a router, it must also match for the delivery to be
1015 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1016 When one of the recipients of a message was unsuitable for the connection
1017 (has no matching addresses), we lost track of needing to mark it
1018 deferred. As a result mail would be lost.
1020 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1022 JH/25 Decoding ACL controls is now done using a binary search; the source code
1023 takes up less space and should be simpler to maintain. Merge the ACL
1024 condition decode tables also, with similar effect.
1026 JH/26 Fix problem with one_time used on a redirect router which returned the
1027 parent address unchanged. A retry would see the parent address marked as
1028 delivered, so not attempt the (identical) child. As a result mail would
1031 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1032 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1033 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1036 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1037 Main config options check_{log,spool}_{inodes,space} are now
1038 100 inodes, 10MB unless set otherwise in the configuration.
1040 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1041 Previously it only applied to the main-section connection policy
1044 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1046 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1047 by me. Added RFC7919 DH primes as an alternative.
1049 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1050 are not in the system include path.
1052 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1053 GnuTLS, when a session startup failed (eg because the client disconnected)
1054 Exim did stdio operations after fclose. This was exposed by a recent
1055 change which nulled out the file handle after the fclose.
1057 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1058 signed directly by the cert-signing cert, rather than an intermediate
1059 OCSP-signing cert. This is the model used by LetsEncrypt.
1061 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1063 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1064 an incoming connection.
1066 HS/02 Bug 1802: Do not half-close the connection after sending a request
1069 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1070 fallback to "prime256v1".
1072 JH/34 SECURITY: Use proper copy of DATA command in error message.
1073 Could leak key material. Remotely exploitable. CVE-2016-9963.
1079 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1080 and 3.4.4 - once the server is enabled to respond to an OCSP request
1081 it does even when not requested, resulting in a stapling non-aware
1082 client dropping the TLS connection.
1084 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1085 support variable-length bit vectors. No functional change.
1087 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1088 The I= interface field on outgoing lines is now after the H= remote
1089 host field, same as incoming lines. There is a separate
1090 outgoing_interface log selector which allows you to disable the
1093 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1094 If not running log_selector +smtp_connection the mainlog would be held
1095 open indefinitely after a "too many connections" event, including to a
1096 deleted file after a log rotate. Leave the per net connection logging
1097 leaving it open for efficiency as that will be quickly detected by the
1098 check on the next write.
1100 HS/01 Bug 1671: Fix post transport crash.
1101 Processing the wait-<transport> messages could crash the delivery
1102 process if the message IDs didn't exist for some reason. When
1103 using 'split_spool_directory=yes' the construction of the spool
1104 file name failed already, exposing the same netto behaviour.
1106 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1107 mime_regex ACL conditions.
1109 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1110 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1111 to HELO, local diagnostic string.
1113 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1114 log line to debug. Even when configured with a tls authenticator many
1115 client connections are expected to not authenticate in this way, so
1116 an authenticate fail is not an error.
1118 HS/02 Add the Exim version string to the process info. This way exiwhat
1119 gives some more detail about the running daemon.
1121 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1122 matter for fast-change records such as DNSBLs.
1124 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1125 retry record, even if constant. There may be multiple transports with
1126 different interface settings and the retry behaviour needs to be kept
1129 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1131 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1133 JH/10 Bug 840: fix log_defer_output option of pipe transport
1135 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1136 hosts_randomize. This matters a lot when combined with mua_wrapper.
1138 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1139 ${quote_pgsql:<string>} operator.
1141 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1143 JH/14 Change header file name for freeradius-client. Relevant if compiling
1144 with Radius support; from the Gentoo tree and checked under Fedora.
1146 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1148 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1149 option transport_filter, meaning no filtering.
1151 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1153 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1154 defaults to "*" (all hosts). The variable is now available when not built
1155 with TLS, default unset, mainly to enable keeping the testsuite sane.
1156 If a server certificate is not supplied (via tls_certificate) an error is
1157 logged, and clients will find TLS connections fail on startup. Presumably
1158 they will retry in-clear.
1159 Packagers of Exim are strongly encouraged to create a server certificate
1160 at installation time.
1162 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1163 with the $config_file variable.
1165 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1166 in transport context, after the attempt, and per-recipient. The latter type
1167 is per host attempted. The event data is the error message, and the errno
1168 information encodes the lookup type (A vs. MX) used for the (first) host,
1169 and the trailing two digits of the smtp 4xx response.
1171 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1172 to write to mainlog (or rejectlog, paniclog) in the window between file
1173 creation and permissions/ownership being changed. Particularly affects
1174 installations where exicyclog is run as root, rather than exim user;
1175 result is that the running daemon panics and dies.
1177 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1179 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1180 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1181 "pri" and "weight". Note that the previous implicit priority given by the
1182 list order is no longer honoured.
1184 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1185 for DKIM processing.
1187 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1188 by defining SUPPORT_SOCKS.
1190 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1191 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1192 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1193 variables are renamed to proxy_{local,external}_{address,port}.
1195 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1196 for a build by defining SUPPORT_I18N
1198 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1199 of the query string, and make ${quote_redis:} do that quoting.
1201 JH/29 Move Events support from Experimental to mainline, enabled by default
1202 and removable for a build by defining DISABLE_EVENT.
1204 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1206 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1207 cached by the daemon.
1209 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1210 by defining LOOKUP_REDIS. The libhiredis library is required.
1212 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1213 keys are given for lookup.
1215 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1216 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1217 only supported when built with TLS support. The PolarSSL SHA routines
1218 are still used when the TLS library is too old for convenient support.
1220 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1221 openssl_options), for security. OpenSSL forces this from version 1.1.0
1222 server-side so match that on older versions.
1224 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1225 allocation for $value could be released as the expansion processing
1226 concluded, but leaving the global pointer active for it.
1228 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1229 and to use the domains and local_parts ACL conditions.
1231 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1232 incorrectly not doubled on cutthrough transmission, hence seen as a
1233 body-termination at the receiving system - resulting in truncated mails.
1234 Commonly the sender saw a TCP-level error, and retransmitted the message
1235 via the normal store-and-forward channel. This could result in duplicates
1236 received - but deduplicating mailstores were liable to retain only the
1237 initial truncated version.
1239 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1241 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1243 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1244 we're in there, support oversigning also; bug 1309.
1246 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1248 HS/04 Add support for keep_environment and add_environment options.
1250 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1251 either intentional arithmetic overflow during PRNG, or testing config-
1254 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1255 delivery resulted in actual delivery. Cancel cutthrough before DATA
1258 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1259 rejects a recipient: pass the reject to the originator.
1261 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1262 Many were false-positives and ignorable, but it's worth fixing the
1265 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1266 for the new environment-manipulation done at startup. Move the routines
1267 from being local to tls.c to being global via the os.c file.
1269 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1270 an extract embedded as result-arg for a map, the first arg for extract
1271 is unavailable so we cannot tell if this is a numbered or keyed
1272 extraction. Accept either.
1278 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1281 JH/02 The smtp transport option "multi_domain" is now expanded.
1283 JH/03 The smtp transport now requests PRDR by default, if the server offers
1286 JH/04 Certificate name checking on server certificates, when exim is a client,
1287 is now done by default. The transport option tls_verify_cert_hostnames
1288 can be used to disable this per-host. The build option
1289 EXPERIMENTAL_CERTNAMES is withdrawn.
1291 JH/05 The value of the tls_verify_certificates smtp transport and main options
1292 default to the word "system" to access the system default CA bundle.
1293 For GnuTLS, only version 3.0.20 or later.
1295 JH/06 Verification of the server certificate for a TLS connection is now tried
1296 (but not required) by default. The verification status is now logged by
1297 default, for both outbound TLS and client-certificate supplying inbound
1300 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1303 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1304 Status Notification (bounce) messages are now MIME format per RFC 3464.
1305 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1306 under the control of the dsn_advertise_hosts option, and routers may
1307 have a dsn_lasthop option.
1309 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1310 default, modifiable by a malware= option. The list separator for
1311 the options can now be changed in the usual way. Bug 68.
1313 JH/10 The smtp_receive_timeout main option is now expanded before use.
1315 JH/11 The incoming_interface log option now also enables logging of the
1316 local interface on delivery outgoing connections.
1318 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1319 if the interface and destination host and port all match.
1321 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1324 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1325 Patch from Andrew Lewis.
1327 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1328 now supports optional time-restrictions, weighting, and priority
1329 modifiers per server. Patch originally by <rommer@active.by>.
1331 JH/16 The spamd_address main option now supports a mixed list of local
1332 and remote servers. Remote servers can be IPv6 addresses, and
1333 specify a port-range.
1335 JH/17 Bug 68: The spamd_address main option now supports an optional
1336 timeout value per server.
1338 JH/18 Bug 1581: Router and transport options headers_add/remove can
1339 now have the list separator specified.
1341 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1344 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1347 JH/21 Support for the A6 type of dns record is withdrawn.
1349 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1350 rather than the verbs used.
1352 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1353 from 255 to 1024 chars.
1355 JH/24 Verification callouts now attempt to use TLS by default.
1357 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1358 are generic router options now. The defaults didn't change.
1360 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1361 Original patch from Alexander Shikoff, worked over by JH.
1363 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1364 files if ZCAT_COMMAND is not executable.
1366 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1368 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1370 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1371 Normally benign, it bites when the pair was led to by a CNAME;
1372 modern usage is to not canonicalize the domain to a CNAME target
1373 (and we were inconsistent anyway for A-only vs AAAA+A).
1375 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1377 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1378 when evaluating $sender_host_dnssec.
1380 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1381 $sender_helo_dnssec variable.
1383 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1385 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1387 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1389 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1390 documented as working, but never had. Support all but $spam_report.
1392 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1393 added for tls authenticator.
1395 HS/03 Add perl_taintmode main config option
1401 TL/01 When running the test suite, the README says that variables such as
1402 no_msglog_check are global and can be placed anywhere in a specific
1403 test's script, however it was observed that placement needed to be near
1404 the beginning for it to behave that way. Changed the runtest perl
1405 script to read through the entire script once to detect and set these
1406 variables, reset to the beginning of the script, and then run through
1407 the script parsing/test process like normal.
1409 TL/02 The BSD's have an arc4random API. One of the functions to induce
1410 adding randomness was arc4random_stir(), but it has been removed in
1411 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1412 function when detected.
1414 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1415 cause callback expansion.
1417 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1418 syntax errors in an expansion can be treated as a string instead of
1419 logging or causing an error, due to the internal use of bool_lax
1420 instead of bool when processing it.
1422 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1423 server certificates when making smtp deliveries.
1425 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1427 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1429 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1431 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1432 Merged patch from Sebastian Wiedenroth.
1434 JH/05 Fix results-pipe from transport process. Several recipients, combined
1435 with certificate use, exposed issues where response data items split
1436 over buffer boundaries were not parsed properly. This eventually
1437 resulted in duplicates being sent. This issue only became common enough
1438 to notice due to the introduction of connection certificate information,
1439 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1441 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1442 size buffer was used, resulting in syntax errors when an expansion
1445 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1446 version 3.3.6 or later.
1448 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1449 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1450 both become "event_action", the variables become $event_name, $event_data
1451 and $event_defer_errno. There is a new variable $verify_mode, usable in
1452 routers, transports and related events. The tls:cert event is now also
1453 raised for inbound connections, if the main configuration event_action
1456 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1457 early OCSP support, but no stapling (appears to be less than 1.0.0).
1459 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1460 server certificate names available under the smtp transport option
1461 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1464 JH/10 Time-related extraction expansions from certificates now use the main
1465 option "timezone" setting for output formatting, and are consistent
1466 between OpenSSL and GnuTLS compilations. Bug 1541.
1468 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1469 encoded parameter in the incoming message. Bug 1558.
1471 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1472 include certificate info, eximon was claiming there were spoolfile
1475 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1477 JH/14 Log delivery-related information more consistently, using the sequence
1478 "H=<name> [<ip>]" wherever possible.
1480 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1481 are problematic for Debian distribution, omit them from the release
1484 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1486 JH/16 Fix string representation of time values on 64bit time_t architectures.
1489 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1490 output list separator was used.
1495 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1496 checkers that were complaining about end of non-void function with no
1499 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1500 This was a regression introduced in 4.83 by another bugfix.
1502 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1504 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1505 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1511 TF/01 Correctly close the server side of TLS when forking for delivery.
1513 When a message was received over SMTP with TLS, Exim failed to clear up
1514 the incoming connection properly after forking off the child process to
1515 deliver the message. In some situations the subsequent outgoing
1516 delivery connection happened to have the same fd number as the incoming
1517 connection previously had. Exim would try to use TLS and fail, logging
1518 a "Bad file descriptor" error.
1520 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1521 utilities have not been installed.
1523 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1524 temporary space as the ACL may create new global variables.
1526 TL/01 LDAP support uses per connection or global context settings, depending
1527 upon the detected version of the libraries at build time.
1529 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1530 to extract and use the src ip:port in logging and expansions as if it
1531 were a direct connection from the outside internet. PPv2 support was
1532 updated based on HAProxy spec change in May 2014.
1534 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1536 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1537 Properly escape header and check for NULL return.
1539 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1542 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1544 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1545 characters in header names, implemented as a verify condition.
1546 Contributed by Michael Fischer v. Mollard.
1548 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1549 results permerror and temperror. Previous values are deprecated but
1550 still accepted. In a future release, err_perm and err_temp will be
1551 completely removed, which will be a backward incompatibility if the
1552 ACL tests for either of these two old results. Patch contributed by
1553 user bes-internal on the mailing list.
1555 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1557 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1558 selectors, in both main and reject logs.
1560 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1563 JH/07 Add malware type "sock" for talking to simple daemon.
1565 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1567 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1568 routers/transports under cutthrough routing.
1570 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1571 numbers. Touch up "bool" conditional to keep the same definition.
1573 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1575 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1577 JH/12 Expand items in router/transport headers_add or headers_remove lists
1578 individually rather than the list as a whole. Bug 1452.
1580 Required for reasonable handling of multiple headers_ options when
1581 they may be empty; requires that headers_remove items with embedded
1582 colons must have them doubled (or the list-separator changed).
1584 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1585 view the policy declared in the DMARC record. Currently, $dmarc_status
1586 is a combined value of both the record presence and the result of the
1589 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1591 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1592 dnslookup router and the smtp transport (applying to the forward
1595 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1596 of ldap servers used for a specific lookup. Patch provided by Heiko
1599 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1600 New variable $lookup_dnssec_authenticated for observability.
1602 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1603 Patch submitted by Lars Timman.
1605 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1607 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1608 Requires trusted mode and valid format message id, aborts otherwise.
1609 Patch contributed by Heiko Schlichting.
1611 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1612 certextract with support for various fields. Bug 1358.
1614 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1615 is requested by default, modifiable by smtp transport option
1618 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1619 operate on certificate variables to give certificate fingerprints
1620 Also new ${sha256:cert_variable}.
1622 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1624 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1625 Christian Aistleitner.
1627 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1629 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1630 file. Patch from Wolfgang Breyha.
1632 JH/25 Expand the coverage of the delivery $host and $host_address to
1633 client authenticators run in verify callout. Bug 1476.
1635 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1636 align with daemon_smtp_ports. Bug 72.
1638 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1639 support and error reporting did not work properly.
1641 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1642 and is readable. Patch from Andrew Colin Kissa.
1644 TL/14 Enhance documentation of ${run expansion and how it parses the
1645 commandline after expansion, particularly in the case when an
1646 unquoted variable expansion results in an empty value.
1648 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1650 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1651 by renaming away from it. Addresses GH issue 15, reported by
1654 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1656 TL/15 SECURITY: prevent double expansion in math comparison functions
1657 (can expand unsanitized data). Not remotely exploitable.
1664 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1666 PP/02 Make -n do something, by making it not do something.
1667 When combined with -bP, the name of an option is not output.
1669 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1672 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1673 $sender_host_name and config options to manage this, and basic check
1676 PP/05 DSCP support for outbound connections and control modifier for inbound.
1678 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1679 (Only plugin which currently uses this is kerberos4, which nobody should
1680 be using, but we should make it available and other future plugins might
1681 conceivably use it, even though it would break NAT; stuff *should* be
1682 using channel bindings instead).
1684 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1685 name; added for Sendmail compatibility; requires admin caller.
1686 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1687 just ignore it); requires trusted caller.
1688 Also parse but ignore: -Ac -Am -X<logfile>
1691 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1693 TL/02 Add +smtp_confirmation as a default logging option.
1695 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1696 Patch by Magnus Holmgren from 2007-02-20.
1698 TL/04 Bugzilla 1281 - Spec typo.
1699 Bugzilla 1283 - Spec typo.
1700 Bugzilla 1290 - Spec grammar fixes.
1702 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1704 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1706 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1707 mailing list by by Dmitry Isaikin.
1709 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1711 JH/02 Support "G" suffix to numbers in ${if comparisons.
1713 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1715 NM/01 Bugzilla 1197 - Spec typo
1716 Bugzilla 1196 - Spec examples corrections
1718 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1720 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1721 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1724 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1725 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1727 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1728 "acl {{name}{arg}...}", and optional args on acl condition
1731 JH/05 Permit multiple router/transport headers_add/remove lines.
1733 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1735 JH/07 Avoid using a waiting database for a single-message-only transport.
1736 Performance patch from Paul Fisher. Bugzilla 1262.
1738 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1741 JH/09 Add $headers_added variable, with content from use of ACL modifier
1742 add_header (but not yet added to the message). Bugzilla 199.
1744 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1745 Pulled from Bugzilla 817 by Wolfgang Breyha.
1747 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1749 (nb: this is the same fix as in Exim 4.80.1)
1751 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1754 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1757 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1759 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1760 advertise SMTP AUTH mechanism to us, instead of a generic
1761 protocol violation error. Also, make Exim more robust to bad
1762 data from the Dovecot auth socket.
1764 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1766 When a queue runner is handling a message, Exim first routes the
1767 recipient addresses, during which it prunes them based on the retry
1768 hints database. After that it attempts to deliver the message to
1769 any remaining recipients. It then updates the hints database using
1772 So if a recipient address works intermittently, it can get repeatedly
1773 deferred at routing time. The retry hints record remains fresh so the
1774 address never reaches the final cutoff time.
1776 This is a fairly common occurrence when a user is bumping up against
1777 their storage quota. Exim had some logic in its local delivery code
1778 to deal with this. However it did not apply to per-recipient defers
1779 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1781 This change adds a proper retry rule check during routing so that the
1782 final cutoff time is checked against the message's age. We only do
1783 this check if there is an address retry record and there is not a
1784 domain retry record; this implies that previous attempts to handle
1785 the address had the retry_use_local_parts option turned on. We use
1786 this as an approximation for the destination being like a local
1787 delivery, as in LMTP.
1789 I suspect this new check makes the old local delivery cutoff check
1790 redundant, but I have not verified this so I left the code in place.
1792 TF/02 Correct gecos expansion when From: is a prefix of the username.
1794 Test 0254 submits a message to Exim with the header
1798 When I ran the test suite under the user fanf2, Exim expanded
1799 the header to contain my full name, whereas it should have added
1800 a Resent-Sender: header. It erroneously treats any prefix of the
1801 username as equal to the username.
1803 This change corrects that bug.
1805 GF/01 DCC debug and logging tidyup
1806 Error conditions log to paniclog rather than rejectlog.
1807 Debug lines prefixed by "DCC: " to remove any ambiguity.
1809 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1811 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1812 Bug spotted by Jeremy Harris; was flawed since initial commit.
1813 Would have resulted in OCSP responses post-SNI triggering an Exim
1814 NULL dereference and crash.
1816 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1818 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1819 Bug detection, analysis and fix by Samuel Thibault.
1820 Bugzilla 1331, Debian bug #698092.
1822 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1824 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1825 Server implementation by Todd Lyons, client by JH.
1826 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1827 config variable "prdr_enable" controls whether the server
1828 advertises the facility. If the client requests PRDR a new
1829 acl_data_smtp_prdr ACL is called once for each recipient, after
1830 the body content is received and before the acl_smtp_data ACL.
1831 The client is controlled by both of: a hosts_try_prdr option
1832 on the smtp transport, and the server advertisement.
1833 Default client logging of deliveries and rejections involving
1834 PRDR are flagged with the string "PRDR".
1836 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1837 fclose(). Diagnosis by Todd Lyons.
1839 PP/17 Update configure.default to handle IPv6 localhost better.
1840 Patch by Alain Williams (plus minor tweaks).
1843 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1844 This is now consistent with GnuTLS, and is now documented: the
1845 previous undocumented portable approach to treating the option as
1846 unset was to force an expansion failure. That still works, and
1847 an empty string is now equivalent.
1849 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1850 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1851 not performing validation itself.
1853 PP/20 Added force_command boolean option to pipe transport.
1854 Patch from Nick Koston, of cPanel Inc.
1856 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1859 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1861 PP/21 Fix eximon continuous updating with timestamped log-files.
1862 Broken in a format-string cleanup in 4.80, missed when I repaired the
1863 other false fix of the same issue.
1864 Report and fix from Heiko Schlichting.
1867 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1868 Report from Prashanth Katuri.
1870 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1871 It's SecureTransport, so affects any MacOS clients which use the
1872 system-integrated TLS libraries, including email clients.
1874 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1875 using a MIME ACL for non-SMTP local injection.
1876 Report and assistance in diagnosis by Warren Baker.
1878 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1880 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1882 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1883 last id that failed so it may be referenced in subsequent ACL's.
1885 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1888 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1889 ldap library initialization, allowing self-signed CA's to be
1890 used. Also properly sets require_cert option later in code by
1891 using NULL (global ldap config) instead of ldap handle (per
1892 session). Bug diagnosis and testing by alxgomz.
1894 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1895 the src/util/ subdirectory.
1897 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1898 renamed to Transport Post Delivery Action by Jeremy Harris, as
1901 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1902 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1903 redis_servers = needs to be configured which will be used by the redis
1904 lookup. Patch from Warren Baker, of The Packet Hub.
1906 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1908 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1909 hostname or reverse DNS when processing a host list. Used suggestions
1910 from multiple comments on this bug.
1912 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1914 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1915 Missed a few lines, added it to make the runtest require no keyboard
1918 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1919 contains upper case chars. Make router use caseful_local_part.
1921 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1922 support when GnuTLS has been built with p11-kit.
1928 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1930 This, or similar/improved, will also be change PP/11 of 4.82.
1936 PP/01 Handle short writes when writing local log-files.
1937 In practice, only affects FreeBSD (8 onwards).
1938 Bugzilla 1053, with thanks to Dmitry Isaikin.
1940 NM/01 Bugzilla 949 - Documentation tweak
1942 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1945 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1947 PP/02 Implemented gsasl authenticator.
1949 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1951 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1952 `pkg-config foo` for cflags/libs.
1954 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1955 with rest of GSASL and with heimdal_gssapi.
1957 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1958 `pkg-config foo` for cflags/libs for the TLS implementation.
1960 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1961 properties get this fed in as external SSF. A number of robustness
1962 and debugging improvements to the cyrus_sasl authenticator.
1964 PP/08 cyrus_sasl server now expands the server_realm option.
1966 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1967 Patch by Jeremy Harris.
1969 PP/10 Added dbmjz lookup type.
1971 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1973 PP/12 MAIL args handles TAB as well as SP, for better interop with
1974 non-compliant senders.
1975 Analysis and variant patch by Todd Lyons.
1977 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1978 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1979 Patch from Dirk Mueller <dmueller@suse.com>
1981 PP/13 tls_peerdn now print-escaped for spool files.
1982 Observed some $tls_peerdn in wild which contained \n, which resulted
1983 in spool file corruption.
1985 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1986 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1987 or write after TLS renegotiation, which otherwise led to messages
1990 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1991 as a tracking header (ie: a signed header comes before the signature).
1992 Patch from Wolfgang Breyha.
1994 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1995 comma-sep list; embedded commas doubled.
1997 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1999 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2001 Report and patch from Dmitry Banschikov.
2003 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2004 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2005 failures. We appear to now support TLS1.1+ with Exim.
2007 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2008 lets Exim select keys and certificates based upon TLS SNI from client.
2009 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2010 before an outbound SMTP session. New log_selector, +tls_sni.
2012 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2013 NULL dereference. Report and patch from Alun Jones.
2015 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2016 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2017 Not seeing resolver debug output on NetBSD, but suspect this is a
2018 resolver implementation change.
2020 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2021 Left warnings. Added "eximon gdb" invocation mode.
2023 PP/21 Defaulting "accept_8bitmime" to true, not false.
2025 PP/22 Added -bw for inetd wait mode support.
2027 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2028 locate the relevant includes and libraries. Made this the default.
2030 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2031 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2033 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2034 This may cause build issues on older platforms.
2036 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2037 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2038 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2039 Added SNI support via GnuTLS too.
2040 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2042 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2044 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2045 Applied second patch from Janne, implementing suggestion to default
2046 multiple-strings-in-record handling to match SPF spec.
2048 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2050 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2051 read-only, out of scope).
2052 Patch from Wolfgang Breyha, report from Stuart Northfield.
2054 PP/29 Fix three issues highlighted by clang analyser static analysis.
2055 Only crash-plausible issue would require the Cambridge-specific
2056 iplookup router and a misconfiguration.
2057 Report from Marcin Mirosław.
2059 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2061 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2062 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2063 As part of this, removing so much warning spew let me fix some minor
2064 real issues in debug logging.
2066 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2067 assignment on my part. Fixed.
2069 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2070 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2071 Janne Snabb (who went above and beyond: thank you).
2073 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2074 string otherwise requires a connection and a bunch more work and it's
2075 relatively easy to get wrong. Should also expose TLS library linkage
2078 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2079 64-bit ${eval} (JH/03).
2081 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2082 GNU libc to support some of the 64-bit stuff, should not lead to
2083 conflicts. Defined before os.h is pulled in, so if a given platform
2084 needs to override this, it can.
2086 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2087 protection layer was required, which is not implemented.
2088 Bugzilla 1254, patch from Wolfgang Breyha.
2090 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2091 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2092 tls_dhparam take prime identifiers. Also unbreak combination of
2093 OpenSSL+DH_params+TLSSNI.
2095 PP/39 Disable SSLv2 by default in OpenSSL support.
2101 PP/01 Solaris build fix for Oracle's LDAP libraries.
2102 Bugzilla 1109, patch from Stephen Usher.
2104 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2106 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2109 TF/02 Fix a couple more cases where we did not log the error message
2110 when unlink() failed. See also change 4.74-TF/03.
2112 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2113 lock up or crash if it happened to be inside a call to libc when it
2114 got a SIGUSR1 from exiwhat.
2116 The SIGUSR1 handler appends the current process status to the process
2117 log which is later printed by exiwhat. It used to use the general
2118 purpose logging code to do this, but several functions it calls are
2119 not safe for signals.
2121 The new output code in the SIGUSR1 handler is specific to the process
2122 log, and simple enough that it's easy to inspect for signal safety.
2123 Removing some special cases also simplifies the general logging code.
2124 Removing the spurious timestamps from the process log simplifies
2127 TF/04 Improved ratelimit ACL condition.
2129 The /noupdate option has been deprecated in favour of /readonly which
2130 has clearer semantics. The /leaky, /strict, and /readonly update modes
2131 are mutually exclusive. The update mode is no longer included in the
2132 database key; it just determines when the database is updated. (This
2133 means that when you upgrade Exim will forget old rate measurements.)
2135 Exim now checks that the per_* options are used with an update mode that
2136 makes sense for the current ACL. For example, when Exim is processing a
2137 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2138 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2139 must specify per_mail/readonly. If you omit the update mode it defaults to
2140 /leaky where that makes sense (as before) or /readonly where required.
2142 The /noupdate option is now undocumented but still supported for
2143 backwards compatibility. It is equivalent to /readonly except that in
2144 ACLs where /readonly is required you may specify /leaky/noupdate or
2145 /strict/noupdate which are treated the same as /readonly.
2147 A useful new feature is the /count= option. This is a generalization
2148 of the per_byte option, so that you can measure the throughput of other
2149 aggregate values. For example, the per_byte option is now equivalent
2150 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2152 The per_rcpt option has been generalized using the /count= mechanism
2153 (though it's more complicated than the per_byte equivalence). When it is
2154 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2155 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2156 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2157 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2158 Note that using per_rcpt with a non-readonly update mode in more than
2159 one ACL will cause the recipients to be double-counted. (The per_mail
2160 and per_byte options don't have this problem.)
2162 The handling of very low rates has changed slightly. If the computed rate
2163 is less than the event's count (usually one) then this event is the first
2164 after a long gap. In this case the rate is set to the same as this event's
2165 count, so that the first message of a spam run is counted properly.
2167 The major new feature is a mechanism for counting the rate of unique
2168 events. The new per_addr option counts the number of different
2169 recipients that someone has sent messages to in the last time period. It
2170 behaves like per_rcpt if all the recipient addresses are different, but
2171 duplicate recipient addresses do not increase the measured rate. Like
2172 the /count= option this is a general mechanism, so the per_addr option
2173 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2174 example, measure the rate that a client uses different sender addresses
2175 with the options per_mail/unique=$sender_address. There are further
2176 details in the main documentation.
2178 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2180 TF/06 Removed a few PCRE remnants.
2182 TF/07 Automatically extract Exim's version number from tags in the git
2183 repository when doing development or release builds.
2185 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2186 Bugzilla 879. Patch from Paul Fisher.
2188 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2189 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2192 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2194 PP/05 Variable $av_failed, true if the AV scanner deferred.
2195 Bugzilla 1078. Patch from John Horne.
2197 PP/06 Stop make process more reliably on build failure.
2198 Bugzilla 1087. Patch from Heiko Schlittermann.
2200 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2201 Bugzilla 1089. Patch from Heiko Schlittermann.
2203 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2204 Bugzilla 1131. Patch from Holger Weiß.
2206 PP/09 Handle IPv6 addresses with SPF.
2207 Bugzilla 860. Patch from Wolfgang Breyha.
2209 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2211 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2214 PP/11 match_* no longer expand right-hand-side by default.
2215 New compile-time build option, EXPAND_LISTMATCH_RHS.
2216 New expansion conditions, "inlist", "inlisti".
2218 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2220 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2222 PP/14 fix log_write() format string regression from TF/03.
2223 Bugzilla 1152. Patch from Dmitry Isaikin.
2229 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2231 PP/02 Harmonised TLS library version reporting; only show if debugging.
2232 Layout now matches that introduced for other libraries in 4.74 PP/03.
2234 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2236 PP/04 New "dns_use_edns0" global option.
2238 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2241 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2242 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2244 TK/01 Updated PolarSSL code to 0.14.2.
2245 Bugzilla 1097. Patch from Andreas Metzler.
2247 PP/07 Catch divide-by-zero in ${eval:...}.
2248 Fixes bugzilla 1102.
2250 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2253 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2254 format-string attack -- SECURITY: remote arbitrary code execution.
2256 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2257 time unintentionally subject to list matching rules, letting the header
2258 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2259 arbitrary string expansion). This allowed for information disclosure.
2261 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2262 INT_MIN/-1 -- value coerced to INT_MAX.
2268 NM/01 Workaround for PCRE version dependency in version reporting
2271 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2272 This fixes portability to compilers other than gcc, notably
2273 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2275 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2276 makefiles for portability to HP-UX and POSIX correctness.
2278 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2279 Also via indented variable definition in the Makefile.
2280 (Debugging by Oliver Heesakkers).
2282 PP/02 Restore caching of spamd results with expanded spamd_address.
2283 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2285 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2286 Improves build reliability. Fix from: Frank Elsner
2288 NM/02 Fix wide character breakage in the rfc2047 coding
2289 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2291 NM/03 Allow underscore in dnslist lookups
2292 Fixes bug 1026. Patch from Graeme Fowler
2294 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2295 Code patches from Adam Ciarcinski of NetBSD.
2297 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2300 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2301 is logged, to avoid truncation. Patch from John Horne.
2303 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2304 Patch from Jakob Hirsch.
2306 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2307 SQL string expansion failure details.
2308 Patch from Andrey Oktyabrski.
2310 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2311 Patch from Simon Arlott.
2313 PP/09 New lookups functionality failed to compile on old gcc which rejects
2314 extern declarations in function scope.
2315 Patch from Oliver Fleischmann
2317 PP/10 Use sig_atomic_t for flags set from signal handlers.
2318 Check getgroups() return and improve debugging.
2319 Fixed developed for diagnosis in bug 927 (which turned out to be
2322 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2323 Patch from Mark Zealey.
2325 PP/12 Bugzilla 1056: Improved spamd server selection.
2326 Patch from Mark Zealey.
2328 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2329 Based on patch from Heiko Schlittermann.
2331 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2332 Patch from Uwe Doering, sign-off by Michael Haardt.
2334 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2335 variable declaration deep within a block. Bug and patch from
2338 PP/15 lookups-Makefile IRIX compatibility coercion.
2340 PP/16 Make DISABLE_DKIM build knob functional.
2342 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2343 Patch by Simon Arlott
2345 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2346 variable argument macros. Our copy now differs from upstream.
2352 TF/01 Failure to get a lock on a hints database can have serious
2353 consequences so log it to the panic log.
2355 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2356 controlled using the smtp_confirmation log selector.
2358 TF/03 Include the error message when we fail to unlink a spool file.
2360 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2361 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2362 for maintaining out-of-tree patches for some time.
2364 PP/01 Bugzilla 139: Documentation and portability issues.
2365 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2366 Handle per-OS dynamic-module compilation flags.
2368 PP/02 Let /dev/null have normal permissions.
2369 The 4.73 fixes were a little too stringent and complained about the
2370 permissions on /dev/null. Exempt it from some checks.
2371 Reported by Andreas M. Kirchwitz.
2373 PP/03 Report version information for many libraries, including
2374 Exim version information for dynamically loaded libraries. Created
2375 version.h, now support a version extension string for distributors
2376 who patch heavily. Dynamic module ABI change.
2378 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2379 privilege escalation vulnerability whereby the Exim run-time user
2380 can cause root to append content of the attacker's choosing to
2383 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2386 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2387 If dropping privileges for untrusted macros, we disabled normal logging
2388 on the basis that it would fail; for the Exim run-time user, this is not
2389 the case, and it resulted in successful deliveries going unlogged.
2390 Fixed. Reported by Andreas Metzler.
2396 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2397 only prepend for the Resent-* case. Fixes regression introduced in
2398 Exim 4.70 by NM/22 for Bugzilla 607.
2400 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2401 increasing numbers of administrators be bitten by this.
2403 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2405 PP/03 Bugzilla 994: added openssl_options main configuration option.
2407 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2409 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2411 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2413 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2414 without a peer certificate, leading to a segfault because of an
2415 assumption that peers always have certificates. Be a little more
2416 paranoid. Problem reported by Martin Tscholak.
2418 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2419 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2420 NB: ClamAV planning to remove STREAM in "middle of 2010".
2421 CL also introduces -bmalware, various -d+acl logging additions and
2422 more caution in buffer sizes.
2424 PP/09 Implemented reverse_ip expansion operator.
2426 PP/10 Bugzilla 937: provide a "debug" ACL control.
2428 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2430 PP/12 Bugzilla 973: Implement --version.
2432 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2434 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2436 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2438 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2439 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2440 ignore trailing whitespace.
2442 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2444 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2447 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2448 Notification from Dr Andrew Aitchison.
2450 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2451 ExtendedDetectionInfo response format.
2452 Notification from John Horne.
2454 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2457 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2458 XSL and documented dependency on system catalogs, with examples of how
2461 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2464 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2465 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2466 configuration file which is writeable by the Exim user or group.
2468 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2469 of configuration files to cover files specified with the -C option if
2470 they are going to be used with root privileges, not just the default
2473 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2474 option (effectively making it always true).
2476 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2477 files to be used while preserving root privileges.
2479 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2480 that rogue child processes cannot use them.
2482 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2483 run-time user, instead of root.
2485 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2486 Exim run-time user without dropping privileges.
2488 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2489 result string, instead of calling string_vformat() twice with the same
2492 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2493 for other users. Others should always drop root privileges if they use
2494 -C on the command line, even for a whitelisted configure file.
2496 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2498 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2504 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2505 $data_path, and $header_path variables; fixed documentation bugs and
2508 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2509 exipick to access non-standard spools, including the "frozen" queue
2512 NM/01 Bugzilla 965: Support mysql stored procedures.
2513 Patch from Alain Williams
2515 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2517 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2518 Patch from Andreas Metzler
2520 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2521 Patch from Kirill Miazine
2523 NM/05 Bugzilla 671: Added umask to procmail example.
2525 JJ/03 installed exipick 20100323.0, fixing doc bug
2527 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2528 directory. Notification and patch from Dan Rosenberg.
2530 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2532 TK/02 Improve log output when DKIM signing operation fails.
2534 MH/01 Treat the transport option dkim_domain as a colon separated
2535 list, not as a single string, and sign the message with each element,
2536 omitting multiple occurences of the same signer.
2538 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2539 Bugzilla 985, 986. Patch by Simon Arlott
2541 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2542 Patch by Simon Arlott
2544 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2545 MBX locking. Notification from Dan Rosenberg.
2551 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2553 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2555 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2557 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2559 NM/04 Bugzilla 915: Flag broken perl installation during build.
2565 TK/01 Added patch by Johannes Berg that expands the main option
2566 "spamd_address" if it starts with a dollar sign.
2568 TK/02 Write list of recipients to X-Envelope-Sender header when building
2569 the mbox-format spool file for content scanning (suggested by Jakob
2572 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2573 (http://www.dcc-servers.net/) support via dccifd. Activated by
2574 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2576 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2577 by Mark Daniel Reidel <mr@df.eu>.
2579 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2580 When building exim an external PCRE library is now needed -
2581 PCRE is a system library on the majority of modern systems.
2582 See entry on PCRE_LIBS in EDITME file.
2584 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2585 conversation. Added nologin parameter to request.
2586 Patch contributed by Kirill Miazine.
2588 TF/01 Do not log submission mode rewrites if they do not change the address.
2590 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2592 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2593 log files in place. Contributed by Roberto Lima.
2595 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2597 TF/03 Bugzilla 615: When checking the local_parts router precondition
2598 after a local_part_suffix or local_part_prefix option, Exim now
2599 does not use the address's named list lookup cache, since this
2600 contains cached lookups for the whole local part.
2602 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2603 Robert Millan. Documentation is in experimental-spec.txt.
2605 TF/04 Bugzilla 668: Fix parallel build (make -j).
2607 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2609 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2610 Patch provided by Jan Srzednicki.
2612 TF/05 Leading white space used to be stripped from $spam_report which
2613 wrecked the formatting. Now it is preserved.
2615 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2616 that they are available at delivery time.
2618 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2620 TF/08 TLS error reporting now respects the incoming_interface and
2621 incoming_port log selectors.
2623 TF/09 Produce a more useful error message if an SMTP transport's hosts
2624 setting expands to an empty string.
2626 NM/06 Bugzilla 744: EXPN did not work under TLS.
2627 Patch provided by Phil Pennock.
2629 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2630 Patch provided by Richard Godbee.
2632 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2633 acl_smtp_notquit, added index entry.
2635 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2636 Patch provided by Eugene Bujak.
2638 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2639 accept(). Patch provided by Maxim Dounin.
2641 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2642 Patch provided by Phil Pennock.
2644 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2646 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2647 Patch provided by Brad "anomie" Jorsch.
2649 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2650 Patch provided by Dean Brooks.
2652 TK/05 Add native DKIM support (does not depend on external libraries).
2654 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2655 Patch provided by Graeme Fowler.
2657 NM/16 Bugzilla 851: Documentation example syntax fix.
2659 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2661 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2664 NM/19 Bugzilla 745: TLS version reporting.
2665 Patch provided by Phil Pennock.
2667 NM/20 Bugzilla 167: bool: condition support.
2668 Patch provided by Phil Pennock.
2670 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2671 clients. Patch provided by Phil Pennock.
2673 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2674 Patch provided by Brad "anomie" Jorsch.
2676 NM/23 Bugzilla 687: Fix misparses in eximstats.
2677 Patch provided by Heiko Schlittermann.
2679 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2680 Patch provided by Heiko Schlittermann.
2682 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2683 plus update to original patch.
2685 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2687 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2688 Patch provided by David Brownlee.
2690 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2692 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2694 NM/30 Bugzilla 888: TLS documentation bugfixes.
2696 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2698 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2699 Unlike the original bugzilla I have changed all shell scripts in src tree.
2701 NM/33 Bugzilla 898: Transport filter timeout fix.
2702 Patch by Todd Rinaldo.
2704 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2705 Patch by Serge Demonchaux.
2707 NM/35 Bugzilla 39: Base64 decode bug fixes.
2708 Patch by Jakob Hirsch.
2710 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2712 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2714 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2716 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2722 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2723 ALT-N's libdkim that I have put here:
2724 http://duncanthrax.net/exim-experimental/
2726 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2727 were called 'true' and it seems that C99 defines that as a reserved
2728 keyword to be used with 'bool' variable types. That means you could
2729 not include C99-style headers which use bools without triggering
2730 build errors in sieve.c.
2732 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2733 as mailq or other aliases. Changed the --help handling significantly
2734 to do whats expected. exim_usage() emits usage/help information.
2736 SC/01 Added the -bylocaldomain option to eximstats.
2738 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2740 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2742 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2748 PH/01 Another patch from the Sieve maintainer.
2750 PH/02 When an IPv6 address is converted to a string for single-key lookup
2751 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2752 dots are used instead of colons so that keys in lsearch files need not
2753 contain colons. This was done some time before quoting was made available
2754 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2755 (notated using the quote facility) so as to distinguish them from IPv4
2756 keys. This meant that lookups for IP addresses in host lists did not work
2757 for iplsearch lookups.
2759 This has been fixed by arranging for IPv6 addresses to be expressed with
2760 colons if the lookup type is iplsearch. This is not incompatible, because
2761 previously such lookups could never work.
2763 The situation is now rather anomalous, since one *can* have colons in
2764 ordinary lsearch keys. However, making the change in all cases is
2765 incompatible and would probably break a number of configurations.
2767 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2770 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2771 conversion specification without a maximum field width, thereby enabling
2772 a rogue spamd server to cause a buffer overflow. While nobody in their
2773 right mind would setup Exim to query an untrusted spamd server, an
2774 attacker that gains access to a server running spamd could potentially
2775 exploit this vulnerability to run arbitrary code as the Exim user.
2777 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2778 $primary_hostname instead of what libspf2 thinks the hosts name is.
2780 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2781 a directory entry by the name of the lookup key. Previously, if a
2782 symlink pointed to a non-existing file or a file in a directory that
2783 Exim lacked permissions to read, a lookup for a key matching that
2784 symlink would fail. Now it is enough that a matching directory entry
2785 exists, symlink or not. (Bugzilla 503.)
2787 PH/03 The body_linecount and body_zerocount variables are now exported in the
2790 PH/04 Added the $dnslist_matched variable.
2792 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2793 This means they are set thereafter only if the connection becomes
2796 PH/06 Added the client_condition to authenticators so that some can be skipped
2797 by clients under certain conditions.
2799 PH/07 The error message for a badly-placed control=no_multiline_responses left
2800 "_responses" off the end of the name.
2802 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2804 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2805 (without spaces) instead of just copying the configuration text.
2807 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2809 PH/11 Added $max_received_linelength.
2811 PH/12 Added +ignore_defer and +include_defer to host lists.
2813 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2814 way in which PCRE > 7.0 is built.
2816 PH/14 Implemented queue_only_load_latch.
2818 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2819 MAIL command. The effect was to mangle the value on 64-bit systems.
2821 PH/16 Another patch from the Sieve maintainer.
2823 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2825 PH/18 If a system quota error occurred while trying to create the file for
2826 a maildir delivery, the message "Mailbox is full" was not appended to the
2827 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2828 only to a quota excession during the actual writing of the file.
2830 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2831 characters?) which causes problems in log lines. The DN values are now
2832 passed through string_printing() before being added to log lines.
2834 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2835 and InterBase are left for another time.)
2837 PH/21 Added message_body_newlines option.
2839 PH/22 Guard against possible overflow in moan_check_errorcopy().
2841 PH/23 POSIX allows open() to be a macro; guard against that.
2843 PH/24 If the recipient of an error message contained an @ in the local part
2844 (suitably quoted, of course), incorrect values were put in $domain and
2845 $local_part during the evaluation of errors_copy.
2851 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2852 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2855 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2856 issue a MAIL command.
2858 PH/02 In an ACL statement such as
2860 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2862 if a client was not listed at all, or was listed with a value other than
2863 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2864 the condition was not true (as it should be), so access was not denied.
2865 The bug was that the ! inversion was incorrectly passed on to the second
2866 item. This has been fixed.
2868 PH/03 Added additional dnslists conditions == and =& which are different from
2869 = and & when the dns lookup returns more than one IP address.
2871 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2872 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2874 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2875 FSYNC, which compiles an option called disable_fsync that allows for
2876 bypassing fsync(). The documentation is heavily laced with warnings.
2878 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2880 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2881 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2882 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2883 including adding "make clean"; (3) Added -fPIC when compiling the test
2884 dynamically loaded module, to get rid of a warning.
2886 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2887 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2888 = 0s. The bug is otherwise harmless.
2890 PH/07 There was a bug in the dovecot authenticator such that the value of
2891 $auth1 could be overwritten, and so not correctly preserved, after a
2892 successful authentication. This usually meant that the value preserved by
2893 the server_setid option was incorrect.
2895 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2897 PH/09 Installed PCRE release 7.0.
2899 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2900 run for batched SMTP input. It is now run at the start of every message
2901 in the batch. While fixing this I discovered that the process information
2902 (output by running exiwhat) was not always getting set for -bs and -bS
2903 input. This is fixed, and it now also says "batched" for BSMTP.
2905 PH/11 Added control=no_pipelining.
2907 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2908 patch, slightly modified), and move the expansion of helo_data till after
2909 the connection is made in the smtp transport (so it can use these
2912 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2914 PH/14 Added log_selector = +pid.
2916 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2918 PH/16 Add ${if forany and ${if forall.
2920 PH/17 Added dsn_from option to vary the From: line in DSNs.
2922 PH/18 Flush SMTP output before performing a callout, unless control =
2923 no_callout_flush is set.
2925 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2926 was true (the default) a successful delivery failed to delete the retry
2927 item, thus causing premature timeout of the address. The bug is now
2930 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2932 PH/21 Long custom messages for fakedefer and fakereject are now split up
2933 into multiline responses in the same way that messages for "deny" and
2934 other ACL rejections are.
2936 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2937 with slight modification.
2939 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2940 draft, changing the syntax and factoring some duplicate code".
2942 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2943 for deliveries of the second and subsequent messages over the same SMTP
2946 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2947 ${reduce, with only minor "tidies".
2949 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2951 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2952 expansion side effects.
2954 PH/27 When a message times out after an over-quota error from an Exim-imposed
2955 quota, the bounce message says "mailbox is full". This message was not
2956 being given when it was a system quota that was exceeded. It now should
2959 MH/03 Made $recipients available in local_scan(). local_scan() already has
2960 better access to the recipient list through recipients_list[], but
2961 $recipients can be useful in postmaster-provided expansion strings.
2963 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2964 in the case of a MAIL command with additional options following the
2965 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2966 were accidentally chopped off.
2968 PH/29 SMTP synchronization checks are implemented when a command is read -
2969 there is a check that no more input is waiting when there shouldn't be
2970 any. However, for some commands, a delay in an ACL can mean that it is
2971 some time before the response is written. In this time, more input might
2972 arrive, invalidly. So now there are extra checks after an ACL has run for
2973 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2974 pipelining has not been advertised.
2976 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2978 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2979 returned in the SMTP response when smtp_return_error_details was false.
2980 This has been fixed.
2982 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2983 instead of the C I/O that was originally supplied, because problems were
2984 reported on Solaris.
2986 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2987 Exim which did not show up earlier: it was assuming that a call to
2988 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2989 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2990 was a macro that became an assignment, so it seemed to work. This has
2991 changed to a proper function call with a void return, hence the compile
2992 error. Exim's code has been fixed.
2994 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2997 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2999 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3001 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3002 because it turns out to be too big or there is a timeout) there is no
3003 "Completed" line in the log. When some messages of this type were
3004 selected by exigrep, they were listed as "not completed". Others were
3005 picked up by some special patterns. I have improved the selection
3006 criteria to be more general.
3008 PH/38 The host_find_failed option in the manualroute router can now be set
3009 to "ignore", to completely ignore a host whose IP address cannot be
3010 found. If all hosts are ignored, the behaviour is controlled by the new
3011 host_all_ignored option.
3013 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3014 homing or because of multiple MX records with /mx) generated more than
3015 one IP address, and the following item turned out to be the local host,
3016 all the secondary addresses of the first item were incorrectly removed
3017 from the list, along with the local host and any following hosts (which
3018 is what is supposed to happen).
3020 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3021 whoever called Exim into the -H file. In the case of the daemon it was
3022 behaving confusingly. When first started, it used values for whoever
3023 started the daemon, but after a SIGHUP it used the Exim user (because it
3024 calls itself on a restart). I have changed the code so that it now always
3027 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3028 message are rejected with the same error (e.g. no authentication or bad
3029 sender address), and a DATA command is nevertheless sent (as can happen
3030 with PIPELINING or a stupid MUA), the error message that was given to the
3031 RCPT commands is included in the rejection of the DATA command. This is
3032 intended to be helpful for MUAs that show only the final error to their
3035 PH/42 Another patch from the Sieve maintainer.
3037 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3038 Eximstats - Fixed some broken HTML links and added missing column headers
3040 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3041 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3043 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3045 PH/43 Yet another patch from the Sieve maintainer.
3047 PH/44 I found a way to check for a TCP/IP connection going away before sending
3048 the response to the final '.' that terminates a message, but only in the
3049 case where the client has not sent further data following the '.'
3050 (unfortunately, this is allowed). However, in many cases there won't be
3051 any further data because there won't be any more messages to send. A call
3052 to select() can be used: if it shows that the input is "ready", there is
3053 either input waiting, or the socket has been closed. An attempt to read
3054 the next input character can distinguish the two cases. Previously, Exim
3055 would have sent an OK response which the client would never have see.
3056 This could lead to message repetition. This fix should cure that, at
3057 least in a lot of common cases.
3059 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3060 advertised in response to EHLO.
3066 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3067 fixed by 4.65/MH/01 (is this a record?) are fixed:
3069 (i) An empty string was always treated as zero by the numeric comparison
3070 operators. This behaviour has been restored.
3072 (ii) It is documented that the numeric comparison operators always treat
3073 their arguments as decimal numbers. This was broken in that numbers
3074 starting with 0 were being interpreted as octal.
3076 While fixing these problems I realized that there was another issue that
3077 hadn't been noticed. Values of message_size_limit (both the global option
3078 and the transport option) were treated as octal if they started with 0.
3079 The documentation was vague. These values are now always treated as
3080 decimal, and I will make that clear in the documentation.
3086 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3087 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3090 MH/01 Don't check that the operands of numeric comparison operators are
3091 integers when their expansion is in "skipping" mode (fixes bug
3092 introduced by 4.64-PH/07).
3094 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3095 child addresses, Exim now panics and dies. Previously, because the count
3096 is held in a short int, deliveries were likely to be lost. As such a
3097 large number of recipients for a single message is ridiculous
3098 (performance will be very, very poor), I have chosen to impose a limit
3099 rather than extend the field.
3105 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3106 leftover -K file (the existence of which was triggered by #402).
3107 While we were at it, introduced process PID as part of the -K
3108 filename. This should rule out race conditions when creating
3111 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3112 processing considerably. Previous code took too long for large mails,
3113 triggering a timeout which in turn triggers #401.
3115 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3116 in the DK code in transports.c. sendfile() is not really portable,
3117 hence the _LINUX specificness.
3119 TF/01 In the add_headers option to the mail command in an Exim filter,
3120 there was a bug that Exim would claim a syntax error in any
3121 header after the first one which had an odd number of characters
3124 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3125 callout verification, Exim cached a "reject" for the entire domain. This
3126 is correct for most verifications, but it is not correct for a recipient
3127 verification with use_sender or use_postmaster set, because in that case
3128 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3129 case of MAIL FROM:<> rejection from other early rejections (e.g.
3130 rejection of HELO). When verifying a recipient using a non-null MAIL
3131 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3132 Whatever the result of the callout, the value of the domain cache is
3133 left unchanged (for any other kind of callout, getting as far as trying
3134 RCPT means that the domain itself is ok).
3136 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3139 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3140 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3141 session, the daemon ignores ECONNECT errors and logs others; it now
3142 ignores EPIPE as well.
3144 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3145 (quoted-printable decoding).
3147 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3148 later the small subsequent patch to fix an introduced bug.
3150 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3152 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3154 PH/08 An error is now given if message_size_limit is specified negative.
3156 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3157 to be given (somewhat) arbitrary names.
3159 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3162 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3163 miscellaneous code fixes
3165 PH/10 Added the log_reject_target ACL modifier to specify where to log
3168 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3169 hostname. This is wrong, because it relates to the incoming message (and
3170 probably the interface on which it is arriving) and not to the outgoing
3171 callout (which could be using a different interface). This has been
3172 changed to use the value of the helo_data option from the smtp transport
3173 instead - this is what is used when a message is actually being sent. If
3174 there is no remote transport (possible with a router that sets up host
3175 addresses), $smtp_active_hostname is used.
3177 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3178 tweaks were necessary in order to get it to work (see also 21 below):
3179 (a) The code assumed that strncpy() returns a negative number on buffer
3180 overflow, which isn't the case. Replaced with Exim's string_format()
3182 (b) There were several signed/unsigned issues. I just did the minimum
3183 hacking in of casts. There is scope for a larger refactoring.
3184 (c) The code used strcasecmp() which is not a standard C function.
3185 Replaced with Exim's strcmpic() function.
3186 (d) The code set only $1; it now sets $auth1 as well.
3187 (e) A simple test gave the error "authentication client didn't specify
3188 service in request". It would seem that Dovecot has changed its
3189 interface. Fortunately there's a specification; I followed it and
3190 changed what the client sends and it appears to be working now.
3192 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3195 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3196 address A is aliased to B and C, where B exists and C does not. Without
3197 -v the output is "A verified" because verification stops after a
3198 successful redirection if more than one address is generated. However,
3199 with -v the child addresses are also verified. Exim was outputting "A
3200 failed to verify" and then showing the successful verification for C,
3201 with its parentage. It now outputs "B failed to verify", showing B's
3202 parentage before showing the successful verification of C.
3204 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3205 look up a TXT record in a specific list after matching in a combined
3208 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3209 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3210 they consult the DNS. I had assumed they would set it the way they
3211 wanted; and indeed my experiments on Linux seem to show that in some
3212 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3213 To be on the safe side, however, I have now made the interface to
3214 host_find_byname() similar to host_find_bydns(), with an argument
3215 containing the DNS resolver options. The host_find_byname() function now
3216 sets these options at its start, just as host_find_bydns() does. The smtp
3217 transport options dns_qualify_single and dns_search_parents are passed to
3218 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3219 of host_find_byname() use the default settings of RES_DEFNAMES
3220 (qualify_single) but not RES_DNSRCH (search_parents).
3222 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3223 spool_read_header() do less string testing, by means of a preliminary
3224 switch on the second character of optional "-foo" lines. (This is
3225 overdue, caused by the large number of possibilities that now exist.
3226 Originally there were few.) While I was there, I also converted the
3227 str(n)cmp tests so they don't re-test the leading "-" and the first
3228 character, in the hope this might squeeze out yet more improvement.
3230 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3231 flag allowing group syntax was set by the header_syntax check but not
3232 turned off, possible causing trouble later; (2) The flag was not being
3233 set at all for the header_verify test, causing "group"-style headers to
3234 be rejected. I have now set it in this case, and also caused header_
3235 verify to ignore an empty address taken from a group. While doing this, I
3236 came across some other cases where the code for allowing group syntax
3237 while scanning a header line wasn't quite right (mostly, not resetting
3238 the flag correctly in the right place). These bugs could have caused
3239 trouble for malformed header lines. I hope it is now all correct.
3241 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3242 with the "reply" argument non-NULL. The code, however (which originally
3243 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3244 but it didn't always do it. This confused somebody who was copying the
3245 code for some other use. I have removed all the tests.
3247 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3248 feature that was used to support insecure browsers during the U.S. crypto
3249 embargo. It requires special client support, and Exim is probably the
3250 only MTA that supported it -- and would never use it because real RSA is
3251 always available. This code has been removed, because it had the bad
3252 effect of slowing Exim down by computing (never used) parameters for the
3253 RSA_EXPORT functionality.
3255 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3256 authenticator to fail if there's a tab character in the incoming data
3257 (there should never be unless someone is messing about, as it's supposed
3258 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3259 if the connection is using TLS or if the remote IP is the same as the
3260 local IP, and the "valid-client-cert option" if a client certificate has
3263 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3264 authenticators. This can be used for authorization after authentication
3265 succeeds. (In the case of plaintext, it servers for both authentication
3268 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3269 if any retry times were supplied.
3271 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3272 connection, where there is no client IP address to check. In this
3273 situation, the verify now always succeeds.
3275 PH/25 Applied John Jetmore's -Mset patch.
3277 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3279 PH/27 In a string expansion for a processed (not raw) header when multiple
3280 headers of the same name were present, leading whitespace was being
3281 removed from all of them, but trailing whitespace was being removed only
3282 from the last one. Now trailing whitespace is removed from each header
3283 before concatenation. Completely empty headers in a concatenation (as
3284 before) are ignored.
3286 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3287 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3289 PH/29 [Removed. This was a change that I later backed out, and forgot to
3290 correct the ChangeLog entry (that I had efficiently created) before
3291 committing the later change.]
3293 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3294 address errors (4xx response to RCPT) over the same connection as other
3295 messages routed to the same hosts. Such deliveries are always "forced",
3296 so retry times are not inspected. This resulted in far too many retries
3297 for the affected addresses. The effect occurred only when there were more
3298 hosts than the hosts_max_try setting in the smtp transport when it had
3299 the 4xx errors. Those hosts that it had tried were not added to the list
3300 of hosts for which the message was waiting, so if all were tried, there
3301 was no problem. Two fixes have been applied:
3303 (i) If there are any address or message errors in an SMTP delivery, none
3304 of the hosts (tried or untried) are now added to the list of hosts
3305 for which the message is waiting, so the message should not be a
3306 candidate for sending over the same connection that was used for a
3307 successful delivery of some other message. This seems entirely
3308 reasonable: after all the message is NOT "waiting for some host".
3309 This is so "obvious" that I'm not sure why it wasn't done
3310 previously. Hope I haven't missed anything, but it can't do any
3311 harm, as the worst effect is to miss an optimization.
3313 (ii) If, despite (i), such a delivery is accidentally attempted, the
3314 routing retry time is respected, so at least it doesn't keep
3315 hammering the server.
3317 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3318 in ${readsocket because some servers need this prod.
3320 PH/32 Added some extra debug output when updating a wait-xxx database.
3322 PH/33 The hint "could be header name not terminated by colon", which has been
3323 given for certain expansion errors for a long time, was not being given
3324 for the ${if def:h_colon_omitted{... case.
3326 PH/34 The spec says: "With one important exception, whenever a domain list is
3327 being scanned, $domain contains the subject domain." There was at least
3328 one case where this was not true.
3330 PH/35 The error "getsockname() failed: connection reset by peer" was being
3331 written to the panic log as well as the main log, but it isn't really
3332 panic-worthy as it just means the connection died rather early on. I have
3333 removed the panic log writing for the ECONNRESET error when getsockname()
3336 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3337 runs only) independently of the message's sender address. This meant
3338 that, if the 4xx error was in fact related to the sender, a different
3339 message to the same recipient with a different sender could confuse
3340 things. In particular, this can happen when sending to a greylisting
3341 server, but other circumstances could also provoke similar problems.
3342 I have changed the default so that the retry time for these errors is now
3343 based a combination of the sender and recipient addresses. This change
3344 can be overridden by setting address_retry_include_sender=false in the
3347 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3348 remote server are returned as part of bounce messages. This was not
3349 happening for LMTP over a pipe (the lmtp transport), but now it is the
3350 same for both kinds of LMTP.
3352 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3353 in header lines that were in fact CNAMEs. This is no longer the case.
3355 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3356 and queue runs started by the daemon processed all messages. This has
3357 been fixed so that -R and -S can now usefully be given with -q<time>.
3359 PH/40 Import PCRE release 6.7 (fixes some bugs).
3361 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3363 PH/42 Give an error if -q is specified more than once.
3365 PH/43 Renamed the variables $interface_address and $interface_port as
3366 $received_ip_address and $received_port, to make it clear that these
3367 values apply to message reception, and not to the outgoing interface when
3368 a message is delivered. (The old names remain recognized, of course.)
3370 PH/44 There was no timeout on the connect() call when using a Unix domain
3371 socket in the ${readsocket expansion. There now is.
3373 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3374 be meaningful with "accept".
3376 SC/01 Eximstats V1.43
3377 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3379 SC/02 Eximstats V1.44
3380 Use a glob alias rather than an array ref in the generated
3381 parser. This improves both readability and performance.
3383 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3384 Collect SpamAssassin and rejection statistics.
3385 Don't display local sender or destination tables unless
3386 there is data to show.
3387 Added average volumes into the top table text output.
3389 SC/04 Eximstats V1.46
3390 Collect data on the number of addresses (recipients)
3391 as well as the number of messages.
3393 SC/05 Eximstats V1.47
3394 Added 'Message too big' to the list of mail rejection
3395 reasons (thanks to Marco Gaiarin).
3397 SC/06 Eximstats V1.48
3398 Mainlog lines which have GMT offsets and are too short to
3399 have a flag are now skipped.
3401 SC/07 Eximstats V1.49 (Alain Williams)
3402 Added the -emptyok flag.
3404 SC/08 Eximstats V1.50
3405 Fixes for obtaining the IP address from reject messages.
3407 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3408 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3409 whitespace changes from 4.64-PH/27
3411 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3414 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3415 are found, allow negative numbers in numeric criteria)
3417 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3419 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3420 to match changes made in 4.64-PH/43
3422 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3424 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3425 "domain" is actually an IP address, and force a failure. This locks out
3426 those revolvers/nameservers that support "A-for-A" lookups, in
3427 contravention of the specifications.
3429 PH/48 When a host name was looked up from an IP address, and the subsequent
3430 forward lookup of the name timed out, the host name was left in
3431 $sender_host_name, contrary to the specification.
3433 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3434 restricted to single-key lookups, Exim was not diagnosing an error if
3435 * or *@ was used with a query-style lookup.
3437 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3439 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3440 long ago, but noone interested enough thought of it. Let's just say that
3441 the "1.1" means that there are some new functions that weren't there at
3442 some point in the past.
3444 PH/51 Error processing for expansion failure of helo_data from an smtp
3445 transport during callout processing was broken.
3447 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3448 tested/used via the -bh/-bhc/-bs options.
3450 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3451 bug, fixed in subsequent PCRE releases).
3453 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3454 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3456 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3462 SC/01 Use a glob alias rather than an array ref in eximstats generated
3463 parser. This improves both readability and performance.
3465 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3466 Don't display local sender or destination tables in eximstats unless
3467 there is data to show.
3468 Added average volumes into the eximstats top table text output.
3470 SC/03 Collect data on the number of addresses (recipients) as well
3471 as the number of messages in eximstats.
3473 TF/01 Correct an error in the documentation for the redirect router. Exim
3474 does (usually) call initgroups() when daemonizing.
3476 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3477 with consistent privilege compared to when running as a daemon.
3479 TF/03 Note in the spec that $authenticated_id is not set for local
3480 submissions from trusted users.
3482 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3483 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3485 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3486 by adding some example configuration directives to the default
3487 configuration file. A little bit of work is required to uncomment the
3488 directives and define how usernames and passwords are checked, but
3489 there is now a framework to start from.
3491 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3492 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3493 without this. I don't know how relevant this is to other LDAP libraries.
3495 PH/02 Add the verb name to the "unknown ACL verb" error.
3497 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3499 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3501 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3502 directory not expanded when it should be if an expanded home directory
3503 was set for the address (which is overridden by the transport).
3505 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3508 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3509 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3510 because it is too late at that time, and has no effect.
3512 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3513 security issue with \' (bugzilla #107). I could not use the
3514 PQescapeStringConn() function, because it needs a PGconn value as one of
3517 PH/08 When testing addresses using -bt, indicate those final addresses that
3518 are duplicates that would not cause an additional delivery. At least one
3519 person was confused, thinking that -bt output corresponded to deliveries.
3520 (Suppressing duplicates isn't a good idea as you lose the information
3521 about possibly different redirections that led to the duplicates.)
3523 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3524 systems where poll() doesn't work, in particular OS X.
3526 PH/10 Added more information to debugging output for retry time not reached.
3528 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3529 operations in malware.c.
3531 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3534 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3535 syslog_duplication set false, log lines that would normally be written
3536 both the the main log and to the reject log were not written to syslog at
3539 PH/14 In the default configuration, change the use of "message" in ACL warn
3540 statements to "add_header".
3542 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3543 not followed by a command (e.g. "seen endif").
3545 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3546 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3549 PH/17 Added extra conditions to the default value of delay_warning_condition
3553 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3554 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3555 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3558 The Auto-Submitted: and various List- headers are standardised, whereas I
3559 don't think Precedence: ever was.
3561 PH/18 Refactored debugging code in route_finduser() to show more information,
3562 in particular, the error code if getpwnam() issues one.
3564 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3565 This is apparently needed in addition to the PH/07 change above to avoid
3566 any possible encoding problems.
3568 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3569 but not after initializing Perl.
3571 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3572 output them only if debugging. By default they are written stderr,
3573 apparently, which is not desirable.
3575 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3578 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3581 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3583 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3584 authenticated or an ident call has been made. Suppress the default
3585 values for $authenticated_id and $authenticated_sender (but permit -oMai
3586 and -oMas) when testing with -bh.
3588 PH/24 Re-jigged the order of the tests in the default configuration so that the
3589 tests for valid domains and recipients precede the DNS black list and CSA
3590 tests, on the grounds that those ones are more expensive.
3592 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3593 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3594 This bug exists in every version of Exim that I still have, right back to
3597 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3598 However, an attempt to turn on case-sensitivity in a regex key by
3599 including (?-i) didn't work because the subject string was already
3600 lowercased, and the effects were non-intuitive. It turns out that a
3601 one-line patch can be used to allow (?-i) to work as expected.
3607 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3608 other effects) broke the use of negated acl sub-conditions.
3610 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3613 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3614 "Deny" causes Exim to reject the incoming connection with a 554 error.
3615 Unfortunately, if there is a major crisis, such as a disk failure,
3616 tcp-wrappers gives "deny", whereas what one would like would be some
3617 kind of temporary error. A kludge has been added to help with this.
3618 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3619 554 error is used if errno is still zero or contains ENOENT (which occurs
3620 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3623 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3625 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3626 errors. Otherwise a message that provokes a temporary error (when other
3627 messages do not) can cause a whole host to time out.
3629 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3630 addresses were routed directly to files or pipes from a redirect router.
3631 File deliveries just didn't batch; pipe deliveries might have suffered
3634 PH/06 A failure to get a lock for a hints database would erroneously always say
3635 "Failed to get write lock", even when it was really a read lock.
3637 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3638 of 0600. This has been changed to use the value of the lockfile_mode
3639 option (which defaults to 0600).
3641 PH/08 Applied small patch from the Sieve maintainer.
3643 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3644 folder from quota calculations, a direct delivery into this folder messed
3645 up the contents of the maildirsize file. This was because the regex was
3646 used only to exclude .Trash (or whatever) when the size of the mailbox
3647 was calculated. There was no check that a delivery was happening into an
3648 excluded directory. This bug has been fixed by ignoring all quota
3649 processing for deliveries into excluded directories.
3651 PH/10 Added the maildirfolder_create_regex option to appendfile.
3657 PH/01 The code for finding all the local interface addresses on a FreeBSD
3658 system running IPv6 was broken. This may well have applied to all BSD
3659 systems, as well as to others that have similar system calls. The broken
3660 code found IPv4 interfaces correctly, but gave incorrect values for the
3661 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3662 that it would not match correctly against @[] and not recognize the IPv6
3665 PH/02 The ipliteral router was not recognizing addresses of the form user@
3666 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3668 PH/03 Added disable_ipv6.
3670 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3671 decoded form, because it is most often used to construct To: headers
3672 lines in autoreplies, and the decoded form may well be syntactically
3673 invalid. However, $reply_address has leading white space removed, and all
3674 newlines turned into spaces so that the autoreply transport does not
3677 PH/05 If group was specified without a user on a router, and no group or user
3678 was specified on a transport, the group from the router was ignored.
3680 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3681 for visible compile-time settings that can be used to change these
3682 numbers, for those that want even more. Backwards compatibility with old
3683 spool files has been maintained. However, going back to a previous Exim
3684 release will lost any variables that are in spool files.
3686 PH/07 Two small changes when running in the test harness: increase delay when
3687 passing a TCP/IP connection to a new process, in case the original
3688 process has to generate a bounce, and remove special handling of
3689 127.0.0.2 (sic), which is no longer necessary.
3691 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3692 be the same on different OS.
3694 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3697 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3698 whether --show-vars was specified or not
3700 JJ/02 exipick: Added support for new ACL variable spool format introduced
3703 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3704 syntactically invalid From: or Reply-to: line, and a filter used this to
3705 generate an autoreply, and therefore failed to obtain an address for the
3706 autoreply, Exim could try to deliver to a non-existent relative file
3707 name, causing unrelated and misleading errors. What now happens is that
3708 it logs this as a hard delivery error, but does not attempt to create a
3711 PH/11 The exinext utility has a -C option for testing purposes, but although
3712 the given file was scanned by exinext itself; it wasn't being passed on
3713 when Exim was called.
3715 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3716 an end-of-file indication when reading a command response.
3718 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3719 compiled. In many other places in Exim, IPv6 addresses are always
3720 recognized, so I have changed this. It also means that IPv4 domain
3721 literals of the form [IPV4:n.n.n.n] are now always recognized.
3723 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3724 used if the router is not running as root, for example, when verifying at
3725 ACL time, or when using -bh. The debugging output from this situation was
3726 non-existent - all you got was a failure to exec. I have made two
3729 (a) Failures to set uid/gid, the current directory, or a process leader
3730 in a subprocess such as that created by queryprogram now generate
3731 suitable debugging output when -d is set.
3733 (b) The queryprogram router detects when it is not running as root,
3734 outputs suitable debugging information if -d is set, and then runs
3735 the subprocess without attempting to change uid/gid.
3737 PH/15 Minor change to Makefile for building test_host (undocumented testing
3740 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3741 additional section of a DNS packet that returns MX or SRV records.
3742 Instead, it always explicitly searches for A/AAAA records. This avoids
3743 major problems that occur when a DNS server includes only records of one
3744 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3745 fixed another bug: if SRV records were looked up and the corresponding
3746 address records were *not* found in the additional section, the port
3747 values from the SRV records were lost.
3749 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3750 using the correct key (the original address) when searching the retry
3751 rules in order to find which one to use for generating the retry hint.
3753 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3754 adding the default one. Similarly, if it contains a Reply-To: header, the
3755 errors_reply_to option, if set, is not used.
3757 PH/19 When calculating a retry time, Exim used to measure the "time since
3758 failure" by looking at the "first failed" field in the retry record. Now
3759 it does not use this if it is later than than the arrival time of the
3760 message. Instead it uses the arrival time. This makes for better
3761 behaviour in cases where some deliveries succeed, thus re-setting the
3762 "first failed" field. An example is a quota failure for a huge message
3763 when small messages continue to be delivered. Without this change, the
3764 "time since failure" will always be short, possible causing more frequent
3765 delivery attempts for the huge message than are intended.
3766 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3768 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3769 $1, $2, $3) because the numerical variables can be reset during some
3770 expansion items (e.g. "match"), thereby losing the authentication data.
3772 PH/21 Make -bV show the size of off_t variables so that the test suite can
3773 decide whether to run tests for quotas > 2G.
3775 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3776 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3777 filecount value is greater than 2G or if a quota value is greater than 2G
3778 on a system where the size of off_t is not greater than 4, a panic error
3781 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3782 never match. The debug and -bh output now contains an explicit error
3783 message indicating a malformed IPv4 address or mask.
3785 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3786 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3787 PH/23 above applies.
3789 PH/25 Do not write to syslog when running in the test harness. The only
3790 occasion when this arises is a failure to open the main or panic logs
3791 (for which there is an explicit test).
3793 PH/26 Added the /no_tell option to "control=freeze".
3795 PH/27 If a host name lookup failed very early in a connection, for example, if
3796 the IP address matched host_lookup and the reverse lookup yielded a name
3797 that did not have a forward lookup, an error message of the form "no IP
3798 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3799 could be logged. Now it outputs the IP address instead of "NULL".
3801 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3802 allows the sender and the authenticated sender to be set when
3803 submitting a message from within Exim. Since child_open_exim() is
3804 documented for local_scan(), the new function should be too.
3806 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3807 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3808 results in an empty string is now treated as unset.
3810 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3812 PH/31 Added sender_verify_fail logging option.
3814 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3815 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3816 tidied the source and removed it altogether.
3818 PH/33 When a queue run was abandoned because the load average was too high, a
3819 log line was always written; now it is written only if the queue_run log
3820 selector is set. In addition, the log line for abandonment now contains
3821 information about the queue run such as the pid. This is always present
3822 in "start" and "stop" lines but was omitted from the "abandon" line.
3824 PH/34 Omit spaces between a header name and the colon in the error message that
3825 is given when verify = headers_syntax fails (if there are lots of them,
3826 the message gets confusing).
3828 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3829 names, as there are now some PTR records that contain slashes. This check
3830 is only to protect against broken name servers that fall over on strange
3831 characters, so the fact that it applies to all lookups doesn't matter.
3833 PH/36 Now that the new test suite is complete, we can remove some of the
3834 special code in Exim that was needed for the old test suite. For example,
3835 sorting DNS records because real resolvers return them in an arbitrary
3836 order. The new test suite's fake resolver always returns records in the
3839 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3840 bounces) except when queue_only is set, to avoid logging races between
3841 the different processes.
3843 PH/38 Panic-die if .include specifies a non-absolute path.
3845 PH/39 A tweak to the "H" retry rule from its user.
3847 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3848 a label. They prevented compilation on older perls.
3850 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3851 a warning to be raised on newish perls.
3853 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3854 on queue. Changes to match documented behaviour of showing count of
3855 messages matching specified criteria.
3857 PH/40 Changed the default ident timeout from 30s to 5s.
3859 PH/41 Added support for the use of login_cap features, on those BSD systems
3860 that have them, for controlling the resources used by pipe deliveries.
3862 PH/42 The content-scanning code uses fopen() to create files in which to put
3863 message data. Previously it was not paying any attention to the mode of
3864 the files. Exim runs with umask(0) because the rest of the code creates
3865 files with open(), and sets the required mode explicitly. Thus, these
3866 files were ending up world-writeable. This was not a big issue, because,
3867 being within the spool directory, they were not world-accessible. I have
3868 created a function called modefopen, which takes an additional mode
3869 argument. It sets umask(777), creates the file, chmods it to the required
3870 mode, then resets the umask. All the relevant calls to fopen() in the
3871 content scanning code have been changed to use this function.
3873 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3874 to 24 hours. This avoids potential overflow problems when processing G
3875 and H retry rules. I suspect nobody ever tinkers with this value.
3877 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3879 PH/45 When the plaintext authenticator is running as a client, the server's
3880 challenges are checked to ensure they are valid base64 strings. By
3881 default, the authentication attempt is cancelled if an invalid string is
3882 received. Setting client_ignore_invalid_base64 true ignores these errors.
3883 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3884 they are received. Thus, the responses can be made to depend on the
3885 challenges. If an invalid string is ignored, an empty string is placed in
3888 PH/46 Messages that are created by the autoreply transport now contains a
3889 References: header, in accordance with RFCs 2822 and 3834.
3891 PH/47 Added authenticated_sender_force to the smtp transport.
3893 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3895 PH/49 Installed latest patch from the Sieve maintainer.
3897 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3898 was also set, the appendfile transport was unnecessarily scanning a
3899 directory of message files (e.g. for maildir delivery) to find the count
3900 of files (along with the size), even though it did not need this
3901 information. It now does the scan only if it needs to find either the
3902 size of the count of files.
3904 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3906 PH/52 Two bugs concerned with error handling when the smtp transport is
3909 (i) Exim was not creating retry information for temporary errors given
3910 for individual recipients after the DATA command when the smtp transport
3911 was used in LMTP mode. This meant that they could be retried too
3912 frequently, and not timed out correctly.
3914 (ii) Exim was setting the flag that allows error details to be returned
3915 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3916 recipients that were returned after the DATA command.
3918 PH/53 This is related to PH/52, but is more general: for any failing address,
3919 when detailed error information was permitted to be returned to the
3920 sender, but the error was temporary, then after the final timeout, only
3921 "retry timeout exceeded" was returned. Now it returns the full error as
3922 well as "retry timeout exceeded".
3924 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3925 do this, and (what is worse) MTAs that accept it.
3927 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3928 will now be deprecated.
3930 PH/56 New os.c-cygwin from the Cygwin maintainer.
3932 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3933 formats (previously only available in exim formats via -bpr, -bpru,
3934 and -bpra. Now also available in native and exiqgrep formats)
3936 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3937 with very large, slow to parse queues
3939 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3941 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3943 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3944 responses to authentication challenges, though it was showing the
3945 challenges; (ii) I've removed the CR characters from the debug output for
3948 PH/58 Allow for the insertion of a newline as well as a space when a string
3949 is turned into more than one encoded-word during RFC 2047 encoding. The
3950 Sieve code now uses this.
3952 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3953 data_4xx, lost_connection, tls_required.
3955 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3956 message was being sent as an SMTP response.
3958 PH/61 Add -l and -k options to exicyclog.
3960 PH/62 When verifying, if an address was redirected to one new address, so that
3961 verification continued, and the new address failed or deferred after
3962 having set something in $address_data, the value of $address_data was not
3963 passed back to the ACL. This was different to the case when no
3964 redirection occurred. The value is now passed back in both cases.
3966 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3967 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3968 use login_cap.h, so on its own it isn't the distinguishing feature. The
3969 new name refers directly to the setclassresources() function.
3971 PH/65 Added configuration files for NetBSD3.
3973 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3975 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3976 is preferred over IPv4.
3978 PH/68 The bounce_return_message and bounce_return_body options were not being
3979 honoured for bounces generated during the reception of non-SMTP messages.
3980 In particular, this applied to messages rejected by the ACL. This bug has
3981 been fixed. However, if bounce_return_message is true and bounce_return_
3982 body is false, the headers that are returned for a non-SMTP message
3983 include only those that have been read before the error was detected.
3984 (In the case of an ACL rejection, they have all been read.)
3986 PH/69 The HTML version of the specification is now built in a directory called
3987 spec_html instead of spec.html, because the latter looks like a path with
3988 a MIME-type, and this confuses some software.
3990 PH/70 Catch two compiler warnings in sieve.c.
3992 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3993 function verify_get_ident() calls ip_connect() to connect a socket, but
3994 if the "connect()" function timed out, ip_connect() used to close the
3995 socket. However, verify_get_ident() also closes the socket later, and in
3996 between Exim writes to the log, which may get opened at this point. When
3997 the socket was closed in ip_connect(), the log could get the same file
3998 descriptor number as the socket. This naturally causes chaos. The fix is
3999 not to close the socket in ip_connect(); the socket should be closed by
4000 the function that creates it. There was only one place in the code where
4001 this was missing, in the iplookup router, which I don't think anybody now
4002 uses, but I've fixed it anyway.
4004 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4005 well as to direct DNS lookups. Otherwise the handling of names in host
4006 lists is inconsistent and therefore confusing.
4012 PH/01 Two changes to the default runtime configuration:
4014 (1) Move the checks for relay_from_hosts and authenticated clients from
4015 after to before the (commented out) DNS black list checks.
4017 (2) Add control=submission to the relay_from_hosts and authenticated
4018 clients checks, on the grounds that messages accepted by these
4019 statements are most likely to be submissions.
4021 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4023 (1) Generate an error if the third argument for the ${prvs expansion is
4026 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4029 (3) Reset the variables that are obtained from the first argument of
4030 ${prvscheck and used in the second argument before leaving the code,
4031 because their memory is reclaimed, so using them afterwards may do
4034 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4035 one (it's much easier than Tom thought :-).
4037 (5) Because of (4), we can now allow for the use of $prvscheck_result
4038 inside the third argument.
4040 PH/03 For some reason, the default setting of PATH when running a command from
4041 a pipe transport was just "/usr/bin". I have changed it to
4044 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4045 anything to be listed in the output from -bV.
4047 PH/05 When a filter generated an autoreply, the entire To: header line was
4048 quoted in the delivery log line, like this:
4050 => >A.N.Other <ano@some.domain> <original@ddress> ...
4052 This has been changed so that it extracts the operative address. There
4053 may be more than one such address. If so, they are comma-separated, like
4056 => >ano@some.domain,ona@other.domain <original@ddress> ...
4058 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4059 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4060 not being looked up in the rDNS to get a host name, Exim was showing the
4061 IP address twice in Received: lines, even though the IP addresses were
4062 identical. For example:
4064 Received: from [1.2.3.4] (helo=[1.2.3.4])
4066 However, if the real host name was known, it was omitting the HELO data
4067 if it matched the actual IP address. This has been tidied up so that it
4068 doesn't show the same IP address twice.
4070 PH/07 When both +timestamp and +memory debugging was on, the value given by
4071 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4072 called by the debug printing, thereby overwriting the timestamp buffer.
4073 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4075 PH/08 When the original message was included in an autoreply transport, it
4076 always said "this is a copy of the message, including all the headers",
4077 even if body_only or headers_only was set. It now gives an appropriate
4080 PH/09 Applied a patch from the Sieve maintainer which:
4082 o fixes some comments
4083 o adds the (disabled) notify extension core
4084 o adds some debug output for the result of if/elsif tests
4085 o points to the current vacation draft in the documentation
4086 and documents the missing references header update
4090 o fixes a bug in processing the envelope test (when testing
4091 multiple envelope elements, the last element determined the
4094 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4095 Electronic Mail") by including:
4097 Auto-submitted: auto-generated
4099 in the messages that it generates (bounce messages and others, such as
4100 warnings). In the case of bounce messages for non-SMTP messages, there was
4101 also a typo: it was using "Auto_submitted" (underscore instead of
4102 hyphen). Since every message generated by Exim is necessarily in response
4103 to another message, thes have all been changed to:
4105 Auto-Submitted: auto-replied
4107 in accordance with these statements in the RFC:
4109 The auto-replied keyword:
4111 - SHOULD be used on messages sent in direct response to another
4112 message by an automatic process,
4114 - MUST NOT be used on manually-generated messages,
4116 - MAY be used on Delivery Status Notifications (DSNs) and Message
4117 Disposition Notifications (MDNs),
4119 - MUST NOT be used on messages generated by automatic or periodic
4120 processes, except for messages which are automatic responses to
4123 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4124 to the default Received: header definition.
4126 PH/12 Added log selector acl_warn_skipped (default on).
4128 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4129 variables because (a) they are in the wrong storage pool and (b) even if
4130 they were copied, it wouldn't work properly because of the caching.
4132 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4133 checking when decoding. Apparently there are clients that generate
4134 overlong encoded strings. Why am I not surprised?
4136 PH/15 If the first argument of "${if match_address" was not empty, but did not
4137 contain an "@" character, Exim crashed. Now it writes a panic log message
4138 and treats the condition as false.
4140 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4142 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4143 extension "envelope-auth". The code is finished and in agreement with
4144 other implementations, but there is no documentation so far and in fact,
4145 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4146 not changing the active code.
4148 Print executed "if" and "elsif" statements when debugging is used. This
4149 helps a great deal to understand what a filter does.
4151 Document more things not specified clearly in RFC3028. I had all this
4152 sorted out, when out of a sudden new issues came to my mind. Oops."
4154 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4157 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4158 canonical form (as documented). However, after a host name lookup from
4159 the IP address, check_host() was doing a simple string comparison with
4160 addresses acquired from the DNS when checking that the found name did
4161 have the original IP as one of its addresses. Since any found IPv6
4162 addresses are likely to be in abbreviated form, the comparison could
4163 fail. Luckily, there already exists a function for doing the comparison
4164 by converting both addresses to binary, so now that is used instead of
4165 the text comparison.
4167 PH/20 There was another similar case to PH/19, when a complete host name was
4168 given in a host list; looking up its IP address could give an abbreviated
4169 form, whereas the current host's name might or might not be abbreviated.
4170 The same fix has been applied.
4176 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4177 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4180 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4181 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4183 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4185 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4186 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4187 Also, for "Auto-Submitted", treat anything other than "no" as
4188 non-personal, in accordance with RFC 3834. (Previously it treated
4189 anything starting "auto-" as non-personal.)
4191 TF/01 The control=submission/name=... option had a problem with syntax
4192 errors if the name included a slash character. The /name= option
4193 now slurps the rest of the string, so it can include any characters
4194 but it must come last in the list of options (after /sender_retain
4197 PH/05 Some modifications to the interface to the fake nameserver for the new
4205 TK/01 Added the "success_on_redirect" address verification option. See
4206 NewStuff for rationale and an example.
4208 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4210 PH/02 Patch to exigrep to allow it to work on syslog lines.
4212 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4213 fread() to skip over the body file's header line, because in Cygwin the
4214 header line is locked and is inaccessible.
4216 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4217 co-exist for some time) to make it clear that it is the Exim ID that is
4218 referenced, not the Message-ID: header line.
4220 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4221 string_format() function, because snprintf() does not exist on all
4224 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4225 ${stat: expansion item.
4227 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4228 protocol synchronization error", to keep the pedants happy.
4230 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4231 well as for IRIX systems, when gcc is being used. See the host.c source
4234 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4236 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4239 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4240 transport and to the smtp transport in LMTP mode.
4242 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4244 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4245 run a filter in a subprocess. This could lead to confusion in subsequent
4246 lookups in the parent process. There should also be a search_tidyup() at
4247 the end of the subprocess.
4249 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4250 only if the host matched helo_try_verify_hosts, which caused the
4251 verification to occur when the EHLO/HELO command was issued. The ACL just
4252 tested the remembered result. Now, if a previous verification attempt has
4253 not happened, "verify = helo" does it there and then.
4255 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4257 TK/03 Fix log output including CR from clamd.
4259 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4260 exist provoked a memory error which could cause a segfault.
4262 PH/15 Installed PCRE 6.2
4264 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4266 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4267 of the problem. Specifically, suggested +O2 rather than +O1 for the
4270 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4272 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4273 record, the port was not correctly logged when the outgoing_port log
4274 selector was set (it logged the transort's default port).
4276 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4277 fallback_hosts, and "hosts" in the smtp transport.
4279 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4280 host errors such as "Connection refused".
4282 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4283 authentication with radiusclient 0.4.9:
4285 - Error returned from rc_read_config was caught wrongly
4286 - Username/password not passed on to radius server due to wrong length.
4288 The presumption is that some radiusclient API changes for 4.51/PH/17
4289 were not taken care of correctly. The code is still untested by me (my
4290 Linux distribution still has 0.3.2 of radiusclient), but it was
4291 contributed by a Radius user.
4293 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4294 expanding the "port" option of the smtp transport.
4296 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4297 while reading a MIME header. Thanks to Tom Hughes for a patch.
4299 PH/24 Include config.h inside local_scan.h so that configuration settings are
4302 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4303 that in an ACL for RCPT (for example), you can examine exactly what was
4306 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4307 commands, but it was not correctly comparing the address with the actual
4308 client host address. Thus, it would show the EHLO address in Received:
4309 header lines when this was not necessary.
4311 PH/27 Added the % operator to ${eval:}.
4313 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4314 it should be ignoring failures (because with -C, for example, it has lost
4315 privilege). It wasn't ignoring creation failures other than "already
4318 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4321 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4322 because an input line was too long, either on its own, or by virtue of
4323 too many continuations, the temporary file was not being removed, and the
4324 return code was incorrect.
4326 PH/31 Missing "BOOL" in function definition in filtertest.c.
4328 PH/32 Applied Sieve patches from the maintainer.
4330 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4332 PH/33 Added "verify = not_blind".
4334 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4335 Local/Makefile (with some defaults set). These are used in built scripts
4336 such as exicyclog, but they have never been used in the exim_install
4337 script (though there are many overriding facilities there). I have
4338 arranged that the exim_install script now takes note of these two
4341 PH/35 Installed configuration files for Dragonfly.
4343 PH/36 When a locally submitted message by a trusted user did not contain a
4344 From: header, and the sender address was obtained from -f or from an SMTP
4345 MAIL command, and the trusted user did not use -F to supply a sender
4346 name, $originator_name was incorrectly used when constructing a From:
4347 header. Furthermore, $originator_name was used for submission mode
4348 messages from external hosts without From: headers in a similar way,
4349 which is clearly wrong.
4351 PH/37 Added control=suppress_local_fixups.
4353 PH/38 When log_selector = +received_sender was set, and the addition of the
4354 sender made the log line's construction buffer exactly full, or one byte
4355 less than full, an overflow happened when the terminating "\n" was
4358 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4359 when the result of a list match is failure because a DNS lookup failed.
4361 PH/40 RM_COMMAND is now used in the building process.
4363 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4364 the "build-* directories that it finds.
4366 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4367 address in a domain literal was a prefix of an interface address.
4369 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4370 when verifying a sender address, unless rewrite_headers is false.
4372 PH/44 Wrote a long comment about why errors_to addresses are verified as
4373 recipients, not senders.
4375 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4376 the ratelimit ACL was added.
4378 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4380 PH/46 Added extra information about PostgreSQL errors to the error string.
4382 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4383 avoiding the need to install special zones in a real server. This is
4384 backwards compatible; if it can't find the fake resolver, it drops back.
4385 Thus, both old and new test suites can be run.
4387 TF/02 Added util/ratelimit.pl
4389 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4390 clock is set back in time.
4392 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4393 Candler <B.Candler@pobox.com>.
4395 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4396 OK for addresses that are the result of redirections.
4398 PH/48 A number of further additions for the benefit of the new test suite,
4399 including a fake gethostbyname() that interfaces to the fake DNS resolver
4402 TF/06 The fix for widen_domains has also been applied to qualify_single and
4403 search_parents which are the other dnslookup options that can cause
4406 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4409 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4411 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4412 service" if it encountered an MX record with an empty target hostname.
4413 The message is now "an MX or SRV record indicated no SMTP service".
4415 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4416 if the DNS of the sending site is misconfigured. This is quite a
4417 common situation. This change restores the behaviour of treating a
4418 helo verification defer as a failure.
4420 PH/51 If self=fail was set on a router, the bounce message did not include the
4421 actual error message.
4427 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4429 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4430 command itself ended in error, the underlying message about the transport
4431 filter timeout was being overwritten with the pipe command error. Now the
4432 underlying error message should be appended to the second error message.
4434 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4436 PH/02 Reduce the amount of output that "make" produces by default. Full output
4437 can still be requested.
4439 PH/03 The warning log line about a condition test deferring for a "warn" verb
4440 was being output only once per connection, rather than after each
4441 occurrence (because it was using the same function as for successful
4442 "warn" verbs). This seems wrong, so I have changed it.
4444 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4445 it should not have, which might have caused a crash in the right
4446 circumstances, but probably never did.
4448 PH/04 Installed a modified version of Tony Finch's patch to make submission
4449 mode fix the return path as well as the Sender: header line, and to
4450 add a /name= option so that you can make the user's friendly name appear
4453 TF/03 Added the control = fakedefer ACL modifier.
4455 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4456 Mark Lowes for thorough testing.
4458 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4460 TK/03 Merged latest SRS patch from Miles Wilton.
4462 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4463 with the definition in sysexits.h (which is #included earlier).
4464 Fortunately, Exim does not actually use EX_OK. The code used to try to
4465 preserve the sysexits.h value, by assuming that macro definitions were
4466 scanned for macro replacements. I have been disabused of this notion,
4467 so now the code just undefines EX_OK before #including unistd.h.
4469 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4470 in the smtp transport. When a block could not be written in a single
4471 write() function, the timeout was being re-applied to each part-write.
4472 This seems wrong - if the receiver was accepting one byte at a time it
4473 would take for ever. The timeout is now adjusted when this happens. It
4474 doesn't have to be particularly precise.
4476 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4477 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4479 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4480 without a domain if the check to <postmaster@domain> fails.
4482 SC/01 Eximstats: added -xls and the ability to specify output files
4483 (patch written by Frank Heydlauf).
4485 SC/02 Eximstats: use FileHandles for outputting results.
4487 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4489 SC/04 Eximstats: fixed display of large numbers with -nvr option
4491 SC/05 Eximstats: fixed merging of reports with empty tables.
4493 SC/06 Eximstats: added the -include_original_destination flag
4495 SC/07 Eximstats: removed tabs and trailing whitespace.
4497 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4499 TK/06 MBOX spool code: Add real "From " MBOX separator line
4500 so the .eml file is really in mbox format (even though
4501 most programs do not really care). Patch from Alex Miller.
4503 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4504 The latter is generated from $received_to and is only set if the
4505 message has one envelope recipient. SA can use these headers,
4506 obviously out-of-the-box. Patch from Alex Miller.
4508 PH/08 The ${def test on a variable was returning false if the variable's
4509 value was "0", contrary to what the specification has always said!
4510 The result should be true unless the variable is empty.
4512 PH/09 The syntax error of a character other than { following "${if
4513 def:variable_name" (after optional whitespace) was not being diagnosed.
4514 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4515 accidental colon was present, for example, could give incorrect results.
4517 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4518 result is used (not including appendfile, where other changes are about
4521 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4522 This involved changing a lot of size variables from int to off_t. It
4523 should work with maildirs and everything.
4525 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4526 spamd dying while we are connected to it.
4528 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4531 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4532 function for BDB 4.3.
4534 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4536 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4537 the sources for PCRE have been re-arranged and more files are now
4540 PH/15 The code I had for printing potentially long long variables in PH/11
4541 above was not the best (it lost precision). The length of off_t variables
4542 is now inspected at build time, and an appropriate printing format (%ld
4543 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4544 to be "long long int" or "long int". This is needed for the internal
4545 formatting function string_vformat().
4547 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4548 the configuration file to be ":syslog", then the script "guesses" where
4549 the logs files are, rather than using the compiled in default. In our
4550 case the guess is not the same as the compiled default, so the script
4551 suddenly stopped working when I started to use syslog. The patch checks
4552 to see if log_file_path is "". If so, it attempts to read it from exim
4553 with no configuration file to get the compiled in version, before it
4554 falls back to the previous guessing code."
4556 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4557 implementing BATV in an Exim configuration. See NewStuff for the gory
4560 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4561 Makefile that are specific to HP-UX.
4563 PH/18 If the "use_postmaster" option was set for a recipient callout together
4564 with the "random" option, the postmaster address was used as the MAIL
4565 FROM address for the random test, but not for the subsequent recipient
4566 test. It is now used for both.
4568 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4569 patch removes a few documentation additions to RFC 3028, because the
4570 latest draft now contains them. It adds the new en;ascii-case comparator
4571 and a new error check for 8bit text in MIME parts. Comparator and
4572 require names are now matched exactly. I enabled the subaddress
4573 extension, but it is not well tested yet (read: it works for me)."
4575 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4576 rework some of the code of TK/09 above to avoid the hardwired use of
4577 "%lld" and "long long". Replaced the call to snprintf() with a call to
4580 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4581 records point to non-existent hosts", "retry timeout exceeded", and
4582 "retry time not reached for any host after a long failure period".
4584 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4585 experimental DomainKeys support:
4587 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4588 (2) On an error such as an illegally used "control", the wrong name for
4589 the control was given.
4591 These problems did NOT occur unless DomainKeys support was compiled.
4593 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4595 PH/24 Added ${if match_ip condition.
4597 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4598 databases so that it will be absolutely obvious if a crash occurs in the
4599 DB library. This is a regular occurrence (often caused by mis-matched
4602 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4603 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4604 picked up on a user's system that detects such things. There doesn't seem
4605 to be a gcc warning option for this - only an attribute that has to be
4606 put on the function's prototype. It seems that in Fedora Core 4 they have
4607 set this on a number of new functions. No doubt there will be more in due
4610 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4611 specify a transport. However, if an address that was verified by such a
4612 router was the subject of a callout, Exim crashed because it tried to
4613 read the rcpt_include_affixes from the non-existent transport. Now it
4614 just assumes that the setting of that option is false. This bug was
4615 introduced by 4.51/PH/31.
4617 PH/28 Changed -d+all to exclude +memory, because that information is very
4618 rarely of interest, but it makes the output a lot bigger. People tend to
4619 do -d+all out of habit.
4621 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4622 code in os-type was giving problems when libc.so lives in lib64, like on
4625 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4626 aren't the modern standard, and it seems that some systems' include files
4627 don't always have them. Exim was already checking for some of the newer
4628 ones like T_AAAA, and defining it itself. I've added checks for all the
4629 record types that Exim uses.
4631 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4632 not automatically generating a new one, as it is supposed to. This
4633 prevented TLS from working. If the file did exist, but contained invalid
4634 data, a new version was generated, as expected. It was only the case of a
4635 non-existent file that was broken.
4637 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4638 with a change in libdomainkeys > 0.64.
4640 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4641 from DNS. If the selector record carries the flag, it now has
4642 precedence over the domain-wide flag.
4644 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4646 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4647 the use of a port name that isn't defined in /etc/services) occurred, the
4648 message was deferred as in a normal delivery, and thus remained on the
4649 spool, instead of being failed because of the mua_wrapper setting. This
4650 is now fixed, and I tidied up some of the mua_wrapper messages at the
4653 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4654 the messages in a hash of arrays rather than using individual hashes.
4655 This is a bit cleaner and results in dramatic memory savings, albeit
4656 at a slight CPU cost.
4658 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4659 as requested by Marc Sherman.
4661 SC/10 Eximstats: added histograms for user specified patterns as requested
4664 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4666 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4667 fopen() in the content-scanning modules that did not already have it.
4673 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4674 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4676 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4678 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4680 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4681 to test Sieve filters that use "vacation".
4683 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4684 that changes the way the GnuTLS parameters are stored in the cache file.
4685 The new format can be generated externally. For backward compatibility,
4686 if the data in the cache doesn't make sense, Exim assumes it has read an
4687 old-format file, and it generates new data and writes a new file. This
4688 means that you can't go back to an older release without removing the
4691 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4692 work if there are any delivery delays because "one_time" forces the
4693 parent to be marked "delivered", so its unseen clone is never tried
4694 again. For this reason, Exim now forbids the simultaneous setting of
4697 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4698 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4699 entry if you want to know the details. The fix, however, neglected to
4700 consider the case when local delivery batching is involved. The test for
4701 "previously delivered" was not happening when checking to see if an
4702 address could be batched with a previous (undelivered) one; under
4703 certain circumstances this could lead to multiple deliveries to the same
4706 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4707 in its include files, and this causes problems building Exim.
4709 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4710 header_syntax) but Exim was just ignoring anything given after a slash.
4711 In particular, this caused confusion with an attempt to use "verify =
4712 reverse_host_lookup/defer_ok". An error is now given when options are
4713 supplied for verify items that do not have them. (Maybe reverse_host_
4714 lookup should have a defer_ok option, but that's a different point.)
4716 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4717 defined by RFC 821) to 2048, because there were problems with some AUTH
4718 commands, and RFC 1869 says the size should be increased for extended
4719 SMTP commands that take arguments.
4721 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4724 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4725 "unknown" error; now it says that the functionality isn't in the binary.
4727 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4728 an address' error message when a string expansion fails (syntax or
4729 whatever). Otherwise the password may appear in the log. Following change
4730 PH/42 below, there is no longer a chance of it appearing in a bounce
4733 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4735 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4736 of its addresses was ever tried. (Bugzilla bug #2.)
4738 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4739 the result incorrectly in the debug output. (It correctly added a newline
4740 to what was transported.)
4742 TF/01 Added $received_time.
4744 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4745 commented out examples of how to interface to a virus scanner and to
4746 SpamAssassin. Also added commented examples of av_scanner and
4747 spamd_address settings.
4749 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4750 and controls are allowed in which ACLs. There were a couple of minor
4751 errors. Some of the entries in the conditions table (which is a table of
4752 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4753 negation of where the condition IS allowed.
4755 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4757 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4758 header file does not have a version number, so I've had to invent a new
4759 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4760 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4761 radiusclient), but it was contributed by a Radius user.
4763 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4764 files or directories, for OpenSSL.
4766 PH/19 When an Exim process that is running as root has to create an Exim log
4767 file, it does so in a subprocess that runs as exim:exim so as to get the
4768 ownership right at creation (otherwise, other Exim processes might see
4769 the file with the wrong ownership). There was no test for failure of this
4770 fork() call, which would lead to the process getting stuck as it waited
4771 for a non-existent subprocess. Forks do occasionally fail when resources
4772 run out. I reviewed all the other calls to fork(); they all seem to check
4775 PH/20 When checking for unexpected SMTP input at connect time (before writing
4776 the banner), Exim was not dealing correctly with a non-positive return
4777 from the read() function. If the client had disconnected by this time,
4778 the result was a log entry for a synchronization error with an empty
4779 string after "input=" when read() returned zero. If read() returned -1
4780 (an event I could not check), uninitialized data bytes were printed.
4781 There were reports of junk text (parts of files, etc) appearing after
4784 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4786 PH/22 Added support for macro redefinition, and (re)definition in between
4787 driver and ACL definitions.
4789 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4790 forgetting to use the resulting value; it was using the unexpanded value.
4792 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4793 hadn't been configured. The fix is from Juergen Kreileder, who
4794 understands it better than I do:
4796 "Here's what I see happening with three configured cyrus_sasl
4797 authenticators configured (plain, login, cram-md5):
4799 On startup auth_cyrus_sasl_init() gets called for each of these.
4800 This means three calls to sasl_listmech() without a specified mech_list.
4801 => SASL tests which mechs of all available mechs actually work
4802 => three warnings about OTP not working
4803 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4805 With the patch, sasl_listmech() also gets called three times. But now
4806 SASL's mech_list option is set to the server_mech specified in the the
4807 authenticator. Or in other words, the answer from sasl_listmech()
4808 gets limited to just the mech you're testing for (which is different
4810 => the return list contains just 'plain' or 'login', 'cram-md5' or
4811 nothing depending on the value of ob->server_mech.
4813 I've just tested the patch: Authentication still works fine,
4814 unavailable mechs specified in the exim configuration are still
4815 caught, and the auth.log warnings about OTP are gone."
4817 PH/25 When debugging is enabled, the contents of the command line are added
4818 to the debugging output, even when log_selector=+arguments is not
4821 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4822 answer is "GNU", and only if the return is "GNU/something" is the answer
4825 PH/27 $acl_verify_message is now set immediately after the failure of a
4826 verification in an ACL, and so is available in subsequent modifiers. In
4827 particular, the message can be preserved by coding like this:
4829 warn !verify = sender
4830 set acl_m0 = $acl_verify_message
4832 Previously, $acl_verify_message was set only while expanding "message"
4833 and "log_message" when a very denied access.
4835 PH/28 Modified OS/os.c-Linux with
4837 -#ifndef OS_LOAD_AVERAGE
4838 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4840 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4841 nomenclature these days.)
4843 PH/29 Installed patch from the Sieve maintainer that adds the options
4844 sieve_useraddress and sieve_subaddress to the redirect router.
4846 PH/30 In these circumstances:
4847 . Two addresses routed to the same list of hosts;
4848 . First host does not offer TLS;
4849 . First host accepts first address;
4850 . First host gives temporary error to second address;
4851 . Second host offers TLS and a TLS session is established;
4852 . Second host accepts second address.
4853 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4854 and peerdn, if requested) that were in fact used only for the second
4857 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4858 attention to any local part prefix or suffix that was matched by the
4859 router that accepted the address. It now behaves in the same way as it
4860 does for delivery: the affixes are removed from the local part unless
4861 rcpt_include_affixes is set on the transport.
4863 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4864 timeout during the DATA phase of an incoming message.
4866 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4867 applied a patch sanctioned by the Sieve maintainer.
4869 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4870 the uid or gid is negative. A case of a negative gid caused this to be
4871 noticed. The fix allows for either to be negative.
4873 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4874 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4877 PH/36 The change PH/12 above was broken. Fixed it.
4879 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4880 the grounds that routing the same address twice would always produce the
4881 same answer. This might have been true once, but it is certainly no
4882 longer true now. Routing a child address may depend on the previous
4883 routing that produced that child. Some complicated redirection strategies
4884 went wrong when messages had multiple recipients, and made Exim's
4885 behaviour dependent on the order in which the addresses were given.
4887 I have moved the duplicate checking until after the routing is complete.
4888 Exim scans the addresses that are assigned to local and remote
4889 transports, and removes any duplicates. This means that more work will be
4890 done, as duplicates will always all be routed, but duplicates are
4891 presumably rare, so I don't expect this is of any significance.
4893 For deliveries to pipes, files, and autoreplies, the duplicate checking
4894 still happens during the routing process, since they are not going to be
4897 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4898 It corrects a timeout issue with spamd. This is Ian's comment: "The
4899 background is that sometimes spamd either never reads data from a
4900 connection it has accepted, or it never writes response data. The exiscan
4901 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4902 blindly assumes that writes won't block so it may never time out."
4904 PH/39 Allow G after quota size as well as K and M.
4906 PH/40 The value set for $authenticated_id in an authenticator may not contain
4907 binary zeroes or newlines because the value is written to log lines and
4908 to spool files. There was no check on this. Now the value is run through
4909 the string_printing() function so that such characters are converted to
4910 printable escape sequences.
4912 PH/41 $message_linecount is a new variable that contains the total number of
4913 lines in the message. Compare $body_linecount, which is the count for the
4916 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4917 bounce and delay warning messages, except in certain special cases, which
4920 (a) An SMTP error message from a remote host;
4921 (b) A message specified in a :fail: redirection;
4922 (c) A message specified in a "fail" command in a system filter;
4923 (d) A message specified in a FAIL return from the queryprogram router;
4924 (e) A message specified by the cannot_route_message router option.
4926 In these cases only, Exim does include the error details in bounce and
4927 warning messages. There are also a few cases where bland messages such
4928 as "unrouteable address" or "local delivery error" are given.
4930 PH/43 $value is now also set for the "else" part of a ${run expansion.
4932 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4933 being worked on, but at least Exim now implements the latest version to
4936 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4937 process to complete was treated as a delivery failure, a timeout while
4938 writing the message to the pipe was logged, but erroneously treated as a
4939 successful delivery. Such timeouts include transport filter timeouts. For
4940 consistency with the overall process timeout, these timeouts are now
4941 treated as errors, giving rise to delivery failures by default. However,
4942 there is now a new Boolean option for the pipe transport called
4943 timeout_defer, which, if set TRUE, converts the failures into defers for
4944 both kinds of timeout. A transport filter timeout is now identified in
4947 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4948 systems where "make" and "gmake" are different, calling "gmake" at top
4949 level broke things. I've arranged for the value of $(MAKE) to be passed
4950 from the Makefile to this script so that it can call the same version of
4954 A note about Exim versions 4.44 and 4.50
4955 ----------------------------------------
4957 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4958 changes of various kinds. As a consequence, a big documentation update was
4959 needed. This delayed the release for rather longer than seemed good, especially
4960 in the light of a couple of (minor) security issues. Therefore, the changes
4961 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4962 release. So 4.44 and 4.50 are in effect two different branches that both start
4965 I have left the 4.50 change log unchanged; it contains all the changes since
4966 4.43. The change log for 4.44 is below; many of its items are identical to
4967 those for 4.50. This seems to be the most sensible way to preserve the
4968 historical information.
4974 1. Minor wording change to the doc/README.SIEVE file.
4976 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4977 computation of the current number of files was incorrect.
4979 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4980 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4983 4. Give more explanation in the error message when the command for a transport
4984 filter fails to execute.
4986 5. There are several places where Exim runs a non-Exim command in a
4987 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4988 was being done only for the command run by the queryprogram router. It is
4989 now done for all such subprocesses. The other cases are: ${run, transport
4990 filters, and the commands run by the lmtp and pipe transports.
4992 6. Added CONFIGURE_GROUP build-time option.
4994 7. Some older OS have a limit of 256 on the maximum number of file
4995 descriptors. Exim was using setrlimit() to set 1000 as a large value
4996 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4997 systems. I've change it so that if it can't get 1000, it tries for 256.
4999 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5000 was an oversight, and furthermore, ever since the addition of extra
5001 controls (e.g. 4.43/32), the checks on when to allow different forms of
5002 "control" were broken. There should now be diagnostics for all cases when a
5003 control that does not make sense is encountered.
5005 9. Added the /retain_sender option to "control=submission".
5007 10. $recipients is now available in the predata ACL (oversight).
5009 11. Tidy the search cache before the fork to do a delivery from a message
5010 received from the command line. Otherwise the child will trigger a lookup
5011 failure and thereby defer the delivery if it tries to use (for example) a
5012 cached ldap connection that the parent has called unbind on.
5014 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5015 of $address_data from the recipient verification was clobbered by the
5016 sender verification.
5018 13. The value of address_data from a sender verification is now available in
5019 $sender_address_data in subsequent conditions in the ACL statement.
5021 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5023 15. Added a new option "connect=<time>" to callout options, to set a different
5026 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5027 was its contents. (It was OK if the option was not defined at all.)
5029 17. A "Completed" log line is now written for messages that are removed from
5030 the spool by the -Mrm option.
5032 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5033 information about exactly what failed.
5035 19. Added -dd to debug only the daemon process.
5037 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5038 handles timeouts, both on the server side and network timeouts. Renamed the
5039 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5041 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5042 It is now set to "smtps".
5044 22. $host_address is now set to the target address during the checking of
5045 ignore_target_hosts.
5047 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5048 being passed; this would have caused $sender_host_name to have been used if
5049 matching the list had actually called for a host name (not very likely,
5050 since this list is usually IP addresses). A host name is now passed as
5053 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5054 code that specifies a non-restarting handler (typically sigaction() in
5055 modern systems) in an attempt to fix a rare and obscure crash bug.
5057 25. Narrowed the window for a race in the daemon that could cause it to ignore
5058 SIGCHLD signals. This is not a major problem, because they are used only to
5059 wake it up if nothing else does.
5061 26. A malformed maildirsize file could cause Exim to calculate negative values
5062 for the mailbox size or file count. Odd effects could occur as a result.
5063 The maildirsize information is now recalculated if the size or filecount
5066 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5067 support for a long time. Removed HAVE_SYS_VFS_H.
5069 28. Installed the latest version of exipick from John Jetmore.
5071 29. In an address list, if the pattern was not a regular expression, an empty
5072 subject address (from a bounce message) matched only if the pattern was an
5073 empty string. Non-empty patterns were not even tested. This was the wrong
5074 because it is perfectly reasonable to use an empty address as part of a
5075 database query. An empty address is now tested by patterns that are
5076 lookups. However, all the other forms of pattern expect the subject to
5077 contain a local part and a domain, and therefore, for them, an empty
5078 address still always fails if the pattern is not itself empty.
5080 30. Exim went into a mad DNS loop when attempting to do a callout where the
5081 host was specified on an smtp transport, and looking it up yielded more
5082 than one IP address.
5084 31. Re-factored the code for checking spool and log partition space into a
5085 function that finds that data and another that does the check. The former
5086 is then used to implement four new variables: $spool_space, $log_space,
5087 $spool_inodes, and $log_inodes.
5089 32. The RFC2047 encoding function was originally intended for short strings
5090 such as real names; it was not keeping to the 75-character limit for
5091 encoded words that the RFC imposes. It now respects the limit, and
5092 generates multiple encoded words if necessary. To be on the safe side, I
5093 have increased the buffer size for the ${rfc2047: expansion operator from
5096 33. It is now permitted to omit both strings after an "if" condition; if the
5097 condition is true, the result is "true". As before, when the second string
5098 is omitted, a false condition yields an empty string. This makes it less
5099 cumbersome to write custom ACL and router conditions.
5101 34. Failure to deliver a bounce message always caused it to be frozen, even if
5102 there was an errors_to setting on the router. The errors_to setting is now
5105 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5106 canonical form (fully expanded) before being placed in
5107 $sender_host_address.
5109 36. The table in the code that translates DNS record types into text (T_A to
5110 "A" for instance) was missing entries for NS and CNAME. It is just possible
5111 that this could have caused confusion if both these types were looked up
5112 for the same domain, because the text type is used as part of Exim's
5113 per-process caching. But the chance of anyone hitting this buglet seems
5116 37. The dnsdb lookup has been extended in a number of ways.
5118 (1) There is a new type, "zns", which walks up the domain tree until it
5119 finds some nameserver records. It should be used with care.
5121 (2) There is a new type, "mxh", which is like "mx" except that it returns
5122 just the host names, not the priorities.
5124 (3) It is now possible to give a list of domains (or IP addresses) to be
5125 looked up. The behaviour when one of the lookups defers can be
5126 controlled by a keyword.
5128 (4) It is now possible to specify the separator character for use when
5129 multiple records are returned.
5131 38. The dnslists ACL condition has been extended: it is now possible to supply
5132 a list of IP addresses and/or domains to be looked up in a particular DNS
5135 39. Added log_selector=+queue_time_overall.
5137 40. When running the queue in the test harness, wait just a tad after forking a
5138 delivery process, to get repeatability of debugging output.
5140 41. Include certificate and key file names in error message when GnuTLS fails
5141 to set them up, because the GnuTLS error message doesn't include the name
5142 of the failing file when there is a problem reading it.
5144 42. Allow both -bf and -bF in the same test run.
5146 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5148 44. The "Exiscan patch" is now merged into the mainline Exim source.
5150 45. Sometimes the final signoff response after QUIT could fail to get
5151 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5152 < 0 before doing a fflush(). This bug looks as though it goes back to the
5153 introduction of TLS in release 3.20, but "sometimes" must have been rare
5154 because the tests only now provoked it.
5156 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5157 (this can affect the format of dates).
5159 47. exim_tidydb, when checking for the continued existence of a message for
5160 which it has found a message-specific retry record, was not finding
5161 messages that were in split spool directories. Consequently, it was
5162 deleting retry records that should have stayed in existence.
5164 48. Steve fixed some bugs in eximstats.
5166 49. The SPA authentication driver was not abandoning authentication and moving
5167 on to the next authenticator when an expansion was forced to fail,
5168 contradicting the general specification for all authenticators. Instead it
5169 was generating a temporary error. It now behaves as specified.
5171 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5172 (the order specifies the preference for clients). The order is now AES256,
5173 AES128, 3DES, ARCFOUR128.
5175 51. Small patch to Sieve code - explicitly set From: when generating an
5178 52. Exim crashed if a remote delivery caused a very long error message to be
5179 recorded - for instance if somebody sent an entire SpamAssassin report back
5180 as a large number of 550 error lines. This bug was coincidentally fixed by
5181 increasing the size of one of Exim's internal buffers (big_buffer) that
5182 happened as part of the Exiscan merge. However, to be on the safe side, I
5183 have made the code more robust (and fixed the comments that describe what
5186 53. Now that there can be additional text after "Completed" in log lines (if
5187 the queue_time_overall log selector is set), a one-byte patch to exigrep
5188 was needed to allow it to recognize "Completed" as not the last thing in
5191 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5192 patch that reportedly fixes this has been added. I am not expert enough to
5193 create a test for it. This is what the patch creator wrote:
5195 "I found a little strange behaviour of ldap code when working with
5196 Windows 2003 AD Domain, where users was placed in more than one
5197 Organization Units. When I tried to give exim partial DN, the exit code
5198 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5199 But simultaneously result of request was absolutely normal ldap result,
5200 so I produce this patch..."
5202 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5203 REFERENCE, so I have modified the code to exclude the patch when that macro
5206 55. Some experimental protocols are using DNS PTR records for new purposes. The
5207 keys for these records are domain names, not reversed IP addresses. The
5208 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5209 leaves it alone. Component reversal etc. now happens only for IP addresses.
5212 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5214 57. Double the size of the debug message buffer (to 2048) so that more of very
5215 long debug lines gets shown.
5217 58. The exicyclog utility now does better if the number of log files to keep
5218 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5220 59. Two changes related to the smtp_active_hostname option:
5222 (1) $smtp_active_hostname is now available as a variable.
5223 (2) The default for smtp_banner uses $smtp_active_hostname instead
5224 of $primary_hostname.
5226 60. The host_aton() function is supposed to be passed a string that is known
5227 to be a valid IP address. However, in the case of IPv6 addresses, it was
5228 not checking this. This is a hostage to fortune. Exim now panics and dies
5229 if the condition is not met. A case was found where this could be provoked
5230 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5231 components; fortuitously, this particular loophole had already been fixed
5232 by change 4.50/55 above.
5234 If there are any other similar loopholes, the new check in host_aton()
5235 itself should stop them being exploited. The report I received stated that
5236 data on the command line could provoke the exploit when Exim was running as
5237 exim, but did not say which command line option was involved. All I could
5238 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5239 running as the user.
5242 61. There was a buffer overflow vulnerability in the SPA authentication code
5243 (which came originally from the Samba project). I have added a test to the
5244 spa_base64_to_bits() function which I hope fixes it.
5247 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5248 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5250 63. The daemon start-up calls getloadavg() while still root for those OS that
5251 need the first call to be done as root, but it missed one case: when
5252 deliver_queue_load_max is set with deliver_drop_privilege. This is
5253 necessary for the benefit of the queue runner, because there is no re-exec
5254 when deliver_drop_privilege is set.
5256 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5257 This has been fixed.
5259 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5260 was in use, was not putting the data itself into the right store pool;
5261 consequently, it could be overwritten for a subsequent message in the same
5262 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5265 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5267 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5268 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5269 the function were treating the return as a boolean value, which happened to
5270 work because 0=false and not-0=true, but is not correct code.
5272 68. The host_aton() function was not handling scoped IPv6 addresses (those
5273 with, for example, "%eth0" on the end) correctly.
5275 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5276 negated items (that is, ~something) in unsigned ints. Some compilers
5277 apparently mutter when there is no cast.
5279 70. If an address verification called from an ACL failed, and did not produce a
5280 user-specific message (i.e. there was only a "system" message), nothing was
5281 put in $acl_verify_message. In this situation, it now puts the system
5284 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5285 session; change 4.31/43 added the unwanted input to the log line - except
5286 that it did not do this in the start of session case. It now does.
5288 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5289 This is wrong and can cause the other end to generate a synchronization
5290 error if it is another Exim or anything else that does the synchronization
5291 check. A QUIT command is no longer sent after a timeout.
5293 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5294 during host lookups.
5296 74. The defer_ok option of callout verification was not working if it was used
5297 when verifying addresses in header lines, that is, for this case:
5299 verify = header_sender/callout=defer_ok
5301 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5302 those file descriptors could be used for SMTP connections. If anything
5303 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5304 could be sent to the SMTP client, causing chaos. The daemon now opens
5305 stdin, stdout, and stderr to /dev/null when it puts itself into the
5308 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5309 log by default. The user can override this with suitable Perl magic.
5311 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5312 the log message when discard triggers, was not working for the DATA ACL or
5313 for the non-SMTP ACL.
5315 78. Error message wording change in sieve.c.
5317 79. If smtp_accept_max_per_host was set, the number of connections could be
5318 restricted to fewer than expected, because the daemon was trying to set up
5319 a new connection before checking whether the processes handling previous
5320 connections had finished. The check for completed processes is now done
5321 earlier. On busy systems, this bug wouldn't be noticed because something
5322 else would have woken the daemon, and it would have reaped the completed
5325 80. If a message was submitted locally by a user whose login name contained one
5326 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5327 It caused a spool format error. I have fixed the spool reading code. A
5328 related problem was that the "from" clause in the Received: line became
5329 illegal because of the space(s). It is now covered by ${quote_local_part.
5331 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5334 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5336 83. Patch from Sieve maintainer to fix unterminated string problem in
5337 "vacation" handling.
5339 84. Some minor changes to the Linux configuration files to help with other
5340 OS variants using glibc.
5342 85. One more patch for Sieve to update vacation handling to latest spec.
5345 ----------------------------------------------------
5346 See the note above about the 4.44 and 4.50 releases.
5347 ----------------------------------------------------
5353 1. Change 4.43/35 introduced a bug that caused file counts to be
5354 incorrectly computed when quota_filecount was set in an appendfile
5357 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5358 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5361 3. Give more explanation in the error message when the command for a transport
5362 filter fails to execute.
5364 4. There are several places where Exim runs a non-Exim command in a
5365 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5366 was being done only for the command run by the queryprogram router. It is
5367 now done for all such subprocesses. The other cases are: ${run, transport
5368 filters, and the commands run by the lmtp and pipe transports.
5370 5. Some older OS have a limit of 256 on the maximum number of file
5371 descriptors. Exim was using setrlimit() to set 1000 as a large value
5372 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5373 systems. I've change it so that if it can't get 1000, it tries for 256.
5375 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5376 was an oversight, and furthermore, ever since the addition of extra
5377 controls (e.g. 4.43/32), the checks on when to allow different forms of
5378 "control" were broken. There should now be diagnostics for all cases when a
5379 control that does not make sense is encountered.
5381 7. $recipients is now available in the predata ACL (oversight).
5383 8. Tidy the search cache before the fork to do a delivery from a message
5384 received from the command line. Otherwise the child will trigger a lookup
5385 failure and thereby defer the delivery if it tries to use (for example) a
5386 cached ldap connection that the parent has called unbind on.
5388 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5389 of $address_data from the recipient verification was clobbered by the
5390 sender verification.
5392 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5393 was its contents. (It was OK if the option was not defined at all.)
5395 11. A "Completed" log line is now written for messages that are removed from
5396 the spool by the -Mrm option.
5398 12. $host_address is now set to the target address during the checking of
5399 ignore_target_hosts.
5401 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5402 being passed; this would have caused $sender_host_name to have been used if
5403 matching the list had actually called for a host name (not very likely,
5404 since this list is usually IP addresses). A host name is now passed as
5407 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5408 code that specifies a non-restarting handler (typically sigaction() in
5409 modern systems) in an attempt to fix a rare and obscure crash bug.
5411 15. Narrowed the window for a race in the daemon that could cause it to ignore
5412 SIGCHLD signals. This is not a major problem, because they are used only to
5413 wake it up if nothing else does.
5415 16. A malformed maildirsize file could cause Exim to calculate negative values
5416 for the mailbox size or file count. Odd effects could occur as a result.
5417 The maildirsize information is now recalculated if the size or filecount
5420 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5421 support for a long time. Removed HAVE_SYS_VFS_H.
5423 18. Updated exipick to current release from John Jetmore.
5425 19. Allow an empty sender to be matched against a lookup in an address list.
5426 Previously the only cases considered were a regular expression, or an
5429 20. Exim went into a mad DNS lookup loop when doing a callout where the
5430 host was specified on the transport, if the DNS lookup yielded more than
5433 21. The RFC2047 encoding function was originally intended for short strings
5434 such as real names; it was not keeping to the 75-character limit for
5435 encoded words that the RFC imposes. It now respects the limit, and
5436 generates multiple encoded words if necessary. To be on the safe side, I
5437 have increased the buffer size for the ${rfc2047: expansion operator from
5440 22. Failure to deliver a bounce message always caused it to be frozen, even if
5441 there was an errors_to setting on the router. The errors_to setting is now
5444 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5445 canonical form (fully expanded) before being placed in
5446 $sender_host_address.
5448 24. Updated eximstats to version 1.33
5450 25. Include certificate and key file names in error message when GnuTLS fails
5451 to set them up, because the GnuTLS error message doesn't include the name
5452 of the failing file when there is a problem reading it.
5454 26. Expand error message when OpenSSL has problems setting up cert/key files.
5457 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5458 (this can affect the format of dates).
5460 28. exim_tidydb, when checking for the continued existence of a message for
5461 which it has found a message-specific retry record, was not finding
5462 messages that were in split spool directories. Consequently, it was
5463 deleting retry records that should have stayed in existence.
5465 29. eximstats updated to version 1.35
5466 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5467 1.35 - bugfix such that pie charts by volume are generated correctly
5469 30. The SPA authentication driver was not abandoning authentication and moving
5470 on to the next authenticator when an expansion was forced to fail,
5471 contradicting the general specification for all authenticators. Instead it
5472 was generating a temporary error. It now behaves as specified.
5474 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5475 (the order specifies the preference for clients). The order is now AES256,
5476 AES128, 3DES, ARCFOUR128.
5478 31. Small patch to Sieve code - explicitly set From: when generating an
5481 32. Exim crashed if a remote delivery caused a very long error message to be
5482 recorded - for instance if somebody sent an entire SpamAssassin report back
5483 as a large number of 550 error lines. This bug was coincidentally fixed by
5484 increasing the size of one of Exim's internal buffers (big_buffer) that
5485 happened as part of the Exiscan merge. However, to be on the safe side, I
5486 have made the code more robust (and fixed the comments that describe what
5489 33. Some experimental protocols are using DNS PTR records for new purposes. The
5490 keys for these records are domain names, not reversed IP addresses. The
5491 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5492 leaves it alone. Component reversal etc. now happens only for IP addresses.
5495 34. The host_aton() function is supposed to be passed a string that is known
5496 to be a valid IP address. However, in the case of IPv6 addresses, it was
5497 not checking this. This is a hostage to fortune. Exim now panics and dies
5498 if the condition is not met. A case was found where this could be provoked
5499 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5500 components; fortuitously, this particular loophole had already been fixed
5501 by change 4.50/55 or 4.44/33 above.
5503 If there are any other similar loopholes, the new check in host_aton()
5504 itself should stop them being exploited. The report I received stated that
5505 data on the command line could provoke the exploit when Exim was running as
5506 exim, but did not say which command line option was involved. All I could
5507 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5508 running as the user.
5511 35. There was a buffer overflow vulnerability in the SPA authentication code
5512 (which came originally from the Samba project). I have added a test to the
5513 spa_base64_to_bits() function which I hope fixes it.
5516 36. The daemon start-up calls getloadavg() while still root for those OS that
5517 need the first call to be done as root, but it missed one case: when
5518 deliver_queue_load_max is set with deliver_drop_privilege. This is
5519 necessary for the benefit of the queue runner, because there is no re-exec
5520 when deliver_drop_privilege is set.
5522 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5523 was in use, was not putting the data itself into the right store pool;
5524 consequently, it could be overwritten for a subsequent message in the same
5525 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5528 38. Sometimes the final signoff response after QUIT could fail to get
5529 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5530 < 0 before doing a fflush(). This bug looks as though it goes back to the
5531 introduction of TLS in release 3.20, but "sometimes" must have been rare
5532 because the tests only now provoked it.
5538 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5539 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5540 Now it returns an integer. A number of calls were still expecting a T/F
5541 return. Fortuitously, in all cases, the tests worked in OK situations,
5542 which is the norm. However, things would have gone wrong on any write
5543 failures on the smtp file descriptor. This function is used when sending
5544 messages over SMTP and also when doing verify callouts.
5546 2. When Exim is called to do synchronous delivery of a locally submitted
5547 message (the -odf or -odi options), it no longer closes stderr before doing
5550 3. Implemented the mua_wrapper option.
5552 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5554 5. Implemented the functions header_remove(), header_testname(),
5555 header_add_at_position(), and receive_remove_recipient(), and exported them
5558 6. If an ACL "warn" statement specified the addition of headers, Exim already
5559 inserted X-ACL-Warn: at the start if there was no header name. However, it
5560 was not making this test for the second and subsequent header lines if
5561 there were newlines in the string. This meant that an invalid header could
5562 be inserted if Exim was badly configured.
5564 7. Allow an ACL "warn" statement to add header lines at the start or after all
5565 the Received: headers, as well as at the end.
5567 8. Added the rcpt_4xx retry error code.
5569 9. Added postmaster_mailfrom=xxx to callout verification option.
5571 10. Added mailfrom=xxxx to the callout verification option, for verify=
5574 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5575 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5577 12. Inserted some casts to stop certain compilers warning when using pointer
5578 differences as field lengths or precisions in printf-type calls (mostly
5579 affecting debugging statements).
5581 13. Added optional readline() support for -be (dynamically loaded).
5583 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5584 same clock tick as a message's arrival, so that its received time was the
5585 same as the "first fail" time on the retry record, and that message
5586 remained on the queue past the ultimate address timeout, every queue runner
5587 would try a delivery (because it was past the ultimate address timeout) but
5588 after another failure, the ultimate address timeout, which should have then
5589 bounced the address, did not kick in. This was a "< instead of <=" error;
5590 in most cases the first failure would have been in the next clock tick
5591 after the received time, and all would be well.
5593 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5594 being recognized when the domain list was tested by the match_domain
5595 condition in an expansion string.
5597 16. Added the ${str2b64: operator.
5599 17. Exim was always calling setrlimit() to set a large limit for the number of
5600 processes, without checking whether the existing limit was already
5601 adequate. (It did check for the limit on file descriptors.) Furthermore,
5602 errors from getrlimit() and setrlimit() were being ignored. Now they are
5603 logged to the main and panic logs, but Exim does carry on, to try to do its
5604 job under whatever limits there are.
5606 18. Imported PCRE 5.0.
5608 19. Trivial typo in log message " temporarily refused connection" (the leading
5611 20. If the log selector return_path_on_delivery was set and an address was
5612 redirected to /dev/null, the delivery process crashed because it assumed
5613 that a return path would always be set for a "successful" delivery. In this
5614 case, the whole delivery is bypassed as an optimization, and therefore no
5617 21. Internal re-arrangement: the function for sending a challenge and reading
5618 a response while authentication was assuming a zero-terminated challenge
5619 string. It's now changed to take a pointer and a length, to allow for
5620 binary data in such strings.
5622 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5624 23. Exim was not respecting finduser_retries when seeking the login of the
5625 uid under which it was called; it was always trying 10 times. (The default
5626 setting of finduser_retries is zero.) Also, it was sleeping after the final
5627 failure, which is pointless.
5629 24. Implemented tls_on_connect_ports.
5631 25. Implemented acl_smtp_predata.
5633 26. If the domain in control=submission is set empty, Exim assumes that the
5634 authenticated id is a complete email address when it generates From: or
5635 Sender: header lines.
5637 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5638 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5639 chown and chgrp in /bin and hostname in /usr/bin.
5641 28. Exim was keeping the "process log" file open after each use, just as it
5642 does for the main log. This opens the possibility of it remaining open for
5643 long periods when the USR1 signal hits a daemon. Occasional processlog
5644 errors were reported, that could have been caused by this. Anyway, it seems
5645 much more sensible not to leave this file open at all, so that is what now
5648 29. The long-running daemon process does not normally write to the log once it
5649 has entered its main loop, and it closes the log before doing so. This is
5650 so that log files can straightforwardly be renamed and moved. However,
5651 there are a couple of unusual error situations where the daemon does write
5652 log entries, and I had neglected to close the log afterwards.
5654 30. The text of an SMTP error response that was received during a remote
5655 delivery was being truncated at 512 bytes. This is too short for some of
5656 the long messages that one sometimes sees. I've increased the limit to
5659 31. It is now possible to make retry rules that apply only when a message has a
5660 specific sender, in particular, an empty sender.
5662 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5663 it possible to be selective about when SMTP synchronization is enforced.
5665 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5667 32. Implemented hosts_connection_nolog.
5669 33. Added an ACL for QUIT.
5671 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5674 35. Added mailbox_size and mailbox_filecount to appendfile.
5676 36. Added control = no_multiline_responses to ACLs.
5678 37. There was a bug in the logic of the code that waits for the clock to tick
5679 in the case where the clock went backwards by a substantial amount such
5680 that the microsecond fraction of "now" was more than the microsecond
5681 fraction of "then" (but the whole seconds number was less).
5683 38. Added support for the libradius Radius client library this is found on
5684 FreeBSD (previously only the radiusclient library was supported).
5690 1. When certain lookups returned multiple values in the form name=value, the
5691 quoting of the values was not always being done properly. Specifically:
5692 (a) If the value started with a double quote, but contained no whitespace,
5694 (b) If the value contained whitespace other than a space character (i.e.
5695 tabs or newlines or carriage returns) it was not quoted.
5696 This fix has been applied to the mysql and pgsql lookups by writing a
5697 separate quoting function and calling it from the lookup code. The fix
5698 should probably also be applied to nisplus, ibase and oracle lookups, but
5699 since I cannot test any of those, I have not disturbed their existing code.
5701 2. A hit in the callout cache for a specific address caused a log line with no
5702 reason for rejecting RCPT. Now it says "Previous (cached) callout
5703 verification failure".
5705 3. There was an off-by-one bug in the queryprogram router. An over-long
5706 return line was truncated at 256 instead of 255 characters, thereby
5707 overflowing its buffer with the terminating zero. As well as fixing this, I
5708 have increased the buffer size to 1024 (and made a note to document this).
5710 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5711 when Exim is waiting for an SMTP response from a remote server, Exim
5712 restarts its select() call on the socket, thereby resetting its timeout.
5713 This is not a problem when such interrupts are rare. Somebody set up a cron
5714 job to run exiwhat every 2 minutes, which is less than the normal select()
5715 timeout (5 or 10 minutes). This meant that the select() timeout never
5716 kicked in because it was always reset. I have fixed this by comparing the
5717 time when an interrupt arrives with the time at the start of the first call
5718 to select(). If more time than the timeout has elapsed, the interrupt is
5719 treated as a timeout.
5721 5. Some internal re-factoring in preparation for the addition of Sieve
5722 extensions (by MH). In particular, the "personal" test is moved to a
5723 separate function, and given an option for scanning Cc: and Bcc: (which is
5724 not set for Exim filters).
5726 6. When Exim created an email address using the login of the caller as the
5727 local part (e.g. when creating a From: or Sender: header line), it was not
5728 quoting the local part when it contained special characters such as @.
5730 7. Installed new OpenBSD configuration files.
5732 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5733 try to make them clearer.
5735 9. Callout options, other than the timeout value, were being ignored when
5736 verifying sender addresses in header lines. For example, when using
5738 verify = header_sender/callout=no_cache
5740 the cache was (incorrectly) being used.
5742 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5743 only the Cygwin environment.
5745 11. When return_path_on_delivery was set as a log selector, if different remote
5746 addresses in the same message used different return paths and parallel
5747 remote delivery occurred, the wrong values would sometimes be logged.
5748 (Whenever a remote delivery process finished, the return path value from
5749 the most recently started remote delivery process was logged.)
5751 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5752 lines when AUTH and/or TLS are in use. This is the "received protocol"
5753 field. Exim used to use "asmtp" for authenticated SMTP, without any
5754 indication (in the protocol name) for TLS use. Now it follows the RFC and
5755 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5756 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5757 names appear in log lines as well as in Received: header lines.
5759 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5760 extensions, and comparison tests, and to fix some bugs.
5762 14. Changes to the "personal" filter test:
5764 (1) The test was buggy in that it was just doing the equivalent of
5765 "contains" tests on header lines. For example, if a user's address was
5766 anne@some.where, the "personal" test would incorrectly be true for
5768 To: susanne@some.where
5770 This test is now done by extracting each address from the header in turn,
5771 and checking the entire address. Other tests that are part of "personal"
5772 are now done using regular expressions (for example, to check local parts
5773 of addresses in From: header lines).
5775 (2) The list of non-personal local parts in From: addresses has been
5776 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5777 taken from the Sieve specification recommendations.
5779 (3) If the message contains any header line starting with "List-" it is
5780 treated as non-personal.
5782 (4) The test for "circular" in the Subject: header line has been removed
5783 because it now seems ill-conceived.
5785 15. Minor typos in src/EDITME comments corrected.
5787 16. Installed latest exipick from John Jetmore.
5789 17. If headers_add on a router specified a text string that was too long for
5790 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5791 of string_sprintf() is now avoided.
5793 18. $message_body_size was not set (it was always zero) when running the DATA
5794 ACL and the local_scan() function.
5796 19. For the "mail" command in an Exim filter, no default was being set for
5797 the once_repeat time, causing a random time value to be used if "once" was
5798 specified. (If the value happened to be <= 0, no repeat happened.) The
5799 default is now 0s, meaning "never repeat". The "vacation" command was OK
5800 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5801 (I found it when inspecting the code).
5803 20. There is now an overall timeout for performing a callout verification. It
5804 defaults to 4 times the callout timeout, which applies to individual SMTP
5805 commands during the callout. The overall timeout applies when there is more
5806 than one host that can be tried. The timeout is checked before trying the
5807 next host. This prevents very long delays if there are a large number of
5808 hosts and all are timing out (e.g. when the network connections are timing
5809 out). The value of the overall timeout can be changed by specifying an
5810 additional sub-option for "callout", called "maxwait". For example:
5812 verify = sender/callout=5s,maxwait=20s
5814 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5815 to the end before writing, but this should make it even safer).
5817 22. Exim was forgetting that it had advertised PIPELINING for the second and
5818 subsequent messages on an SMTP connection. It was also not resetting its
5819 memory on STARTTLS and an internal HELO.
5821 23. When Exim logs an SMTP synchronization error within a session, it now
5822 records whether PIPELINING has been advertised or not.
5824 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5825 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5826 rather than long int.
5828 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5830 26. Added the never_mail option to autoreply.
5836 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5837 crash if the getsockname() call failed; this can happen if a connection is
5838 closed very soon after it is established. The problem was simply in the
5839 order in which certain operations were done, causing Exim to try to write
5840 to the SMTP stream before it had set up the file descriptor. The bug has
5841 been fixed by making things happen in the correct order.
5847 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5848 before the connection was closed, thus losing the rejection response.
5850 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5851 some early Solaris releases, but causes trouble in current releases where
5852 socklen_t is defined.
5854 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5857 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5860 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5861 that releasing the top of it at the end releases what was used for sub-
5862 expansions (unless the block got too big). However, discard this block if
5863 the first thing is a variable or header, so that we can use its block when
5864 it is dynamic (useful for very large $message_headers, for example).
5866 6. Lookups now cache *every* query, not just the most recent. A new, separate
5867 store pool is used for this. It can be recovered when all lookup caches are
5868 flushed. Lookups now release memory at the end of their result strings.
5869 This has involved some general refactoring of the lookup sources.
5871 7. Some code has been added to the store_xxx() functions to reduce the amount
5872 of flapping under certain conditions.
5874 8. log_incoming_interface used to affect only the <= reception log lines. Now
5875 it causes the local interface and port to be added to several more SMTP log
5876 lines, for example "SMTP connection from", and rejection lines.
5878 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5880 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5882 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5883 could be overwritten at the end of the current message (or the start of a
5884 new message if it was set in a HELO ACL). The value is now preserved for
5885 the duration of the SMTP connection.
5887 12. If a transport had a headers_rewrite setting, and a matching header line
5888 contained an unqualified address, that address was qualified, even if it
5889 did not match any rewriting rules. The underlying bug was that the values
5890 of the flags that permit the existence of unqualified sender and recipient
5891 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5892 non-local messages, and by -bnq for local messages) were not being
5893 preserved with the message after it was received.
5895 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5896 "next input=" as part of the text comprising the host identity instead of
5897 the correct text. The code was using the same buffer for two different
5898 strings. However, depending on which order the printing function evaluated
5899 its arguments, the bug did not always show up. Under Linux, for example, my
5900 test suite worked just fine.
5902 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5903 doesn't work with some older versions of Perl. It has been changed to "my",
5904 which in any case is probably the better facility to use.
5906 15. A really picky compiler found some instances of statements for creating
5907 error messages that either had too many or two few arguments for the format
5910 16. The size of the buffer for calls to the DNS resolver has been increased
5911 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5912 for addresses that have a lot of PTR records. This alleviates a problem; it
5913 does not fully solve it.
5915 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5916 buffer now truncates the list and logs the incident, which is the same
5917 action as happens when Exim is looking up a host name and its aliases.
5918 Previously in this situation something unpredictable would happen;
5919 sometimes it was "internal error: store_reset failed".
5921 18. If a server dropped the connection unexpectedly when an Exim client was
5922 using GnuTLS and trying to read a response, the client delivery process
5923 crashed while trying to generate an error log message.
5925 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5926 string, for example:
5928 warn message = H1: something\nH2: something
5930 the text was added as a single header line from Exim's point of view
5931 though it ended up OK in the delivered message. However, searching for the
5932 second and subsequent header lines using $h_h2: did not work. This has been
5933 fixed. Similarly, if a system filter added multiple headers in this way,
5934 the routers could not see them.
5936 20. Expanded the error message when iplsearch is called with an invalid key to
5937 suggest using net-iplsearch in a host list.
5939 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5940 ACLs are no longer actually imposed (and a message to that effect is
5943 22. If a "gecos" field in a passwd entry contained escaped characters, in
5944 particular, if it contained a \" sequence, Exim got it wrong when building
5945 a From: or a Sender: header from that name. A second bug also caused
5946 incorrect handling when an unquoted " was present following a character
5947 that needed quoting.
5949 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5950 was not being matched caselessly.
5952 24. Arranged for all hyphens in the exim.8 source to be escaped with
5955 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5956 properly. Recipient callout cache records were still being keyed to include
5957 the sender, even when use_sender was set false. This led to far more
5958 callouts that were necessary. The sender is no longer included in the key
5959 when use_sender is false.
5961 26. Added "control = submission" modifier to ACLs.
5963 27. Added the ${base62d: operator to decode base 62 numbers.
5965 28. dnsdb lookups can now access SRV records.
5967 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5968 the configuration file.
5970 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5971 (-v) mode. This makes the output for a verbose queue run more intelligible.
5973 31. Added a use_postmaster feature to recipient callouts.
5975 32. Added the $body_zerocount variable, containing the number of binary zero
5976 bytes in the message body.
5978 33. The time of last modification of the "new" subdirectory is now used as the
5979 "mailbox time last read" when there is a quota error for a maildir
5982 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5984 35. Added +ignore_unknown as a special item in host lists.
5986 36. Code for decoding IPv6 addresses in host lists is now included, even if
5987 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5988 address was recognized as an IP address, but was then not correctly decoded
5989 into binary, causing unexpected and incorrect effects when compared with
5996 1. Very minor rewording of debugging text in manualroute to say "list of
5997 hosts" instead of "hostlist".
5999 2. If verify=header_syntax was set, and a header line with an unqualified
6000 address (no domain) and a large number of spaces between the end of the
6001 name and the colon was received, the reception process suffered a buffer
6002 overflow, and (when I tested it) crashed. This was caused by some obsolete
6003 code that should have been removed. The fix is to remove it!
6005 3. When running in the test harness, delay a bit after writing a bounce
6006 message to get a bit more predictability in the log output.
6008 4. Added a call to search_tidyup() just before forking a reception process. In
6009 theory, someone could use a lookup in the expansion of smtp_accept_max_
6010 per_host which, without the tidyup, could leave open a database connection.
6012 5. Added the variables $recipient_data and $sender_data which get set from a
6013 lookup success in an ACL "recipients" or "senders" condition, or a router
6014 "senders" option, similar to $domain_data and $local_part_data.
6016 6. Moved the writing of debug_print from before to after the "senders" test
6019 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6020 problems for message scanning, either using a data ACL, or using
6021 local_scan() because the Received: header was not generated till after they
6022 were called (in order to set the time as the time of reception completion).
6023 I have revised the way this works. The header is now generated after the
6024 body is received, but before the ACL or local_scan() are called. After they
6025 are run, the timestamp in the header is updated.
6031 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6032 before starting a queue runner without re-exec. This happened only when
6033 deliver_drop_privilege was set or when the Exim user was set to root. The
6034 effect of the bug was that timeouts during subsequent deliveries caused
6035 crashes instead of being properly handled. The handler is now left at its
6036 default (and expected) setting.
6038 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6039 message, again when deliver_drop_privilege is set or Exim is run as root.
6040 The bug described in (1) was not present in this case, but the tidying up
6041 of the other signals was missing. I have made the two cases consistent.
6043 3. The ignore_target_hosts setting on a manualroute router was being ignored
6044 for hosts that were looked up using the /MX notation.
6046 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6049 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6050 operated on the sender address. After changing the $sender_address to <>
6051 for the sender address verify, Exim was re-instated it as the original
6052 (before rewriting) address, but remembering that it had rewritten it, so it
6053 wasn't rewriting it again. This bug also had the effect of breaking the
6054 sender address verification caching when the sender address was rewritten.
6056 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6057 This has been changed so that if the ip literal address matches
6058 ignore_target_hosts, the router declines.
6060 7. Added expansion conditions match_domain, match_address, and match_local_
6061 part (NOT match_host).
6063 8. The placeholder for the Received: header didn't have a length field set.
6065 9. Added code to Exim itself and to exim_lock to test for a specific race
6066 condition that could lead to file corruption when using MBX delivery. The
6067 issue is with the lockfile that is created in /tmp. If this file is removed
6068 after a process has opened it but before that process has acquired a lock,
6069 there is the potential for a second process to recreate the file and also
6070 acquire a lock. This could lead to two Exim processes writing to the file
6071 at the same time. The added code performs the same test as UW imapd; it
6072 checks after acquiring the lock that its file descriptor still refers to
6073 the same named file.
6075 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6076 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6079 11. Added the smtp_active_hostname option. If used, this will typically be made
6080 to depend on the incoming interface address. Because $interface_address is
6081 not set up until the daemon has forked a reception process, error responses
6082 that can happen earlier (such as "too many connections") no longer contain
6085 12. If an expansion in a condition on a "warn" statement fails because a lookup
6086 defers, the "warn" statement is abandoned, and the next ACL statement is
6087 processed. Previously this caused the whole ACL to be aborted.
6089 13. Added the iplsearch lookup type.
6091 14. Added ident_timeout as a log selector.
6093 15. Added tls_certificate_verified as a log selector.
6095 16. Added a global option tls_require_ciphers (compare the smtp transport
6096 option of the same name). This controls incoming TLS connections.
6098 17. I finally figured out how to make tls_require_ciphers do a similar thing
6099 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6100 before starting the TLS session.
6102 18. Tabs are now shown as \t in -bP output.
6104 19. If the log selector return_path_on_delivery was set, Exim crashed when
6105 bouncing a message because it had too many Received: header lines.
6107 20. If two routers both had headers_remove settings, and the first one included
6108 a superfluous trailing colon, the final name in the first list and the
6109 first name in the second list were incorrectly joined into one item (with a
6110 colon in the middle).
6116 1. Added -C and -D options to the exinext utility, mainly to make it easier
6117 to include in the automated testing, but these could be helpful when
6118 multiple configurations are in use.
6120 2. The exinext utility was not formatting the output nicely when there was
6121 an alternate port involved in the retry record key, nor when there was a
6122 message id as well (for retries that were specific to a specific message
6123 and a specific host). It was also confused by IPv6 addresses, because of
6124 the additional colons they contain. I have fixed the IPv4 problem, and
6125 patched it up to do a reasonable job for IPv6.
6127 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6128 delivery, the log line now contains "pipelined" if PIPELINING was used.
6130 4. An SMTP transport process used to panic and die if the bind() call to set
6131 an explicit outgoing interface failed. This has been changed; it is now
6132 treated in the same way as a connect() failure.
6134 5. A reference to $sender_host_name in the part of a conditional expansion
6135 that was being skipped was still causing a DNS lookup. This no longer
6138 6. The def: expansion condition was not recognizing references to header lines
6139 that used bh_ and bheader_.
6141 7. Added the _cache feature to named lists.
6143 8. The code for checking quota_filecount in the appendfile transport was
6144 allowing one more file than it should have been.
6146 9. For compatibility with Sendmail, the command line option
6154 and sets the incoming protocol and host name (for trusted callers). The
6155 host name and its colon can be omitted when only the protocol is to be set.
6156 Note the Exim already has two private options, -pd and -ps, that refer to
6157 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6158 "s", but I don't think that's a major issue.
6160 10. A number of refactoring changes to the code, none of which should affect
6163 (a) The number of logging options was getting close to filling up the
6164 32-bit word that was used as a bit map. I have split them into two classes:
6165 those that are passed in the argument to log_write(), and those that are
6166 only ever tested independently outside of that function. These are now in
6167 separate 32-bit words, so there is plenty of room for expansion again.
6168 There is no change in the user interface or the logging behaviour.
6170 (b) When building, for example, log lines, the code previously used a
6171 macro that called string_cat() twice, in order to add two strings. This is
6172 not really sufficiently general. Furthermore, there was one instance where
6173 it was actually wrong because one of the argument was used twice, and in
6174 one call a function was used. (As it happened, calling the function twice
6175 did not affect the overall behaviour.) The macro has been replaced by a
6176 function that can join an arbitrary number of extra strings onto a growing
6179 (c) The code for expansion conditions now uses a table and a binary chop
6180 instead of a serial search (which was left over from when there were very
6181 few conditions). Also, it now recognizes conditions like "pam" even when
6182 the relevant support is not compiled in: a suitably worded error message is
6183 given if an attempt is made to use such a condition.
6185 11. Added ${time_interval:xxxxx}.
6187 12. A bug was causing one of the ddress fields not to be passed back correctly
6188 from remote delivery subprocesses. The field in question was not being
6189 subsequently used, so this caused to problems in practice.
6191 13. Added new log selectors queue_time and deliver_time.
6193 14. Might have fixed a bug in maildirsizefile handling that threw up
6194 "unexpected character" debug warnings, and recalculated the data
6195 unnecessarily. In any case, I expanded the warning message to give more
6198 15. Added the message "Restricted characters in address" to the statements in
6199 the default ACL that block characters like @ and % in local parts.
6201 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6202 Three changes have been made:
6204 (a) There was a serious bug; a negative response to MAIL caused the whole
6205 recipient domain to be cached as invalid, thereby blocking all messages
6206 to all local parts at the same domain, from all senders. This bug has
6207 been fixed. The domain is no longer cached after a negative response to
6208 MAIL if the sender used is not empty.
6210 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6213 (c) A new callout option, "use_sender" has been added for people who want
6214 the modified behaviour.
6220 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6223 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6224 indeed breaks things for older releases.
6226 3. Added additional logging to the case where there is a problem reading data
6227 from a filter that is running in a subprocess using a pipe, in order to
6228 try to track down a specific problem.
6230 4. Testing facility fudge: when running in the test harness and attempting
6231 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6232 getting "No route to host". Convert this to a timeout.
6234 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6237 6. Some OS don't have socklen_t but use size_t instead. This affects the
6238 fifth argument of getsockopt() amongst other things. This is now
6239 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6240 can be set for individual OS. I have set it for SunOS5, OSF1, and
6241 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6242 some earlier ones do not.
6244 7. Change 4.30/15 was not doing the test caselessly.
6246 8. The standard form for an IPv6 address literal was being rejected by address
6247 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6248 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6249 this, as well as the form without the "IPv6" on the front (but only when
6250 address literals are enabled, of course).
6252 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6254 10. Exim crashed if a message with an empty sender address specified by -f
6255 encountered a router with an errors_to setting. This could be provoked only
6256 by a command such as
6260 where an empty string was supplied; "<>" did not hit this bug.
6262 11. Installed PCRE release 4.5.
6264 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6265 remained set. It is now erased.
6267 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6268 times from message ids (which are base 36 rather than the normal 62).
6270 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6271 were being counted as actual protocol errors, and logged if the log
6272 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6273 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6274 and DATA following a set of rejected RCPTs do not count as protocol errors.
6275 In other words, Exim assumes they were pipelined, though this may not
6276 actually be the case. Of course, in all cases the client gets an
6277 appropriate error code.
6279 15. If a lookup fails in an ACL condition, a message about the failure may
6280 be available; it is used if testing the ACL cannot continue, because most
6281 such messages specify what the cause of the deferral is. However, some
6282 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6283 that caused an old message to be retained and used if a later statement
6284 caused a defer, replacing the real cause of the deferral.
6286 16. If an IP address had so many PTR records that the DNS lookup buffer
6287 was not large enough to hold them, Exim could crash while trying to process
6288 the truncated data. It now detects and logs this case.
6290 17. Further to 4.21/58, another change has been made: if (and only if) the
6291 first line of a message (the first header line) ends with CRLF, a bare LF
6292 in a subsequent header line has a space inserted after it, so as not to
6293 terminate the header.
6295 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6296 unnecessarily, used atoi() instead of strtol(), and didn't check the
6297 termination when getting file sizes from file names by regex.
6299 19. Completely re-implemented the support for maildirsize files, in the light
6300 of a number of problems with the previous contributed implementation
6301 (4.30/29). In particular:
6303 . If the quota is zero, the maildirsize file is maintained, but no quota is
6306 . If the maildir directory does not exist, it is created before any attempt
6307 to write a maildirsize file.
6309 . The quota value in the file is just a cache; if the quota is changed in
6310 the transport, the new value overrides.
6312 . A regular expression is available for excluding directories from the
6315 20. The autoreply transport checks the characters in options that define the
6316 message's headers; it allows continued headers, but it was checking with
6317 isspace() after an embedded newline instead of explicitly looking for a
6320 21. If all the "regular" hosts to which an address was routed had passed their
6321 expiry times, and had not reached their retry times, the address was
6322 bounced, even if fallback hosts were defined. Now Exim should go on to try
6325 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6326 equivalent code in the SMTP transport. Some hosts send humungous responses
6327 to HELO/EHLO, more than 1024 it seems.
6329 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6330 gives compiler warnings in some environments. I've now done it "properly",
6333 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6334 (because of problems with the built-in one) was declared to return uschar *
6335 instead of char *, causing compiler failure.
6337 25. Fixed a file descriptor leak when processing alias/forward files.
6339 26. Fixed a minor format string issue in dbfn.c.
6341 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6343 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6344 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6347 29. When a sender address is verified, it is cached, to save repeating the test
6348 when there is more than one recipient in a message. However, when the
6349 verification involves a callout, it is possible for different callout
6350 options to be set for different recipients. It is too complicated to keep
6351 track of this in the cache, so now Exim always runs a verification when a
6352 callout is required, relying on the callout cache for the optimization.
6353 The overhead is duplication of the address routing, but this should not be
6356 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6357 to be verified with callout=postmaster, and the main callout worked but the
6358 postmaster check failed, the verification correctly failed. However, if a
6359 subsequent RCPT command asked for sender verification *without* the
6360 postmaster check, incorrect caching caused this verification also to fail,
6363 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6364 it was not caching the DNS options (qualify_single, search_parents) that
6365 were used when the lookup failed. A subsequent lookup with different
6366 options therefore always gave the same answer, though there were cases
6367 where it should not have. (Example: a "domains = !$mx_any" option on a
6368 dnslookup router: the "domains" option is always processed without any
6369 widening, but the router might have qualify_single set.) Now Exim uses the
6370 cached value only when the same options are set.
6372 32. Added John Jetmore's "exipick" utility to the distribution.
6374 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6375 than a timeout (e.g. a certificate is required, and is not provided), an
6376 Exim server now closes the connection immediately. Previously it waited for
6377 the client to close - but if the client is SSL, it seems that they each
6378 wait for each other, leading to a delay before one of them times out.
6380 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6381 maintained 0.8.x compatibility because I don't think many are using it, and
6382 it is clearly obsolete.
6384 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6387 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6388 client certificate was expired. A simple patch fixes this, though I don't
6389 understand the full logic of why the verify callback is called multiple
6392 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6393 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6394 which causes problems with some clients (such as the Certicom SSL Plus
6395 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6396 disables the coutermeasure allowing Eudora to connect."
6398 38. Exim was not checking that a write() to a log file succeeded. This could
6399 lead to Bad Things if a log got too big, in particular if it hit a file
6400 size limit. Exim now panics and dies if it cannot write to a log file, just
6401 as it does if it cannot open a log file.
6403 39. Modified OS/Makefile-Linux so that it now contains
6405 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6407 The two -D definitions ensure that Exim is compiled with large file
6408 support, which makes it possible to handle log files that are bigger than
6411 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6412 instance) a domain was checked against a named list that involved a lookup,
6413 causing $domain_data to be set, then another domain was checked against the
6414 same list, then the first domain was re-checked, the value of $domain_data
6415 after the final check could be wrong. In particular, if the second check
6416 failed, it could be set empty. This bug probably also applied to
6419 41. The strip_trailing_dot option was not being applied to the address given
6420 with the -f command-line option.
6422 42. The code for reading a message's header from the spool was incrementing
6423 $received_count, but never initializing it. This meant that the value was
6424 incorrect (doubled) while delivering a message in the same process in which
6425 it was received. In the most common configuration of Exim, this never
6426 happens - a fresh exec is done - but it can happen when
6427 deliver_drop_privilege is set.
6429 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6430 it now includes up to 150 characters of the unexpected data in the log
6433 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6434 and building data strings. The size of both of these buffers was 10 000
6435 bytes - far larger than anybody would *ever* want, thought I. Needless to
6436 say, somebody hit the limit. I have increased the maximum line length to
6437 20 000 and the maximum data length of concatenated lines to 100 000. I have
6438 also fixed two bugs, because there was no checking on these buffers. Tsk,
6439 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6440 buffer is too small.
6442 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6443 lsearch lookups. Now it does.
6445 46. When parsing a route_list item in a manualroute router, a fixed-length
6446 buffer was used for the list of hosts. I made this 1024 bytes long,
6447 thinking that nobody would ever have a list of hosts that long. Wrong.
6448 Somebody had a whole pile of complicated expansion conditions, and the
6449 string was silently truncated, leading to an expansion error. It turns out
6450 that it is easier to change to an unlimited length (owing to other changes
6451 that have happened since this code was originally written) than to build
6452 structure for giving a limitation error. The length of the item that
6453 expands into the list of hosts is now unlimited.
6455 47. The lsearch lookup could not handle data where the length of text line was
6456 more than 4095 characters. Such lines were truncated, leading to shortened
6457 data being returned. It should now handle lines of any length.
6459 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6460 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6463 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6464 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6465 no longer happens in comment lines. A list of replacements is now placed
6466 at the head of all of the source files, except those whose only change is
6467 to replace PERL_COMMAND in the very first #! line.
6469 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6470 messages on the queue, with a bottom-up merge sort, using code contributed
6471 by Michael Haardt. This should make operations like -bp somewhat faster on
6472 large queues. It won't affect queue runners, except when queue_run_in_order
6475 51. Installed eximstats 1.31 in the distribution.
6477 52. Added support for SRV lookups to the dnslookup router.
6479 53. If an ACL referred to $message_body or $message_body_end, the value was not
6480 reset for any messages that followed in the same SMTP session.
6482 54. The store-handling optimization for building very long strings was not
6483 differentiating between the different store pools. I don't think this
6484 actually made any difference in practice, but I've tidied it.
6486 55. While running the routers to verify a sender address, $sender_address
6487 was still set to the sender address. This is wrong, because when routing to
6488 send a bounce to the sender, it would be empty. Therefore, I have changed
6489 it so that, while verifying a sender address, $sender_address is set to <>.
6490 (There is no change to what happens when verifying a recipient address.)
6492 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6493 target A or AAAA records (if not already returned) without resetting the
6494 qualify_single or search_parents options of the DNS resolver. These are
6495 inappropriate in this case because the targets of MX and SRV records must
6496 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6497 target that, when qualified, matched something in the local domain. These
6498 two options are now turned off when doing these lookups.
6500 57. It seems that at least some releases of Reiserfs (which does not have the
6501 concept of a fixed number of inodes) returns zero and not -1 for the
6502 number of available inodes. This interacted badly with check_spool_inodes,
6503 which assumed that -1 was the "no such thing" setting. What I have done is
6504 to check that the total number of inodes is greater than zero before doing
6505 the test of how many are available.
6507 58. When a "warn" ACL statement has a log_message modifier, the message is
6508 remembered, and not repeated. This is to avoid a lot of repetition when a
6509 message has many recipients that cause the same warning to be written.
6510 However, Exim was preserving the list of already written lines for an
6511 entire SMTP session, which doesn't seem right. The memory is now reset if a
6512 new message is started.
6514 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6515 debugging output unless log_rewrite was also set.
6517 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6518 of (int)(handle) when we know that handle contains (void *)(-1).
6520 61. The Exim daemon panic-logs an error return when it closes the incoming
6521 connection. However "connection reset by peer" seems to be common, and
6522 isn't really an error worthy of noting specially, so that particular error
6525 62. When Exim is trying to find all the local interfaces, it used to panic and
6526 die if the ioctl to get the interface flags failed. However, it seems that
6527 on at least one OS (Solaris 9) it is possible to have an interface that is
6528 included in the list of interfaces, but for which you get a failure error
6529 for this call. This happens when the interface is not "plumbed" into a
6530 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6531 failure of the "get flags" call assumes that the interface is down.
6533 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6534 makes life easier for people who are doing arithmetic on fields extracted
6535 from dates, where you often get leading zeros that should not be
6536 interpreted as octal.
6538 64. Added qualify_domain to the redirect router, to override the global
6541 65. If a pathologically long header line contained very many addresses (the
6542 report of this problem mentioned 10 000) and each of them was rewritten,
6543 Exim could use up a very large amount of memory. (It kept on making new
6544 copies of the header line as it rewrote, and never released the old ones.)
6545 At the expense of a bit more processing, the header rewriting function has
6546 been changed so that it no longer eats memory in this way.
6548 66. The generation of the Received: header has been moved from the time that a
6549 message starts to be received, to the time that it finishes. The timestamp
6550 in the Received: header should now be very close to that of the <= log
6551 line. There are two side-effects of this change:
6553 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6554 logged header lines no longer include the local Received: line, because
6555 it has not yet been created. The same applies to a copy of the message
6556 that is returned to a non-SMTP sender when a message is rejected.
6558 (b) When a filter file is tested using -bf, no additional Received: header
6559 is added to the test message. After some thought, I decided that this
6562 This change does not affect the value of $received_for. It is still set
6563 after address rewriting, but before local_scan() is called.
6565 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6567 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6568 gave an unhelpful panic error message, and a defer error. I have managed to
6569 change this behaviour so that it now rejects any supplied certificate,
6570 which seems right, as the list of acceptable certificates is empty.
6572 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6573 gave an unhelpful defer error. I have not managed to make this reject any
6574 supplied certificates, but the error message it gives is "no certificate
6575 supplied", which is not helpful.
6577 70. exigrep's output now also includes lines that are not associated with any
6578 message, but which match the given pattern. Implemented by a patch from
6579 Martin Sluka, which also tidied up the Perl a bit.
6581 71. Recipient callout verification, like sender verification, was using <> in
6582 the MAIL FROM command. This isn't really the right thing, since the actual
6583 sender may affect whether the remote host accepts the recipient or not. I
6584 have changed it to use the actual sender in the callout; this means that
6585 the cache record is now keyed on a recipient/sender pair, not just the
6586 recipient address. There doesn't seem to be a real danger of callout loops,
6587 since a callout by the remote host to check the sender would use <>.
6588 [SEE ABOVE: changed after hitting problems.]
6590 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6591 temporary errors. However, in the case of such a code being given after
6592 the end of a data transmission (i.e. after ".") Exim was failing to write
6593 a retry record for the message. (Yes, there was some broken host that was
6594 actually sending 8xx at this point.)
6596 73. An unknown lookup type in a host list could cause Exim to panic-die when
6597 the list was checked. (An example that provoked this was putting <; in the
6598 middle of a list instead of at the start.) If this happened during a DATA
6599 ACL check, a -D file could be left lying around. This kind of configuration
6600 error no longer causes Exim to die; instead it causes a defer error. The
6601 incident is still logged to the main and panic logs.
6603 74. Buglet left over from Exim 3 conversion. The message "too many messages
6604 in one connection" was written to the rejectlog but not the mainlog, except
6605 when address rewriting (yes!) was being logged.
6607 75. Added write_rejectlog option.
6609 76. When a system filter was run not as root (that is, when system_filter_user
6610 was set), the values of the $n variables were not being returned to the
6611 main process; thus, they were not subsequently available in the $sn
6614 77. Added +return_path_on_delivery log selector.
6616 78. A connection timeout was being treated differently from recipients deferred
6617 when testing hosts_max_try with a message that was older than the host's
6618 retry timeout. (The host should not be counted, thus allowing all hosts to
6619 be tried at least once before bouncing.) This may have been the cause of an
6620 occasionally reported bug whereby a message would remain on the queue
6621 longer than the retry timeout, but would be bounced if a delivery was
6622 forced. I say "may" because I never totally pinned down the problem;
6623 setting up timeout/retry tests is difficult. See also the next item.
6625 79. The ultimate address timeout was not being applied to errors that involved
6626 a combination of host plus message (for example, a timeout on a MAIL
6627 command). When an address resolved to a number of possible hosts, and they
6628 were not all tried for each delivery (e.g. because of hosts_max_try), a
6629 message could remain on the queue longer than the retry timeout.
6631 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6634 81. Fixed an obscure SMTP outgoing bug which required at least the following
6635 conditions: (a) there was another message waiting for the same server;
6636 (b) the server returned 5xx to all RCPT commands in the first message so
6637 that the message was not completed; (c) the server dropped the connection
6638 or gave a negative response to the RSET that Exim sends to abort the
6639 transaction. The observed case was a dropped connection after DATA that had
6640 been sent in pipelining mode. That is, the server had advertised PIPELINING
6641 but was not implementing it correctly. The effect of the bug was incorrect
6642 behaviour, such as trying another host, and this could lead to a crash.
6648 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6649 and daemon.c were passed as pointers to ints; they should have been
6650 pointers to socklen_t variables (which are typically unsigned ints).
6652 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6655 3. Fixed a really odd bug that affected only the testing scheme; patching a
6656 certain fixed string in the binary changed the value of another string that
6657 happened to be identical to the end of the original first string.
6659 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6660 name", it returns that address as the IP address. On some operating
6661 systems (e.g. Solaris), it also passes back the IP address string as the
6662 "host name". However, on others (e.g. Linux), it passes back an empty
6663 string. Exim wasn't checking for this, and was changing the host name to an
6664 empty string, assuming it had been canonicalized.
6666 5. Although rare, it is permitted to have more than one PTR record for a given
6667 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6668 all the names associated with an address, because they do in Solaris.
6669 However, it seems that they do not in Linux for data that comes from the
6670 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6671 I found this out when I moved to a new Linux workstation and tried to run
6672 the Exim test suite.
6674 To get round this problem I have changed the code so that it now does its
6675 own call to the DNS to look up PTR records when searching for a host name.
6676 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6677 addresses that are only in /etc/hosts are still found.
6679 This behaviour is, however, controlled by an option called host_lookup_
6680 order, which defaults to "bydns:byaddr". If people want to use the other
6681 order, or indeed, just use one or the other means of lookup, they can
6682 specify it in this variable.
6684 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6685 some operating systems, this comes back from gethostbyaddr() as an empty
6686 string, and this is what Exim used to test for. However, it seems that in
6687 other systems, "." is yielded. Exim now tests for this case too.
6689 7. The values of check_spool_space and check_log_space are now held internally
6690 as a number of kilobytes instead of an absolute number of bytes. If a
6691 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6692 kilobyte. This means that much larger values can be stored.
6694 8. Exim monitor: an attempt to get the action menu when not actually pointing
6695 at a message produces an empty menu entitled "No message selected". This
6696 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6697 no entries in it ("Shell widget menu has zero width and/or height"). So I
6698 have added a single, blank menu entry in this case.
6700 9. Added ${quote_local_part.
6702 10. MIME decoding is now applied to the contents of Subject: header lines when
6705 11. Now that a reference to $sender_host_address automatically causes a reverse
6706 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6707 host lookup before query-style lookups in lists that might use this
6708 variable. This has therefore been abolished, and the "net-" prefix is no
6709 longer necessary for query-style lookups.
6711 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6712 have been a typo for LFLAGS, so it has been changed.
6714 13. The install script calls Exim with "-C /dev/null" in order to find the
6715 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6716 to be output. However, since Exim outputs its version number before the
6717 error, it didn't break the script. It just looked ugly. I fixed this by
6718 always allowing "-C /dev/null" if the caller is root.
6720 14. Ignore overlarge ACL variable number when reading spool file - insurance
6721 against a later release with more variables having written the file.
6723 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6724 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6725 this, as well as the form without the "IPv6" on the front.
6727 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6728 OS/Makefile-Darwin file.
6730 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6731 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6733 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6734 variable. However, some OS use ints for inodes. I've added cast to long int
6735 to get rid of the compiler warning.
6737 19. I had forgotten to lock out "/../" in configuration file names when
6738 ALT_CONFIG_PREFIX was set.
6740 20. Routers used for verification do not need to specify transports. However,
6741 if such a router generated a host list, and callout was configured, Exim
6742 crashed, because it could not find a port number from the (non-existent)
6743 transport. It now assumes port 25 in this circumstance.
6745 21. Added the -t option to exigrep.
6747 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6748 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6749 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6750 them is set, LOOKUP_LSEARCH is forced.
6752 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6754 24. Added sender and host information to the "rejected by local_scan()" log
6755 line; previously there was no indication of these.
6757 25. Added .include_if_exists.
6759 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6760 Linux. It turns out that not all file systems support this. Apparently some
6761 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6762 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6763 not supported on the file descriptor, is now ignored when Exim is trying to
6764 sync a directory. This applies only to Linux.
6766 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6768 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6769 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6772 29. Added support for maildirsize files from supplied patch (modified a bit).
6774 30. The use of :fail: followed by an empty string could lead Exim to respond to
6775 sender verification failures with (e.g.):
6777 550 Verification failed for <xxx>
6778 550 Sender verify failed
6780 where the first response line was missing the '-' that indicates it is not
6781 the final line of the response.
6783 31. The loop for finding the name of the user that called Exim had a hardwired
6784 limit of 10; it now uses the value of finduser_retries, which is used for
6785 all other user lookups.
6787 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6790 33. Exim was neglecting to zero errno before one call of strtol() when
6791 expanding a string and expecting an integer value. On some systems this
6792 resulted in spurious "integer overflow" errors. Also, it was casting the
6793 result into an int without checking.
6795 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6796 did not work. The code looks as if it has *never* worked, though it appears
6797 to have been documented since at least release 1.62. I have made it work.
6799 35. The "timeout_DNS" error in retry rules, also documented since at least
6800 1.62, also never worked. As it isn't clear exactly what this means, and
6801 clearly it isn't a major issue, I have abolished the feature by treating it
6802 as "timeout", and writing a warning to the main and panic logs.
6804 36. The display of retry rules for -brt wasn't always showing the error code
6807 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6808 timeout_connect_A, timeout_connect_MX.
6810 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6811 to the empty sender.
6813 39. The daemon was not analysing the content of -oX till after it had closed
6814 stderr and disconnected from the controlling terminal. This meant that any
6815 syntax errors were only noted on the panic log, and the return code from
6816 the command was 0. By re-arranging the code a little, I've made the
6817 decoding happen first, so such errors now appear on stderr, and the return
6818 code is 1. However, the actual setting up of the sockets still happens in
6819 the disconnected process, so errors there are still only recorded on the
6822 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6823 connections (as happens on some IP stacks) was logged at start up time as
6824 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6825 it from "IPv6 and IPv4", which means that two separate sockets are being
6828 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6829 says whether AF_INET or AF_INET6 was passed as an argument.
6831 42. Exiwhat output was messed up when time zones were included in log
6834 43. Exiwhat now gives more information about the daemon's listening ports,
6835 and whether -tls-on-connect was used.
6837 44. The "port" option of the smtp transport is now expanded.
6839 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6840 silently ignored. Now an error message is written to the main and panic
6843 46. There's a new ACL modifier called "logwrite" which writes to a log file
6844 as soon as it is encountered.
6846 47. Added $local_user_uid and $local_user_gid at routing time.
6848 48. Exim crashed when trying to verify a sender address that was being
6851 49. Exim was recognizing only a space character after ".include". It now also
6852 recognizes a tab character.
6854 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6855 extracting the relevant information from the specification. The man page no
6856 longer contains scrambled data for the -d option, and I've added a section
6857 at the front about calling Exim under different names.
6859 51. Added "extra_headers" argument to the "mail" command in filter files.
6861 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6864 53. Installed eximstats 1.29.
6866 54. Added transport_filter_timeout as a generic transport option.
6868 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6869 Cc: header lines. This was required by RFC 822, but it not required by RFC
6872 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6873 incoming messages that did not have them. Now it does so only if the
6874 message originates locally, that is, if there is no associated remote host
6875 address. When Resent- header lines are present, this applies to the Resent-
6876 lines rather than the non-Resent- lines.
6878 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6879 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6881 58. Messages for configuration errors now include the name of the main
6882 configuration file - useful now that there may be more than one file in a
6883 list (.included file names were always shown).
6885 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6886 for those rare installations that do not start the daemon as root or run it
6887 setuid root. I've cut out the call to initgroups() if the daemon is not
6890 60. The Exim user and group can now be bound into the binary as text strings
6891 that are looked up at the start of Exim's processing.
6893 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6895 62. Added $mailstore_basename variable.
6897 63. Installed patch to sieve.c from Michael Haardt.
6899 64. When Exim failed to open the panic log after failing to open the main log,
6900 the original message it was trying to log was written to stderr and debug
6901 output, but if they were not available (the usual case in production), it
6902 was lost. Now it is written to syslog before the two lines that record the
6903 failures to open the logs.
6905 65. Users' Exim filters run in subprocesses under the user's uid. It is
6906 possible for a "deliver" command or an alias in a "personal" command to
6907 provoke an address rewrite. If logging of address rewriting is configured,
6908 this fails because the process is not running as root or exim. There may be
6909 a better way of dealing with this, but for the moment (because 4.30 needs
6910 to be released), I have disabled address rewrite logging when running a
6911 filter in a non-root, non-exim process.
6917 1. The buildconfig auxiliary program wasn't quoting the value set for
6918 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6919 not defined. This bug was masked in 4.22 by the effect that was fixed in
6922 2. Some messages that were rejected after a message id was allocated were
6923 shown as "incomplete" by exigrep. It no longer does this for messages that
6924 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6926 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6927 have allow_domain_literals set, the ID did not get logged in the <= line.
6928 Domain literals are now always recognized in Message-ID: header lines.
6930 4. The first argument for a ${extract expansion item is the key name or field
6931 number. Leading and trailing spaces in this item were not being ignored,
6932 causing some misleading effects.
6934 5. When deliver_drop_privilege was set, single queue runner processes started
6935 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6936 same command in the process it spins off) were not dropping privilege.
6938 6. When the daemon running as "exim" started a queue runner, it always
6939 re-executed Exim in the spun-off process. This is a waste of effort when
6940 deliver_drop_privilege is set. The new process now just calls the
6941 queue-runner function directly.
6947 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6950 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6951 ignored. Though the use of -oP was forcing the writing of a pid file, it
6952 was always written to the default place.
6954 3. If the message "no IP address found for host xxxx" is generated during
6955 incoming verification, it is now followed by identification of the incoming
6956 connection (so you can more easily find what provoked it).
6958 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6960 5. Added some features to "harden" Exim a bit more against certain attacks:
6962 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6963 be put in Local/Makefile. This is like the never_users runtime option,
6964 but it cannot be overridden. The default setting is "root".
6966 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6967 prefix string with which any file named in a -C command line option
6970 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6971 is retained for -C and -D only if the caller of Exim is root. Without
6972 it, the exim user may also use -C and -D and retain privilege.
6974 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6975 command line option is disabled.
6977 6. Macro names set by the -D option must start with an upper case letter, just
6978 like macro names defined in the configuration file.
6980 7. Added "dereference=" facility to LDAP.
6982 8. Two instances of the typo "uknown" in the source files are fixed.
6984 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6985 the Configure-Makefile script screwed up while processing it.
6987 10. Incorporated PCRE 4.4.
6989 11. The SMTP synchronization check was not operating right at the start of an
6990 SMTP session. For example, it could not catch a HELO sent before the client
6991 waited for the greeting. There is now a check for outstanding input at the
6992 point when the greeting is written. Because of the duplex, asynchronous
6993 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6994 way, but not yet received, when the check is performed.
6996 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6997 on TCP/IP sockets, because this apparently causes some broken clients to
7000 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7001 unchanged) from the Cygwin maintainer.
7003 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7004 was supported instead of testing for mbx. Effectively a typo.
7006 15. The spa authenticator server code was not checking that the input it
7007 received was valid base64.
7009 16. The debug output line for the "set" modifier in ACLs was not showing the
7010 name of the variable that was being set.
7012 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7014 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7015 lookup on its own. Something else was needed to trigger the lookup. For
7016 example, a match in host_lookup or the need for a host name in a host list.
7017 Now, if $sender_host_name is referenced and the host name has not yet been
7018 looked up, a lookup is performed. If the lookup fails, the variable remains
7019 empty, and $host_lookup_failed is set to "1".
7021 19. Added "eqi" as a case-independent comparison operator.
7023 20. The saslauthd authentication condition could segfault if neither service
7024 nor realm was specified.
7026 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7027 error message that was logged was misleading, and incoming SMTP
7028 connections were dropped. The message is now more accurate, and temporary
7029 errors are given to SMTP connections.
7031 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7032 (except RSET) with a 421 error, until QUIT is received. However, it was
7033 failing to send a response to QUIT.
7035 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7036 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7037 ACL is now run after the helo_try_verify_hosts code.
7039 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7040 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7041 case-independent, so other case variants are also recognized). Apparently
7042 some systems use these upper case variants.
7044 25. If more than two messages were waiting for the same host, and a transport
7045 filter was specified for the transport, Exim sent two messages over the
7046 same TCP/IP connection, and then failed with "socket operation on non-
7047 socket" when it tried to send the third.
7049 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7051 27. The extern definition of crypt16() in expand.c was not being excluded when
7052 the OS had its own crypt16() function.
7054 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7055 as a preferred synonym for return_size_limit, both as an option and as an
7058 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7060 30. Changed the default configuration ACL to relax the local part checking rule
7061 for addresses that are not in any local domains. For these addresses,
7062 slashes and pipe symbols are allowed within local parts, but the sequence
7063 /../ is explicitly forbidden.
7065 31. SPA server authentication was not clearing the challenge buffer before
7068 32. log_message in a "warn" ACL statement was writing to the reject log as
7069 well as to the main log, which contradicts the documentation and doesn't
7070 seem right (because no rejection is happening). So I have stopped it.
7072 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7073 However, I am unable to do any testing of this.
7075 34. Fixed an infelicity in the appendfile transport. When checking directories
7076 for a mailbox, to see if any needed to be created, it was accidentally
7077 using path names with one or more superfluous leading slashes; tracing
7078 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7080 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7081 contents are added to the log line that is written for every discarded
7082 recipient. (Previously a log_message setting was ignored.)
7084 36. The ${quote: operator now quotes the string if it is empty.
7086 37. The install script runs exim in order to find its version number. If for
7087 some reason other than non-existence or emptiness, which it checks, it
7088 could not run './exim', it was installing it with an empty version number,
7089 i.e. as "exim-". This error state is now caught, and the installation is
7092 38. An argument was missing from the function that creates an error message
7093 when Exim fails to connect to the socket for saslauthd authentication.
7094 This could cause Exim to crash, or give a corrupted message.
7096 39. Added isip, isip4, and isip6 to ${if conditions.
7098 40. The ACL variables $acl_xx are now saved with the message, and can be
7099 accessed later in routers, transports, and filters.
7101 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7102 strings in the file are not string-expanded.
7104 42. If a MAIL command specified a SIZE value that was too large to fit into an
7105 int variable, the check against message_size_limit failed. Such values are
7106 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7107 day this will have to be increased, but I don't think I want to be around
7108 when emails are that large.
7115 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7116 iconv() is not standard in FreeBSD.
7118 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7119 IPv6 enabled. The observed symptom was a segmentation fault on return from
7120 the function os_common_find_running_interfaces() in src/os.c.
7122 3. In the check_special_case() function in daemon.c I had used "errno" as an
7123 argument name, which causes warnings on some systems. This was basically a
7124 typo, since it was named "eno" in the comments!
7126 4. The code that waits for the clock to tick (at a resolution of some fraction
7127 of a second) so as to ensure message-id uniqueness was always waiting for
7128 at least one whole tick, when it could have waited for less. [This is
7129 almost certainly not relevant at current processor speeds, where it is
7130 unlikely to ever wait at all. But we try to future-proof.]
7132 5. The function that sleeps for a time interval that includes fractions of a
7133 second contained a race. It did not block SIGALRM between setting the
7134 timer, and suspending (a couple of lines later). If the interval was short
7135 and the sigsuspend() was delayed until after it had expired, the suspension
7136 never ended. On busy systems this could lead to processes getting stuck for
7139 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7140 process, before it forks any delivery processes. The open lookup caching
7141 mechanism meant that the open file or database connection was passed into
7142 the delivery process. The problem was that delivery processes always tidy
7143 up cached lookup data. This could cause a problem for the next delivery
7144 process started by the queue runner, because the external queue runner
7145 process does not know about the closure. So the next delivery process
7146 still has data in the lookup cache. In the case of a file lookup, there was
7147 no problem because closing a file descriptor in a subprocess doesn't affect
7148 the parent. However, if the lookup was caching a connection to a database,
7149 the connection was closed, and the second delivery process was likely to
7150 see errors such as "PGSQL: query failed: server closed the connection
7151 unexpectedly". The problem has been fixed by closing all cached lookups
7152 in a queue runner before running a delivery process.
7154 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7155 seem to have the "const" qualifier which it has on other OS. I've
7158 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7159 *of the same type* (client or server) with the same public name that an
7160 error should be diagnosed.
7162 9. When Exim looked up a host name for an IP address, but failed to find the
7163 original IP address when looking up the host name (a safety check), it
7164 output the message "<ip address> does not match any IP for NULL", which was
7165 confusing, to say the least. The bug was that the host name should have
7166 appeared instead of "NULL".
7168 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7169 user that is built into the binary, and the -C or -D options is used, root
7170 privilege is dropped before the configuration file is read. In addition,
7171 logging is switched to stderr instead of the normal log files. If the
7172 configuration then re-defines the Exim user, the unprivileged environment
7173 is probably not what is expected, so Exim logs a panic warning message (but
7176 However, if deliver_drop_privilege is set, the unprivileged state may well
7177 be exactly what is intended, so the warning has been cut out in that case,
7178 and Exim is allowed to try to write to its normal log files.
7184 1. smtp_return_error_details was not giving details for temporary sender
7185 or receiver verification errors.
7187 2. Diagnose a configuration error if two authenticators have the same public
7190 3. Exim used not to create the message log file for a message until the first
7191 delivery attempt. This could be confusing when incoming messages were held
7192 for policy or load reasons. The message log file is now created at the time
7193 the message is received, and an initial "Received" line is written to it.
7195 4. The automatically generated man page for command line options had a minor
7196 bug that caused no ill effects; however, a more serious problem was that
7197 the procedure for building the man page automatically didn't always
7198 operate. Consequently, release 4.20 contains an out-of-date version. This
7199 shouldn't happen again.
7201 5. When building Exim with embedded Perl support, the script that builds the
7202 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7203 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7205 6. The freeze_tell option was not being used for messages that were frozen on
7206 arrival, either by an ACL or by local_scan().
7208 7. Added the smtp_incomplete_transaction log selector.
7210 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7211 was accepting AUTH without a new EHLO.
7213 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7214 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7217 10. Logging of TCP/IP connections (when configured) now happens in the main
7218 daemon process instead of the child process, so that the TCP/IP connection
7219 count is more accurate (but it can never be perfect).
7221 11. The use of "drop" in a nested ACL was not being handled correctly in the
7222 outer ACL. Now, if condition failure induced by the nested "drop" causes
7223 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7224 or "require"), the connection is dropped.
7226 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7227 that yield "discard" can now be used with an "accept" or a "discard" verb,
7228 but an error is generated for any others (because I can't see a useful way
7229 to define what should happen).
7231 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7232 are now processed in the same way as lines in the Exim configuration file.
7233 In particular, continuation lines are supported.
7235 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7237 15. Added -ti meaning -t -i.
7239 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7240 domains, and warn by logging if others are found.
7242 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7243 returned from GIFCONF when Exim is trying to find the list of interfaces on
7244 a host. The code in os.c has been modified to copy each ifreq to an aligned
7245 structure in all cases.
7247 Also, in some cases, the returned ifreq's were being copied to a 'struct
7248 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7249 means the last couple of bytes of an IPv6 address could be chopped if the
7250 ifreq contained only a normal sockaddr (14 bytes storage).
7252 18. Named domain lists were not supported in the hosts_treat_as_local option.
7253 An entry such as +xxxx was not recognized, and was treated as a literal
7256 19. Ensure that header lines added by a DATA ACL are included in the reject log
7257 if the ACL subsequently rejects the message.
7259 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7260 MD5 (which is deprecated).
7262 21. When testing a filter file using -bf, Exim was writing a message when it
7263 took the sender from a "From " line in the message, but it was not doing so
7264 when it took $return_path from a Return-Path: header line. It now does.
7266 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7267 with a valid header line field name (a series of printing characters
7268 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7270 23. Changed "disc" in the source to "disk" to conform to the documentation and
7271 the book and for uniformity.
7273 24. Ignore Sendmail's -Ooption=value command line item.
7275 25. When execve() failed while trying to run a command in a pipe transport,
7276 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7277 could be confused with a return value of 69 from the command itself. This
7278 has been changed to 127, the value the shell returns if it is asked to run
7279 a non-existent command. The wording for the related log line suggests a
7280 non-existent command as the problem.
7282 26. If received_header_text expands to an empty string, do not add a Received:
7283 header line to the message. (Well, it adds a token one on the spool, but
7284 marks it "old" so that it doesn't get used or transmitted.)
7286 27. Installed eximstats 1.28 (addition of -nt option).
7288 28. There was no check for failure on the call to getsockname() in the daemon
7289 code. This can fail if there is a shortage of resources on the system, with
7290 ENOMEM, for example. A temporary error is now given on failure.
7292 29. Contrary to the C standard, it seems that in some environments, the
7293 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7294 program. Exim now does this explicitly; it affects the formatting of
7295 timestamps using strftime().
7297 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7298 complaints. I've now initialized all the variables, to avoid this.
7300 32. Header lines added by a system filter were not being "seen" during
7301 transport-time rewrites.
7303 33. The info_callback() function passed to OpenSSL is set up with type void
7304 (*)(SSL *, int, int), as described somewhere. However, when calling the
7305 function (actually a macro) that sets it up, the type void(*)() is
7306 expected. I've put in a cast to prevent warnings from picky compilers.
7308 34. If a DNS black list lookup found a CNAME record, but there were no A
7309 records associated with the domain it pointed at, Exim crashed.
7311 35. If a DNS black list lookup returned more than one A record, Exim ignored
7312 all but the first. It now scans all returned addresses if a particular IP
7313 value is being sought. In this situation, the contents of the
7314 $dnslist_value variable are a list of all the addresses, separated by a
7317 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7318 to accept a host name and all its aliases if the forward lookup for any of
7319 them yielded the IP address of the incoming connection. Now it accepts only
7320 those names whose forward lookup yields the correct IP address. Any other
7321 names are discarded. This closes a loophole whereby a rogue DNS
7322 administrator could create reverse DNS records to break through a
7323 wildcarded host restriction in an ACL.
7325 37. If a user filter or a system filter that ran in a subprocess used any of
7326 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7327 the wrong values were passed to the pipe command ($thisaddress had the
7328 value of $0, $0 had the value of $1, etc). This bug was introduced by
7329 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7331 38. Improved the line breaking for long SMTP error messages from ACLs.
7332 Previously, if there was no break point between 40 and 75 characters, Exim
7333 left the rest of the message alone. Two changes have been made: (a) I've
7334 reduced the minimum length to 35 characters; (b) if it can't find a break
7335 point between 35 and 75 characters, it looks ahead and uses the first one
7336 that it finds. This may give the occasional overlong line, but at least the
7337 remaining text gets split now.
7339 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7340 file descriptors might be low, and that setting 1000 would always raise it.
7341 It turns out that in some environments, the limit is already over 1000 and
7342 that lowering it causes trouble. So now Exim takes care not to decrease it.
7344 40. When delivering a message, the value of $return_path is set to $sender_
7345 address at the start of routing (routers may change the value). By an
7346 oversight, this default was not being set up when an address was tested by
7347 -bt or -bv, which affected the outcome if any router or filter referred to
7350 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7351 log only when "message" or "log_message" are set. However, if one of the
7352 conditions was an address verification, or a call to a nested ACL, the
7353 messages generated by the underlying test were being passed through. This
7354 no longer happens. The underlying message is available in $acl_verify_
7355 message for both "message" and "log_message" expansions, so it can be
7356 passed through if needed.
7358 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7359 new expansion $bh_ to give the encoded byte string without charset
7360 translation. Translation happens only if iconv() is available; HAVE_ICONV
7361 indicates this at build time. HEADERS_CHARSET gives the charset to
7362 translate to; headers_charset can change it in the configuration, and
7363 "headers charset" can change it in an individual filter file.
7365 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7366 that creates RFC 2047 encoded "words" labels them as that charset instead
7367 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7368 expansion operator; (ii) when Exim creates a From: line for a local
7369 message; (iii) when a header line is rewritten to include a "phrase" part.
7371 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7372 buggy, causing it to skip the first lines of messages whose message ID
7373 ended in 'D'. This would not have bitten before Exim release 4.14, because
7374 message IDs were unlikely to end in 'D' before then. The effect was to have
7375 incorrect size information for certain domains.
7377 45. #include "config.h" was missing at the start of the crypt16.c module. This
7378 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7381 46. If there was a timeout during a "random" callout check, Exim treated it as
7382 a failure of the random address, and carried on sending RSET and the real
7383 address. If the delay was just some slowness somewhere, the response to the
7384 original RCPT would be taken as a response to RSET and so on, causing
7385 mayhem of various kinds.
7387 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7388 when I implemented it. It didn't allow for the fact that some option values
7389 may legitimately be negative (e.g. size_addition), and it didn't even do
7390 the right test for positive values.
7392 48. Domain names in DNS records are case-independent. Exim always looks them up
7393 in lower case. Some resolvers return domain names in exactly the case they
7394 appear in the zone file, that is, they may contain uppercase letters. Not
7395 all resolvers do this - some return always lower case. Exim was treating a
7396 change of case by a resolver as a change of domain, similar to a widening
7397 of a domain abbreviation. This triggered its re-routing code and so it was
7398 trying to route what was effectively the same domain again. This normally
7399 caused routing to fail (because the router wouldn't handle the domain
7400 twice). Now Exim checks for this case specially, and just changes the
7401 casing of the domain that it ultimately uses when it transmits the message
7404 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7407 50. If a filter generated a file delivery with a non-absolute name (possible if
7408 no home directory exists for the router), the forbid_file option was not
7411 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7412 the existing equality matching.
7414 52. Exim was using ints instead of ino_t variables in some places where it was
7415 dealing with inode numbers.
7417 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7418 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7419 called TMPDIR, and if it finds it is different, it changes its value.
7421 54. The smtp_printf() function is now made available to local_scan() so
7422 additional output lines can be written before returning. There is also an
7423 smtp_fflush() function to enable the detection of a dropped connection.
7424 The variables smtp_input and smtp_batched_input are exported to
7427 55. Changed the default runtime configuration: the message "Unknown user"
7428 has been removed from the ACL, and instead placed on the localuser router,
7429 using the cannot_route_message feature. This means that any verification
7430 failures that generate their own messages won't get overridden. Similarly,
7431 the "Unrouteable address" message that was in the ACL for unverifiable
7432 relay addresses has also been removed.
7434 56. Added hosts_avoid_esmtp to the smtp transport.
7436 57. The exicyclog script was not checking for the esoteric option
7437 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7438 will work only if exicyclog is run under the appropriate euid.
7440 58. Following a discussion on the list, the rules by which Exim recognises line
7441 endings on incoming messages have been changed. The -dropcr and drop_cr
7442 options are now no-ops, retained only for backwards compatibility. The
7443 following line terminators are recognized: LF CRLF CR. However, special
7444 processing applies to CR:
7446 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7447 nor a local message in the state where . is a terminator.
7449 (ii) If a bare CR is encountered in a header line, an extra space is added
7450 after the line terminator so as not to end the header. The reasoning
7451 behind this is that bare CRs in header lines are most likely either
7452 to be mistakes, or people trying to play silly games.
7454 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7455 was being incorrectly given as 18 bytes larger than it should have been.
7456 This is a VOB (very old bug).
7458 60. This may never have affected anything current, but just in case it has:
7459 When the local host is found other than at the start of a list of hosts,
7460 the local host, those with the same MX, and any that follow, are discarded.
7461 When the list in question was part of a longer list of hosts, the following
7462 hosts (not currently being processed) were also being discarded. This no
7463 longer happens. I'm not sure if this situation could ever has previously
7466 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7469 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7470 header. This is something that is recommended in a new Internet Draft, and
7471 is something that is documented as being done by Sendmail. There are two
7472 possible values. For messages generated by the autoreply transport, Exim
7475 Auto-Submitted: auto-replied
7477 whereas for all other generated messages (e.g. bounces) it adds
7479 Auto-Submitted: auto-generated
7481 63. The "personal" condition in filters now includes a test for the
7482 Auto-Submitted: header. If it contains the string "auto-" the message it
7483 not considered personal.
7485 64. Added rcpt_include_affixes as a generic transport option.
7487 65. Added queue_only_override (default true).
7489 66. Added the syslog_duplication option.
7491 67. If what should have been the first header line of a message consisted of
7492 a space followed by a colon, Exim was mis-interpreting it as a header line.
7493 It isn't of course - it is syntactically invalid and should therefore be
7494 treated as the start of the message body. The misbehaviour could have
7495 caused a number of strange effects, including loss of data in subsequent
7496 header lines, and spool format errors.
7498 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7499 client host had authenticated. This control can now be exercised by an ACL
7500 for more flexibility.
7502 69. By default, callouts do not happen when testing with -bh. There is now a
7503 variant, -bhc, which does actually run the callout code, including
7504 consulting and updating the callout cache.
7506 70. Added support for saslauthd authentication, courtesy of Alexander
7509 71. If statvfs() failed on the spool or log directories while checking their
7510 size for availability, Exim confusingly gave the error "space shortage".
7511 Furthermore, in debugging mode it crashed with a floating point exception.
7512 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7513 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7514 serious problem, Exim now writes to the main and panic logs when this
7515 happens, with details of the failure. It then refuses to accept the
7516 incoming message, giving the message "spool directory problem" or "log
7517 directory problem" with a 421 code for SMTP messages.
7519 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7520 0, 1, and 2 exist, because some OS complain for execs without them (see
7521 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7522 descriptors. However, the code omitted to check that the open succeeded,
7523 causing mysterious errors if for some reason the permissions on /dev/null
7524 got screwed. Now Exim writes a message to the main and panic logs, and
7525 bombs out if it can't open /dev/null.
7527 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7528 interact so that it is all more flexible. It is supposed to remain
7529 backwards compatible. Also added extra_local_interfaces.
7531 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7532 to bomb out with an assertion failure - to the client this appears as a
7533 connection drop. This problem occurs in the part of the code that was taken
7534 from the Samba project. Fortunately, the assertion is in a very simple
7535 function, so I have fixed this by reproducing the function inline in the
7536 one place where it is called, and arranging for authentication to fail
7537 instead of killing the process with assert().
7539 75. The SPA client code was not working when the server requested OEM rather
7540 than Unicode encoding.
7542 76. Added code to make require_files with a specific uid setting more usable in
7543 the case where statting the file as root fails - usually a non-root-mounted
7544 NFS file system. When this happens and the failure is EACCES, Exim now
7545 forks a subprocess and does the per-uid checking as the relevant uid.
7547 77. Added process_log_path.
7549 78. If log_file_path was not explicitly set, a setting of check_log_space or
7550 check_log_inodes was ignored.
7552 79. If a space check for the spool or log partitions fails, the incident is now
7553 logged. Of course, in the latter case the data may get lost...
7555 80. Added the %p formatting code to string_format() so that it can be used to
7556 print addresses in debug_print(). Adjusted all the address printing in the
7557 debugging in store.c to use %p rather than %d.
7559 81. There was a concern that a line of code in smtp_in.c could overflow a
7560 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7561 initially expressed, the concern was not well-founded, because trailing
7562 spaces are removed early. However, if the trailing spaces were followed by
7563 a NULL, they did not get removed, so the overflow was possible. Two fixes
7566 (a) I re-wrote the offending code in a cleaner fashion.
7567 (b) If an incoming SMTP command contains a NULL character, it is rejected
7570 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7571 runs initgroups(), so that if the Exim user is in any additional groups,
7572 they will be used during message reception.
7578 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob