1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
9 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
10 side, to decode a TLS 1.3 packet capture.
12 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
13 Previously the default library behaviour applied, sending two, each in
16 JH/03 Debug output for ACL now gives the config file name and line number for
19 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
21 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
23 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
24 buffer overrun for (non-chunking) other transports.
26 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
27 TLS1.3, means that a server rejecting a client certificate is not visible
28 to the client until the first read of encrypted data (typically the
29 response to EHLO). Add detection for that case and treat it as a failed
30 TLS connection attempt, so that the normal retry-in-clear can work (if
33 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
34 and/or domain. Found and fixed by Jason Betts.
36 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
37 configuration). If a CNAME target was not a wellformed name pattern, a
40 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
41 the OS reports them interleaved with other addresses.
43 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
44 used both for input and for a verify callout, both encrypted, SMTP
45 responses being sent by the server could be lost. This resulted in
46 dropped connections and sometimes bounces generated by a peer sending
49 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
50 string. Previously it was possible to cause undefined behaviour in a
51 library routine (usually a crash). Found by "zerons".
53 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
56 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
57 API was removed, so update to use the newer ones.
59 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
60 any timeout set, is taking a long time. Previously we would hang on to a
61 rotated logfile "forever" if the input was arriving with long gaps
62 (a previous attempt to fix addressed lack, for a long time, of initial
65 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
66 shared (NFS) environment. The length of the tempfile name is now
67 4 + 16 ("hdr.$message_exim_id") which might break on file
68 systems which restrict the file name length to lower values.
71 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
72 shared (NFS) environment.
74 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
75 did for all versions <4.90). Notably -M, -m, --invert, -I may be
78 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
79 on some platforms for bit 31.
81 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
82 to changes apparently associated with TLS1.3 handling some of the APIs
83 previously used were either nonfunctional or inappropriate. Strings
84 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
85 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
86 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
87 This affects log line X= elements, the $tls_{in,out}_cipher variables,
88 and the use of specific cipher names in the encrypted= ACL condition.
90 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
92 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
93 verification result was not updated unless hosts_require_ocsp applied.
95 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
96 queue_list_requires_admin set to false, non-admin users were denied the
99 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
100 directory-of-certs mode. Previously they were advertised despite the
103 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
104 A single TCP connection by a client will now hold a TLS connection open
105 for multiple message deliveries, by default. Previoud the default was to
108 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
109 default. If built with the facility, DANE will be used. The facility
110 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
112 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
113 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
114 must be defined and you must still, unless you define DISABLE_TLS, manage
115 the the include-dir and library-file requirements that go with that
116 choice. Non-TLS builds are still supported.
118 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
121 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
122 default. If the platform supports and has the facility enabled, it will
123 be requested on all coneections.
125 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
126 controlled by the build-time option SUPPORT_PIPE_CONNECT.
128 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
130 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
131 success-DSN messages. Previously the From: header was always the default
132 one for these; the option was ignored.
134 JH/28 Fix the timeout on smtp response to apply to the whole response.
135 Previously it was reset for every read, so a teergrubing peer sending
136 single bytes within the time limit could extend the connection for a
137 long time. Credit to Qualsys Security Advisory Team for the discovery.
139 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
140 delivery address, which leaked information of the results of local
141 forwarding. Change to the original envelope recipient address, per
144 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
145 requested. Previously not bounce was generated and a log entry of
146 error ignored was made.
148 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
150 JH/32 Introduce a general tainting mechanism for values read from the input
151 channel, and values derived from them. Refuse to expand any tainted
152 values, to catch one form of exploit.
154 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
155 was unused and the unexpanded text used for the test. Found and
156 fixed by Ruben Jenster.
158 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
159 an attempt to use a TLS library read routine dereffed a nul pointer,
162 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
163 them in our smtp response.
165 JH/36 Have the general router option retry_use_local_part default to true when
166 any of the restrictive preconditions are set (to anything). Previously it
167 was only for check_local user. The change removes one item of manual
168 configuration which is required for proper retries when a remote router
169 handles a subset of addresses for a domain.
171 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
172 link count into consideration.
174 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
175 caused the extension of big_buffer, the following lines were ignored.
177 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
178 accordance with RFC 2308. Previously there was no expiry, so a longlived
179 receive process (eg. due to ACL delays) versus a short SOA value could
182 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
184 JH/39 Promote DMARC support to mainline.
186 JH/40 Bug 2452: Add a References: header to DSNs.
188 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
189 parameters. The relevant library call is documented as "Deprecated: This
190 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
191 3.6.0, DH parameters are negotiated following RFC7919."
193 HS/06 Change the default of dnssec_request_domains to "*"
195 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
196 carried on and emitted a BDAT command, even when PIPELINING was not
199 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
200 buffer was used for the filename, resulting in a trap when tainted
201 arguments (eg. $domain) were used.
203 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
204 recommended to avoid a possible server-load attack. The feature can be
205 re-enabled via the openssl_options main cofiguration option.
207 JH/45 local_scan API: documented the current smtp_printf() call. This changed
208 for version 4.90 - adding a "more data" boolean to the arguments.
209 Bumped the ABI version number also, this having been missed previously;
210 release versions 4.90 to 4.92.3 inclusive were effectively broken in
211 respect of usage of smtp_printf() by either local_scan code or libraries
212 accessed via the ${dlfunc } expansion item. Both will need coding
213 adjustment for any calls to smtp_printf() to match the new function
214 signature; a FALSE value for the new argument is always safe.
216 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
217 the file-offset (which the Linux syscall does, and exim expects); this
218 resulted in an indefinite loop.
220 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
221 to do ARC verification. The Authentication-Results: header line added
222 by the configuration then had no ARC item.
228 JH/01 Remove code calling the customisable local_scan function, unless a new
229 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
231 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
232 non-signal-safe functions being used.
234 JH/03 Bug 2269: When presented with a received message having a stupidly large
235 number of DKIM-Signature headers, disable DKIM verification to avoid
236 a resource-consumption attack. The limit is set at twenty.
238 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
239 report of oldest_pass in ${authres } in consequence, and separate out
240 some descriptions of reasons for verification fail.
242 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
243 files in the spool were present and unlocked. A queue-runner could spot
244 them, resulting in a duplicate delivery. Fix that by doing the unlock
245 after the unlink. Investigation by Tim Stewart. Take the opportunity to
246 add more error-checking on spoolfile handling while that code is being
249 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
250 No known attacks, no CVE, this is defensive hardening.
252 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
253 a queue-runner could start a delivery while other operations were ongoing.
254 Cutthrough delivery was a common victim, resulting in duplicate delivery.
255 Found and investigated by Tim Stewart. Fix by using the open message data
256 file handle rather than opening another, and not locally closing it (which
257 releases a lock) for that case, while creating the temporary .eml format
258 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
260 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
261 $sender_verify_failure/$recipient_verify_failure to "random".
263 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
266 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
267 Previously this would segfault.
269 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
272 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
273 like zero, since the resolver should be doing this for us, But we need one
274 as a CNAME but no MX presence gets the CNAME returned; we need to check
275 that doesn't point to an MX to declare it "no MX returned" rather than
276 "error, loop". A new main option is added so the older capability of
277 following some limited number of chain links is maintained.
279 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
281 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
282 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
283 now-deprecated earlier definitions used only specified the range up to TLS
284 1.2 (in the older-version library docs).
286 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
288 JH/15 Rework TLS client-side context management. Stop using a global, and
289 explicitly pass a context around. This enables future use of TLS for
290 connections to service-daemons (eg. malware scanning) while a client smtp
291 connection is using TLS; with cutthrough connections this is quite likely.
293 JH/16 Fix ARC verification to do AS checks in reverse order.
295 JH/17 Support a "tls" option on the ${readsocket } expansion item.
297 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
298 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
299 Previously the "utf8" would be re-prepended for every additional message.
301 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
302 Previously thery were accepted, resulting in issues when attempting to
303 forward messages to a non-supporting MTA.
305 PP/02 Let -n work with printing macros too, not just options.
307 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
308 one parent address was copied, and bogus data was used at delivery-logging
309 time. Either a crash (after delivery) or bogus log data could result.
310 Discovery and analysis by Tim Stewart.
312 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
313 Previously if the string ended mid-character, we did not insert the
314 promised '?' replacement.
316 PP/04 Documentation: current string operators work on bytes, not codepoints.
318 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
319 should pack well giving a smaller memory footprint so better caching and
320 therefore performance. Group the declarations where this can't be done so
321 that the byte-sized flag variables are not interspersed among pointer
322 variables, giving a better chance of good packing by the compiler.
324 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
325 non-null, to avoid issues with sites running BATV. Previously reports were
326 sent with an empty envelope sender so looked like bounces.
328 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
329 The ignore_error flag wasn't being returned from the filter subprocess so
330 was not set for later routers. Investigation and fix by Matthias Kurz.
332 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
333 and a msg:complete for the whole, when a message is manually removed using
334 -Mrm. Developement by Matthias Kurz, hacked on by JH.
336 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
337 a "Gnu special" function, asprintf() in the DB utility binary builds; I
338 hope that is portable enough.
340 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
341 requiring a known-CA anchor certificate; make it now rely entirely on the
342 TLSA as an anchor. Checking the name on the leaf cert against the name
343 on the A-record for the host is still done for TA (but not for EE mode).
345 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
346 included in delivery lines for non-proxied connections, when compiled with
347 SUPPORT_SOCKS and running with proxy logging enabled.
349 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
350 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
351 move the existing event to fire before the normal logging of message
352 failure so that custom logging is bracketed by normal logging.
354 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
355 msg:fail:internal event. Developement by Matthias Kurz.
357 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
358 far too small for todays use of crypto signatures stored there. Go all
359 the way to the max DNS message size of 64kB, even though this might be
360 overmuch for IOT constrained device use.
362 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
363 copy a string over itself. The library routine is documented as not
364 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
366 JH/32 For main options check_spool_space and check_inode_space, where the
367 platform supports 64b integers, support more than the previous 2^31 kB
368 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
369 the previous G, M, k.
371 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
372 $authenticated_fail_id variable on authentication failure. Previously
375 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
376 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
377 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
378 GNUTLS_SEC_PARAM_MEDIUM.
380 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
381 side. Previously we would continue as if no SNI had been received.
383 JH/36 Harden the handling of string-lists. When a list consisted of a sole
384 "<" character, which should be a list-separator specification, we walked
385 off past the nul-terimation.
387 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
388 causes) even when the retry time is not yet met. Previously they were
389 not, meaning that when (say) an account was over-quota and temp-rejecting,
390 and multiple senders' messages were queued, only one sender would get
391 notified on each configured delay_warning cycle.
393 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
395 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
396 cipher-suites, an error can be left on the stack even for a succeeding
397 accept; this results in impossible error messages when a later operation
400 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
401 return error codes indicating retry. Under TLS1.3 this becomes required.
403 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
404 it only wrote the new authenticators, resulting in a lack of tracking of
405 peer changes of ESMTP extensions until the next cache flush.
407 JH/41 Fix the loop reading a message header line to check for integer overflow,
408 and more-often against header_maxsize. Previously a crafted message could
409 induce a crash of the recive process; now the message is cleanly rejected.
411 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
412 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
418 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
419 When redis_servers is set to a list of > 1 element, and the Redis servers
420 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
421 case of MOVED into a DEFER case instead, thus moving the query onto the
422 next server in the list. For a cluster of N elements, all N servers must
423 be defined in redis_servers.
425 GF/02 Catch and remove uninitialized value warning in exiqsumm
426 Check for existence of @ARGV before looking at $ARGV[0]
428 JH/01 Replace the store_release() internal interface with store_newblock(),
429 which internalises the check required to safely use the old one, plus
430 the allocate and data copy operations duplicated in both (!) of the
431 extant use locations.
433 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
434 modifier. This matches the restriction on the commandline.
436 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
437 Previously only the last row was returned.
439 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
440 we assumed that tags in the header were well-formed, and parsed the
441 element content after inspecting only the first char of the tag.
442 Assumptions at that stage could crash the receive process on malformed
445 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
446 While running the DKIM ACL we operate on the Permanent memory pool so that
447 variables created with "set" persist to the DATA ACL. Also (at any time)
448 DNS lookups that fail create cache records using the Permanent pool. But
449 expansions release any allocations made on the current pool - so a dnsdb
450 lookup expansion done in the DKIM ACL releases the memory used for the
451 DNS negative-cache, and bad things result. Solution is to switch to the
452 Main pool for expansions.
453 While we're in that code, add checks on the DNS cache during store_reset,
454 active in the testsuite.
455 Problem spotted, and debugging aided, by Wolfgang Breyha.
457 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
458 When none of the hosts presented to a transport match an already-open
459 connection, close it and proceed with the list. Previously we would
460 queue the message. Spotted by Lena with Yahoo, probably involving
463 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
464 Previously a spurious "250 OK id=" response was appended to the proper
467 JH/08 The "support for" informational output now, which built with Content
468 Scanning support, has a line for the malware scanner interfaces compiled
469 in. Interface can be individually included or not at build time.
471 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
472 by the template makefile "src/EDITME". The "STREAM" support for an older
473 ClamAV interface method is removed.
475 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
476 rows affected is given instead).
478 JH/11 The runtime Berkeley DB library version is now additionally output by
479 "exim -d -bV". Previously only the compile-time version was shown.
481 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
482 SMTP connection. Previously, when one had more receipients than the
483 first, an abortive onward connection was made. Move to full support for
484 multiple onward connections in sequence, handling cutthrough connection
485 for all multi-message initiating connections.
487 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
488 routers. Previously, a multi-recipient message would fail to match the
489 onward-connection opened for the first recipient, and cause its closure.
491 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
492 a timeout on read on a GnuTLS initiating connection, resulting in the
493 initiating connection being dropped. This mattered most when the callout
494 was marked defer_ok. Fix to keep the two timeout-detection methods
497 JH/15 Relax results from ACL control request to enable cutthrough, in
498 unsupported situations, from error to silently (except under debug)
499 ignoring. This covers use with PRDR, frozen messages, queue-only and
502 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
504 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
505 metadata, resulting in a crash in free().
507 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
508 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
509 Broken also in d185889f4, with init system revamp.
511 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
512 Previously we abruptly closed the connection after reading a malware-
513 found indication; now we go on to read the "scan ok" response line,
516 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
517 ACL. Previously, a crash would result.
519 JH/19 Speed up macro lookups during configuration file read, by skipping non-
520 macro text after a replacement (previously it was only once per line) and
521 by skipping builtin macros when searching for an uppercase lead character.
523 JH/20 DANE support moved from Experimental to mainline. The Makefile control
524 for the build is renamed.
526 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
527 was allocated for every new TLS startup, meaning one per message. Fix
528 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
530 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
531 reported the original. Fix to report (as far as possible) the ACL
532 result replacing the original.
534 JH/23 Fix memory leak during multi-message connections using STARTTLS under
535 OpenSSL. Certificate information is loaded for every new TLS startup,
536 and the resources needed to be freed.
538 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
540 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
543 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
544 DATA response info to the (existing) per-recipient response info for
545 the "C=" log element. It can have useful tracking info from the
546 destination system. Patch from Simon Arlott.
548 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
549 length value. Previously this would segfault.
551 HS/02 Support Avast multiline protoocol, this allows passing flags to
552 newer versions of the scanner.
554 JH/28 Ensure that variables possibly set during message acceptance are marked
555 dead before release of memory in the daemon loop. This stops complaints
556 about them when the debug_store option is enabled. Discovered specifically
557 for sender_rate_period, but applies to a whole set of variables.
558 Do the same for the queue-runner and queue-list loops, for variables set
559 from spool message files. Do the same for the SMTP per-message loop, for
560 certain variables indirectly set in ACL operations.
562 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
563 as a multi-recipient message from a mailinglist manager). The coding had
564 an arbitrary cutoff number of characters while checking for more input;
565 enforced by writing a NUL into the buffer. This corrupted long / fast
566 input. The problem was exposed more widely when more pipelineing of SMTP
567 responses was introduced, and one Exim system was feeding another.
568 The symptom is log complaints of SMTP syntax error (NUL chars) on the
569 receiving system, and refused recipients seen by the sending system
570 (propating to people being dropped from mailing lists).
571 Discovered and pinpointed by David Carter.
573 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
574 replaced by the ${authresults } expansion.
576 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
578 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
579 allows proper process termination in container environments.
581 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
582 Previously the "final dot" had a newline after it; ensure it is CR,LF.
584 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
585 and "err_perm", deprecated since 4.83 when the RFC-defined words
586 "temperror" and "permerror" were introduced.
588 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
589 transport-filters or DKIM-signing. The restriction was lost in the
590 consolidation of verify-callout and delivery SMTP handling.
591 Extend the restriction to also cover ARC-signing.
593 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
594 in defer=pass mode supply a 450 to the initiator. Previously the message
597 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
598 tls_require_ciphers is used as before.
600 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
601 "pass_unscanned". Only tmpfails from the scanner are written to
602 the paniclog, as they may require admin intervention (permission
603 denied, license issues). Other scanner errors (like decompression
604 bombs) do not cause a paniclog entry.
606 JH/36 Fix reinitialisation of DKIM logging variable between messages.
607 Previously it was possible to log spurious information in receive log
610 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
611 triggered odd behaviour from Outlook Express clients.
613 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
616 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
617 since the IETF WG has not yet settled on that versus the original
618 "bare" representation.
620 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
621 Previously the millisecond value corrupted the output.
622 Fix also for syslog_pid=no and log_selector +pid, for which the pid
623 corrupted the output.
629 JH/01 Rework error string handling in TLS interface so that the caller in
630 more cases is responsible for logging. This permits library-sourced
631 string to be attached to addresses during delivery, and collapses
632 pairs of long lines into single ones.
634 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
635 during configuration. Wildcards are allowed and expanded.
637 JH/02 Rework error string handling in DKIM to pass more info back to callers.
638 This permits better logging.
640 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
641 do not close it down and have the child transport start it up again on
642 the passed-on TCP connection. Instead, proxy the child (and any
643 subsequent ones) for TLS via a unix-domain socket channel. Logging is
644 affected: the continued delivery log lines do not have any DNSSEC, TLS
645 Certificate or OCSP information. TLS cipher information is still logged.
647 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
648 identical IP addresses on different listening ports. Will also affect
651 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
652 add noisy ifdef guards to special-case this sillyness.
653 Patch from Bernd Kuhls.
655 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
656 than 255 are no longer allowed.
658 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
659 Disable the session-cache too, which might reduce our load. Since we
660 currrectly use a new context for every connection, both as server and
661 client, there is no benefit for these.
662 GnuTLS appears to not support tickets server-side by default (we don't
663 call gnutls_session_ticket_enable_server()) but client side is enabled
664 by default on recent versions (3.1.3 +) unless the PFS priority string
667 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
668 <https://reproducible-builds.org/specs/source-date-epoch/>.
670 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
671 the check for any unsuccessful recipients did not notice the limit, and
672 erroneously found still-pending ones.
674 JH/08 Pipeline CHUNKING command and data together, on kernels that support
675 MSG_MORE. Only in-clear (not on TLS connections).
677 JH/09 Avoid using a temporary file during transport using dkim. Unless a
678 transport-filter is involved we can buffer the headers in memory for
679 creating the signature, and read the spool data file once for the
680 signature and again for transmission.
682 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
683 4.77 as the kernel support then wasn't solid, having issues in 64bit
684 mode. Now, it's been long enough. Add support for FreeBSD also.
686 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
687 case where the routing stage had gathered several addresses to send to
688 a host before calling the transport for the first, we previously failed
689 to close down TLS in the old transport process before passing the TCP
690 connection to the new process. The new one sent a STARTTLS command
691 which naturally failed, giving a failed delivery and bloating the retry
692 database. Investigation and fix prototype from Wolfgang Breyha.
694 JH/12 Fix check on SMTP command input synchronisation. Previously there were
695 false-negatives in the check that the sender had not preempted a response
696 or prompt from Exim (running as a server), due to that code's lack of
697 awareness of the SMTP input buffering.
699 PP/04 Add commandline_checks_require_admin option.
700 Exim drops privileges sanely, various checks such as -be aren't a
701 security problem, as long as you trust local users with access to their
702 own account. When invoked by services which pass untrusted data to
703 Exim, this might be an issue. Set this option in main configuration
704 AND make fixes to the calling application, such as using `--` to stop
707 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
708 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
709 responses to those, into a single TLS record each way (this usually means
710 a single packet). As a side issue, smtp_enforce_sync now works on TLS
713 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
714 affects you only if you're dancing at the edge of the param size limits.
715 If you are, and this message makes sense to you, then: raise the
716 configured limit or use OpenSSL 1.1. Nothing we can do for older
719 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
720 cmdline element to get the documented default one. Previously it was
723 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
724 are unsuable for non-message-specific reasons - eg. connection timeout,
725 banner-time rejection.
727 JH/16 Fix logging of delivery remote port, when specified by router, under
730 PP/06 Repair manualroute's ability to take options in any order, even if one
731 is the name of a transport.
734 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
736 JH/17 Change the list-building routines interface to use the expanding-string
737 triplet model, for better allocation and copying behaviour.
739 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
740 Previously it was constructed the first time a possibly-matching string
741 was met in the configuration file input during startup; now it is done
744 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
745 compatible one, to avoid the (poorly documented) possibility of a config
746 file in the working directory redirecting the DB files, possibly correpting
747 some existing file. CVE-2017-10140 assigned for BDB.
749 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
750 cache-hot. Previously, although the result was properly cached, the
751 initial verify call returned a defer.
753 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
754 the main verify for receipient in uncached-mode.
756 JH/22 Retire historical build files to an "unsupported" subdir. These are
757 defined as "ones for which we have no current evidence of testing".
759 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
760 if present. Previously it was ignored.
762 JH/24 Start using specified-initialisers in C structure init coding. This is
763 a C99 feature (it's 2017, so now considered safe).
765 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
766 if was a fixed-sized field and bitmask ops via macros; it is now more
769 PP/07 GitHub PR 56: Apply MariaDB build fix.
770 Patch provided by Jaroslav Škarvada.
772 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
773 during Coverity cleanups [4.87 JH/47]
774 Diagnosis and fix provided by Michael Fischer v. Mollard.
776 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
777 the right size to place the terminating semicolon on its own folded
778 line, the header hash was calculated to an incorrect value thanks to
779 the (relaxed) space the fold became.
781 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
782 and confused the parent.
784 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
785 which could crash as a result. This could lead to undeliverable messages.
787 JH/28 Logging: "next input sent too soon" now shows where input was truncated
790 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
791 matters on fast-turnover and PID-randomising systems, which were getting
792 out-of-order delivery.
794 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
795 a possibly-overlapping copy. The symptom was that "Remote host closed
796 connection in response to HELO" was logged instead of the actual 4xx
799 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
800 Previously only that bufferd was discarded, resulting in SYMTP command
803 JH/32 DKIM: when a message has multiple signatures matching an identity given
804 in dkim_verify_signers, run the dkim acl once for each. Previously only
805 one run was done. Bug 2189.
807 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
808 "panic the current process" to "deliberately defer". The panic log is
809 still written with the problem list name; the mail and reject logs now
810 get a temp-reject line for the message that was being handled, saying
811 something like "domains check lookup or other defer". The SMTP 451
812 message is still "Temporary local problem".
814 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
815 A crafted sequence of BDAT commands could result in in-use memory beeing
816 freed. CVE-2017-16943.
818 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
819 from SMTP input. Previously it was always done; now only done for DATA
820 and not BDAT commands. CVE-2017-16944.
822 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
823 to the message (such as an overlong header line). Previously this was
824 not done and we did not exit BDAT mode. Followon from the previous item
825 though a different problem.
831 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
832 than -2003 did; needs libidn2 in addition to libidn.
834 JH/02 The path option on a pipe transport is now expanded before use.
836 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
837 Patch provided by "Björn", documentation fix added too.
839 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
840 missing a wire-to-host endian conversion.
842 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
843 close after a BDAT command line could be taken as a following command,
844 giving a synch failure. Fix by only checking for synch immediately
845 before acknowledging the chunk.
847 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
848 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
849 macro. Patches provided by Josh Soref.
851 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
852 Previously we did not; the RFC seems ambiguous and VRFY is not listed
853 by IANA as a service extension. However, John Klensin suggests that we
856 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
857 the dkim code may be unix-mode line endings rather than smtp wire-format
858 CRLF, so prepend a CR to any bare LF.
860 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
861 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
863 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
864 the first were themselves being wrongly included in the feed into dkim
865 processing; with most chunk sizes in use this resulted in an incorrect
866 body hash calculated value.
868 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
869 DKIM signature block, for verification. Although advised against by
870 standards it is specifically not ruled illegal.
872 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
874 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
875 missing a body hash (the bh= tag).
877 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
878 It seems that HAProxy sends the Proxy Protocol information in clear and
879 only then does a TLS startup, so do the same.
881 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
882 TCP connections (such as for Spamd) unless the daemon successfully set
883 Fast Open mode on its listening sockets. This fixes breakage seen on
884 too-old kernels or those not configured for Fast Open, at the cost of
885 requiring both directions being enabled for TFO, and TFO never being used
886 by non-daemon-related Exim processes.
888 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
889 endings, at least on the first header line. Try to canonify any that get
890 past that check, despite the cost.
892 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
893 now limited to an arbitrary five deep, while parsing addresses with the
894 strip_excess_angle_brackets option enabled.
896 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
897 instead leave the unprompted TLS handshake in socket buffer for the
898 TLS library to consume.
900 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
902 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
904 JH/16 Drop variables when they go out of scope. Memory management drops a whole
905 region in one operation, for speed, and this leaves assigned pointers
906 dangling. Add checks run only under the testsuite which checks all
907 variables at a store-reset and panics on a dangling pointer; add code
908 explicitly nulling out all the variables discovered. Fixes one known
909 bug: a transport crash, where a dangling pointer for $sending_ip_address
910 originally assigned in a verify callout, is re-used.
912 PP/06 Drop '.' from @INC in various Perl scripts.
914 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
916 PP/08 Reduce a number of compilation warnings under clang; building with
917 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
918 should be warning-free.
920 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
922 HS/01 Fix portability problems introduced by PP/08 for platforms where
923 realloc(NULL) is not equivalent to malloc() [SunOS et al].
925 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
926 chunk. This allows us to accept broken chunked messages. We need a more
927 general solution here.
929 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
930 already-broken messages in the queue.
932 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
934 JH/19 Fix reference counting bug in routing-generated-address tracking.
940 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
941 supports it and a size is available (ie. the sending peer gave us one).
943 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
944 years of being deprecated). The replacements are the ACLs
945 acl_smtp_mime and acl_not_smtp_mime.
947 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
948 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
949 or even an in-clear connection were permitted. Now, if the host lookup
950 was dnssec and dane was requested then the host is only used if the
951 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
952 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
953 if one fails this test.
954 This means that a poorly-configured remote DNS will make it incommunicado;
955 but it protects against a DNS-interception attack on it.
957 JH/04 Bug 1810: make continued-use of an open smtp transport connection
958 non-noisy when a race steals the message being considered.
960 JH/05 If main configuration option tls_certificate is unset, generate a
961 self-signed certificate for inbound TLS connections.
963 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
964 in rewrites and routers.
966 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
967 and logged a warning sing 4.83; now they are a configuration file error.
969 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
970 (lacking @domain). Apply the same qualification processing as RCPT.
972 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
974 JH/10 Support ${sha256:} applied to a string (as well as the previous
977 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
978 a cutthrough deliver is pending, as we always want to make a connection.
979 This also avoids re-routing the message when later placing the cutthrough
980 connection after a verify cache hit.
981 Do not update it with the verify result either.
983 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
984 when routing results in more than one destination address.
986 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
987 signing (which inhibits the cutthrough capability). Previously only
988 the presence of an option was tested; now an expansion evaluating as
989 empty is permissible (obviously it should depend only on data available
990 when the cutthrough connection is made).
992 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
993 the relevant preceding SMTP command did not note the pipelining mode.
995 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
996 Previously they were not counted.
998 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
999 as one having no matching records. Previously we deferred the message
1000 that needed the lookup.
1002 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1003 distinguished as "(=".
1005 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1006 for missing MX records. Previously it only worked for missing A records.
1008 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1010 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1011 after the data-go-ahead and data-ack. Patch from Jason Betts.
1013 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1014 even for a "none" policy. Patch from Tony Meyer.
1016 JH/22 Fix continued use of a connection for further deliveries. If a port was
1017 specified by a router, it must also match for the delivery to be
1020 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1021 When one of the recipients of a message was unsuitable for the connection
1022 (has no matching addresses), we lost track of needing to mark it
1023 deferred. As a result mail would be lost.
1025 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1027 JH/25 Decoding ACL controls is now done using a binary search; the source code
1028 takes up less space and should be simpler to maintain. Merge the ACL
1029 condition decode tables also, with similar effect.
1031 JH/26 Fix problem with one_time used on a redirect router which returned the
1032 parent address unchanged. A retry would see the parent address marked as
1033 delivered, so not attempt the (identical) child. As a result mail would
1036 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1037 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1038 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1041 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1042 Main config options check_{log,spool}_{inodes,space} are now
1043 100 inodes, 10MB unless set otherwise in the configuration.
1045 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1046 Previously it only applied to the main-section connection policy
1049 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1051 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1052 by me. Added RFC7919 DH primes as an alternative.
1054 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1055 are not in the system include path.
1057 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1058 GnuTLS, when a session startup failed (eg because the client disconnected)
1059 Exim did stdio operations after fclose. This was exposed by a recent
1060 change which nulled out the file handle after the fclose.
1062 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1063 signed directly by the cert-signing cert, rather than an intermediate
1064 OCSP-signing cert. This is the model used by LetsEncrypt.
1066 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1068 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1069 an incoming connection.
1071 HS/02 Bug 1802: Do not half-close the connection after sending a request
1074 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1075 fallback to "prime256v1".
1077 JH/34 SECURITY: Use proper copy of DATA command in error message.
1078 Could leak key material. Remotely exploitable. CVE-2016-9963.
1084 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1085 and 3.4.4 - once the server is enabled to respond to an OCSP request
1086 it does even when not requested, resulting in a stapling non-aware
1087 client dropping the TLS connection.
1089 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1090 support variable-length bit vectors. No functional change.
1092 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1093 The I= interface field on outgoing lines is now after the H= remote
1094 host field, same as incoming lines. There is a separate
1095 outgoing_interface log selector which allows you to disable the
1098 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1099 If not running log_selector +smtp_connection the mainlog would be held
1100 open indefinitely after a "too many connections" event, including to a
1101 deleted file after a log rotate. Leave the per net connection logging
1102 leaving it open for efficiency as that will be quickly detected by the
1103 check on the next write.
1105 HS/01 Bug 1671: Fix post transport crash.
1106 Processing the wait-<transport> messages could crash the delivery
1107 process if the message IDs didn't exist for some reason. When
1108 using 'split_spool_directory=yes' the construction of the spool
1109 file name failed already, exposing the same netto behaviour.
1111 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1112 mime_regex ACL conditions.
1114 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1115 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1116 to HELO, local diagnostic string.
1118 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1119 log line to debug. Even when configured with a tls authenticator many
1120 client connections are expected to not authenticate in this way, so
1121 an authenticate fail is not an error.
1123 HS/02 Add the Exim version string to the process info. This way exiwhat
1124 gives some more detail about the running daemon.
1126 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1127 matter for fast-change records such as DNSBLs.
1129 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1130 retry record, even if constant. There may be multiple transports with
1131 different interface settings and the retry behaviour needs to be kept
1134 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1136 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1138 JH/10 Bug 840: fix log_defer_output option of pipe transport
1140 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1141 hosts_randomize. This matters a lot when combined with mua_wrapper.
1143 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1144 ${quote_pgsql:<string>} operator.
1146 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1148 JH/14 Change header file name for freeradius-client. Relevant if compiling
1149 with Radius support; from the Gentoo tree and checked under Fedora.
1151 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1153 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1154 option transport_filter, meaning no filtering.
1156 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1158 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1159 defaults to "*" (all hosts). The variable is now available when not built
1160 with TLS, default unset, mainly to enable keeping the testsuite sane.
1161 If a server certificate is not supplied (via tls_certificate) an error is
1162 logged, and clients will find TLS connections fail on startup. Presumably
1163 they will retry in-clear.
1164 Packagers of Exim are strongly encouraged to create a server certificate
1165 at installation time.
1167 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1168 with the $config_file variable.
1170 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1171 in transport context, after the attempt, and per-recipient. The latter type
1172 is per host attempted. The event data is the error message, and the errno
1173 information encodes the lookup type (A vs. MX) used for the (first) host,
1174 and the trailing two digits of the smtp 4xx response.
1176 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1177 to write to mainlog (or rejectlog, paniclog) in the window between file
1178 creation and permissions/ownership being changed. Particularly affects
1179 installations where exicyclog is run as root, rather than exim user;
1180 result is that the running daemon panics and dies.
1182 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1184 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1185 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1186 "pri" and "weight". Note that the previous implicit priority given by the
1187 list order is no longer honoured.
1189 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1190 for DKIM processing.
1192 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1193 by defining SUPPORT_SOCKS.
1195 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1196 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1197 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1198 variables are renamed to proxy_{local,external}_{address,port}.
1200 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1201 for a build by defining SUPPORT_I18N
1203 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1204 of the query string, and make ${quote_redis:} do that quoting.
1206 JH/29 Move Events support from Experimental to mainline, enabled by default
1207 and removable for a build by defining DISABLE_EVENT.
1209 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1211 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1212 cached by the daemon.
1214 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1215 by defining LOOKUP_REDIS. The libhiredis library is required.
1217 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1218 keys are given for lookup.
1220 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1221 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1222 only supported when built with TLS support. The PolarSSL SHA routines
1223 are still used when the TLS library is too old for convenient support.
1225 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1226 openssl_options), for security. OpenSSL forces this from version 1.1.0
1227 server-side so match that on older versions.
1229 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1230 allocation for $value could be released as the expansion processing
1231 concluded, but leaving the global pointer active for it.
1233 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1234 and to use the domains and local_parts ACL conditions.
1236 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1237 incorrectly not doubled on cutthrough transmission, hence seen as a
1238 body-termination at the receiving system - resulting in truncated mails.
1239 Commonly the sender saw a TCP-level error, and retransmitted the message
1240 via the normal store-and-forward channel. This could result in duplicates
1241 received - but deduplicating mailstores were liable to retain only the
1242 initial truncated version.
1244 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1246 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1248 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1249 we're in there, support oversigning also; bug 1309.
1251 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1253 HS/04 Add support for keep_environment and add_environment options.
1255 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1256 either intentional arithmetic overflow during PRNG, or testing config-
1259 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1260 delivery resulted in actual delivery. Cancel cutthrough before DATA
1263 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1264 rejects a recipient: pass the reject to the originator.
1266 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1267 Many were false-positives and ignorable, but it's worth fixing the
1270 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1271 for the new environment-manipulation done at startup. Move the routines
1272 from being local to tls.c to being global via the os.c file.
1274 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1275 an extract embedded as result-arg for a map, the first arg for extract
1276 is unavailable so we cannot tell if this is a numbered or keyed
1277 extraction. Accept either.
1283 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1286 JH/02 The smtp transport option "multi_domain" is now expanded.
1288 JH/03 The smtp transport now requests PRDR by default, if the server offers
1291 JH/04 Certificate name checking on server certificates, when exim is a client,
1292 is now done by default. The transport option tls_verify_cert_hostnames
1293 can be used to disable this per-host. The build option
1294 EXPERIMENTAL_CERTNAMES is withdrawn.
1296 JH/05 The value of the tls_verify_certificates smtp transport and main options
1297 default to the word "system" to access the system default CA bundle.
1298 For GnuTLS, only version 3.0.20 or later.
1300 JH/06 Verification of the server certificate for a TLS connection is now tried
1301 (but not required) by default. The verification status is now logged by
1302 default, for both outbound TLS and client-certificate supplying inbound
1305 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1308 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1309 Status Notification (bounce) messages are now MIME format per RFC 3464.
1310 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1311 under the control of the dsn_advertise_hosts option, and routers may
1312 have a dsn_lasthop option.
1314 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1315 default, modifiable by a malware= option. The list separator for
1316 the options can now be changed in the usual way. Bug 68.
1318 JH/10 The smtp_receive_timeout main option is now expanded before use.
1320 JH/11 The incoming_interface log option now also enables logging of the
1321 local interface on delivery outgoing connections.
1323 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1324 if the interface and destination host and port all match.
1326 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1329 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1330 Patch from Andrew Lewis.
1332 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1333 now supports optional time-restrictions, weighting, and priority
1334 modifiers per server. Patch originally by <rommer@active.by>.
1336 JH/16 The spamd_address main option now supports a mixed list of local
1337 and remote servers. Remote servers can be IPv6 addresses, and
1338 specify a port-range.
1340 JH/17 Bug 68: The spamd_address main option now supports an optional
1341 timeout value per server.
1343 JH/18 Bug 1581: Router and transport options headers_add/remove can
1344 now have the list separator specified.
1346 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1349 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1352 JH/21 Support for the A6 type of dns record is withdrawn.
1354 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1355 rather than the verbs used.
1357 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1358 from 255 to 1024 chars.
1360 JH/24 Verification callouts now attempt to use TLS by default.
1362 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1363 are generic router options now. The defaults didn't change.
1365 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1366 Original patch from Alexander Shikoff, worked over by JH.
1368 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1369 files if ZCAT_COMMAND is not executable.
1371 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1373 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1375 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1376 Normally benign, it bites when the pair was led to by a CNAME;
1377 modern usage is to not canonicalize the domain to a CNAME target
1378 (and we were inconsistent anyway for A-only vs AAAA+A).
1380 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1382 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1383 when evaluating $sender_host_dnssec.
1385 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1386 $sender_helo_dnssec variable.
1388 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1390 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1392 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1394 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1395 documented as working, but never had. Support all but $spam_report.
1397 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1398 added for tls authenticator.
1400 HS/03 Add perl_taintmode main config option
1406 TL/01 When running the test suite, the README says that variables such as
1407 no_msglog_check are global and can be placed anywhere in a specific
1408 test's script, however it was observed that placement needed to be near
1409 the beginning for it to behave that way. Changed the runtest perl
1410 script to read through the entire script once to detect and set these
1411 variables, reset to the beginning of the script, and then run through
1412 the script parsing/test process like normal.
1414 TL/02 The BSD's have an arc4random API. One of the functions to induce
1415 adding randomness was arc4random_stir(), but it has been removed in
1416 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1417 function when detected.
1419 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1420 cause callback expansion.
1422 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1423 syntax errors in an expansion can be treated as a string instead of
1424 logging or causing an error, due to the internal use of bool_lax
1425 instead of bool when processing it.
1427 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1428 server certificates when making smtp deliveries.
1430 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1432 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1434 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1436 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1437 Merged patch from Sebastian Wiedenroth.
1439 JH/05 Fix results-pipe from transport process. Several recipients, combined
1440 with certificate use, exposed issues where response data items split
1441 over buffer boundaries were not parsed properly. This eventually
1442 resulted in duplicates being sent. This issue only became common enough
1443 to notice due to the introduction of connection certificate information,
1444 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1446 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1447 size buffer was used, resulting in syntax errors when an expansion
1450 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1451 version 3.3.6 or later.
1453 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1454 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1455 both become "event_action", the variables become $event_name, $event_data
1456 and $event_defer_errno. There is a new variable $verify_mode, usable in
1457 routers, transports and related events. The tls:cert event is now also
1458 raised for inbound connections, if the main configuration event_action
1461 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1462 early OCSP support, but no stapling (appears to be less than 1.0.0).
1464 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1465 server certificate names available under the smtp transport option
1466 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1469 JH/10 Time-related extraction expansions from certificates now use the main
1470 option "timezone" setting for output formatting, and are consistent
1471 between OpenSSL and GnuTLS compilations. Bug 1541.
1473 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1474 encoded parameter in the incoming message. Bug 1558.
1476 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1477 include certificate info, eximon was claiming there were spoolfile
1480 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1482 JH/14 Log delivery-related information more consistently, using the sequence
1483 "H=<name> [<ip>]" wherever possible.
1485 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1486 are problematic for Debian distribution, omit them from the release
1489 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1491 JH/16 Fix string representation of time values on 64bit time_t architectures.
1494 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1495 output list separator was used.
1500 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1501 checkers that were complaining about end of non-void function with no
1504 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1505 This was a regression introduced in 4.83 by another bugfix.
1507 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1509 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1510 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1516 TF/01 Correctly close the server side of TLS when forking for delivery.
1518 When a message was received over SMTP with TLS, Exim failed to clear up
1519 the incoming connection properly after forking off the child process to
1520 deliver the message. In some situations the subsequent outgoing
1521 delivery connection happened to have the same fd number as the incoming
1522 connection previously had. Exim would try to use TLS and fail, logging
1523 a "Bad file descriptor" error.
1525 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1526 utilities have not been installed.
1528 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1529 temporary space as the ACL may create new global variables.
1531 TL/01 LDAP support uses per connection or global context settings, depending
1532 upon the detected version of the libraries at build time.
1534 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1535 to extract and use the src ip:port in logging and expansions as if it
1536 were a direct connection from the outside internet. PPv2 support was
1537 updated based on HAProxy spec change in May 2014.
1539 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1541 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1542 Properly escape header and check for NULL return.
1544 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1547 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1549 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1550 characters in header names, implemented as a verify condition.
1551 Contributed by Michael Fischer v. Mollard.
1553 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1554 results permerror and temperror. Previous values are deprecated but
1555 still accepted. In a future release, err_perm and err_temp will be
1556 completely removed, which will be a backward incompatibility if the
1557 ACL tests for either of these two old results. Patch contributed by
1558 user bes-internal on the mailing list.
1560 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1562 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1563 selectors, in both main and reject logs.
1565 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1568 JH/07 Add malware type "sock" for talking to simple daemon.
1570 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1572 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1573 routers/transports under cutthrough routing.
1575 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1576 numbers. Touch up "bool" conditional to keep the same definition.
1578 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1580 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1582 JH/12 Expand items in router/transport headers_add or headers_remove lists
1583 individually rather than the list as a whole. Bug 1452.
1585 Required for reasonable handling of multiple headers_ options when
1586 they may be empty; requires that headers_remove items with embedded
1587 colons must have them doubled (or the list-separator changed).
1589 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1590 view the policy declared in the DMARC record. Currently, $dmarc_status
1591 is a combined value of both the record presence and the result of the
1594 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1596 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1597 dnslookup router and the smtp transport (applying to the forward
1600 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1601 of ldap servers used for a specific lookup. Patch provided by Heiko
1604 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1605 New variable $lookup_dnssec_authenticated for observability.
1607 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1608 Patch submitted by Lars Timman.
1610 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1612 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1613 Requires trusted mode and valid format message id, aborts otherwise.
1614 Patch contributed by Heiko Schlichting.
1616 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1617 certextract with support for various fields. Bug 1358.
1619 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1620 is requested by default, modifiable by smtp transport option
1623 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1624 operate on certificate variables to give certificate fingerprints
1625 Also new ${sha256:cert_variable}.
1627 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1629 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1630 Christian Aistleitner.
1632 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1634 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1635 file. Patch from Wolfgang Breyha.
1637 JH/25 Expand the coverage of the delivery $host and $host_address to
1638 client authenticators run in verify callout. Bug 1476.
1640 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1641 align with daemon_smtp_ports. Bug 72.
1643 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1644 support and error reporting did not work properly.
1646 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1647 and is readable. Patch from Andrew Colin Kissa.
1649 TL/14 Enhance documentation of ${run expansion and how it parses the
1650 commandline after expansion, particularly in the case when an
1651 unquoted variable expansion results in an empty value.
1653 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1655 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1656 by renaming away from it. Addresses GH issue 15, reported by
1659 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1661 TL/15 SECURITY: prevent double expansion in math comparison functions
1662 (can expand unsanitized data). Not remotely exploitable.
1669 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1671 PP/02 Make -n do something, by making it not do something.
1672 When combined with -bP, the name of an option is not output.
1674 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1677 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1678 $sender_host_name and config options to manage this, and basic check
1681 PP/05 DSCP support for outbound connections and control modifier for inbound.
1683 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1684 (Only plugin which currently uses this is kerberos4, which nobody should
1685 be using, but we should make it available and other future plugins might
1686 conceivably use it, even though it would break NAT; stuff *should* be
1687 using channel bindings instead).
1689 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1690 name; added for Sendmail compatibility; requires admin caller.
1691 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1692 just ignore it); requires trusted caller.
1693 Also parse but ignore: -Ac -Am -X<logfile>
1696 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1698 TL/02 Add +smtp_confirmation as a default logging option.
1700 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1701 Patch by Magnus Holmgren from 2007-02-20.
1703 TL/04 Bugzilla 1281 - Spec typo.
1704 Bugzilla 1283 - Spec typo.
1705 Bugzilla 1290 - Spec grammar fixes.
1707 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1709 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1711 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1712 mailing list by by Dmitry Isaikin.
1714 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1716 JH/02 Support "G" suffix to numbers in ${if comparisons.
1718 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1720 NM/01 Bugzilla 1197 - Spec typo
1721 Bugzilla 1196 - Spec examples corrections
1723 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1725 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1726 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1729 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1730 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1732 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1733 "acl {{name}{arg}...}", and optional args on acl condition
1736 JH/05 Permit multiple router/transport headers_add/remove lines.
1738 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1740 JH/07 Avoid using a waiting database for a single-message-only transport.
1741 Performance patch from Paul Fisher. Bugzilla 1262.
1743 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1746 JH/09 Add $headers_added variable, with content from use of ACL modifier
1747 add_header (but not yet added to the message). Bugzilla 199.
1749 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1750 Pulled from Bugzilla 817 by Wolfgang Breyha.
1752 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1754 (nb: this is the same fix as in Exim 4.80.1)
1756 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1759 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1762 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1764 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1765 advertise SMTP AUTH mechanism to us, instead of a generic
1766 protocol violation error. Also, make Exim more robust to bad
1767 data from the Dovecot auth socket.
1769 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1771 When a queue runner is handling a message, Exim first routes the
1772 recipient addresses, during which it prunes them based on the retry
1773 hints database. After that it attempts to deliver the message to
1774 any remaining recipients. It then updates the hints database using
1777 So if a recipient address works intermittently, it can get repeatedly
1778 deferred at routing time. The retry hints record remains fresh so the
1779 address never reaches the final cutoff time.
1781 This is a fairly common occurrence when a user is bumping up against
1782 their storage quota. Exim had some logic in its local delivery code
1783 to deal with this. However it did not apply to per-recipient defers
1784 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1786 This change adds a proper retry rule check during routing so that the
1787 final cutoff time is checked against the message's age. We only do
1788 this check if there is an address retry record and there is not a
1789 domain retry record; this implies that previous attempts to handle
1790 the address had the retry_use_local_parts option turned on. We use
1791 this as an approximation for the destination being like a local
1792 delivery, as in LMTP.
1794 I suspect this new check makes the old local delivery cutoff check
1795 redundant, but I have not verified this so I left the code in place.
1797 TF/02 Correct gecos expansion when From: is a prefix of the username.
1799 Test 0254 submits a message to Exim with the header
1803 When I ran the test suite under the user fanf2, Exim expanded
1804 the header to contain my full name, whereas it should have added
1805 a Resent-Sender: header. It erroneously treats any prefix of the
1806 username as equal to the username.
1808 This change corrects that bug.
1810 GF/01 DCC debug and logging tidyup
1811 Error conditions log to paniclog rather than rejectlog.
1812 Debug lines prefixed by "DCC: " to remove any ambiguity.
1814 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1816 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1817 Bug spotted by Jeremy Harris; was flawed since initial commit.
1818 Would have resulted in OCSP responses post-SNI triggering an Exim
1819 NULL dereference and crash.
1821 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1823 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1824 Bug detection, analysis and fix by Samuel Thibault.
1825 Bugzilla 1331, Debian bug #698092.
1827 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1829 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1830 Server implementation by Todd Lyons, client by JH.
1831 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1832 config variable "prdr_enable" controls whether the server
1833 advertises the facility. If the client requests PRDR a new
1834 acl_data_smtp_prdr ACL is called once for each recipient, after
1835 the body content is received and before the acl_smtp_data ACL.
1836 The client is controlled by both of: a hosts_try_prdr option
1837 on the smtp transport, and the server advertisement.
1838 Default client logging of deliveries and rejections involving
1839 PRDR are flagged with the string "PRDR".
1841 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1842 fclose(). Diagnosis by Todd Lyons.
1844 PP/17 Update configure.default to handle IPv6 localhost better.
1845 Patch by Alain Williams (plus minor tweaks).
1848 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1849 This is now consistent with GnuTLS, and is now documented: the
1850 previous undocumented portable approach to treating the option as
1851 unset was to force an expansion failure. That still works, and
1852 an empty string is now equivalent.
1854 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1855 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1856 not performing validation itself.
1858 PP/20 Added force_command boolean option to pipe transport.
1859 Patch from Nick Koston, of cPanel Inc.
1861 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1864 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1866 PP/21 Fix eximon continuous updating with timestamped log-files.
1867 Broken in a format-string cleanup in 4.80, missed when I repaired the
1868 other false fix of the same issue.
1869 Report and fix from Heiko Schlichting.
1872 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1873 Report from Prashanth Katuri.
1875 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1876 It's SecureTransport, so affects any MacOS clients which use the
1877 system-integrated TLS libraries, including email clients.
1879 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1880 using a MIME ACL for non-SMTP local injection.
1881 Report and assistance in diagnosis by Warren Baker.
1883 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1885 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1887 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1888 last id that failed so it may be referenced in subsequent ACL's.
1890 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1893 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1894 ldap library initialization, allowing self-signed CA's to be
1895 used. Also properly sets require_cert option later in code by
1896 using NULL (global ldap config) instead of ldap handle (per
1897 session). Bug diagnosis and testing by alxgomz.
1899 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1900 the src/util/ subdirectory.
1902 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1903 renamed to Transport Post Delivery Action by Jeremy Harris, as
1906 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1907 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1908 redis_servers = needs to be configured which will be used by the redis
1909 lookup. Patch from Warren Baker, of The Packet Hub.
1911 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1913 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1914 hostname or reverse DNS when processing a host list. Used suggestions
1915 from multiple comments on this bug.
1917 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1919 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1920 Missed a few lines, added it to make the runtest require no keyboard
1923 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1924 contains upper case chars. Make router use caseful_local_part.
1926 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1927 support when GnuTLS has been built with p11-kit.
1933 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1935 This, or similar/improved, will also be change PP/11 of 4.82.
1941 PP/01 Handle short writes when writing local log-files.
1942 In practice, only affects FreeBSD (8 onwards).
1943 Bugzilla 1053, with thanks to Dmitry Isaikin.
1945 NM/01 Bugzilla 949 - Documentation tweak
1947 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1950 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1952 PP/02 Implemented gsasl authenticator.
1954 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1956 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1957 `pkg-config foo` for cflags/libs.
1959 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1960 with rest of GSASL and with heimdal_gssapi.
1962 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1963 `pkg-config foo` for cflags/libs for the TLS implementation.
1965 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1966 properties get this fed in as external SSF. A number of robustness
1967 and debugging improvements to the cyrus_sasl authenticator.
1969 PP/08 cyrus_sasl server now expands the server_realm option.
1971 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1972 Patch by Jeremy Harris.
1974 PP/10 Added dbmjz lookup type.
1976 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1978 PP/12 MAIL args handles TAB as well as SP, for better interop with
1979 non-compliant senders.
1980 Analysis and variant patch by Todd Lyons.
1982 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1983 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1984 Patch from Dirk Mueller <dmueller@suse.com>
1986 PP/13 tls_peerdn now print-escaped for spool files.
1987 Observed some $tls_peerdn in wild which contained \n, which resulted
1988 in spool file corruption.
1990 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1991 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1992 or write after TLS renegotiation, which otherwise led to messages
1995 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1996 as a tracking header (ie: a signed header comes before the signature).
1997 Patch from Wolfgang Breyha.
1999 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2000 comma-sep list; embedded commas doubled.
2002 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2004 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2006 Report and patch from Dmitry Banschikov.
2008 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2009 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2010 failures. We appear to now support TLS1.1+ with Exim.
2012 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2013 lets Exim select keys and certificates based upon TLS SNI from client.
2014 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2015 before an outbound SMTP session. New log_selector, +tls_sni.
2017 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2018 NULL dereference. Report and patch from Alun Jones.
2020 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2021 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2022 Not seeing resolver debug output on NetBSD, but suspect this is a
2023 resolver implementation change.
2025 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2026 Left warnings. Added "eximon gdb" invocation mode.
2028 PP/21 Defaulting "accept_8bitmime" to true, not false.
2030 PP/22 Added -bw for inetd wait mode support.
2032 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2033 locate the relevant includes and libraries. Made this the default.
2035 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2036 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2038 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2039 This may cause build issues on older platforms.
2041 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2042 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2043 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2044 Added SNI support via GnuTLS too.
2045 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2047 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2049 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2050 Applied second patch from Janne, implementing suggestion to default
2051 multiple-strings-in-record handling to match SPF spec.
2053 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2055 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2056 read-only, out of scope).
2057 Patch from Wolfgang Breyha, report from Stuart Northfield.
2059 PP/29 Fix three issues highlighted by clang analyser static analysis.
2060 Only crash-plausible issue would require the Cambridge-specific
2061 iplookup router and a misconfiguration.
2062 Report from Marcin Mirosław.
2064 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2066 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2067 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2068 As part of this, removing so much warning spew let me fix some minor
2069 real issues in debug logging.
2071 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2072 assignment on my part. Fixed.
2074 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2075 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2076 Janne Snabb (who went above and beyond: thank you).
2078 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2079 string otherwise requires a connection and a bunch more work and it's
2080 relatively easy to get wrong. Should also expose TLS library linkage
2083 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2084 64-bit ${eval} (JH/03).
2086 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2087 GNU libc to support some of the 64-bit stuff, should not lead to
2088 conflicts. Defined before os.h is pulled in, so if a given platform
2089 needs to override this, it can.
2091 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2092 protection layer was required, which is not implemented.
2093 Bugzilla 1254, patch from Wolfgang Breyha.
2095 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2096 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2097 tls_dhparam take prime identifiers. Also unbreak combination of
2098 OpenSSL+DH_params+TLSSNI.
2100 PP/39 Disable SSLv2 by default in OpenSSL support.
2106 PP/01 Solaris build fix for Oracle's LDAP libraries.
2107 Bugzilla 1109, patch from Stephen Usher.
2109 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2111 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2114 TF/02 Fix a couple more cases where we did not log the error message
2115 when unlink() failed. See also change 4.74-TF/03.
2117 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2118 lock up or crash if it happened to be inside a call to libc when it
2119 got a SIGUSR1 from exiwhat.
2121 The SIGUSR1 handler appends the current process status to the process
2122 log which is later printed by exiwhat. It used to use the general
2123 purpose logging code to do this, but several functions it calls are
2124 not safe for signals.
2126 The new output code in the SIGUSR1 handler is specific to the process
2127 log, and simple enough that it's easy to inspect for signal safety.
2128 Removing some special cases also simplifies the general logging code.
2129 Removing the spurious timestamps from the process log simplifies
2132 TF/04 Improved ratelimit ACL condition.
2134 The /noupdate option has been deprecated in favour of /readonly which
2135 has clearer semantics. The /leaky, /strict, and /readonly update modes
2136 are mutually exclusive. The update mode is no longer included in the
2137 database key; it just determines when the database is updated. (This
2138 means that when you upgrade Exim will forget old rate measurements.)
2140 Exim now checks that the per_* options are used with an update mode that
2141 makes sense for the current ACL. For example, when Exim is processing a
2142 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2143 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2144 must specify per_mail/readonly. If you omit the update mode it defaults to
2145 /leaky where that makes sense (as before) or /readonly where required.
2147 The /noupdate option is now undocumented but still supported for
2148 backwards compatibility. It is equivalent to /readonly except that in
2149 ACLs where /readonly is required you may specify /leaky/noupdate or
2150 /strict/noupdate which are treated the same as /readonly.
2152 A useful new feature is the /count= option. This is a generalization
2153 of the per_byte option, so that you can measure the throughput of other
2154 aggregate values. For example, the per_byte option is now equivalent
2155 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2157 The per_rcpt option has been generalized using the /count= mechanism
2158 (though it's more complicated than the per_byte equivalence). When it is
2159 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2160 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2161 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2162 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2163 Note that using per_rcpt with a non-readonly update mode in more than
2164 one ACL will cause the recipients to be double-counted. (The per_mail
2165 and per_byte options don't have this problem.)
2167 The handling of very low rates has changed slightly. If the computed rate
2168 is less than the event's count (usually one) then this event is the first
2169 after a long gap. In this case the rate is set to the same as this event's
2170 count, so that the first message of a spam run is counted properly.
2172 The major new feature is a mechanism for counting the rate of unique
2173 events. The new per_addr option counts the number of different
2174 recipients that someone has sent messages to in the last time period. It
2175 behaves like per_rcpt if all the recipient addresses are different, but
2176 duplicate recipient addresses do not increase the measured rate. Like
2177 the /count= option this is a general mechanism, so the per_addr option
2178 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2179 example, measure the rate that a client uses different sender addresses
2180 with the options per_mail/unique=$sender_address. There are further
2181 details in the main documentation.
2183 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2185 TF/06 Removed a few PCRE remnants.
2187 TF/07 Automatically extract Exim's version number from tags in the git
2188 repository when doing development or release builds.
2190 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2191 Bugzilla 879. Patch from Paul Fisher.
2193 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2194 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2197 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2199 PP/05 Variable $av_failed, true if the AV scanner deferred.
2200 Bugzilla 1078. Patch from John Horne.
2202 PP/06 Stop make process more reliably on build failure.
2203 Bugzilla 1087. Patch from Heiko Schlittermann.
2205 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2206 Bugzilla 1089. Patch from Heiko Schlittermann.
2208 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2209 Bugzilla 1131. Patch from Holger Weiß.
2211 PP/09 Handle IPv6 addresses with SPF.
2212 Bugzilla 860. Patch from Wolfgang Breyha.
2214 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2216 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2219 PP/11 match_* no longer expand right-hand-side by default.
2220 New compile-time build option, EXPAND_LISTMATCH_RHS.
2221 New expansion conditions, "inlist", "inlisti".
2223 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2225 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2227 PP/14 fix log_write() format string regression from TF/03.
2228 Bugzilla 1152. Patch from Dmitry Isaikin.
2234 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2236 PP/02 Harmonised TLS library version reporting; only show if debugging.
2237 Layout now matches that introduced for other libraries in 4.74 PP/03.
2239 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2241 PP/04 New "dns_use_edns0" global option.
2243 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2246 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2247 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2249 TK/01 Updated PolarSSL code to 0.14.2.
2250 Bugzilla 1097. Patch from Andreas Metzler.
2252 PP/07 Catch divide-by-zero in ${eval:...}.
2253 Fixes bugzilla 1102.
2255 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2258 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2259 format-string attack -- SECURITY: remote arbitrary code execution.
2261 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2262 time unintentionally subject to list matching rules, letting the header
2263 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2264 arbitrary string expansion). This allowed for information disclosure.
2266 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2267 INT_MIN/-1 -- value coerced to INT_MAX.
2273 NM/01 Workaround for PCRE version dependency in version reporting
2276 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2277 This fixes portability to compilers other than gcc, notably
2278 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2280 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2281 makefiles for portability to HP-UX and POSIX correctness.
2283 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2284 Also via indented variable definition in the Makefile.
2285 (Debugging by Oliver Heesakkers).
2287 PP/02 Restore caching of spamd results with expanded spamd_address.
2288 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2290 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2291 Improves build reliability. Fix from: Frank Elsner
2293 NM/02 Fix wide character breakage in the rfc2047 coding
2294 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2296 NM/03 Allow underscore in dnslist lookups
2297 Fixes bug 1026. Patch from Graeme Fowler
2299 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2300 Code patches from Adam Ciarcinski of NetBSD.
2302 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2305 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2306 is logged, to avoid truncation. Patch from John Horne.
2308 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2309 Patch from Jakob Hirsch.
2311 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2312 SQL string expansion failure details.
2313 Patch from Andrey Oktyabrski.
2315 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2316 Patch from Simon Arlott.
2318 PP/09 New lookups functionality failed to compile on old gcc which rejects
2319 extern declarations in function scope.
2320 Patch from Oliver Fleischmann
2322 PP/10 Use sig_atomic_t for flags set from signal handlers.
2323 Check getgroups() return and improve debugging.
2324 Fixed developed for diagnosis in bug 927 (which turned out to be
2327 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2328 Patch from Mark Zealey.
2330 PP/12 Bugzilla 1056: Improved spamd server selection.
2331 Patch from Mark Zealey.
2333 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2334 Based on patch from Heiko Schlittermann.
2336 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2337 Patch from Uwe Doering, sign-off by Michael Haardt.
2339 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2340 variable declaration deep within a block. Bug and patch from
2343 PP/15 lookups-Makefile IRIX compatibility coercion.
2345 PP/16 Make DISABLE_DKIM build knob functional.
2347 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2348 Patch by Simon Arlott
2350 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2351 variable argument macros. Our copy now differs from upstream.
2357 TF/01 Failure to get a lock on a hints database can have serious
2358 consequences so log it to the panic log.
2360 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2361 controlled using the smtp_confirmation log selector.
2363 TF/03 Include the error message when we fail to unlink a spool file.
2365 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2366 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2367 for maintaining out-of-tree patches for some time.
2369 PP/01 Bugzilla 139: Documentation and portability issues.
2370 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2371 Handle per-OS dynamic-module compilation flags.
2373 PP/02 Let /dev/null have normal permissions.
2374 The 4.73 fixes were a little too stringent and complained about the
2375 permissions on /dev/null. Exempt it from some checks.
2376 Reported by Andreas M. Kirchwitz.
2378 PP/03 Report version information for many libraries, including
2379 Exim version information for dynamically loaded libraries. Created
2380 version.h, now support a version extension string for distributors
2381 who patch heavily. Dynamic module ABI change.
2383 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2384 privilege escalation vulnerability whereby the Exim run-time user
2385 can cause root to append content of the attacker's choosing to
2388 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2391 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2392 If dropping privileges for untrusted macros, we disabled normal logging
2393 on the basis that it would fail; for the Exim run-time user, this is not
2394 the case, and it resulted in successful deliveries going unlogged.
2395 Fixed. Reported by Andreas Metzler.
2401 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2402 only prepend for the Resent-* case. Fixes regression introduced in
2403 Exim 4.70 by NM/22 for Bugzilla 607.
2405 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2406 increasing numbers of administrators be bitten by this.
2408 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2410 PP/03 Bugzilla 994: added openssl_options main configuration option.
2412 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2414 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2416 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2418 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2419 without a peer certificate, leading to a segfault because of an
2420 assumption that peers always have certificates. Be a little more
2421 paranoid. Problem reported by Martin Tscholak.
2423 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2424 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2425 NB: ClamAV planning to remove STREAM in "middle of 2010".
2426 CL also introduces -bmalware, various -d+acl logging additions and
2427 more caution in buffer sizes.
2429 PP/09 Implemented reverse_ip expansion operator.
2431 PP/10 Bugzilla 937: provide a "debug" ACL control.
2433 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2435 PP/12 Bugzilla 973: Implement --version.
2437 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2439 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2441 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2443 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2444 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2445 ignore trailing whitespace.
2447 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2449 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2452 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2453 Notification from Dr Andrew Aitchison.
2455 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2456 ExtendedDetectionInfo response format.
2457 Notification from John Horne.
2459 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2462 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2463 XSL and documented dependency on system catalogs, with examples of how
2466 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2469 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2470 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2471 configuration file which is writeable by the Exim user or group.
2473 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2474 of configuration files to cover files specified with the -C option if
2475 they are going to be used with root privileges, not just the default
2478 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2479 option (effectively making it always true).
2481 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2482 files to be used while preserving root privileges.
2484 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2485 that rogue child processes cannot use them.
2487 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2488 run-time user, instead of root.
2490 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2491 Exim run-time user without dropping privileges.
2493 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2494 result string, instead of calling string_vformat() twice with the same
2497 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2498 for other users. Others should always drop root privileges if they use
2499 -C on the command line, even for a whitelisted configure file.
2501 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2503 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2509 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2510 $data_path, and $header_path variables; fixed documentation bugs and
2513 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2514 exipick to access non-standard spools, including the "frozen" queue
2517 NM/01 Bugzilla 965: Support mysql stored procedures.
2518 Patch from Alain Williams
2520 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2522 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2523 Patch from Andreas Metzler
2525 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2526 Patch from Kirill Miazine
2528 NM/05 Bugzilla 671: Added umask to procmail example.
2530 JJ/03 installed exipick 20100323.0, fixing doc bug
2532 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2533 directory. Notification and patch from Dan Rosenberg.
2535 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2537 TK/02 Improve log output when DKIM signing operation fails.
2539 MH/01 Treat the transport option dkim_domain as a colon separated
2540 list, not as a single string, and sign the message with each element,
2541 omitting multiple occurences of the same signer.
2543 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2544 Bugzilla 985, 986. Patch by Simon Arlott
2546 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2547 Patch by Simon Arlott
2549 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2550 MBX locking. Notification from Dan Rosenberg.
2556 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2558 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2560 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2562 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2564 NM/04 Bugzilla 915: Flag broken perl installation during build.
2570 TK/01 Added patch by Johannes Berg that expands the main option
2571 "spamd_address" if it starts with a dollar sign.
2573 TK/02 Write list of recipients to X-Envelope-Sender header when building
2574 the mbox-format spool file for content scanning (suggested by Jakob
2577 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2578 (http://www.dcc-servers.net/) support via dccifd. Activated by
2579 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2581 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2582 by Mark Daniel Reidel <mr@df.eu>.
2584 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2585 When building exim an external PCRE library is now needed -
2586 PCRE is a system library on the majority of modern systems.
2587 See entry on PCRE_LIBS in EDITME file.
2589 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2590 conversation. Added nologin parameter to request.
2591 Patch contributed by Kirill Miazine.
2593 TF/01 Do not log submission mode rewrites if they do not change the address.
2595 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2597 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2598 log files in place. Contributed by Roberto Lima.
2600 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2602 TF/03 Bugzilla 615: When checking the local_parts router precondition
2603 after a local_part_suffix or local_part_prefix option, Exim now
2604 does not use the address's named list lookup cache, since this
2605 contains cached lookups for the whole local part.
2607 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2608 Robert Millan. Documentation is in experimental-spec.txt.
2610 TF/04 Bugzilla 668: Fix parallel build (make -j).
2612 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2614 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2615 Patch provided by Jan Srzednicki.
2617 TF/05 Leading white space used to be stripped from $spam_report which
2618 wrecked the formatting. Now it is preserved.
2620 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2621 that they are available at delivery time.
2623 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2625 TF/08 TLS error reporting now respects the incoming_interface and
2626 incoming_port log selectors.
2628 TF/09 Produce a more useful error message if an SMTP transport's hosts
2629 setting expands to an empty string.
2631 NM/06 Bugzilla 744: EXPN did not work under TLS.
2632 Patch provided by Phil Pennock.
2634 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2635 Patch provided by Richard Godbee.
2637 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2638 acl_smtp_notquit, added index entry.
2640 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2641 Patch provided by Eugene Bujak.
2643 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2644 accept(). Patch provided by Maxim Dounin.
2646 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2647 Patch provided by Phil Pennock.
2649 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2651 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2652 Patch provided by Brad "anomie" Jorsch.
2654 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2655 Patch provided by Dean Brooks.
2657 TK/05 Add native DKIM support (does not depend on external libraries).
2659 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2660 Patch provided by Graeme Fowler.
2662 NM/16 Bugzilla 851: Documentation example syntax fix.
2664 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2666 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2669 NM/19 Bugzilla 745: TLS version reporting.
2670 Patch provided by Phil Pennock.
2672 NM/20 Bugzilla 167: bool: condition support.
2673 Patch provided by Phil Pennock.
2675 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2676 clients. Patch provided by Phil Pennock.
2678 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2679 Patch provided by Brad "anomie" Jorsch.
2681 NM/23 Bugzilla 687: Fix misparses in eximstats.
2682 Patch provided by Heiko Schlittermann.
2684 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2685 Patch provided by Heiko Schlittermann.
2687 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2688 plus update to original patch.
2690 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2692 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2693 Patch provided by David Brownlee.
2695 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2697 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2699 NM/30 Bugzilla 888: TLS documentation bugfixes.
2701 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2703 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2704 Unlike the original bugzilla I have changed all shell scripts in src tree.
2706 NM/33 Bugzilla 898: Transport filter timeout fix.
2707 Patch by Todd Rinaldo.
2709 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2710 Patch by Serge Demonchaux.
2712 NM/35 Bugzilla 39: Base64 decode bug fixes.
2713 Patch by Jakob Hirsch.
2715 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2717 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2719 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2721 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2727 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2728 ALT-N's libdkim that I have put here:
2729 http://duncanthrax.net/exim-experimental/
2731 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2732 were called 'true' and it seems that C99 defines that as a reserved
2733 keyword to be used with 'bool' variable types. That means you could
2734 not include C99-style headers which use bools without triggering
2735 build errors in sieve.c.
2737 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2738 as mailq or other aliases. Changed the --help handling significantly
2739 to do whats expected. exim_usage() emits usage/help information.
2741 SC/01 Added the -bylocaldomain option to eximstats.
2743 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2745 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2747 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2753 PH/01 Another patch from the Sieve maintainer.
2755 PH/02 When an IPv6 address is converted to a string for single-key lookup
2756 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2757 dots are used instead of colons so that keys in lsearch files need not
2758 contain colons. This was done some time before quoting was made available
2759 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2760 (notated using the quote facility) so as to distinguish them from IPv4
2761 keys. This meant that lookups for IP addresses in host lists did not work
2762 for iplsearch lookups.
2764 This has been fixed by arranging for IPv6 addresses to be expressed with
2765 colons if the lookup type is iplsearch. This is not incompatible, because
2766 previously such lookups could never work.
2768 The situation is now rather anomalous, since one *can* have colons in
2769 ordinary lsearch keys. However, making the change in all cases is
2770 incompatible and would probably break a number of configurations.
2772 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2775 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2776 conversion specification without a maximum field width, thereby enabling
2777 a rogue spamd server to cause a buffer overflow. While nobody in their
2778 right mind would setup Exim to query an untrusted spamd server, an
2779 attacker that gains access to a server running spamd could potentially
2780 exploit this vulnerability to run arbitrary code as the Exim user.
2782 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2783 $primary_hostname instead of what libspf2 thinks the hosts name is.
2785 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2786 a directory entry by the name of the lookup key. Previously, if a
2787 symlink pointed to a non-existing file or a file in a directory that
2788 Exim lacked permissions to read, a lookup for a key matching that
2789 symlink would fail. Now it is enough that a matching directory entry
2790 exists, symlink or not. (Bugzilla 503.)
2792 PH/03 The body_linecount and body_zerocount variables are now exported in the
2795 PH/04 Added the $dnslist_matched variable.
2797 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2798 This means they are set thereafter only if the connection becomes
2801 PH/06 Added the client_condition to authenticators so that some can be skipped
2802 by clients under certain conditions.
2804 PH/07 The error message for a badly-placed control=no_multiline_responses left
2805 "_responses" off the end of the name.
2807 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2809 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2810 (without spaces) instead of just copying the configuration text.
2812 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2814 PH/11 Added $max_received_linelength.
2816 PH/12 Added +ignore_defer and +include_defer to host lists.
2818 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2819 way in which PCRE > 7.0 is built.
2821 PH/14 Implemented queue_only_load_latch.
2823 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2824 MAIL command. The effect was to mangle the value on 64-bit systems.
2826 PH/16 Another patch from the Sieve maintainer.
2828 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2830 PH/18 If a system quota error occurred while trying to create the file for
2831 a maildir delivery, the message "Mailbox is full" was not appended to the
2832 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2833 only to a quota excession during the actual writing of the file.
2835 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2836 characters?) which causes problems in log lines. The DN values are now
2837 passed through string_printing() before being added to log lines.
2839 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2840 and InterBase are left for another time.)
2842 PH/21 Added message_body_newlines option.
2844 PH/22 Guard against possible overflow in moan_check_errorcopy().
2846 PH/23 POSIX allows open() to be a macro; guard against that.
2848 PH/24 If the recipient of an error message contained an @ in the local part
2849 (suitably quoted, of course), incorrect values were put in $domain and
2850 $local_part during the evaluation of errors_copy.
2856 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2857 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2860 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2861 issue a MAIL command.
2863 PH/02 In an ACL statement such as
2865 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2867 if a client was not listed at all, or was listed with a value other than
2868 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2869 the condition was not true (as it should be), so access was not denied.
2870 The bug was that the ! inversion was incorrectly passed on to the second
2871 item. This has been fixed.
2873 PH/03 Added additional dnslists conditions == and =& which are different from
2874 = and & when the dns lookup returns more than one IP address.
2876 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2877 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2879 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2880 FSYNC, which compiles an option called disable_fsync that allows for
2881 bypassing fsync(). The documentation is heavily laced with warnings.
2883 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2885 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2886 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2887 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2888 including adding "make clean"; (3) Added -fPIC when compiling the test
2889 dynamically loaded module, to get rid of a warning.
2891 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2892 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2893 = 0s. The bug is otherwise harmless.
2895 PH/07 There was a bug in the dovecot authenticator such that the value of
2896 $auth1 could be overwritten, and so not correctly preserved, after a
2897 successful authentication. This usually meant that the value preserved by
2898 the server_setid option was incorrect.
2900 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2902 PH/09 Installed PCRE release 7.0.
2904 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2905 run for batched SMTP input. It is now run at the start of every message
2906 in the batch. While fixing this I discovered that the process information
2907 (output by running exiwhat) was not always getting set for -bs and -bS
2908 input. This is fixed, and it now also says "batched" for BSMTP.
2910 PH/11 Added control=no_pipelining.
2912 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2913 patch, slightly modified), and move the expansion of helo_data till after
2914 the connection is made in the smtp transport (so it can use these
2917 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2919 PH/14 Added log_selector = +pid.
2921 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2923 PH/16 Add ${if forany and ${if forall.
2925 PH/17 Added dsn_from option to vary the From: line in DSNs.
2927 PH/18 Flush SMTP output before performing a callout, unless control =
2928 no_callout_flush is set.
2930 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2931 was true (the default) a successful delivery failed to delete the retry
2932 item, thus causing premature timeout of the address. The bug is now
2935 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2937 PH/21 Long custom messages for fakedefer and fakereject are now split up
2938 into multiline responses in the same way that messages for "deny" and
2939 other ACL rejections are.
2941 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2942 with slight modification.
2944 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2945 draft, changing the syntax and factoring some duplicate code".
2947 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2948 for deliveries of the second and subsequent messages over the same SMTP
2951 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2952 ${reduce, with only minor "tidies".
2954 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2956 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2957 expansion side effects.
2959 PH/27 When a message times out after an over-quota error from an Exim-imposed
2960 quota, the bounce message says "mailbox is full". This message was not
2961 being given when it was a system quota that was exceeded. It now should
2964 MH/03 Made $recipients available in local_scan(). local_scan() already has
2965 better access to the recipient list through recipients_list[], but
2966 $recipients can be useful in postmaster-provided expansion strings.
2968 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2969 in the case of a MAIL command with additional options following the
2970 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2971 were accidentally chopped off.
2973 PH/29 SMTP synchronization checks are implemented when a command is read -
2974 there is a check that no more input is waiting when there shouldn't be
2975 any. However, for some commands, a delay in an ACL can mean that it is
2976 some time before the response is written. In this time, more input might
2977 arrive, invalidly. So now there are extra checks after an ACL has run for
2978 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2979 pipelining has not been advertised.
2981 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2983 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2984 returned in the SMTP response when smtp_return_error_details was false.
2985 This has been fixed.
2987 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2988 instead of the C I/O that was originally supplied, because problems were
2989 reported on Solaris.
2991 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2992 Exim which did not show up earlier: it was assuming that a call to
2993 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2994 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2995 was a macro that became an assignment, so it seemed to work. This has
2996 changed to a proper function call with a void return, hence the compile
2997 error. Exim's code has been fixed.
2999 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3002 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3004 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3006 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3007 because it turns out to be too big or there is a timeout) there is no
3008 "Completed" line in the log. When some messages of this type were
3009 selected by exigrep, they were listed as "not completed". Others were
3010 picked up by some special patterns. I have improved the selection
3011 criteria to be more general.
3013 PH/38 The host_find_failed option in the manualroute router can now be set
3014 to "ignore", to completely ignore a host whose IP address cannot be
3015 found. If all hosts are ignored, the behaviour is controlled by the new
3016 host_all_ignored option.
3018 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3019 homing or because of multiple MX records with /mx) generated more than
3020 one IP address, and the following item turned out to be the local host,
3021 all the secondary addresses of the first item were incorrectly removed
3022 from the list, along with the local host and any following hosts (which
3023 is what is supposed to happen).
3025 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3026 whoever called Exim into the -H file. In the case of the daemon it was
3027 behaving confusingly. When first started, it used values for whoever
3028 started the daemon, but after a SIGHUP it used the Exim user (because it
3029 calls itself on a restart). I have changed the code so that it now always
3032 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3033 message are rejected with the same error (e.g. no authentication or bad
3034 sender address), and a DATA command is nevertheless sent (as can happen
3035 with PIPELINING or a stupid MUA), the error message that was given to the
3036 RCPT commands is included in the rejection of the DATA command. This is
3037 intended to be helpful for MUAs that show only the final error to their
3040 PH/42 Another patch from the Sieve maintainer.
3042 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3043 Eximstats - Fixed some broken HTML links and added missing column headers
3045 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3046 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3048 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3050 PH/43 Yet another patch from the Sieve maintainer.
3052 PH/44 I found a way to check for a TCP/IP connection going away before sending
3053 the response to the final '.' that terminates a message, but only in the
3054 case where the client has not sent further data following the '.'
3055 (unfortunately, this is allowed). However, in many cases there won't be
3056 any further data because there won't be any more messages to send. A call
3057 to select() can be used: if it shows that the input is "ready", there is
3058 either input waiting, or the socket has been closed. An attempt to read
3059 the next input character can distinguish the two cases. Previously, Exim
3060 would have sent an OK response which the client would never have see.
3061 This could lead to message repetition. This fix should cure that, at
3062 least in a lot of common cases.
3064 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3065 advertised in response to EHLO.
3071 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3072 fixed by 4.65/MH/01 (is this a record?) are fixed:
3074 (i) An empty string was always treated as zero by the numeric comparison
3075 operators. This behaviour has been restored.
3077 (ii) It is documented that the numeric comparison operators always treat
3078 their arguments as decimal numbers. This was broken in that numbers
3079 starting with 0 were being interpreted as octal.
3081 While fixing these problems I realized that there was another issue that
3082 hadn't been noticed. Values of message_size_limit (both the global option
3083 and the transport option) were treated as octal if they started with 0.
3084 The documentation was vague. These values are now always treated as
3085 decimal, and I will make that clear in the documentation.
3091 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3092 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3095 MH/01 Don't check that the operands of numeric comparison operators are
3096 integers when their expansion is in "skipping" mode (fixes bug
3097 introduced by 4.64-PH/07).
3099 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3100 child addresses, Exim now panics and dies. Previously, because the count
3101 is held in a short int, deliveries were likely to be lost. As such a
3102 large number of recipients for a single message is ridiculous
3103 (performance will be very, very poor), I have chosen to impose a limit
3104 rather than extend the field.
3110 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3111 leftover -K file (the existence of which was triggered by #402).
3112 While we were at it, introduced process PID as part of the -K
3113 filename. This should rule out race conditions when creating
3116 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3117 processing considerably. Previous code took too long for large mails,
3118 triggering a timeout which in turn triggers #401.
3120 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3121 in the DK code in transports.c. sendfile() is not really portable,
3122 hence the _LINUX specificness.
3124 TF/01 In the add_headers option to the mail command in an Exim filter,
3125 there was a bug that Exim would claim a syntax error in any
3126 header after the first one which had an odd number of characters
3129 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3130 callout verification, Exim cached a "reject" for the entire domain. This
3131 is correct for most verifications, but it is not correct for a recipient
3132 verification with use_sender or use_postmaster set, because in that case
3133 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3134 case of MAIL FROM:<> rejection from other early rejections (e.g.
3135 rejection of HELO). When verifying a recipient using a non-null MAIL
3136 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3137 Whatever the result of the callout, the value of the domain cache is
3138 left unchanged (for any other kind of callout, getting as far as trying
3139 RCPT means that the domain itself is ok).
3141 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3144 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3145 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3146 session, the daemon ignores ECONNECT errors and logs others; it now
3147 ignores EPIPE as well.
3149 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3150 (quoted-printable decoding).
3152 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3153 later the small subsequent patch to fix an introduced bug.
3155 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3157 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3159 PH/08 An error is now given if message_size_limit is specified negative.
3161 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3162 to be given (somewhat) arbitrary names.
3164 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3167 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3168 miscellaneous code fixes
3170 PH/10 Added the log_reject_target ACL modifier to specify where to log
3173 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3174 hostname. This is wrong, because it relates to the incoming message (and
3175 probably the interface on which it is arriving) and not to the outgoing
3176 callout (which could be using a different interface). This has been
3177 changed to use the value of the helo_data option from the smtp transport
3178 instead - this is what is used when a message is actually being sent. If
3179 there is no remote transport (possible with a router that sets up host
3180 addresses), $smtp_active_hostname is used.
3182 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3183 tweaks were necessary in order to get it to work (see also 21 below):
3184 (a) The code assumed that strncpy() returns a negative number on buffer
3185 overflow, which isn't the case. Replaced with Exim's string_format()
3187 (b) There were several signed/unsigned issues. I just did the minimum
3188 hacking in of casts. There is scope for a larger refactoring.
3189 (c) The code used strcasecmp() which is not a standard C function.
3190 Replaced with Exim's strcmpic() function.
3191 (d) The code set only $1; it now sets $auth1 as well.
3192 (e) A simple test gave the error "authentication client didn't specify
3193 service in request". It would seem that Dovecot has changed its
3194 interface. Fortunately there's a specification; I followed it and
3195 changed what the client sends and it appears to be working now.
3197 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3200 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3201 address A is aliased to B and C, where B exists and C does not. Without
3202 -v the output is "A verified" because verification stops after a
3203 successful redirection if more than one address is generated. However,
3204 with -v the child addresses are also verified. Exim was outputting "A
3205 failed to verify" and then showing the successful verification for C,
3206 with its parentage. It now outputs "B failed to verify", showing B's
3207 parentage before showing the successful verification of C.
3209 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3210 look up a TXT record in a specific list after matching in a combined
3213 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3214 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3215 they consult the DNS. I had assumed they would set it the way they
3216 wanted; and indeed my experiments on Linux seem to show that in some
3217 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3218 To be on the safe side, however, I have now made the interface to
3219 host_find_byname() similar to host_find_bydns(), with an argument
3220 containing the DNS resolver options. The host_find_byname() function now
3221 sets these options at its start, just as host_find_bydns() does. The smtp
3222 transport options dns_qualify_single and dns_search_parents are passed to
3223 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3224 of host_find_byname() use the default settings of RES_DEFNAMES
3225 (qualify_single) but not RES_DNSRCH (search_parents).
3227 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3228 spool_read_header() do less string testing, by means of a preliminary
3229 switch on the second character of optional "-foo" lines. (This is
3230 overdue, caused by the large number of possibilities that now exist.
3231 Originally there were few.) While I was there, I also converted the
3232 str(n)cmp tests so they don't re-test the leading "-" and the first
3233 character, in the hope this might squeeze out yet more improvement.
3235 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3236 flag allowing group syntax was set by the header_syntax check but not
3237 turned off, possible causing trouble later; (2) The flag was not being
3238 set at all for the header_verify test, causing "group"-style headers to
3239 be rejected. I have now set it in this case, and also caused header_
3240 verify to ignore an empty address taken from a group. While doing this, I
3241 came across some other cases where the code for allowing group syntax
3242 while scanning a header line wasn't quite right (mostly, not resetting
3243 the flag correctly in the right place). These bugs could have caused
3244 trouble for malformed header lines. I hope it is now all correct.
3246 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3247 with the "reply" argument non-NULL. The code, however (which originally
3248 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3249 but it didn't always do it. This confused somebody who was copying the
3250 code for some other use. I have removed all the tests.
3252 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3253 feature that was used to support insecure browsers during the U.S. crypto
3254 embargo. It requires special client support, and Exim is probably the
3255 only MTA that supported it -- and would never use it because real RSA is
3256 always available. This code has been removed, because it had the bad
3257 effect of slowing Exim down by computing (never used) parameters for the
3258 RSA_EXPORT functionality.
3260 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3261 authenticator to fail if there's a tab character in the incoming data
3262 (there should never be unless someone is messing about, as it's supposed
3263 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3264 if the connection is using TLS or if the remote IP is the same as the
3265 local IP, and the "valid-client-cert option" if a client certificate has
3268 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3269 authenticators. This can be used for authorization after authentication
3270 succeeds. (In the case of plaintext, it servers for both authentication
3273 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3274 if any retry times were supplied.
3276 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3277 connection, where there is no client IP address to check. In this
3278 situation, the verify now always succeeds.
3280 PH/25 Applied John Jetmore's -Mset patch.
3282 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3284 PH/27 In a string expansion for a processed (not raw) header when multiple
3285 headers of the same name were present, leading whitespace was being
3286 removed from all of them, but trailing whitespace was being removed only
3287 from the last one. Now trailing whitespace is removed from each header
3288 before concatenation. Completely empty headers in a concatenation (as
3289 before) are ignored.
3291 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3292 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3294 PH/29 [Removed. This was a change that I later backed out, and forgot to
3295 correct the ChangeLog entry (that I had efficiently created) before
3296 committing the later change.]
3298 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3299 address errors (4xx response to RCPT) over the same connection as other
3300 messages routed to the same hosts. Such deliveries are always "forced",
3301 so retry times are not inspected. This resulted in far too many retries
3302 for the affected addresses. The effect occurred only when there were more
3303 hosts than the hosts_max_try setting in the smtp transport when it had
3304 the 4xx errors. Those hosts that it had tried were not added to the list
3305 of hosts for which the message was waiting, so if all were tried, there
3306 was no problem. Two fixes have been applied:
3308 (i) If there are any address or message errors in an SMTP delivery, none
3309 of the hosts (tried or untried) are now added to the list of hosts
3310 for which the message is waiting, so the message should not be a
3311 candidate for sending over the same connection that was used for a
3312 successful delivery of some other message. This seems entirely
3313 reasonable: after all the message is NOT "waiting for some host".
3314 This is so "obvious" that I'm not sure why it wasn't done
3315 previously. Hope I haven't missed anything, but it can't do any
3316 harm, as the worst effect is to miss an optimization.
3318 (ii) If, despite (i), such a delivery is accidentally attempted, the
3319 routing retry time is respected, so at least it doesn't keep
3320 hammering the server.
3322 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3323 in ${readsocket because some servers need this prod.
3325 PH/32 Added some extra debug output when updating a wait-xxx database.
3327 PH/33 The hint "could be header name not terminated by colon", which has been
3328 given for certain expansion errors for a long time, was not being given
3329 for the ${if def:h_colon_omitted{... case.
3331 PH/34 The spec says: "With one important exception, whenever a domain list is
3332 being scanned, $domain contains the subject domain." There was at least
3333 one case where this was not true.
3335 PH/35 The error "getsockname() failed: connection reset by peer" was being
3336 written to the panic log as well as the main log, but it isn't really
3337 panic-worthy as it just means the connection died rather early on. I have
3338 removed the panic log writing for the ECONNRESET error when getsockname()
3341 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3342 runs only) independently of the message's sender address. This meant
3343 that, if the 4xx error was in fact related to the sender, a different
3344 message to the same recipient with a different sender could confuse
3345 things. In particular, this can happen when sending to a greylisting
3346 server, but other circumstances could also provoke similar problems.
3347 I have changed the default so that the retry time for these errors is now
3348 based a combination of the sender and recipient addresses. This change
3349 can be overridden by setting address_retry_include_sender=false in the
3352 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3353 remote server are returned as part of bounce messages. This was not
3354 happening for LMTP over a pipe (the lmtp transport), but now it is the
3355 same for both kinds of LMTP.
3357 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3358 in header lines that were in fact CNAMEs. This is no longer the case.
3360 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3361 and queue runs started by the daemon processed all messages. This has
3362 been fixed so that -R and -S can now usefully be given with -q<time>.
3364 PH/40 Import PCRE release 6.7 (fixes some bugs).
3366 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3368 PH/42 Give an error if -q is specified more than once.
3370 PH/43 Renamed the variables $interface_address and $interface_port as
3371 $received_ip_address and $received_port, to make it clear that these
3372 values apply to message reception, and not to the outgoing interface when
3373 a message is delivered. (The old names remain recognized, of course.)
3375 PH/44 There was no timeout on the connect() call when using a Unix domain
3376 socket in the ${readsocket expansion. There now is.
3378 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3379 be meaningful with "accept".
3381 SC/01 Eximstats V1.43
3382 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3384 SC/02 Eximstats V1.44
3385 Use a glob alias rather than an array ref in the generated
3386 parser. This improves both readability and performance.
3388 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3389 Collect SpamAssassin and rejection statistics.
3390 Don't display local sender or destination tables unless
3391 there is data to show.
3392 Added average volumes into the top table text output.
3394 SC/04 Eximstats V1.46
3395 Collect data on the number of addresses (recipients)
3396 as well as the number of messages.
3398 SC/05 Eximstats V1.47
3399 Added 'Message too big' to the list of mail rejection
3400 reasons (thanks to Marco Gaiarin).
3402 SC/06 Eximstats V1.48
3403 Mainlog lines which have GMT offsets and are too short to
3404 have a flag are now skipped.
3406 SC/07 Eximstats V1.49 (Alain Williams)
3407 Added the -emptyok flag.
3409 SC/08 Eximstats V1.50
3410 Fixes for obtaining the IP address from reject messages.
3412 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3413 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3414 whitespace changes from 4.64-PH/27
3416 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3419 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3420 are found, allow negative numbers in numeric criteria)
3422 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3424 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3425 to match changes made in 4.64-PH/43
3427 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3429 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3430 "domain" is actually an IP address, and force a failure. This locks out
3431 those revolvers/nameservers that support "A-for-A" lookups, in
3432 contravention of the specifications.
3434 PH/48 When a host name was looked up from an IP address, and the subsequent
3435 forward lookup of the name timed out, the host name was left in
3436 $sender_host_name, contrary to the specification.
3438 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3439 restricted to single-key lookups, Exim was not diagnosing an error if
3440 * or *@ was used with a query-style lookup.
3442 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3444 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3445 long ago, but noone interested enough thought of it. Let's just say that
3446 the "1.1" means that there are some new functions that weren't there at
3447 some point in the past.
3449 PH/51 Error processing for expansion failure of helo_data from an smtp
3450 transport during callout processing was broken.
3452 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3453 tested/used via the -bh/-bhc/-bs options.
3455 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3456 bug, fixed in subsequent PCRE releases).
3458 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3459 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3461 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3467 SC/01 Use a glob alias rather than an array ref in eximstats generated
3468 parser. This improves both readability and performance.
3470 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3471 Don't display local sender or destination tables in eximstats unless
3472 there is data to show.
3473 Added average volumes into the eximstats top table text output.
3475 SC/03 Collect data on the number of addresses (recipients) as well
3476 as the number of messages in eximstats.
3478 TF/01 Correct an error in the documentation for the redirect router. Exim
3479 does (usually) call initgroups() when daemonizing.
3481 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3482 with consistent privilege compared to when running as a daemon.
3484 TF/03 Note in the spec that $authenticated_id is not set for local
3485 submissions from trusted users.
3487 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3488 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3490 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3491 by adding some example configuration directives to the default
3492 configuration file. A little bit of work is required to uncomment the
3493 directives and define how usernames and passwords are checked, but
3494 there is now a framework to start from.
3496 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3497 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3498 without this. I don't know how relevant this is to other LDAP libraries.
3500 PH/02 Add the verb name to the "unknown ACL verb" error.
3502 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3504 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3506 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3507 directory not expanded when it should be if an expanded home directory
3508 was set for the address (which is overridden by the transport).
3510 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3513 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3514 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3515 because it is too late at that time, and has no effect.
3517 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3518 security issue with \' (bugzilla #107). I could not use the
3519 PQescapeStringConn() function, because it needs a PGconn value as one of
3522 PH/08 When testing addresses using -bt, indicate those final addresses that
3523 are duplicates that would not cause an additional delivery. At least one
3524 person was confused, thinking that -bt output corresponded to deliveries.
3525 (Suppressing duplicates isn't a good idea as you lose the information
3526 about possibly different redirections that led to the duplicates.)
3528 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3529 systems where poll() doesn't work, in particular OS X.
3531 PH/10 Added more information to debugging output for retry time not reached.
3533 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3534 operations in malware.c.
3536 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3539 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3540 syslog_duplication set false, log lines that would normally be written
3541 both the the main log and to the reject log were not written to syslog at
3544 PH/14 In the default configuration, change the use of "message" in ACL warn
3545 statements to "add_header".
3547 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3548 not followed by a command (e.g. "seen endif").
3550 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3551 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3554 PH/17 Added extra conditions to the default value of delay_warning_condition
3558 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3559 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3560 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3563 The Auto-Submitted: and various List- headers are standardised, whereas I
3564 don't think Precedence: ever was.
3566 PH/18 Refactored debugging code in route_finduser() to show more information,
3567 in particular, the error code if getpwnam() issues one.
3569 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3570 This is apparently needed in addition to the PH/07 change above to avoid
3571 any possible encoding problems.
3573 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3574 but not after initializing Perl.
3576 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3577 output them only if debugging. By default they are written stderr,
3578 apparently, which is not desirable.
3580 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3583 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3586 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3588 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3589 authenticated or an ident call has been made. Suppress the default
3590 values for $authenticated_id and $authenticated_sender (but permit -oMai
3591 and -oMas) when testing with -bh.
3593 PH/24 Re-jigged the order of the tests in the default configuration so that the
3594 tests for valid domains and recipients precede the DNS black list and CSA
3595 tests, on the grounds that those ones are more expensive.
3597 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3598 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3599 This bug exists in every version of Exim that I still have, right back to
3602 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3603 However, an attempt to turn on case-sensitivity in a regex key by
3604 including (?-i) didn't work because the subject string was already
3605 lowercased, and the effects were non-intuitive. It turns out that a
3606 one-line patch can be used to allow (?-i) to work as expected.
3612 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3613 other effects) broke the use of negated acl sub-conditions.
3615 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3618 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3619 "Deny" causes Exim to reject the incoming connection with a 554 error.
3620 Unfortunately, if there is a major crisis, such as a disk failure,
3621 tcp-wrappers gives "deny", whereas what one would like would be some
3622 kind of temporary error. A kludge has been added to help with this.
3623 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3624 554 error is used if errno is still zero or contains ENOENT (which occurs
3625 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3628 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3630 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3631 errors. Otherwise a message that provokes a temporary error (when other
3632 messages do not) can cause a whole host to time out.
3634 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3635 addresses were routed directly to files or pipes from a redirect router.
3636 File deliveries just didn't batch; pipe deliveries might have suffered
3639 PH/06 A failure to get a lock for a hints database would erroneously always say
3640 "Failed to get write lock", even when it was really a read lock.
3642 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3643 of 0600. This has been changed to use the value of the lockfile_mode
3644 option (which defaults to 0600).
3646 PH/08 Applied small patch from the Sieve maintainer.
3648 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3649 folder from quota calculations, a direct delivery into this folder messed
3650 up the contents of the maildirsize file. This was because the regex was
3651 used only to exclude .Trash (or whatever) when the size of the mailbox
3652 was calculated. There was no check that a delivery was happening into an
3653 excluded directory. This bug has been fixed by ignoring all quota
3654 processing for deliveries into excluded directories.
3656 PH/10 Added the maildirfolder_create_regex option to appendfile.
3662 PH/01 The code for finding all the local interface addresses on a FreeBSD
3663 system running IPv6 was broken. This may well have applied to all BSD
3664 systems, as well as to others that have similar system calls. The broken
3665 code found IPv4 interfaces correctly, but gave incorrect values for the
3666 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3667 that it would not match correctly against @[] and not recognize the IPv6
3670 PH/02 The ipliteral router was not recognizing addresses of the form user@
3671 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3673 PH/03 Added disable_ipv6.
3675 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3676 decoded form, because it is most often used to construct To: headers
3677 lines in autoreplies, and the decoded form may well be syntactically
3678 invalid. However, $reply_address has leading white space removed, and all
3679 newlines turned into spaces so that the autoreply transport does not
3682 PH/05 If group was specified without a user on a router, and no group or user
3683 was specified on a transport, the group from the router was ignored.
3685 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3686 for visible compile-time settings that can be used to change these
3687 numbers, for those that want even more. Backwards compatibility with old
3688 spool files has been maintained. However, going back to a previous Exim
3689 release will lost any variables that are in spool files.
3691 PH/07 Two small changes when running in the test harness: increase delay when
3692 passing a TCP/IP connection to a new process, in case the original
3693 process has to generate a bounce, and remove special handling of
3694 127.0.0.2 (sic), which is no longer necessary.
3696 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3697 be the same on different OS.
3699 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3702 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3703 whether --show-vars was specified or not
3705 JJ/02 exipick: Added support for new ACL variable spool format introduced
3708 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3709 syntactically invalid From: or Reply-to: line, and a filter used this to
3710 generate an autoreply, and therefore failed to obtain an address for the
3711 autoreply, Exim could try to deliver to a non-existent relative file
3712 name, causing unrelated and misleading errors. What now happens is that
3713 it logs this as a hard delivery error, but does not attempt to create a
3716 PH/11 The exinext utility has a -C option for testing purposes, but although
3717 the given file was scanned by exinext itself; it wasn't being passed on
3718 when Exim was called.
3720 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3721 an end-of-file indication when reading a command response.
3723 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3724 compiled. In many other places in Exim, IPv6 addresses are always
3725 recognized, so I have changed this. It also means that IPv4 domain
3726 literals of the form [IPV4:n.n.n.n] are now always recognized.
3728 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3729 used if the router is not running as root, for example, when verifying at
3730 ACL time, or when using -bh. The debugging output from this situation was
3731 non-existent - all you got was a failure to exec. I have made two
3734 (a) Failures to set uid/gid, the current directory, or a process leader
3735 in a subprocess such as that created by queryprogram now generate
3736 suitable debugging output when -d is set.
3738 (b) The queryprogram router detects when it is not running as root,
3739 outputs suitable debugging information if -d is set, and then runs
3740 the subprocess without attempting to change uid/gid.
3742 PH/15 Minor change to Makefile for building test_host (undocumented testing
3745 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3746 additional section of a DNS packet that returns MX or SRV records.
3747 Instead, it always explicitly searches for A/AAAA records. This avoids
3748 major problems that occur when a DNS server includes only records of one
3749 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3750 fixed another bug: if SRV records were looked up and the corresponding
3751 address records were *not* found in the additional section, the port
3752 values from the SRV records were lost.
3754 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3755 using the correct key (the original address) when searching the retry
3756 rules in order to find which one to use for generating the retry hint.
3758 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3759 adding the default one. Similarly, if it contains a Reply-To: header, the
3760 errors_reply_to option, if set, is not used.
3762 PH/19 When calculating a retry time, Exim used to measure the "time since
3763 failure" by looking at the "first failed" field in the retry record. Now
3764 it does not use this if it is later than than the arrival time of the
3765 message. Instead it uses the arrival time. This makes for better
3766 behaviour in cases where some deliveries succeed, thus re-setting the
3767 "first failed" field. An example is a quota failure for a huge message
3768 when small messages continue to be delivered. Without this change, the
3769 "time since failure" will always be short, possible causing more frequent
3770 delivery attempts for the huge message than are intended.
3771 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3773 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3774 $1, $2, $3) because the numerical variables can be reset during some
3775 expansion items (e.g. "match"), thereby losing the authentication data.
3777 PH/21 Make -bV show the size of off_t variables so that the test suite can
3778 decide whether to run tests for quotas > 2G.
3780 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3781 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3782 filecount value is greater than 2G or if a quota value is greater than 2G
3783 on a system where the size of off_t is not greater than 4, a panic error
3786 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3787 never match. The debug and -bh output now contains an explicit error
3788 message indicating a malformed IPv4 address or mask.
3790 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3791 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3792 PH/23 above applies.
3794 PH/25 Do not write to syslog when running in the test harness. The only
3795 occasion when this arises is a failure to open the main or panic logs
3796 (for which there is an explicit test).
3798 PH/26 Added the /no_tell option to "control=freeze".
3800 PH/27 If a host name lookup failed very early in a connection, for example, if
3801 the IP address matched host_lookup and the reverse lookup yielded a name
3802 that did not have a forward lookup, an error message of the form "no IP
3803 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3804 could be logged. Now it outputs the IP address instead of "NULL".
3806 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3807 allows the sender and the authenticated sender to be set when
3808 submitting a message from within Exim. Since child_open_exim() is
3809 documented for local_scan(), the new function should be too.
3811 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3812 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3813 results in an empty string is now treated as unset.
3815 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3817 PH/31 Added sender_verify_fail logging option.
3819 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3820 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3821 tidied the source and removed it altogether.
3823 PH/33 When a queue run was abandoned because the load average was too high, a
3824 log line was always written; now it is written only if the queue_run log
3825 selector is set. In addition, the log line for abandonment now contains
3826 information about the queue run such as the pid. This is always present
3827 in "start" and "stop" lines but was omitted from the "abandon" line.
3829 PH/34 Omit spaces between a header name and the colon in the error message that
3830 is given when verify = headers_syntax fails (if there are lots of them,
3831 the message gets confusing).
3833 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3834 names, as there are now some PTR records that contain slashes. This check
3835 is only to protect against broken name servers that fall over on strange
3836 characters, so the fact that it applies to all lookups doesn't matter.
3838 PH/36 Now that the new test suite is complete, we can remove some of the
3839 special code in Exim that was needed for the old test suite. For example,
3840 sorting DNS records because real resolvers return them in an arbitrary
3841 order. The new test suite's fake resolver always returns records in the
3844 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3845 bounces) except when queue_only is set, to avoid logging races between
3846 the different processes.
3848 PH/38 Panic-die if .include specifies a non-absolute path.
3850 PH/39 A tweak to the "H" retry rule from its user.
3852 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3853 a label. They prevented compilation on older perls.
3855 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3856 a warning to be raised on newish perls.
3858 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3859 on queue. Changes to match documented behaviour of showing count of
3860 messages matching specified criteria.
3862 PH/40 Changed the default ident timeout from 30s to 5s.
3864 PH/41 Added support for the use of login_cap features, on those BSD systems
3865 that have them, for controlling the resources used by pipe deliveries.
3867 PH/42 The content-scanning code uses fopen() to create files in which to put
3868 message data. Previously it was not paying any attention to the mode of
3869 the files. Exim runs with umask(0) because the rest of the code creates
3870 files with open(), and sets the required mode explicitly. Thus, these
3871 files were ending up world-writeable. This was not a big issue, because,
3872 being within the spool directory, they were not world-accessible. I have
3873 created a function called modefopen, which takes an additional mode
3874 argument. It sets umask(777), creates the file, chmods it to the required
3875 mode, then resets the umask. All the relevant calls to fopen() in the
3876 content scanning code have been changed to use this function.
3878 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3879 to 24 hours. This avoids potential overflow problems when processing G
3880 and H retry rules. I suspect nobody ever tinkers with this value.
3882 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3884 PH/45 When the plaintext authenticator is running as a client, the server's
3885 challenges are checked to ensure they are valid base64 strings. By
3886 default, the authentication attempt is cancelled if an invalid string is
3887 received. Setting client_ignore_invalid_base64 true ignores these errors.
3888 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3889 they are received. Thus, the responses can be made to depend on the
3890 challenges. If an invalid string is ignored, an empty string is placed in
3893 PH/46 Messages that are created by the autoreply transport now contains a
3894 References: header, in accordance with RFCs 2822 and 3834.
3896 PH/47 Added authenticated_sender_force to the smtp transport.
3898 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3900 PH/49 Installed latest patch from the Sieve maintainer.
3902 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3903 was also set, the appendfile transport was unnecessarily scanning a
3904 directory of message files (e.g. for maildir delivery) to find the count
3905 of files (along with the size), even though it did not need this
3906 information. It now does the scan only if it needs to find either the
3907 size of the count of files.
3909 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3911 PH/52 Two bugs concerned with error handling when the smtp transport is
3914 (i) Exim was not creating retry information for temporary errors given
3915 for individual recipients after the DATA command when the smtp transport
3916 was used in LMTP mode. This meant that they could be retried too
3917 frequently, and not timed out correctly.
3919 (ii) Exim was setting the flag that allows error details to be returned
3920 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3921 recipients that were returned after the DATA command.
3923 PH/53 This is related to PH/52, but is more general: for any failing address,
3924 when detailed error information was permitted to be returned to the
3925 sender, but the error was temporary, then after the final timeout, only
3926 "retry timeout exceeded" was returned. Now it returns the full error as
3927 well as "retry timeout exceeded".
3929 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3930 do this, and (what is worse) MTAs that accept it.
3932 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3933 will now be deprecated.
3935 PH/56 New os.c-cygwin from the Cygwin maintainer.
3937 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3938 formats (previously only available in exim formats via -bpr, -bpru,
3939 and -bpra. Now also available in native and exiqgrep formats)
3941 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3942 with very large, slow to parse queues
3944 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3946 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3948 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3949 responses to authentication challenges, though it was showing the
3950 challenges; (ii) I've removed the CR characters from the debug output for
3953 PH/58 Allow for the insertion of a newline as well as a space when a string
3954 is turned into more than one encoded-word during RFC 2047 encoding. The
3955 Sieve code now uses this.
3957 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3958 data_4xx, lost_connection, tls_required.
3960 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3961 message was being sent as an SMTP response.
3963 PH/61 Add -l and -k options to exicyclog.
3965 PH/62 When verifying, if an address was redirected to one new address, so that
3966 verification continued, and the new address failed or deferred after
3967 having set something in $address_data, the value of $address_data was not
3968 passed back to the ACL. This was different to the case when no
3969 redirection occurred. The value is now passed back in both cases.
3971 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3972 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3973 use login_cap.h, so on its own it isn't the distinguishing feature. The
3974 new name refers directly to the setclassresources() function.
3976 PH/65 Added configuration files for NetBSD3.
3978 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3980 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3981 is preferred over IPv4.
3983 PH/68 The bounce_return_message and bounce_return_body options were not being
3984 honoured for bounces generated during the reception of non-SMTP messages.
3985 In particular, this applied to messages rejected by the ACL. This bug has
3986 been fixed. However, if bounce_return_message is true and bounce_return_
3987 body is false, the headers that are returned for a non-SMTP message
3988 include only those that have been read before the error was detected.
3989 (In the case of an ACL rejection, they have all been read.)
3991 PH/69 The HTML version of the specification is now built in a directory called
3992 spec_html instead of spec.html, because the latter looks like a path with
3993 a MIME-type, and this confuses some software.
3995 PH/70 Catch two compiler warnings in sieve.c.
3997 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3998 function verify_get_ident() calls ip_connect() to connect a socket, but
3999 if the "connect()" function timed out, ip_connect() used to close the
4000 socket. However, verify_get_ident() also closes the socket later, and in
4001 between Exim writes to the log, which may get opened at this point. When
4002 the socket was closed in ip_connect(), the log could get the same file
4003 descriptor number as the socket. This naturally causes chaos. The fix is
4004 not to close the socket in ip_connect(); the socket should be closed by
4005 the function that creates it. There was only one place in the code where
4006 this was missing, in the iplookup router, which I don't think anybody now
4007 uses, but I've fixed it anyway.
4009 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4010 well as to direct DNS lookups. Otherwise the handling of names in host
4011 lists is inconsistent and therefore confusing.
4017 PH/01 Two changes to the default runtime configuration:
4019 (1) Move the checks for relay_from_hosts and authenticated clients from
4020 after to before the (commented out) DNS black list checks.
4022 (2) Add control=submission to the relay_from_hosts and authenticated
4023 clients checks, on the grounds that messages accepted by these
4024 statements are most likely to be submissions.
4026 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4028 (1) Generate an error if the third argument for the ${prvs expansion is
4031 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4034 (3) Reset the variables that are obtained from the first argument of
4035 ${prvscheck and used in the second argument before leaving the code,
4036 because their memory is reclaimed, so using them afterwards may do
4039 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4040 one (it's much easier than Tom thought :-).
4042 (5) Because of (4), we can now allow for the use of $prvscheck_result
4043 inside the third argument.
4045 PH/03 For some reason, the default setting of PATH when running a command from
4046 a pipe transport was just "/usr/bin". I have changed it to
4049 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4050 anything to be listed in the output from -bV.
4052 PH/05 When a filter generated an autoreply, the entire To: header line was
4053 quoted in the delivery log line, like this:
4055 => >A.N.Other <ano@some.domain> <original@ddress> ...
4057 This has been changed so that it extracts the operative address. There
4058 may be more than one such address. If so, they are comma-separated, like
4061 => >ano@some.domain,ona@other.domain <original@ddress> ...
4063 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4064 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4065 not being looked up in the rDNS to get a host name, Exim was showing the
4066 IP address twice in Received: lines, even though the IP addresses were
4067 identical. For example:
4069 Received: from [1.2.3.4] (helo=[1.2.3.4])
4071 However, if the real host name was known, it was omitting the HELO data
4072 if it matched the actual IP address. This has been tidied up so that it
4073 doesn't show the same IP address twice.
4075 PH/07 When both +timestamp and +memory debugging was on, the value given by
4076 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4077 called by the debug printing, thereby overwriting the timestamp buffer.
4078 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4080 PH/08 When the original message was included in an autoreply transport, it
4081 always said "this is a copy of the message, including all the headers",
4082 even if body_only or headers_only was set. It now gives an appropriate
4085 PH/09 Applied a patch from the Sieve maintainer which:
4087 o fixes some comments
4088 o adds the (disabled) notify extension core
4089 o adds some debug output for the result of if/elsif tests
4090 o points to the current vacation draft in the documentation
4091 and documents the missing references header update
4095 o fixes a bug in processing the envelope test (when testing
4096 multiple envelope elements, the last element determined the
4099 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4100 Electronic Mail") by including:
4102 Auto-submitted: auto-generated
4104 in the messages that it generates (bounce messages and others, such as
4105 warnings). In the case of bounce messages for non-SMTP messages, there was
4106 also a typo: it was using "Auto_submitted" (underscore instead of
4107 hyphen). Since every message generated by Exim is necessarily in response
4108 to another message, thes have all been changed to:
4110 Auto-Submitted: auto-replied
4112 in accordance with these statements in the RFC:
4114 The auto-replied keyword:
4116 - SHOULD be used on messages sent in direct response to another
4117 message by an automatic process,
4119 - MUST NOT be used on manually-generated messages,
4121 - MAY be used on Delivery Status Notifications (DSNs) and Message
4122 Disposition Notifications (MDNs),
4124 - MUST NOT be used on messages generated by automatic or periodic
4125 processes, except for messages which are automatic responses to
4128 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4129 to the default Received: header definition.
4131 PH/12 Added log selector acl_warn_skipped (default on).
4133 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4134 variables because (a) they are in the wrong storage pool and (b) even if
4135 they were copied, it wouldn't work properly because of the caching.
4137 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4138 checking when decoding. Apparently there are clients that generate
4139 overlong encoded strings. Why am I not surprised?
4141 PH/15 If the first argument of "${if match_address" was not empty, but did not
4142 contain an "@" character, Exim crashed. Now it writes a panic log message
4143 and treats the condition as false.
4145 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4147 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4148 extension "envelope-auth". The code is finished and in agreement with
4149 other implementations, but there is no documentation so far and in fact,
4150 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4151 not changing the active code.
4153 Print executed "if" and "elsif" statements when debugging is used. This
4154 helps a great deal to understand what a filter does.
4156 Document more things not specified clearly in RFC3028. I had all this
4157 sorted out, when out of a sudden new issues came to my mind. Oops."
4159 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4162 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4163 canonical form (as documented). However, after a host name lookup from
4164 the IP address, check_host() was doing a simple string comparison with
4165 addresses acquired from the DNS when checking that the found name did
4166 have the original IP as one of its addresses. Since any found IPv6
4167 addresses are likely to be in abbreviated form, the comparison could
4168 fail. Luckily, there already exists a function for doing the comparison
4169 by converting both addresses to binary, so now that is used instead of
4170 the text comparison.
4172 PH/20 There was another similar case to PH/19, when a complete host name was
4173 given in a host list; looking up its IP address could give an abbreviated
4174 form, whereas the current host's name might or might not be abbreviated.
4175 The same fix has been applied.
4181 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4182 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4185 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4186 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4188 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4190 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4191 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4192 Also, for "Auto-Submitted", treat anything other than "no" as
4193 non-personal, in accordance with RFC 3834. (Previously it treated
4194 anything starting "auto-" as non-personal.)
4196 TF/01 The control=submission/name=... option had a problem with syntax
4197 errors if the name included a slash character. The /name= option
4198 now slurps the rest of the string, so it can include any characters
4199 but it must come last in the list of options (after /sender_retain
4202 PH/05 Some modifications to the interface to the fake nameserver for the new
4210 TK/01 Added the "success_on_redirect" address verification option. See
4211 NewStuff for rationale and an example.
4213 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4215 PH/02 Patch to exigrep to allow it to work on syslog lines.
4217 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4218 fread() to skip over the body file's header line, because in Cygwin the
4219 header line is locked and is inaccessible.
4221 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4222 co-exist for some time) to make it clear that it is the Exim ID that is
4223 referenced, not the Message-ID: header line.
4225 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4226 string_format() function, because snprintf() does not exist on all
4229 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4230 ${stat: expansion item.
4232 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4233 protocol synchronization error", to keep the pedants happy.
4235 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4236 well as for IRIX systems, when gcc is being used. See the host.c source
4239 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4241 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4244 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4245 transport and to the smtp transport in LMTP mode.
4247 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4249 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4250 run a filter in a subprocess. This could lead to confusion in subsequent
4251 lookups in the parent process. There should also be a search_tidyup() at
4252 the end of the subprocess.
4254 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4255 only if the host matched helo_try_verify_hosts, which caused the
4256 verification to occur when the EHLO/HELO command was issued. The ACL just
4257 tested the remembered result. Now, if a previous verification attempt has
4258 not happened, "verify = helo" does it there and then.
4260 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4262 TK/03 Fix log output including CR from clamd.
4264 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4265 exist provoked a memory error which could cause a segfault.
4267 PH/15 Installed PCRE 6.2
4269 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4271 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4272 of the problem. Specifically, suggested +O2 rather than +O1 for the
4275 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4277 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4278 record, the port was not correctly logged when the outgoing_port log
4279 selector was set (it logged the transort's default port).
4281 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4282 fallback_hosts, and "hosts" in the smtp transport.
4284 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4285 host errors such as "Connection refused".
4287 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4288 authentication with radiusclient 0.4.9:
4290 - Error returned from rc_read_config was caught wrongly
4291 - Username/password not passed on to radius server due to wrong length.
4293 The presumption is that some radiusclient API changes for 4.51/PH/17
4294 were not taken care of correctly. The code is still untested by me (my
4295 Linux distribution still has 0.3.2 of radiusclient), but it was
4296 contributed by a Radius user.
4298 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4299 expanding the "port" option of the smtp transport.
4301 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4302 while reading a MIME header. Thanks to Tom Hughes for a patch.
4304 PH/24 Include config.h inside local_scan.h so that configuration settings are
4307 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4308 that in an ACL for RCPT (for example), you can examine exactly what was
4311 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4312 commands, but it was not correctly comparing the address with the actual
4313 client host address. Thus, it would show the EHLO address in Received:
4314 header lines when this was not necessary.
4316 PH/27 Added the % operator to ${eval:}.
4318 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4319 it should be ignoring failures (because with -C, for example, it has lost
4320 privilege). It wasn't ignoring creation failures other than "already
4323 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4326 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4327 because an input line was too long, either on its own, or by virtue of
4328 too many continuations, the temporary file was not being removed, and the
4329 return code was incorrect.
4331 PH/31 Missing "BOOL" in function definition in filtertest.c.
4333 PH/32 Applied Sieve patches from the maintainer.
4335 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4337 PH/33 Added "verify = not_blind".
4339 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4340 Local/Makefile (with some defaults set). These are used in built scripts
4341 such as exicyclog, but they have never been used in the exim_install
4342 script (though there are many overriding facilities there). I have
4343 arranged that the exim_install script now takes note of these two
4346 PH/35 Installed configuration files for Dragonfly.
4348 PH/36 When a locally submitted message by a trusted user did not contain a
4349 From: header, and the sender address was obtained from -f or from an SMTP
4350 MAIL command, and the trusted user did not use -F to supply a sender
4351 name, $originator_name was incorrectly used when constructing a From:
4352 header. Furthermore, $originator_name was used for submission mode
4353 messages from external hosts without From: headers in a similar way,
4354 which is clearly wrong.
4356 PH/37 Added control=suppress_local_fixups.
4358 PH/38 When log_selector = +received_sender was set, and the addition of the
4359 sender made the log line's construction buffer exactly full, or one byte
4360 less than full, an overflow happened when the terminating "\n" was
4363 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4364 when the result of a list match is failure because a DNS lookup failed.
4366 PH/40 RM_COMMAND is now used in the building process.
4368 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4369 the "build-* directories that it finds.
4371 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4372 address in a domain literal was a prefix of an interface address.
4374 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4375 when verifying a sender address, unless rewrite_headers is false.
4377 PH/44 Wrote a long comment about why errors_to addresses are verified as
4378 recipients, not senders.
4380 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4381 the ratelimit ACL was added.
4383 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4385 PH/46 Added extra information about PostgreSQL errors to the error string.
4387 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4388 avoiding the need to install special zones in a real server. This is
4389 backwards compatible; if it can't find the fake resolver, it drops back.
4390 Thus, both old and new test suites can be run.
4392 TF/02 Added util/ratelimit.pl
4394 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4395 clock is set back in time.
4397 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4398 Candler <B.Candler@pobox.com>.
4400 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4401 OK for addresses that are the result of redirections.
4403 PH/48 A number of further additions for the benefit of the new test suite,
4404 including a fake gethostbyname() that interfaces to the fake DNS resolver
4407 TF/06 The fix for widen_domains has also been applied to qualify_single and
4408 search_parents which are the other dnslookup options that can cause
4411 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4414 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4416 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4417 service" if it encountered an MX record with an empty target hostname.
4418 The message is now "an MX or SRV record indicated no SMTP service".
4420 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4421 if the DNS of the sending site is misconfigured. This is quite a
4422 common situation. This change restores the behaviour of treating a
4423 helo verification defer as a failure.
4425 PH/51 If self=fail was set on a router, the bounce message did not include the
4426 actual error message.
4432 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4434 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4435 command itself ended in error, the underlying message about the transport
4436 filter timeout was being overwritten with the pipe command error. Now the
4437 underlying error message should be appended to the second error message.
4439 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4441 PH/02 Reduce the amount of output that "make" produces by default. Full output
4442 can still be requested.
4444 PH/03 The warning log line about a condition test deferring for a "warn" verb
4445 was being output only once per connection, rather than after each
4446 occurrence (because it was using the same function as for successful
4447 "warn" verbs). This seems wrong, so I have changed it.
4449 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4450 it should not have, which might have caused a crash in the right
4451 circumstances, but probably never did.
4453 PH/04 Installed a modified version of Tony Finch's patch to make submission
4454 mode fix the return path as well as the Sender: header line, and to
4455 add a /name= option so that you can make the user's friendly name appear
4458 TF/03 Added the control = fakedefer ACL modifier.
4460 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4461 Mark Lowes for thorough testing.
4463 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4465 TK/03 Merged latest SRS patch from Miles Wilton.
4467 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4468 with the definition in sysexits.h (which is #included earlier).
4469 Fortunately, Exim does not actually use EX_OK. The code used to try to
4470 preserve the sysexits.h value, by assuming that macro definitions were
4471 scanned for macro replacements. I have been disabused of this notion,
4472 so now the code just undefines EX_OK before #including unistd.h.
4474 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4475 in the smtp transport. When a block could not be written in a single
4476 write() function, the timeout was being re-applied to each part-write.
4477 This seems wrong - if the receiver was accepting one byte at a time it
4478 would take for ever. The timeout is now adjusted when this happens. It
4479 doesn't have to be particularly precise.
4481 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4482 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4484 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4485 without a domain if the check to <postmaster@domain> fails.
4487 SC/01 Eximstats: added -xls and the ability to specify output files
4488 (patch written by Frank Heydlauf).
4490 SC/02 Eximstats: use FileHandles for outputting results.
4492 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4494 SC/04 Eximstats: fixed display of large numbers with -nvr option
4496 SC/05 Eximstats: fixed merging of reports with empty tables.
4498 SC/06 Eximstats: added the -include_original_destination flag
4500 SC/07 Eximstats: removed tabs and trailing whitespace.
4502 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4504 TK/06 MBOX spool code: Add real "From " MBOX separator line
4505 so the .eml file is really in mbox format (even though
4506 most programs do not really care). Patch from Alex Miller.
4508 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4509 The latter is generated from $received_to and is only set if the
4510 message has one envelope recipient. SA can use these headers,
4511 obviously out-of-the-box. Patch from Alex Miller.
4513 PH/08 The ${def test on a variable was returning false if the variable's
4514 value was "0", contrary to what the specification has always said!
4515 The result should be true unless the variable is empty.
4517 PH/09 The syntax error of a character other than { following "${if
4518 def:variable_name" (after optional whitespace) was not being diagnosed.
4519 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4520 accidental colon was present, for example, could give incorrect results.
4522 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4523 result is used (not including appendfile, where other changes are about
4526 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4527 This involved changing a lot of size variables from int to off_t. It
4528 should work with maildirs and everything.
4530 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4531 spamd dying while we are connected to it.
4533 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4536 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4537 function for BDB 4.3.
4539 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4541 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4542 the sources for PCRE have been re-arranged and more files are now
4545 PH/15 The code I had for printing potentially long long variables in PH/11
4546 above was not the best (it lost precision). The length of off_t variables
4547 is now inspected at build time, and an appropriate printing format (%ld
4548 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4549 to be "long long int" or "long int". This is needed for the internal
4550 formatting function string_vformat().
4552 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4553 the configuration file to be ":syslog", then the script "guesses" where
4554 the logs files are, rather than using the compiled in default. In our
4555 case the guess is not the same as the compiled default, so the script
4556 suddenly stopped working when I started to use syslog. The patch checks
4557 to see if log_file_path is "". If so, it attempts to read it from exim
4558 with no configuration file to get the compiled in version, before it
4559 falls back to the previous guessing code."
4561 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4562 implementing BATV in an Exim configuration. See NewStuff for the gory
4565 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4566 Makefile that are specific to HP-UX.
4568 PH/18 If the "use_postmaster" option was set for a recipient callout together
4569 with the "random" option, the postmaster address was used as the MAIL
4570 FROM address for the random test, but not for the subsequent recipient
4571 test. It is now used for both.
4573 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4574 patch removes a few documentation additions to RFC 3028, because the
4575 latest draft now contains them. It adds the new en;ascii-case comparator
4576 and a new error check for 8bit text in MIME parts. Comparator and
4577 require names are now matched exactly. I enabled the subaddress
4578 extension, but it is not well tested yet (read: it works for me)."
4580 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4581 rework some of the code of TK/09 above to avoid the hardwired use of
4582 "%lld" and "long long". Replaced the call to snprintf() with a call to
4585 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4586 records point to non-existent hosts", "retry timeout exceeded", and
4587 "retry time not reached for any host after a long failure period".
4589 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4590 experimental DomainKeys support:
4592 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4593 (2) On an error such as an illegally used "control", the wrong name for
4594 the control was given.
4596 These problems did NOT occur unless DomainKeys support was compiled.
4598 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4600 PH/24 Added ${if match_ip condition.
4602 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4603 databases so that it will be absolutely obvious if a crash occurs in the
4604 DB library. This is a regular occurrence (often caused by mis-matched
4607 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4608 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4609 picked up on a user's system that detects such things. There doesn't seem
4610 to be a gcc warning option for this - only an attribute that has to be
4611 put on the function's prototype. It seems that in Fedora Core 4 they have
4612 set this on a number of new functions. No doubt there will be more in due
4615 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4616 specify a transport. However, if an address that was verified by such a
4617 router was the subject of a callout, Exim crashed because it tried to
4618 read the rcpt_include_affixes from the non-existent transport. Now it
4619 just assumes that the setting of that option is false. This bug was
4620 introduced by 4.51/PH/31.
4622 PH/28 Changed -d+all to exclude +memory, because that information is very
4623 rarely of interest, but it makes the output a lot bigger. People tend to
4624 do -d+all out of habit.
4626 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4627 code in os-type was giving problems when libc.so lives in lib64, like on
4630 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4631 aren't the modern standard, and it seems that some systems' include files
4632 don't always have them. Exim was already checking for some of the newer
4633 ones like T_AAAA, and defining it itself. I've added checks for all the
4634 record types that Exim uses.
4636 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4637 not automatically generating a new one, as it is supposed to. This
4638 prevented TLS from working. If the file did exist, but contained invalid
4639 data, a new version was generated, as expected. It was only the case of a
4640 non-existent file that was broken.
4642 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4643 with a change in libdomainkeys > 0.64.
4645 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4646 from DNS. If the selector record carries the flag, it now has
4647 precedence over the domain-wide flag.
4649 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4651 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4652 the use of a port name that isn't defined in /etc/services) occurred, the
4653 message was deferred as in a normal delivery, and thus remained on the
4654 spool, instead of being failed because of the mua_wrapper setting. This
4655 is now fixed, and I tidied up some of the mua_wrapper messages at the
4658 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4659 the messages in a hash of arrays rather than using individual hashes.
4660 This is a bit cleaner and results in dramatic memory savings, albeit
4661 at a slight CPU cost.
4663 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4664 as requested by Marc Sherman.
4666 SC/10 Eximstats: added histograms for user specified patterns as requested
4669 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4671 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4672 fopen() in the content-scanning modules that did not already have it.
4678 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4679 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4681 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4683 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4685 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4686 to test Sieve filters that use "vacation".
4688 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4689 that changes the way the GnuTLS parameters are stored in the cache file.
4690 The new format can be generated externally. For backward compatibility,
4691 if the data in the cache doesn't make sense, Exim assumes it has read an
4692 old-format file, and it generates new data and writes a new file. This
4693 means that you can't go back to an older release without removing the
4696 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4697 work if there are any delivery delays because "one_time" forces the
4698 parent to be marked "delivered", so its unseen clone is never tried
4699 again. For this reason, Exim now forbids the simultaneous setting of
4702 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4703 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4704 entry if you want to know the details. The fix, however, neglected to
4705 consider the case when local delivery batching is involved. The test for
4706 "previously delivered" was not happening when checking to see if an
4707 address could be batched with a previous (undelivered) one; under
4708 certain circumstances this could lead to multiple deliveries to the same
4711 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4712 in its include files, and this causes problems building Exim.
4714 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4715 header_syntax) but Exim was just ignoring anything given after a slash.
4716 In particular, this caused confusion with an attempt to use "verify =
4717 reverse_host_lookup/defer_ok". An error is now given when options are
4718 supplied for verify items that do not have them. (Maybe reverse_host_
4719 lookup should have a defer_ok option, but that's a different point.)
4721 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4722 defined by RFC 821) to 2048, because there were problems with some AUTH
4723 commands, and RFC 1869 says the size should be increased for extended
4724 SMTP commands that take arguments.
4726 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4729 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4730 "unknown" error; now it says that the functionality isn't in the binary.
4732 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4733 an address' error message when a string expansion fails (syntax or
4734 whatever). Otherwise the password may appear in the log. Following change
4735 PH/42 below, there is no longer a chance of it appearing in a bounce
4738 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4740 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4741 of its addresses was ever tried. (Bugzilla bug #2.)
4743 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4744 the result incorrectly in the debug output. (It correctly added a newline
4745 to what was transported.)
4747 TF/01 Added $received_time.
4749 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4750 commented out examples of how to interface to a virus scanner and to
4751 SpamAssassin. Also added commented examples of av_scanner and
4752 spamd_address settings.
4754 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4755 and controls are allowed in which ACLs. There were a couple of minor
4756 errors. Some of the entries in the conditions table (which is a table of
4757 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4758 negation of where the condition IS allowed.
4760 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4762 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4763 header file does not have a version number, so I've had to invent a new
4764 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4765 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4766 radiusclient), but it was contributed by a Radius user.
4768 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4769 files or directories, for OpenSSL.
4771 PH/19 When an Exim process that is running as root has to create an Exim log
4772 file, it does so in a subprocess that runs as exim:exim so as to get the
4773 ownership right at creation (otherwise, other Exim processes might see
4774 the file with the wrong ownership). There was no test for failure of this
4775 fork() call, which would lead to the process getting stuck as it waited
4776 for a non-existent subprocess. Forks do occasionally fail when resources
4777 run out. I reviewed all the other calls to fork(); they all seem to check
4780 PH/20 When checking for unexpected SMTP input at connect time (before writing
4781 the banner), Exim was not dealing correctly with a non-positive return
4782 from the read() function. If the client had disconnected by this time,
4783 the result was a log entry for a synchronization error with an empty
4784 string after "input=" when read() returned zero. If read() returned -1
4785 (an event I could not check), uninitialized data bytes were printed.
4786 There were reports of junk text (parts of files, etc) appearing after
4789 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4791 PH/22 Added support for macro redefinition, and (re)definition in between
4792 driver and ACL definitions.
4794 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4795 forgetting to use the resulting value; it was using the unexpanded value.
4797 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4798 hadn't been configured. The fix is from Juergen Kreileder, who
4799 understands it better than I do:
4801 "Here's what I see happening with three configured cyrus_sasl
4802 authenticators configured (plain, login, cram-md5):
4804 On startup auth_cyrus_sasl_init() gets called for each of these.
4805 This means three calls to sasl_listmech() without a specified mech_list.
4806 => SASL tests which mechs of all available mechs actually work
4807 => three warnings about OTP not working
4808 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4810 With the patch, sasl_listmech() also gets called three times. But now
4811 SASL's mech_list option is set to the server_mech specified in the the
4812 authenticator. Or in other words, the answer from sasl_listmech()
4813 gets limited to just the mech you're testing for (which is different
4815 => the return list contains just 'plain' or 'login', 'cram-md5' or
4816 nothing depending on the value of ob->server_mech.
4818 I've just tested the patch: Authentication still works fine,
4819 unavailable mechs specified in the exim configuration are still
4820 caught, and the auth.log warnings about OTP are gone."
4822 PH/25 When debugging is enabled, the contents of the command line are added
4823 to the debugging output, even when log_selector=+arguments is not
4826 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4827 answer is "GNU", and only if the return is "GNU/something" is the answer
4830 PH/27 $acl_verify_message is now set immediately after the failure of a
4831 verification in an ACL, and so is available in subsequent modifiers. In
4832 particular, the message can be preserved by coding like this:
4834 warn !verify = sender
4835 set acl_m0 = $acl_verify_message
4837 Previously, $acl_verify_message was set only while expanding "message"
4838 and "log_message" when a very denied access.
4840 PH/28 Modified OS/os.c-Linux with
4842 -#ifndef OS_LOAD_AVERAGE
4843 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4845 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4846 nomenclature these days.)
4848 PH/29 Installed patch from the Sieve maintainer that adds the options
4849 sieve_useraddress and sieve_subaddress to the redirect router.
4851 PH/30 In these circumstances:
4852 . Two addresses routed to the same list of hosts;
4853 . First host does not offer TLS;
4854 . First host accepts first address;
4855 . First host gives temporary error to second address;
4856 . Second host offers TLS and a TLS session is established;
4857 . Second host accepts second address.
4858 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4859 and peerdn, if requested) that were in fact used only for the second
4862 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4863 attention to any local part prefix or suffix that was matched by the
4864 router that accepted the address. It now behaves in the same way as it
4865 does for delivery: the affixes are removed from the local part unless
4866 rcpt_include_affixes is set on the transport.
4868 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4869 timeout during the DATA phase of an incoming message.
4871 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4872 applied a patch sanctioned by the Sieve maintainer.
4874 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4875 the uid or gid is negative. A case of a negative gid caused this to be
4876 noticed. The fix allows for either to be negative.
4878 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4879 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4882 PH/36 The change PH/12 above was broken. Fixed it.
4884 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4885 the grounds that routing the same address twice would always produce the
4886 same answer. This might have been true once, but it is certainly no
4887 longer true now. Routing a child address may depend on the previous
4888 routing that produced that child. Some complicated redirection strategies
4889 went wrong when messages had multiple recipients, and made Exim's
4890 behaviour dependent on the order in which the addresses were given.
4892 I have moved the duplicate checking until after the routing is complete.
4893 Exim scans the addresses that are assigned to local and remote
4894 transports, and removes any duplicates. This means that more work will be
4895 done, as duplicates will always all be routed, but duplicates are
4896 presumably rare, so I don't expect this is of any significance.
4898 For deliveries to pipes, files, and autoreplies, the duplicate checking
4899 still happens during the routing process, since they are not going to be
4902 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4903 It corrects a timeout issue with spamd. This is Ian's comment: "The
4904 background is that sometimes spamd either never reads data from a
4905 connection it has accepted, or it never writes response data. The exiscan
4906 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4907 blindly assumes that writes won't block so it may never time out."
4909 PH/39 Allow G after quota size as well as K and M.
4911 PH/40 The value set for $authenticated_id in an authenticator may not contain
4912 binary zeroes or newlines because the value is written to log lines and
4913 to spool files. There was no check on this. Now the value is run through
4914 the string_printing() function so that such characters are converted to
4915 printable escape sequences.
4917 PH/41 $message_linecount is a new variable that contains the total number of
4918 lines in the message. Compare $body_linecount, which is the count for the
4921 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4922 bounce and delay warning messages, except in certain special cases, which
4925 (a) An SMTP error message from a remote host;
4926 (b) A message specified in a :fail: redirection;
4927 (c) A message specified in a "fail" command in a system filter;
4928 (d) A message specified in a FAIL return from the queryprogram router;
4929 (e) A message specified by the cannot_route_message router option.
4931 In these cases only, Exim does include the error details in bounce and
4932 warning messages. There are also a few cases where bland messages such
4933 as "unrouteable address" or "local delivery error" are given.
4935 PH/43 $value is now also set for the "else" part of a ${run expansion.
4937 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4938 being worked on, but at least Exim now implements the latest version to
4941 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4942 process to complete was treated as a delivery failure, a timeout while
4943 writing the message to the pipe was logged, but erroneously treated as a
4944 successful delivery. Such timeouts include transport filter timeouts. For
4945 consistency with the overall process timeout, these timeouts are now
4946 treated as errors, giving rise to delivery failures by default. However,
4947 there is now a new Boolean option for the pipe transport called
4948 timeout_defer, which, if set TRUE, converts the failures into defers for
4949 both kinds of timeout. A transport filter timeout is now identified in
4952 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4953 systems where "make" and "gmake" are different, calling "gmake" at top
4954 level broke things. I've arranged for the value of $(MAKE) to be passed
4955 from the Makefile to this script so that it can call the same version of
4959 A note about Exim versions 4.44 and 4.50
4960 ----------------------------------------
4962 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4963 changes of various kinds. As a consequence, a big documentation update was
4964 needed. This delayed the release for rather longer than seemed good, especially
4965 in the light of a couple of (minor) security issues. Therefore, the changes
4966 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4967 release. So 4.44 and 4.50 are in effect two different branches that both start
4970 I have left the 4.50 change log unchanged; it contains all the changes since
4971 4.43. The change log for 4.44 is below; many of its items are identical to
4972 those for 4.50. This seems to be the most sensible way to preserve the
4973 historical information.
4979 1. Minor wording change to the doc/README.SIEVE file.
4981 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4982 computation of the current number of files was incorrect.
4984 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4985 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4988 4. Give more explanation in the error message when the command for a transport
4989 filter fails to execute.
4991 5. There are several places where Exim runs a non-Exim command in a
4992 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4993 was being done only for the command run by the queryprogram router. It is
4994 now done for all such subprocesses. The other cases are: ${run, transport
4995 filters, and the commands run by the lmtp and pipe transports.
4997 6. Added CONFIGURE_GROUP build-time option.
4999 7. Some older OS have a limit of 256 on the maximum number of file
5000 descriptors. Exim was using setrlimit() to set 1000 as a large value
5001 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5002 systems. I've change it so that if it can't get 1000, it tries for 256.
5004 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5005 was an oversight, and furthermore, ever since the addition of extra
5006 controls (e.g. 4.43/32), the checks on when to allow different forms of
5007 "control" were broken. There should now be diagnostics for all cases when a
5008 control that does not make sense is encountered.
5010 9. Added the /retain_sender option to "control=submission".
5012 10. $recipients is now available in the predata ACL (oversight).
5014 11. Tidy the search cache before the fork to do a delivery from a message
5015 received from the command line. Otherwise the child will trigger a lookup
5016 failure and thereby defer the delivery if it tries to use (for example) a
5017 cached ldap connection that the parent has called unbind on.
5019 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5020 of $address_data from the recipient verification was clobbered by the
5021 sender verification.
5023 13. The value of address_data from a sender verification is now available in
5024 $sender_address_data in subsequent conditions in the ACL statement.
5026 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5028 15. Added a new option "connect=<time>" to callout options, to set a different
5031 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5032 was its contents. (It was OK if the option was not defined at all.)
5034 17. A "Completed" log line is now written for messages that are removed from
5035 the spool by the -Mrm option.
5037 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5038 information about exactly what failed.
5040 19. Added -dd to debug only the daemon process.
5042 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5043 handles timeouts, both on the server side and network timeouts. Renamed the
5044 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5046 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5047 It is now set to "smtps".
5049 22. $host_address is now set to the target address during the checking of
5050 ignore_target_hosts.
5052 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5053 being passed; this would have caused $sender_host_name to have been used if
5054 matching the list had actually called for a host name (not very likely,
5055 since this list is usually IP addresses). A host name is now passed as
5058 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5059 code that specifies a non-restarting handler (typically sigaction() in
5060 modern systems) in an attempt to fix a rare and obscure crash bug.
5062 25. Narrowed the window for a race in the daemon that could cause it to ignore
5063 SIGCHLD signals. This is not a major problem, because they are used only to
5064 wake it up if nothing else does.
5066 26. A malformed maildirsize file could cause Exim to calculate negative values
5067 for the mailbox size or file count. Odd effects could occur as a result.
5068 The maildirsize information is now recalculated if the size or filecount
5071 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5072 support for a long time. Removed HAVE_SYS_VFS_H.
5074 28. Installed the latest version of exipick from John Jetmore.
5076 29. In an address list, if the pattern was not a regular expression, an empty
5077 subject address (from a bounce message) matched only if the pattern was an
5078 empty string. Non-empty patterns were not even tested. This was the wrong
5079 because it is perfectly reasonable to use an empty address as part of a
5080 database query. An empty address is now tested by patterns that are
5081 lookups. However, all the other forms of pattern expect the subject to
5082 contain a local part and a domain, and therefore, for them, an empty
5083 address still always fails if the pattern is not itself empty.
5085 30. Exim went into a mad DNS loop when attempting to do a callout where the
5086 host was specified on an smtp transport, and looking it up yielded more
5087 than one IP address.
5089 31. Re-factored the code for checking spool and log partition space into a
5090 function that finds that data and another that does the check. The former
5091 is then used to implement four new variables: $spool_space, $log_space,
5092 $spool_inodes, and $log_inodes.
5094 32. The RFC2047 encoding function was originally intended for short strings
5095 such as real names; it was not keeping to the 75-character limit for
5096 encoded words that the RFC imposes. It now respects the limit, and
5097 generates multiple encoded words if necessary. To be on the safe side, I
5098 have increased the buffer size for the ${rfc2047: expansion operator from
5101 33. It is now permitted to omit both strings after an "if" condition; if the
5102 condition is true, the result is "true". As before, when the second string
5103 is omitted, a false condition yields an empty string. This makes it less
5104 cumbersome to write custom ACL and router conditions.
5106 34. Failure to deliver a bounce message always caused it to be frozen, even if
5107 there was an errors_to setting on the router. The errors_to setting is now
5110 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5111 canonical form (fully expanded) before being placed in
5112 $sender_host_address.
5114 36. The table in the code that translates DNS record types into text (T_A to
5115 "A" for instance) was missing entries for NS and CNAME. It is just possible
5116 that this could have caused confusion if both these types were looked up
5117 for the same domain, because the text type is used as part of Exim's
5118 per-process caching. But the chance of anyone hitting this buglet seems
5121 37. The dnsdb lookup has been extended in a number of ways.
5123 (1) There is a new type, "zns", which walks up the domain tree until it
5124 finds some nameserver records. It should be used with care.
5126 (2) There is a new type, "mxh", which is like "mx" except that it returns
5127 just the host names, not the priorities.
5129 (3) It is now possible to give a list of domains (or IP addresses) to be
5130 looked up. The behaviour when one of the lookups defers can be
5131 controlled by a keyword.
5133 (4) It is now possible to specify the separator character for use when
5134 multiple records are returned.
5136 38. The dnslists ACL condition has been extended: it is now possible to supply
5137 a list of IP addresses and/or domains to be looked up in a particular DNS
5140 39. Added log_selector=+queue_time_overall.
5142 40. When running the queue in the test harness, wait just a tad after forking a
5143 delivery process, to get repeatability of debugging output.
5145 41. Include certificate and key file names in error message when GnuTLS fails
5146 to set them up, because the GnuTLS error message doesn't include the name
5147 of the failing file when there is a problem reading it.
5149 42. Allow both -bf and -bF in the same test run.
5151 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5153 44. The "Exiscan patch" is now merged into the mainline Exim source.
5155 45. Sometimes the final signoff response after QUIT could fail to get
5156 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5157 < 0 before doing a fflush(). This bug looks as though it goes back to the
5158 introduction of TLS in release 3.20, but "sometimes" must have been rare
5159 because the tests only now provoked it.
5161 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5162 (this can affect the format of dates).
5164 47. exim_tidydb, when checking for the continued existence of a message for
5165 which it has found a message-specific retry record, was not finding
5166 messages that were in split spool directories. Consequently, it was
5167 deleting retry records that should have stayed in existence.
5169 48. Steve fixed some bugs in eximstats.
5171 49. The SPA authentication driver was not abandoning authentication and moving
5172 on to the next authenticator when an expansion was forced to fail,
5173 contradicting the general specification for all authenticators. Instead it
5174 was generating a temporary error. It now behaves as specified.
5176 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5177 (the order specifies the preference for clients). The order is now AES256,
5178 AES128, 3DES, ARCFOUR128.
5180 51. Small patch to Sieve code - explicitly set From: when generating an
5183 52. Exim crashed if a remote delivery caused a very long error message to be
5184 recorded - for instance if somebody sent an entire SpamAssassin report back
5185 as a large number of 550 error lines. This bug was coincidentally fixed by
5186 increasing the size of one of Exim's internal buffers (big_buffer) that
5187 happened as part of the Exiscan merge. However, to be on the safe side, I
5188 have made the code more robust (and fixed the comments that describe what
5191 53. Now that there can be additional text after "Completed" in log lines (if
5192 the queue_time_overall log selector is set), a one-byte patch to exigrep
5193 was needed to allow it to recognize "Completed" as not the last thing in
5196 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5197 patch that reportedly fixes this has been added. I am not expert enough to
5198 create a test for it. This is what the patch creator wrote:
5200 "I found a little strange behaviour of ldap code when working with
5201 Windows 2003 AD Domain, where users was placed in more than one
5202 Organization Units. When I tried to give exim partial DN, the exit code
5203 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5204 But simultaneously result of request was absolutely normal ldap result,
5205 so I produce this patch..."
5207 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5208 REFERENCE, so I have modified the code to exclude the patch when that macro
5211 55. Some experimental protocols are using DNS PTR records for new purposes. The
5212 keys for these records are domain names, not reversed IP addresses. The
5213 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5214 leaves it alone. Component reversal etc. now happens only for IP addresses.
5217 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5219 57. Double the size of the debug message buffer (to 2048) so that more of very
5220 long debug lines gets shown.
5222 58. The exicyclog utility now does better if the number of log files to keep
5223 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5225 59. Two changes related to the smtp_active_hostname option:
5227 (1) $smtp_active_hostname is now available as a variable.
5228 (2) The default for smtp_banner uses $smtp_active_hostname instead
5229 of $primary_hostname.
5231 60. The host_aton() function is supposed to be passed a string that is known
5232 to be a valid IP address. However, in the case of IPv6 addresses, it was
5233 not checking this. This is a hostage to fortune. Exim now panics and dies
5234 if the condition is not met. A case was found where this could be provoked
5235 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5236 components; fortuitously, this particular loophole had already been fixed
5237 by change 4.50/55 above.
5239 If there are any other similar loopholes, the new check in host_aton()
5240 itself should stop them being exploited. The report I received stated that
5241 data on the command line could provoke the exploit when Exim was running as
5242 exim, but did not say which command line option was involved. All I could
5243 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5244 running as the user.
5247 61. There was a buffer overflow vulnerability in the SPA authentication code
5248 (which came originally from the Samba project). I have added a test to the
5249 spa_base64_to_bits() function which I hope fixes it.
5252 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5253 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5255 63. The daemon start-up calls getloadavg() while still root for those OS that
5256 need the first call to be done as root, but it missed one case: when
5257 deliver_queue_load_max is set with deliver_drop_privilege. This is
5258 necessary for the benefit of the queue runner, because there is no re-exec
5259 when deliver_drop_privilege is set.
5261 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5262 This has been fixed.
5264 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5265 was in use, was not putting the data itself into the right store pool;
5266 consequently, it could be overwritten for a subsequent message in the same
5267 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5270 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5272 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5273 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5274 the function were treating the return as a boolean value, which happened to
5275 work because 0=false and not-0=true, but is not correct code.
5277 68. The host_aton() function was not handling scoped IPv6 addresses (those
5278 with, for example, "%eth0" on the end) correctly.
5280 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5281 negated items (that is, ~something) in unsigned ints. Some compilers
5282 apparently mutter when there is no cast.
5284 70. If an address verification called from an ACL failed, and did not produce a
5285 user-specific message (i.e. there was only a "system" message), nothing was
5286 put in $acl_verify_message. In this situation, it now puts the system
5289 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5290 session; change 4.31/43 added the unwanted input to the log line - except
5291 that it did not do this in the start of session case. It now does.
5293 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5294 This is wrong and can cause the other end to generate a synchronization
5295 error if it is another Exim or anything else that does the synchronization
5296 check. A QUIT command is no longer sent after a timeout.
5298 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5299 during host lookups.
5301 74. The defer_ok option of callout verification was not working if it was used
5302 when verifying addresses in header lines, that is, for this case:
5304 verify = header_sender/callout=defer_ok
5306 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5307 those file descriptors could be used for SMTP connections. If anything
5308 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5309 could be sent to the SMTP client, causing chaos. The daemon now opens
5310 stdin, stdout, and stderr to /dev/null when it puts itself into the
5313 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5314 log by default. The user can override this with suitable Perl magic.
5316 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5317 the log message when discard triggers, was not working for the DATA ACL or
5318 for the non-SMTP ACL.
5320 78. Error message wording change in sieve.c.
5322 79. If smtp_accept_max_per_host was set, the number of connections could be
5323 restricted to fewer than expected, because the daemon was trying to set up
5324 a new connection before checking whether the processes handling previous
5325 connections had finished. The check for completed processes is now done
5326 earlier. On busy systems, this bug wouldn't be noticed because something
5327 else would have woken the daemon, and it would have reaped the completed
5330 80. If a message was submitted locally by a user whose login name contained one
5331 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5332 It caused a spool format error. I have fixed the spool reading code. A
5333 related problem was that the "from" clause in the Received: line became
5334 illegal because of the space(s). It is now covered by ${quote_local_part.
5336 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5339 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5341 83. Patch from Sieve maintainer to fix unterminated string problem in
5342 "vacation" handling.
5344 84. Some minor changes to the Linux configuration files to help with other
5345 OS variants using glibc.
5347 85. One more patch for Sieve to update vacation handling to latest spec.
5350 ----------------------------------------------------
5351 See the note above about the 4.44 and 4.50 releases.
5352 ----------------------------------------------------
5358 1. Change 4.43/35 introduced a bug that caused file counts to be
5359 incorrectly computed when quota_filecount was set in an appendfile
5362 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5363 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5366 3. Give more explanation in the error message when the command for a transport
5367 filter fails to execute.
5369 4. There are several places where Exim runs a non-Exim command in a
5370 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5371 was being done only for the command run by the queryprogram router. It is
5372 now done for all such subprocesses. The other cases are: ${run, transport
5373 filters, and the commands run by the lmtp and pipe transports.
5375 5. Some older OS have a limit of 256 on the maximum number of file
5376 descriptors. Exim was using setrlimit() to set 1000 as a large value
5377 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5378 systems. I've change it so that if it can't get 1000, it tries for 256.
5380 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5381 was an oversight, and furthermore, ever since the addition of extra
5382 controls (e.g. 4.43/32), the checks on when to allow different forms of
5383 "control" were broken. There should now be diagnostics for all cases when a
5384 control that does not make sense is encountered.
5386 7. $recipients is now available in the predata ACL (oversight).
5388 8. Tidy the search cache before the fork to do a delivery from a message
5389 received from the command line. Otherwise the child will trigger a lookup
5390 failure and thereby defer the delivery if it tries to use (for example) a
5391 cached ldap connection that the parent has called unbind on.
5393 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5394 of $address_data from the recipient verification was clobbered by the
5395 sender verification.
5397 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5398 was its contents. (It was OK if the option was not defined at all.)
5400 11. A "Completed" log line is now written for messages that are removed from
5401 the spool by the -Mrm option.
5403 12. $host_address is now set to the target address during the checking of
5404 ignore_target_hosts.
5406 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5407 being passed; this would have caused $sender_host_name to have been used if
5408 matching the list had actually called for a host name (not very likely,
5409 since this list is usually IP addresses). A host name is now passed as
5412 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5413 code that specifies a non-restarting handler (typically sigaction() in
5414 modern systems) in an attempt to fix a rare and obscure crash bug.
5416 15. Narrowed the window for a race in the daemon that could cause it to ignore
5417 SIGCHLD signals. This is not a major problem, because they are used only to
5418 wake it up if nothing else does.
5420 16. A malformed maildirsize file could cause Exim to calculate negative values
5421 for the mailbox size or file count. Odd effects could occur as a result.
5422 The maildirsize information is now recalculated if the size or filecount
5425 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5426 support for a long time. Removed HAVE_SYS_VFS_H.
5428 18. Updated exipick to current release from John Jetmore.
5430 19. Allow an empty sender to be matched against a lookup in an address list.
5431 Previously the only cases considered were a regular expression, or an
5434 20. Exim went into a mad DNS lookup loop when doing a callout where the
5435 host was specified on the transport, if the DNS lookup yielded more than
5438 21. The RFC2047 encoding function was originally intended for short strings
5439 such as real names; it was not keeping to the 75-character limit for
5440 encoded words that the RFC imposes. It now respects the limit, and
5441 generates multiple encoded words if necessary. To be on the safe side, I
5442 have increased the buffer size for the ${rfc2047: expansion operator from
5445 22. Failure to deliver a bounce message always caused it to be frozen, even if
5446 there was an errors_to setting on the router. The errors_to setting is now
5449 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5450 canonical form (fully expanded) before being placed in
5451 $sender_host_address.
5453 24. Updated eximstats to version 1.33
5455 25. Include certificate and key file names in error message when GnuTLS fails
5456 to set them up, because the GnuTLS error message doesn't include the name
5457 of the failing file when there is a problem reading it.
5459 26. Expand error message when OpenSSL has problems setting up cert/key files.
5462 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5463 (this can affect the format of dates).
5465 28. exim_tidydb, when checking for the continued existence of a message for
5466 which it has found a message-specific retry record, was not finding
5467 messages that were in split spool directories. Consequently, it was
5468 deleting retry records that should have stayed in existence.
5470 29. eximstats updated to version 1.35
5471 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5472 1.35 - bugfix such that pie charts by volume are generated correctly
5474 30. The SPA authentication driver was not abandoning authentication and moving
5475 on to the next authenticator when an expansion was forced to fail,
5476 contradicting the general specification for all authenticators. Instead it
5477 was generating a temporary error. It now behaves as specified.
5479 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5480 (the order specifies the preference for clients). The order is now AES256,
5481 AES128, 3DES, ARCFOUR128.
5483 31. Small patch to Sieve code - explicitly set From: when generating an
5486 32. Exim crashed if a remote delivery caused a very long error message to be
5487 recorded - for instance if somebody sent an entire SpamAssassin report back
5488 as a large number of 550 error lines. This bug was coincidentally fixed by
5489 increasing the size of one of Exim's internal buffers (big_buffer) that
5490 happened as part of the Exiscan merge. However, to be on the safe side, I
5491 have made the code more robust (and fixed the comments that describe what
5494 33. Some experimental protocols are using DNS PTR records for new purposes. The
5495 keys for these records are domain names, not reversed IP addresses. The
5496 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5497 leaves it alone. Component reversal etc. now happens only for IP addresses.
5500 34. The host_aton() function is supposed to be passed a string that is known
5501 to be a valid IP address. However, in the case of IPv6 addresses, it was
5502 not checking this. This is a hostage to fortune. Exim now panics and dies
5503 if the condition is not met. A case was found where this could be provoked
5504 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5505 components; fortuitously, this particular loophole had already been fixed
5506 by change 4.50/55 or 4.44/33 above.
5508 If there are any other similar loopholes, the new check in host_aton()
5509 itself should stop them being exploited. The report I received stated that
5510 data on the command line could provoke the exploit when Exim was running as
5511 exim, but did not say which command line option was involved. All I could
5512 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5513 running as the user.
5516 35. There was a buffer overflow vulnerability in the SPA authentication code
5517 (which came originally from the Samba project). I have added a test to the
5518 spa_base64_to_bits() function which I hope fixes it.
5521 36. The daemon start-up calls getloadavg() while still root for those OS that
5522 need the first call to be done as root, but it missed one case: when
5523 deliver_queue_load_max is set with deliver_drop_privilege. This is
5524 necessary for the benefit of the queue runner, because there is no re-exec
5525 when deliver_drop_privilege is set.
5527 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5528 was in use, was not putting the data itself into the right store pool;
5529 consequently, it could be overwritten for a subsequent message in the same
5530 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5533 38. Sometimes the final signoff response after QUIT could fail to get
5534 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5535 < 0 before doing a fflush(). This bug looks as though it goes back to the
5536 introduction of TLS in release 3.20, but "sometimes" must have been rare
5537 because the tests only now provoked it.
5543 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5544 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5545 Now it returns an integer. A number of calls were still expecting a T/F
5546 return. Fortuitously, in all cases, the tests worked in OK situations,
5547 which is the norm. However, things would have gone wrong on any write
5548 failures on the smtp file descriptor. This function is used when sending
5549 messages over SMTP and also when doing verify callouts.
5551 2. When Exim is called to do synchronous delivery of a locally submitted
5552 message (the -odf or -odi options), it no longer closes stderr before doing
5555 3. Implemented the mua_wrapper option.
5557 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5559 5. Implemented the functions header_remove(), header_testname(),
5560 header_add_at_position(), and receive_remove_recipient(), and exported them
5563 6. If an ACL "warn" statement specified the addition of headers, Exim already
5564 inserted X-ACL-Warn: at the start if there was no header name. However, it
5565 was not making this test for the second and subsequent header lines if
5566 there were newlines in the string. This meant that an invalid header could
5567 be inserted if Exim was badly configured.
5569 7. Allow an ACL "warn" statement to add header lines at the start or after all
5570 the Received: headers, as well as at the end.
5572 8. Added the rcpt_4xx retry error code.
5574 9. Added postmaster_mailfrom=xxx to callout verification option.
5576 10. Added mailfrom=xxxx to the callout verification option, for verify=
5579 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5580 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5582 12. Inserted some casts to stop certain compilers warning when using pointer
5583 differences as field lengths or precisions in printf-type calls (mostly
5584 affecting debugging statements).
5586 13. Added optional readline() support for -be (dynamically loaded).
5588 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5589 same clock tick as a message's arrival, so that its received time was the
5590 same as the "first fail" time on the retry record, and that message
5591 remained on the queue past the ultimate address timeout, every queue runner
5592 would try a delivery (because it was past the ultimate address timeout) but
5593 after another failure, the ultimate address timeout, which should have then
5594 bounced the address, did not kick in. This was a "< instead of <=" error;
5595 in most cases the first failure would have been in the next clock tick
5596 after the received time, and all would be well.
5598 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5599 being recognized when the domain list was tested by the match_domain
5600 condition in an expansion string.
5602 16. Added the ${str2b64: operator.
5604 17. Exim was always calling setrlimit() to set a large limit for the number of
5605 processes, without checking whether the existing limit was already
5606 adequate. (It did check for the limit on file descriptors.) Furthermore,
5607 errors from getrlimit() and setrlimit() were being ignored. Now they are
5608 logged to the main and panic logs, but Exim does carry on, to try to do its
5609 job under whatever limits there are.
5611 18. Imported PCRE 5.0.
5613 19. Trivial typo in log message " temporarily refused connection" (the leading
5616 20. If the log selector return_path_on_delivery was set and an address was
5617 redirected to /dev/null, the delivery process crashed because it assumed
5618 that a return path would always be set for a "successful" delivery. In this
5619 case, the whole delivery is bypassed as an optimization, and therefore no
5622 21. Internal re-arrangement: the function for sending a challenge and reading
5623 a response while authentication was assuming a zero-terminated challenge
5624 string. It's now changed to take a pointer and a length, to allow for
5625 binary data in such strings.
5627 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5629 23. Exim was not respecting finduser_retries when seeking the login of the
5630 uid under which it was called; it was always trying 10 times. (The default
5631 setting of finduser_retries is zero.) Also, it was sleeping after the final
5632 failure, which is pointless.
5634 24. Implemented tls_on_connect_ports.
5636 25. Implemented acl_smtp_predata.
5638 26. If the domain in control=submission is set empty, Exim assumes that the
5639 authenticated id is a complete email address when it generates From: or
5640 Sender: header lines.
5642 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5643 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5644 chown and chgrp in /bin and hostname in /usr/bin.
5646 28. Exim was keeping the "process log" file open after each use, just as it
5647 does for the main log. This opens the possibility of it remaining open for
5648 long periods when the USR1 signal hits a daemon. Occasional processlog
5649 errors were reported, that could have been caused by this. Anyway, it seems
5650 much more sensible not to leave this file open at all, so that is what now
5653 29. The long-running daemon process does not normally write to the log once it
5654 has entered its main loop, and it closes the log before doing so. This is
5655 so that log files can straightforwardly be renamed and moved. However,
5656 there are a couple of unusual error situations where the daemon does write
5657 log entries, and I had neglected to close the log afterwards.
5659 30. The text of an SMTP error response that was received during a remote
5660 delivery was being truncated at 512 bytes. This is too short for some of
5661 the long messages that one sometimes sees. I've increased the limit to
5664 31. It is now possible to make retry rules that apply only when a message has a
5665 specific sender, in particular, an empty sender.
5667 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5668 it possible to be selective about when SMTP synchronization is enforced.
5670 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5672 32. Implemented hosts_connection_nolog.
5674 33. Added an ACL for QUIT.
5676 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5679 35. Added mailbox_size and mailbox_filecount to appendfile.
5681 36. Added control = no_multiline_responses to ACLs.
5683 37. There was a bug in the logic of the code that waits for the clock to tick
5684 in the case where the clock went backwards by a substantial amount such
5685 that the microsecond fraction of "now" was more than the microsecond
5686 fraction of "then" (but the whole seconds number was less).
5688 38. Added support for the libradius Radius client library this is found on
5689 FreeBSD (previously only the radiusclient library was supported).
5695 1. When certain lookups returned multiple values in the form name=value, the
5696 quoting of the values was not always being done properly. Specifically:
5697 (a) If the value started with a double quote, but contained no whitespace,
5699 (b) If the value contained whitespace other than a space character (i.e.
5700 tabs or newlines or carriage returns) it was not quoted.
5701 This fix has been applied to the mysql and pgsql lookups by writing a
5702 separate quoting function and calling it from the lookup code. The fix
5703 should probably also be applied to nisplus, ibase and oracle lookups, but
5704 since I cannot test any of those, I have not disturbed their existing code.
5706 2. A hit in the callout cache for a specific address caused a log line with no
5707 reason for rejecting RCPT. Now it says "Previous (cached) callout
5708 verification failure".
5710 3. There was an off-by-one bug in the queryprogram router. An over-long
5711 return line was truncated at 256 instead of 255 characters, thereby
5712 overflowing its buffer with the terminating zero. As well as fixing this, I
5713 have increased the buffer size to 1024 (and made a note to document this).
5715 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5716 when Exim is waiting for an SMTP response from a remote server, Exim
5717 restarts its select() call on the socket, thereby resetting its timeout.
5718 This is not a problem when such interrupts are rare. Somebody set up a cron
5719 job to run exiwhat every 2 minutes, which is less than the normal select()
5720 timeout (5 or 10 minutes). This meant that the select() timeout never
5721 kicked in because it was always reset. I have fixed this by comparing the
5722 time when an interrupt arrives with the time at the start of the first call
5723 to select(). If more time than the timeout has elapsed, the interrupt is
5724 treated as a timeout.
5726 5. Some internal re-factoring in preparation for the addition of Sieve
5727 extensions (by MH). In particular, the "personal" test is moved to a
5728 separate function, and given an option for scanning Cc: and Bcc: (which is
5729 not set for Exim filters).
5731 6. When Exim created an email address using the login of the caller as the
5732 local part (e.g. when creating a From: or Sender: header line), it was not
5733 quoting the local part when it contained special characters such as @.
5735 7. Installed new OpenBSD configuration files.
5737 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5738 try to make them clearer.
5740 9. Callout options, other than the timeout value, were being ignored when
5741 verifying sender addresses in header lines. For example, when using
5743 verify = header_sender/callout=no_cache
5745 the cache was (incorrectly) being used.
5747 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5748 only the Cygwin environment.
5750 11. When return_path_on_delivery was set as a log selector, if different remote
5751 addresses in the same message used different return paths and parallel
5752 remote delivery occurred, the wrong values would sometimes be logged.
5753 (Whenever a remote delivery process finished, the return path value from
5754 the most recently started remote delivery process was logged.)
5756 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5757 lines when AUTH and/or TLS are in use. This is the "received protocol"
5758 field. Exim used to use "asmtp" for authenticated SMTP, without any
5759 indication (in the protocol name) for TLS use. Now it follows the RFC and
5760 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5761 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5762 names appear in log lines as well as in Received: header lines.
5764 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5765 extensions, and comparison tests, and to fix some bugs.
5767 14. Changes to the "personal" filter test:
5769 (1) The test was buggy in that it was just doing the equivalent of
5770 "contains" tests on header lines. For example, if a user's address was
5771 anne@some.where, the "personal" test would incorrectly be true for
5773 To: susanne@some.where
5775 This test is now done by extracting each address from the header in turn,
5776 and checking the entire address. Other tests that are part of "personal"
5777 are now done using regular expressions (for example, to check local parts
5778 of addresses in From: header lines).
5780 (2) The list of non-personal local parts in From: addresses has been
5781 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5782 taken from the Sieve specification recommendations.
5784 (3) If the message contains any header line starting with "List-" it is
5785 treated as non-personal.
5787 (4) The test for "circular" in the Subject: header line has been removed
5788 because it now seems ill-conceived.
5790 15. Minor typos in src/EDITME comments corrected.
5792 16. Installed latest exipick from John Jetmore.
5794 17. If headers_add on a router specified a text string that was too long for
5795 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5796 of string_sprintf() is now avoided.
5798 18. $message_body_size was not set (it was always zero) when running the DATA
5799 ACL and the local_scan() function.
5801 19. For the "mail" command in an Exim filter, no default was being set for
5802 the once_repeat time, causing a random time value to be used if "once" was
5803 specified. (If the value happened to be <= 0, no repeat happened.) The
5804 default is now 0s, meaning "never repeat". The "vacation" command was OK
5805 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5806 (I found it when inspecting the code).
5808 20. There is now an overall timeout for performing a callout verification. It
5809 defaults to 4 times the callout timeout, which applies to individual SMTP
5810 commands during the callout. The overall timeout applies when there is more
5811 than one host that can be tried. The timeout is checked before trying the
5812 next host. This prevents very long delays if there are a large number of
5813 hosts and all are timing out (e.g. when the network connections are timing
5814 out). The value of the overall timeout can be changed by specifying an
5815 additional sub-option for "callout", called "maxwait". For example:
5817 verify = sender/callout=5s,maxwait=20s
5819 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5820 to the end before writing, but this should make it even safer).
5822 22. Exim was forgetting that it had advertised PIPELINING for the second and
5823 subsequent messages on an SMTP connection. It was also not resetting its
5824 memory on STARTTLS and an internal HELO.
5826 23. When Exim logs an SMTP synchronization error within a session, it now
5827 records whether PIPELINING has been advertised or not.
5829 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5830 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5831 rather than long int.
5833 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5835 26. Added the never_mail option to autoreply.
5841 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5842 crash if the getsockname() call failed; this can happen if a connection is
5843 closed very soon after it is established. The problem was simply in the
5844 order in which certain operations were done, causing Exim to try to write
5845 to the SMTP stream before it had set up the file descriptor. The bug has
5846 been fixed by making things happen in the correct order.
5852 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5853 before the connection was closed, thus losing the rejection response.
5855 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5856 some early Solaris releases, but causes trouble in current releases where
5857 socklen_t is defined.
5859 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5862 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5865 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5866 that releasing the top of it at the end releases what was used for sub-
5867 expansions (unless the block got too big). However, discard this block if
5868 the first thing is a variable or header, so that we can use its block when
5869 it is dynamic (useful for very large $message_headers, for example).
5871 6. Lookups now cache *every* query, not just the most recent. A new, separate
5872 store pool is used for this. It can be recovered when all lookup caches are
5873 flushed. Lookups now release memory at the end of their result strings.
5874 This has involved some general refactoring of the lookup sources.
5876 7. Some code has been added to the store_xxx() functions to reduce the amount
5877 of flapping under certain conditions.
5879 8. log_incoming_interface used to affect only the <= reception log lines. Now
5880 it causes the local interface and port to be added to several more SMTP log
5881 lines, for example "SMTP connection from", and rejection lines.
5883 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5885 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5887 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5888 could be overwritten at the end of the current message (or the start of a
5889 new message if it was set in a HELO ACL). The value is now preserved for
5890 the duration of the SMTP connection.
5892 12. If a transport had a headers_rewrite setting, and a matching header line
5893 contained an unqualified address, that address was qualified, even if it
5894 did not match any rewriting rules. The underlying bug was that the values
5895 of the flags that permit the existence of unqualified sender and recipient
5896 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5897 non-local messages, and by -bnq for local messages) were not being
5898 preserved with the message after it was received.
5900 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5901 "next input=" as part of the text comprising the host identity instead of
5902 the correct text. The code was using the same buffer for two different
5903 strings. However, depending on which order the printing function evaluated
5904 its arguments, the bug did not always show up. Under Linux, for example, my
5905 test suite worked just fine.
5907 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5908 doesn't work with some older versions of Perl. It has been changed to "my",
5909 which in any case is probably the better facility to use.
5911 15. A really picky compiler found some instances of statements for creating
5912 error messages that either had too many or two few arguments for the format
5915 16. The size of the buffer for calls to the DNS resolver has been increased
5916 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5917 for addresses that have a lot of PTR records. This alleviates a problem; it
5918 does not fully solve it.
5920 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5921 buffer now truncates the list and logs the incident, which is the same
5922 action as happens when Exim is looking up a host name and its aliases.
5923 Previously in this situation something unpredictable would happen;
5924 sometimes it was "internal error: store_reset failed".
5926 18. If a server dropped the connection unexpectedly when an Exim client was
5927 using GnuTLS and trying to read a response, the client delivery process
5928 crashed while trying to generate an error log message.
5930 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5931 string, for example:
5933 warn message = H1: something\nH2: something
5935 the text was added as a single header line from Exim's point of view
5936 though it ended up OK in the delivered message. However, searching for the
5937 second and subsequent header lines using $h_h2: did not work. This has been
5938 fixed. Similarly, if a system filter added multiple headers in this way,
5939 the routers could not see them.
5941 20. Expanded the error message when iplsearch is called with an invalid key to
5942 suggest using net-iplsearch in a host list.
5944 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5945 ACLs are no longer actually imposed (and a message to that effect is
5948 22. If a "gecos" field in a passwd entry contained escaped characters, in
5949 particular, if it contained a \" sequence, Exim got it wrong when building
5950 a From: or a Sender: header from that name. A second bug also caused
5951 incorrect handling when an unquoted " was present following a character
5952 that needed quoting.
5954 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5955 was not being matched caselessly.
5957 24. Arranged for all hyphens in the exim.8 source to be escaped with
5960 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5961 properly. Recipient callout cache records were still being keyed to include
5962 the sender, even when use_sender was set false. This led to far more
5963 callouts that were necessary. The sender is no longer included in the key
5964 when use_sender is false.
5966 26. Added "control = submission" modifier to ACLs.
5968 27. Added the ${base62d: operator to decode base 62 numbers.
5970 28. dnsdb lookups can now access SRV records.
5972 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5973 the configuration file.
5975 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5976 (-v) mode. This makes the output for a verbose queue run more intelligible.
5978 31. Added a use_postmaster feature to recipient callouts.
5980 32. Added the $body_zerocount variable, containing the number of binary zero
5981 bytes in the message body.
5983 33. The time of last modification of the "new" subdirectory is now used as the
5984 "mailbox time last read" when there is a quota error for a maildir
5987 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5989 35. Added +ignore_unknown as a special item in host lists.
5991 36. Code for decoding IPv6 addresses in host lists is now included, even if
5992 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5993 address was recognized as an IP address, but was then not correctly decoded
5994 into binary, causing unexpected and incorrect effects when compared with
6001 1. Very minor rewording of debugging text in manualroute to say "list of
6002 hosts" instead of "hostlist".
6004 2. If verify=header_syntax was set, and a header line with an unqualified
6005 address (no domain) and a large number of spaces between the end of the
6006 name and the colon was received, the reception process suffered a buffer
6007 overflow, and (when I tested it) crashed. This was caused by some obsolete
6008 code that should have been removed. The fix is to remove it!
6010 3. When running in the test harness, delay a bit after writing a bounce
6011 message to get a bit more predictability in the log output.
6013 4. Added a call to search_tidyup() just before forking a reception process. In
6014 theory, someone could use a lookup in the expansion of smtp_accept_max_
6015 per_host which, without the tidyup, could leave open a database connection.
6017 5. Added the variables $recipient_data and $sender_data which get set from a
6018 lookup success in an ACL "recipients" or "senders" condition, or a router
6019 "senders" option, similar to $domain_data and $local_part_data.
6021 6. Moved the writing of debug_print from before to after the "senders" test
6024 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6025 problems for message scanning, either using a data ACL, or using
6026 local_scan() because the Received: header was not generated till after they
6027 were called (in order to set the time as the time of reception completion).
6028 I have revised the way this works. The header is now generated after the
6029 body is received, but before the ACL or local_scan() are called. After they
6030 are run, the timestamp in the header is updated.
6036 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6037 before starting a queue runner without re-exec. This happened only when
6038 deliver_drop_privilege was set or when the Exim user was set to root. The
6039 effect of the bug was that timeouts during subsequent deliveries caused
6040 crashes instead of being properly handled. The handler is now left at its
6041 default (and expected) setting.
6043 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6044 message, again when deliver_drop_privilege is set or Exim is run as root.
6045 The bug described in (1) was not present in this case, but the tidying up
6046 of the other signals was missing. I have made the two cases consistent.
6048 3. The ignore_target_hosts setting on a manualroute router was being ignored
6049 for hosts that were looked up using the /MX notation.
6051 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6054 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6055 operated on the sender address. After changing the $sender_address to <>
6056 for the sender address verify, Exim was re-instated it as the original
6057 (before rewriting) address, but remembering that it had rewritten it, so it
6058 wasn't rewriting it again. This bug also had the effect of breaking the
6059 sender address verification caching when the sender address was rewritten.
6061 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6062 This has been changed so that if the ip literal address matches
6063 ignore_target_hosts, the router declines.
6065 7. Added expansion conditions match_domain, match_address, and match_local_
6066 part (NOT match_host).
6068 8. The placeholder for the Received: header didn't have a length field set.
6070 9. Added code to Exim itself and to exim_lock to test for a specific race
6071 condition that could lead to file corruption when using MBX delivery. The
6072 issue is with the lockfile that is created in /tmp. If this file is removed
6073 after a process has opened it but before that process has acquired a lock,
6074 there is the potential for a second process to recreate the file and also
6075 acquire a lock. This could lead to two Exim processes writing to the file
6076 at the same time. The added code performs the same test as UW imapd; it
6077 checks after acquiring the lock that its file descriptor still refers to
6078 the same named file.
6080 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6081 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6084 11. Added the smtp_active_hostname option. If used, this will typically be made
6085 to depend on the incoming interface address. Because $interface_address is
6086 not set up until the daemon has forked a reception process, error responses
6087 that can happen earlier (such as "too many connections") no longer contain
6090 12. If an expansion in a condition on a "warn" statement fails because a lookup
6091 defers, the "warn" statement is abandoned, and the next ACL statement is
6092 processed. Previously this caused the whole ACL to be aborted.
6094 13. Added the iplsearch lookup type.
6096 14. Added ident_timeout as a log selector.
6098 15. Added tls_certificate_verified as a log selector.
6100 16. Added a global option tls_require_ciphers (compare the smtp transport
6101 option of the same name). This controls incoming TLS connections.
6103 17. I finally figured out how to make tls_require_ciphers do a similar thing
6104 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6105 before starting the TLS session.
6107 18. Tabs are now shown as \t in -bP output.
6109 19. If the log selector return_path_on_delivery was set, Exim crashed when
6110 bouncing a message because it had too many Received: header lines.
6112 20. If two routers both had headers_remove settings, and the first one included
6113 a superfluous trailing colon, the final name in the first list and the
6114 first name in the second list were incorrectly joined into one item (with a
6115 colon in the middle).
6121 1. Added -C and -D options to the exinext utility, mainly to make it easier
6122 to include in the automated testing, but these could be helpful when
6123 multiple configurations are in use.
6125 2. The exinext utility was not formatting the output nicely when there was
6126 an alternate port involved in the retry record key, nor when there was a
6127 message id as well (for retries that were specific to a specific message
6128 and a specific host). It was also confused by IPv6 addresses, because of
6129 the additional colons they contain. I have fixed the IPv4 problem, and
6130 patched it up to do a reasonable job for IPv6.
6132 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6133 delivery, the log line now contains "pipelined" if PIPELINING was used.
6135 4. An SMTP transport process used to panic and die if the bind() call to set
6136 an explicit outgoing interface failed. This has been changed; it is now
6137 treated in the same way as a connect() failure.
6139 5. A reference to $sender_host_name in the part of a conditional expansion
6140 that was being skipped was still causing a DNS lookup. This no longer
6143 6. The def: expansion condition was not recognizing references to header lines
6144 that used bh_ and bheader_.
6146 7. Added the _cache feature to named lists.
6148 8. The code for checking quota_filecount in the appendfile transport was
6149 allowing one more file than it should have been.
6151 9. For compatibility with Sendmail, the command line option
6159 and sets the incoming protocol and host name (for trusted callers). The
6160 host name and its colon can be omitted when only the protocol is to be set.
6161 Note the Exim already has two private options, -pd and -ps, that refer to
6162 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6163 "s", but I don't think that's a major issue.
6165 10. A number of refactoring changes to the code, none of which should affect
6168 (a) The number of logging options was getting close to filling up the
6169 32-bit word that was used as a bit map. I have split them into two classes:
6170 those that are passed in the argument to log_write(), and those that are
6171 only ever tested independently outside of that function. These are now in
6172 separate 32-bit words, so there is plenty of room for expansion again.
6173 There is no change in the user interface or the logging behaviour.
6175 (b) When building, for example, log lines, the code previously used a
6176 macro that called string_cat() twice, in order to add two strings. This is
6177 not really sufficiently general. Furthermore, there was one instance where
6178 it was actually wrong because one of the argument was used twice, and in
6179 one call a function was used. (As it happened, calling the function twice
6180 did not affect the overall behaviour.) The macro has been replaced by a
6181 function that can join an arbitrary number of extra strings onto a growing
6184 (c) The code for expansion conditions now uses a table and a binary chop
6185 instead of a serial search (which was left over from when there were very
6186 few conditions). Also, it now recognizes conditions like "pam" even when
6187 the relevant support is not compiled in: a suitably worded error message is
6188 given if an attempt is made to use such a condition.
6190 11. Added ${time_interval:xxxxx}.
6192 12. A bug was causing one of the ddress fields not to be passed back correctly
6193 from remote delivery subprocesses. The field in question was not being
6194 subsequently used, so this caused to problems in practice.
6196 13. Added new log selectors queue_time and deliver_time.
6198 14. Might have fixed a bug in maildirsizefile handling that threw up
6199 "unexpected character" debug warnings, and recalculated the data
6200 unnecessarily. In any case, I expanded the warning message to give more
6203 15. Added the message "Restricted characters in address" to the statements in
6204 the default ACL that block characters like @ and % in local parts.
6206 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6207 Three changes have been made:
6209 (a) There was a serious bug; a negative response to MAIL caused the whole
6210 recipient domain to be cached as invalid, thereby blocking all messages
6211 to all local parts at the same domain, from all senders. This bug has
6212 been fixed. The domain is no longer cached after a negative response to
6213 MAIL if the sender used is not empty.
6215 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6218 (c) A new callout option, "use_sender" has been added for people who want
6219 the modified behaviour.
6225 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6228 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6229 indeed breaks things for older releases.
6231 3. Added additional logging to the case where there is a problem reading data
6232 from a filter that is running in a subprocess using a pipe, in order to
6233 try to track down a specific problem.
6235 4. Testing facility fudge: when running in the test harness and attempting
6236 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6237 getting "No route to host". Convert this to a timeout.
6239 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6242 6. Some OS don't have socklen_t but use size_t instead. This affects the
6243 fifth argument of getsockopt() amongst other things. This is now
6244 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6245 can be set for individual OS. I have set it for SunOS5, OSF1, and
6246 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6247 some earlier ones do not.
6249 7. Change 4.30/15 was not doing the test caselessly.
6251 8. The standard form for an IPv6 address literal was being rejected by address
6252 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6253 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6254 this, as well as the form without the "IPv6" on the front (but only when
6255 address literals are enabled, of course).
6257 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6259 10. Exim crashed if a message with an empty sender address specified by -f
6260 encountered a router with an errors_to setting. This could be provoked only
6261 by a command such as
6265 where an empty string was supplied; "<>" did not hit this bug.
6267 11. Installed PCRE release 4.5.
6269 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6270 remained set. It is now erased.
6272 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6273 times from message ids (which are base 36 rather than the normal 62).
6275 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6276 were being counted as actual protocol errors, and logged if the log
6277 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6278 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6279 and DATA following a set of rejected RCPTs do not count as protocol errors.
6280 In other words, Exim assumes they were pipelined, though this may not
6281 actually be the case. Of course, in all cases the client gets an
6282 appropriate error code.
6284 15. If a lookup fails in an ACL condition, a message about the failure may
6285 be available; it is used if testing the ACL cannot continue, because most
6286 such messages specify what the cause of the deferral is. However, some
6287 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6288 that caused an old message to be retained and used if a later statement
6289 caused a defer, replacing the real cause of the deferral.
6291 16. If an IP address had so many PTR records that the DNS lookup buffer
6292 was not large enough to hold them, Exim could crash while trying to process
6293 the truncated data. It now detects and logs this case.
6295 17. Further to 4.21/58, another change has been made: if (and only if) the
6296 first line of a message (the first header line) ends with CRLF, a bare LF
6297 in a subsequent header line has a space inserted after it, so as not to
6298 terminate the header.
6300 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6301 unnecessarily, used atoi() instead of strtol(), and didn't check the
6302 termination when getting file sizes from file names by regex.
6304 19. Completely re-implemented the support for maildirsize files, in the light
6305 of a number of problems with the previous contributed implementation
6306 (4.30/29). In particular:
6308 . If the quota is zero, the maildirsize file is maintained, but no quota is
6311 . If the maildir directory does not exist, it is created before any attempt
6312 to write a maildirsize file.
6314 . The quota value in the file is just a cache; if the quota is changed in
6315 the transport, the new value overrides.
6317 . A regular expression is available for excluding directories from the
6320 20. The autoreply transport checks the characters in options that define the
6321 message's headers; it allows continued headers, but it was checking with
6322 isspace() after an embedded newline instead of explicitly looking for a
6325 21. If all the "regular" hosts to which an address was routed had passed their
6326 expiry times, and had not reached their retry times, the address was
6327 bounced, even if fallback hosts were defined. Now Exim should go on to try
6330 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6331 equivalent code in the SMTP transport. Some hosts send humungous responses
6332 to HELO/EHLO, more than 1024 it seems.
6334 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6335 gives compiler warnings in some environments. I've now done it "properly",
6338 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6339 (because of problems with the built-in one) was declared to return uschar *
6340 instead of char *, causing compiler failure.
6342 25. Fixed a file descriptor leak when processing alias/forward files.
6344 26. Fixed a minor format string issue in dbfn.c.
6346 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6348 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6349 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6352 29. When a sender address is verified, it is cached, to save repeating the test
6353 when there is more than one recipient in a message. However, when the
6354 verification involves a callout, it is possible for different callout
6355 options to be set for different recipients. It is too complicated to keep
6356 track of this in the cache, so now Exim always runs a verification when a
6357 callout is required, relying on the callout cache for the optimization.
6358 The overhead is duplication of the address routing, but this should not be
6361 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6362 to be verified with callout=postmaster, and the main callout worked but the
6363 postmaster check failed, the verification correctly failed. However, if a
6364 subsequent RCPT command asked for sender verification *without* the
6365 postmaster check, incorrect caching caused this verification also to fail,
6368 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6369 it was not caching the DNS options (qualify_single, search_parents) that
6370 were used when the lookup failed. A subsequent lookup with different
6371 options therefore always gave the same answer, though there were cases
6372 where it should not have. (Example: a "domains = !$mx_any" option on a
6373 dnslookup router: the "domains" option is always processed without any
6374 widening, but the router might have qualify_single set.) Now Exim uses the
6375 cached value only when the same options are set.
6377 32. Added John Jetmore's "exipick" utility to the distribution.
6379 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6380 than a timeout (e.g. a certificate is required, and is not provided), an
6381 Exim server now closes the connection immediately. Previously it waited for
6382 the client to close - but if the client is SSL, it seems that they each
6383 wait for each other, leading to a delay before one of them times out.
6385 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6386 maintained 0.8.x compatibility because I don't think many are using it, and
6387 it is clearly obsolete.
6389 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6392 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6393 client certificate was expired. A simple patch fixes this, though I don't
6394 understand the full logic of why the verify callback is called multiple
6397 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6398 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6399 which causes problems with some clients (such as the Certicom SSL Plus
6400 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6401 disables the coutermeasure allowing Eudora to connect."
6403 38. Exim was not checking that a write() to a log file succeeded. This could
6404 lead to Bad Things if a log got too big, in particular if it hit a file
6405 size limit. Exim now panics and dies if it cannot write to a log file, just
6406 as it does if it cannot open a log file.
6408 39. Modified OS/Makefile-Linux so that it now contains
6410 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6412 The two -D definitions ensure that Exim is compiled with large file
6413 support, which makes it possible to handle log files that are bigger than
6416 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6417 instance) a domain was checked against a named list that involved a lookup,
6418 causing $domain_data to be set, then another domain was checked against the
6419 same list, then the first domain was re-checked, the value of $domain_data
6420 after the final check could be wrong. In particular, if the second check
6421 failed, it could be set empty. This bug probably also applied to
6424 41. The strip_trailing_dot option was not being applied to the address given
6425 with the -f command-line option.
6427 42. The code for reading a message's header from the spool was incrementing
6428 $received_count, but never initializing it. This meant that the value was
6429 incorrect (doubled) while delivering a message in the same process in which
6430 it was received. In the most common configuration of Exim, this never
6431 happens - a fresh exec is done - but it can happen when
6432 deliver_drop_privilege is set.
6434 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6435 it now includes up to 150 characters of the unexpected data in the log
6438 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6439 and building data strings. The size of both of these buffers was 10 000
6440 bytes - far larger than anybody would *ever* want, thought I. Needless to
6441 say, somebody hit the limit. I have increased the maximum line length to
6442 20 000 and the maximum data length of concatenated lines to 100 000. I have
6443 also fixed two bugs, because there was no checking on these buffers. Tsk,
6444 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6445 buffer is too small.
6447 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6448 lsearch lookups. Now it does.
6450 46. When parsing a route_list item in a manualroute router, a fixed-length
6451 buffer was used for the list of hosts. I made this 1024 bytes long,
6452 thinking that nobody would ever have a list of hosts that long. Wrong.
6453 Somebody had a whole pile of complicated expansion conditions, and the
6454 string was silently truncated, leading to an expansion error. It turns out
6455 that it is easier to change to an unlimited length (owing to other changes
6456 that have happened since this code was originally written) than to build
6457 structure for giving a limitation error. The length of the item that
6458 expands into the list of hosts is now unlimited.
6460 47. The lsearch lookup could not handle data where the length of text line was
6461 more than 4095 characters. Such lines were truncated, leading to shortened
6462 data being returned. It should now handle lines of any length.
6464 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6465 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6468 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6469 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6470 no longer happens in comment lines. A list of replacements is now placed
6471 at the head of all of the source files, except those whose only change is
6472 to replace PERL_COMMAND in the very first #! line.
6474 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6475 messages on the queue, with a bottom-up merge sort, using code contributed
6476 by Michael Haardt. This should make operations like -bp somewhat faster on
6477 large queues. It won't affect queue runners, except when queue_run_in_order
6480 51. Installed eximstats 1.31 in the distribution.
6482 52. Added support for SRV lookups to the dnslookup router.
6484 53. If an ACL referred to $message_body or $message_body_end, the value was not
6485 reset for any messages that followed in the same SMTP session.
6487 54. The store-handling optimization for building very long strings was not
6488 differentiating between the different store pools. I don't think this
6489 actually made any difference in practice, but I've tidied it.
6491 55. While running the routers to verify a sender address, $sender_address
6492 was still set to the sender address. This is wrong, because when routing to
6493 send a bounce to the sender, it would be empty. Therefore, I have changed
6494 it so that, while verifying a sender address, $sender_address is set to <>.
6495 (There is no change to what happens when verifying a recipient address.)
6497 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6498 target A or AAAA records (if not already returned) without resetting the
6499 qualify_single or search_parents options of the DNS resolver. These are
6500 inappropriate in this case because the targets of MX and SRV records must
6501 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6502 target that, when qualified, matched something in the local domain. These
6503 two options are now turned off when doing these lookups.
6505 57. It seems that at least some releases of Reiserfs (which does not have the
6506 concept of a fixed number of inodes) returns zero and not -1 for the
6507 number of available inodes. This interacted badly with check_spool_inodes,
6508 which assumed that -1 was the "no such thing" setting. What I have done is
6509 to check that the total number of inodes is greater than zero before doing
6510 the test of how many are available.
6512 58. When a "warn" ACL statement has a log_message modifier, the message is
6513 remembered, and not repeated. This is to avoid a lot of repetition when a
6514 message has many recipients that cause the same warning to be written.
6515 However, Exim was preserving the list of already written lines for an
6516 entire SMTP session, which doesn't seem right. The memory is now reset if a
6517 new message is started.
6519 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6520 debugging output unless log_rewrite was also set.
6522 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6523 of (int)(handle) when we know that handle contains (void *)(-1).
6525 61. The Exim daemon panic-logs an error return when it closes the incoming
6526 connection. However "connection reset by peer" seems to be common, and
6527 isn't really an error worthy of noting specially, so that particular error
6530 62. When Exim is trying to find all the local interfaces, it used to panic and
6531 die if the ioctl to get the interface flags failed. However, it seems that
6532 on at least one OS (Solaris 9) it is possible to have an interface that is
6533 included in the list of interfaces, but for which you get a failure error
6534 for this call. This happens when the interface is not "plumbed" into a
6535 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6536 failure of the "get flags" call assumes that the interface is down.
6538 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6539 makes life easier for people who are doing arithmetic on fields extracted
6540 from dates, where you often get leading zeros that should not be
6541 interpreted as octal.
6543 64. Added qualify_domain to the redirect router, to override the global
6546 65. If a pathologically long header line contained very many addresses (the
6547 report of this problem mentioned 10 000) and each of them was rewritten,
6548 Exim could use up a very large amount of memory. (It kept on making new
6549 copies of the header line as it rewrote, and never released the old ones.)
6550 At the expense of a bit more processing, the header rewriting function has
6551 been changed so that it no longer eats memory in this way.
6553 66. The generation of the Received: header has been moved from the time that a
6554 message starts to be received, to the time that it finishes. The timestamp
6555 in the Received: header should now be very close to that of the <= log
6556 line. There are two side-effects of this change:
6558 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6559 logged header lines no longer include the local Received: line, because
6560 it has not yet been created. The same applies to a copy of the message
6561 that is returned to a non-SMTP sender when a message is rejected.
6563 (b) When a filter file is tested using -bf, no additional Received: header
6564 is added to the test message. After some thought, I decided that this
6567 This change does not affect the value of $received_for. It is still set
6568 after address rewriting, but before local_scan() is called.
6570 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6572 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6573 gave an unhelpful panic error message, and a defer error. I have managed to
6574 change this behaviour so that it now rejects any supplied certificate,
6575 which seems right, as the list of acceptable certificates is empty.
6577 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6578 gave an unhelpful defer error. I have not managed to make this reject any
6579 supplied certificates, but the error message it gives is "no certificate
6580 supplied", which is not helpful.
6582 70. exigrep's output now also includes lines that are not associated with any
6583 message, but which match the given pattern. Implemented by a patch from
6584 Martin Sluka, which also tidied up the Perl a bit.
6586 71. Recipient callout verification, like sender verification, was using <> in
6587 the MAIL FROM command. This isn't really the right thing, since the actual
6588 sender may affect whether the remote host accepts the recipient or not. I
6589 have changed it to use the actual sender in the callout; this means that
6590 the cache record is now keyed on a recipient/sender pair, not just the
6591 recipient address. There doesn't seem to be a real danger of callout loops,
6592 since a callout by the remote host to check the sender would use <>.
6593 [SEE ABOVE: changed after hitting problems.]
6595 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6596 temporary errors. However, in the case of such a code being given after
6597 the end of a data transmission (i.e. after ".") Exim was failing to write
6598 a retry record for the message. (Yes, there was some broken host that was
6599 actually sending 8xx at this point.)
6601 73. An unknown lookup type in a host list could cause Exim to panic-die when
6602 the list was checked. (An example that provoked this was putting <; in the
6603 middle of a list instead of at the start.) If this happened during a DATA
6604 ACL check, a -D file could be left lying around. This kind of configuration
6605 error no longer causes Exim to die; instead it causes a defer error. The
6606 incident is still logged to the main and panic logs.
6608 74. Buglet left over from Exim 3 conversion. The message "too many messages
6609 in one connection" was written to the rejectlog but not the mainlog, except
6610 when address rewriting (yes!) was being logged.
6612 75. Added write_rejectlog option.
6614 76. When a system filter was run not as root (that is, when system_filter_user
6615 was set), the values of the $n variables were not being returned to the
6616 main process; thus, they were not subsequently available in the $sn
6619 77. Added +return_path_on_delivery log selector.
6621 78. A connection timeout was being treated differently from recipients deferred
6622 when testing hosts_max_try with a message that was older than the host's
6623 retry timeout. (The host should not be counted, thus allowing all hosts to
6624 be tried at least once before bouncing.) This may have been the cause of an
6625 occasionally reported bug whereby a message would remain on the queue
6626 longer than the retry timeout, but would be bounced if a delivery was
6627 forced. I say "may" because I never totally pinned down the problem;
6628 setting up timeout/retry tests is difficult. See also the next item.
6630 79. The ultimate address timeout was not being applied to errors that involved
6631 a combination of host plus message (for example, a timeout on a MAIL
6632 command). When an address resolved to a number of possible hosts, and they
6633 were not all tried for each delivery (e.g. because of hosts_max_try), a
6634 message could remain on the queue longer than the retry timeout.
6636 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6639 81. Fixed an obscure SMTP outgoing bug which required at least the following
6640 conditions: (a) there was another message waiting for the same server;
6641 (b) the server returned 5xx to all RCPT commands in the first message so
6642 that the message was not completed; (c) the server dropped the connection
6643 or gave a negative response to the RSET that Exim sends to abort the
6644 transaction. The observed case was a dropped connection after DATA that had
6645 been sent in pipelining mode. That is, the server had advertised PIPELINING
6646 but was not implementing it correctly. The effect of the bug was incorrect
6647 behaviour, such as trying another host, and this could lead to a crash.
6653 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6654 and daemon.c were passed as pointers to ints; they should have been
6655 pointers to socklen_t variables (which are typically unsigned ints).
6657 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6660 3. Fixed a really odd bug that affected only the testing scheme; patching a
6661 certain fixed string in the binary changed the value of another string that
6662 happened to be identical to the end of the original first string.
6664 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6665 name", it returns that address as the IP address. On some operating
6666 systems (e.g. Solaris), it also passes back the IP address string as the
6667 "host name". However, on others (e.g. Linux), it passes back an empty
6668 string. Exim wasn't checking for this, and was changing the host name to an
6669 empty string, assuming it had been canonicalized.
6671 5. Although rare, it is permitted to have more than one PTR record for a given
6672 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6673 all the names associated with an address, because they do in Solaris.
6674 However, it seems that they do not in Linux for data that comes from the
6675 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6676 I found this out when I moved to a new Linux workstation and tried to run
6677 the Exim test suite.
6679 To get round this problem I have changed the code so that it now does its
6680 own call to the DNS to look up PTR records when searching for a host name.
6681 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6682 addresses that are only in /etc/hosts are still found.
6684 This behaviour is, however, controlled by an option called host_lookup_
6685 order, which defaults to "bydns:byaddr". If people want to use the other
6686 order, or indeed, just use one or the other means of lookup, they can
6687 specify it in this variable.
6689 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6690 some operating systems, this comes back from gethostbyaddr() as an empty
6691 string, and this is what Exim used to test for. However, it seems that in
6692 other systems, "." is yielded. Exim now tests for this case too.
6694 7. The values of check_spool_space and check_log_space are now held internally
6695 as a number of kilobytes instead of an absolute number of bytes. If a
6696 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6697 kilobyte. This means that much larger values can be stored.
6699 8. Exim monitor: an attempt to get the action menu when not actually pointing
6700 at a message produces an empty menu entitled "No message selected". This
6701 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6702 no entries in it ("Shell widget menu has zero width and/or height"). So I
6703 have added a single, blank menu entry in this case.
6705 9. Added ${quote_local_part.
6707 10. MIME decoding is now applied to the contents of Subject: header lines when
6710 11. Now that a reference to $sender_host_address automatically causes a reverse
6711 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6712 host lookup before query-style lookups in lists that might use this
6713 variable. This has therefore been abolished, and the "net-" prefix is no
6714 longer necessary for query-style lookups.
6716 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6717 have been a typo for LFLAGS, so it has been changed.
6719 13. The install script calls Exim with "-C /dev/null" in order to find the
6720 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6721 to be output. However, since Exim outputs its version number before the
6722 error, it didn't break the script. It just looked ugly. I fixed this by
6723 always allowing "-C /dev/null" if the caller is root.
6725 14. Ignore overlarge ACL variable number when reading spool file - insurance
6726 against a later release with more variables having written the file.
6728 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6729 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6730 this, as well as the form without the "IPv6" on the front.
6732 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6733 OS/Makefile-Darwin file.
6735 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6736 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6738 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6739 variable. However, some OS use ints for inodes. I've added cast to long int
6740 to get rid of the compiler warning.
6742 19. I had forgotten to lock out "/../" in configuration file names when
6743 ALT_CONFIG_PREFIX was set.
6745 20. Routers used for verification do not need to specify transports. However,
6746 if such a router generated a host list, and callout was configured, Exim
6747 crashed, because it could not find a port number from the (non-existent)
6748 transport. It now assumes port 25 in this circumstance.
6750 21. Added the -t option to exigrep.
6752 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6753 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6754 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6755 them is set, LOOKUP_LSEARCH is forced.
6757 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6759 24. Added sender and host information to the "rejected by local_scan()" log
6760 line; previously there was no indication of these.
6762 25. Added .include_if_exists.
6764 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6765 Linux. It turns out that not all file systems support this. Apparently some
6766 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6767 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6768 not supported on the file descriptor, is now ignored when Exim is trying to
6769 sync a directory. This applies only to Linux.
6771 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6773 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6774 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6777 29. Added support for maildirsize files from supplied patch (modified a bit).
6779 30. The use of :fail: followed by an empty string could lead Exim to respond to
6780 sender verification failures with (e.g.):
6782 550 Verification failed for <xxx>
6783 550 Sender verify failed
6785 where the first response line was missing the '-' that indicates it is not
6786 the final line of the response.
6788 31. The loop for finding the name of the user that called Exim had a hardwired
6789 limit of 10; it now uses the value of finduser_retries, which is used for
6790 all other user lookups.
6792 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6795 33. Exim was neglecting to zero errno before one call of strtol() when
6796 expanding a string and expecting an integer value. On some systems this
6797 resulted in spurious "integer overflow" errors. Also, it was casting the
6798 result into an int without checking.
6800 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6801 did not work. The code looks as if it has *never* worked, though it appears
6802 to have been documented since at least release 1.62. I have made it work.
6804 35. The "timeout_DNS" error in retry rules, also documented since at least
6805 1.62, also never worked. As it isn't clear exactly what this means, and
6806 clearly it isn't a major issue, I have abolished the feature by treating it
6807 as "timeout", and writing a warning to the main and panic logs.
6809 36. The display of retry rules for -brt wasn't always showing the error code
6812 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6813 timeout_connect_A, timeout_connect_MX.
6815 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6816 to the empty sender.
6818 39. The daemon was not analysing the content of -oX till after it had closed
6819 stderr and disconnected from the controlling terminal. This meant that any
6820 syntax errors were only noted on the panic log, and the return code from
6821 the command was 0. By re-arranging the code a little, I've made the
6822 decoding happen first, so such errors now appear on stderr, and the return
6823 code is 1. However, the actual setting up of the sockets still happens in
6824 the disconnected process, so errors there are still only recorded on the
6827 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6828 connections (as happens on some IP stacks) was logged at start up time as
6829 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6830 it from "IPv6 and IPv4", which means that two separate sockets are being
6833 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6834 says whether AF_INET or AF_INET6 was passed as an argument.
6836 42. Exiwhat output was messed up when time zones were included in log
6839 43. Exiwhat now gives more information about the daemon's listening ports,
6840 and whether -tls-on-connect was used.
6842 44. The "port" option of the smtp transport is now expanded.
6844 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6845 silently ignored. Now an error message is written to the main and panic
6848 46. There's a new ACL modifier called "logwrite" which writes to a log file
6849 as soon as it is encountered.
6851 47. Added $local_user_uid and $local_user_gid at routing time.
6853 48. Exim crashed when trying to verify a sender address that was being
6856 49. Exim was recognizing only a space character after ".include". It now also
6857 recognizes a tab character.
6859 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6860 extracting the relevant information from the specification. The man page no
6861 longer contains scrambled data for the -d option, and I've added a section
6862 at the front about calling Exim under different names.
6864 51. Added "extra_headers" argument to the "mail" command in filter files.
6866 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6869 53. Installed eximstats 1.29.
6871 54. Added transport_filter_timeout as a generic transport option.
6873 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6874 Cc: header lines. This was required by RFC 822, but it not required by RFC
6877 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6878 incoming messages that did not have them. Now it does so only if the
6879 message originates locally, that is, if there is no associated remote host
6880 address. When Resent- header lines are present, this applies to the Resent-
6881 lines rather than the non-Resent- lines.
6883 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6884 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6886 58. Messages for configuration errors now include the name of the main
6887 configuration file - useful now that there may be more than one file in a
6888 list (.included file names were always shown).
6890 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6891 for those rare installations that do not start the daemon as root or run it
6892 setuid root. I've cut out the call to initgroups() if the daemon is not
6895 60. The Exim user and group can now be bound into the binary as text strings
6896 that are looked up at the start of Exim's processing.
6898 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6900 62. Added $mailstore_basename variable.
6902 63. Installed patch to sieve.c from Michael Haardt.
6904 64. When Exim failed to open the panic log after failing to open the main log,
6905 the original message it was trying to log was written to stderr and debug
6906 output, but if they were not available (the usual case in production), it
6907 was lost. Now it is written to syslog before the two lines that record the
6908 failures to open the logs.
6910 65. Users' Exim filters run in subprocesses under the user's uid. It is
6911 possible for a "deliver" command or an alias in a "personal" command to
6912 provoke an address rewrite. If logging of address rewriting is configured,
6913 this fails because the process is not running as root or exim. There may be
6914 a better way of dealing with this, but for the moment (because 4.30 needs
6915 to be released), I have disabled address rewrite logging when running a
6916 filter in a non-root, non-exim process.
6922 1. The buildconfig auxiliary program wasn't quoting the value set for
6923 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6924 not defined. This bug was masked in 4.22 by the effect that was fixed in
6927 2. Some messages that were rejected after a message id was allocated were
6928 shown as "incomplete" by exigrep. It no longer does this for messages that
6929 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6931 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6932 have allow_domain_literals set, the ID did not get logged in the <= line.
6933 Domain literals are now always recognized in Message-ID: header lines.
6935 4. The first argument for a ${extract expansion item is the key name or field
6936 number. Leading and trailing spaces in this item were not being ignored,
6937 causing some misleading effects.
6939 5. When deliver_drop_privilege was set, single queue runner processes started
6940 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6941 same command in the process it spins off) were not dropping privilege.
6943 6. When the daemon running as "exim" started a queue runner, it always
6944 re-executed Exim in the spun-off process. This is a waste of effort when
6945 deliver_drop_privilege is set. The new process now just calls the
6946 queue-runner function directly.
6952 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6955 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6956 ignored. Though the use of -oP was forcing the writing of a pid file, it
6957 was always written to the default place.
6959 3. If the message "no IP address found for host xxxx" is generated during
6960 incoming verification, it is now followed by identification of the incoming
6961 connection (so you can more easily find what provoked it).
6963 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6965 5. Added some features to "harden" Exim a bit more against certain attacks:
6967 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6968 be put in Local/Makefile. This is like the never_users runtime option,
6969 but it cannot be overridden. The default setting is "root".
6971 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6972 prefix string with which any file named in a -C command line option
6975 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6976 is retained for -C and -D only if the caller of Exim is root. Without
6977 it, the exim user may also use -C and -D and retain privilege.
6979 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6980 command line option is disabled.
6982 6. Macro names set by the -D option must start with an upper case letter, just
6983 like macro names defined in the configuration file.
6985 7. Added "dereference=" facility to LDAP.
6987 8. Two instances of the typo "uknown" in the source files are fixed.
6989 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6990 the Configure-Makefile script screwed up while processing it.
6992 10. Incorporated PCRE 4.4.
6994 11. The SMTP synchronization check was not operating right at the start of an
6995 SMTP session. For example, it could not catch a HELO sent before the client
6996 waited for the greeting. There is now a check for outstanding input at the
6997 point when the greeting is written. Because of the duplex, asynchronous
6998 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6999 way, but not yet received, when the check is performed.
7001 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7002 on TCP/IP sockets, because this apparently causes some broken clients to
7005 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7006 unchanged) from the Cygwin maintainer.
7008 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7009 was supported instead of testing for mbx. Effectively a typo.
7011 15. The spa authenticator server code was not checking that the input it
7012 received was valid base64.
7014 16. The debug output line for the "set" modifier in ACLs was not showing the
7015 name of the variable that was being set.
7017 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7019 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7020 lookup on its own. Something else was needed to trigger the lookup. For
7021 example, a match in host_lookup or the need for a host name in a host list.
7022 Now, if $sender_host_name is referenced and the host name has not yet been
7023 looked up, a lookup is performed. If the lookup fails, the variable remains
7024 empty, and $host_lookup_failed is set to "1".
7026 19. Added "eqi" as a case-independent comparison operator.
7028 20. The saslauthd authentication condition could segfault if neither service
7029 nor realm was specified.
7031 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7032 error message that was logged was misleading, and incoming SMTP
7033 connections were dropped. The message is now more accurate, and temporary
7034 errors are given to SMTP connections.
7036 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7037 (except RSET) with a 421 error, until QUIT is received. However, it was
7038 failing to send a response to QUIT.
7040 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7041 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7042 ACL is now run after the helo_try_verify_hosts code.
7044 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7045 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7046 case-independent, so other case variants are also recognized). Apparently
7047 some systems use these upper case variants.
7049 25. If more than two messages were waiting for the same host, and a transport
7050 filter was specified for the transport, Exim sent two messages over the
7051 same TCP/IP connection, and then failed with "socket operation on non-
7052 socket" when it tried to send the third.
7054 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7056 27. The extern definition of crypt16() in expand.c was not being excluded when
7057 the OS had its own crypt16() function.
7059 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7060 as a preferred synonym for return_size_limit, both as an option and as an
7063 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7065 30. Changed the default configuration ACL to relax the local part checking rule
7066 for addresses that are not in any local domains. For these addresses,
7067 slashes and pipe symbols are allowed within local parts, but the sequence
7068 /../ is explicitly forbidden.
7070 31. SPA server authentication was not clearing the challenge buffer before
7073 32. log_message in a "warn" ACL statement was writing to the reject log as
7074 well as to the main log, which contradicts the documentation and doesn't
7075 seem right (because no rejection is happening). So I have stopped it.
7077 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7078 However, I am unable to do any testing of this.
7080 34. Fixed an infelicity in the appendfile transport. When checking directories
7081 for a mailbox, to see if any needed to be created, it was accidentally
7082 using path names with one or more superfluous leading slashes; tracing
7083 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7085 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7086 contents are added to the log line that is written for every discarded
7087 recipient. (Previously a log_message setting was ignored.)
7089 36. The ${quote: operator now quotes the string if it is empty.
7091 37. The install script runs exim in order to find its version number. If for
7092 some reason other than non-existence or emptiness, which it checks, it
7093 could not run './exim', it was installing it with an empty version number,
7094 i.e. as "exim-". This error state is now caught, and the installation is
7097 38. An argument was missing from the function that creates an error message
7098 when Exim fails to connect to the socket for saslauthd authentication.
7099 This could cause Exim to crash, or give a corrupted message.
7101 39. Added isip, isip4, and isip6 to ${if conditions.
7103 40. The ACL variables $acl_xx are now saved with the message, and can be
7104 accessed later in routers, transports, and filters.
7106 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7107 strings in the file are not string-expanded.
7109 42. If a MAIL command specified a SIZE value that was too large to fit into an
7110 int variable, the check against message_size_limit failed. Such values are
7111 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7112 day this will have to be increased, but I don't think I want to be around
7113 when emails are that large.
7120 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7121 iconv() is not standard in FreeBSD.
7123 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7124 IPv6 enabled. The observed symptom was a segmentation fault on return from
7125 the function os_common_find_running_interfaces() in src/os.c.
7127 3. In the check_special_case() function in daemon.c I had used "errno" as an
7128 argument name, which causes warnings on some systems. This was basically a
7129 typo, since it was named "eno" in the comments!
7131 4. The code that waits for the clock to tick (at a resolution of some fraction
7132 of a second) so as to ensure message-id uniqueness was always waiting for
7133 at least one whole tick, when it could have waited for less. [This is
7134 almost certainly not relevant at current processor speeds, where it is
7135 unlikely to ever wait at all. But we try to future-proof.]
7137 5. The function that sleeps for a time interval that includes fractions of a
7138 second contained a race. It did not block SIGALRM between setting the
7139 timer, and suspending (a couple of lines later). If the interval was short
7140 and the sigsuspend() was delayed until after it had expired, the suspension
7141 never ended. On busy systems this could lead to processes getting stuck for
7144 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7145 process, before it forks any delivery processes. The open lookup caching
7146 mechanism meant that the open file or database connection was passed into
7147 the delivery process. The problem was that delivery processes always tidy
7148 up cached lookup data. This could cause a problem for the next delivery
7149 process started by the queue runner, because the external queue runner
7150 process does not know about the closure. So the next delivery process
7151 still has data in the lookup cache. In the case of a file lookup, there was
7152 no problem because closing a file descriptor in a subprocess doesn't affect
7153 the parent. However, if the lookup was caching a connection to a database,
7154 the connection was closed, and the second delivery process was likely to
7155 see errors such as "PGSQL: query failed: server closed the connection
7156 unexpectedly". The problem has been fixed by closing all cached lookups
7157 in a queue runner before running a delivery process.
7159 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7160 seem to have the "const" qualifier which it has on other OS. I've
7163 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7164 *of the same type* (client or server) with the same public name that an
7165 error should be diagnosed.
7167 9. When Exim looked up a host name for an IP address, but failed to find the
7168 original IP address when looking up the host name (a safety check), it
7169 output the message "<ip address> does not match any IP for NULL", which was
7170 confusing, to say the least. The bug was that the host name should have
7171 appeared instead of "NULL".
7173 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7174 user that is built into the binary, and the -C or -D options is used, root
7175 privilege is dropped before the configuration file is read. In addition,
7176 logging is switched to stderr instead of the normal log files. If the
7177 configuration then re-defines the Exim user, the unprivileged environment
7178 is probably not what is expected, so Exim logs a panic warning message (but
7181 However, if deliver_drop_privilege is set, the unprivileged state may well
7182 be exactly what is intended, so the warning has been cut out in that case,
7183 and Exim is allowed to try to write to its normal log files.
7189 1. smtp_return_error_details was not giving details for temporary sender
7190 or receiver verification errors.
7192 2. Diagnose a configuration error if two authenticators have the same public
7195 3. Exim used not to create the message log file for a message until the first
7196 delivery attempt. This could be confusing when incoming messages were held
7197 for policy or load reasons. The message log file is now created at the time
7198 the message is received, and an initial "Received" line is written to it.
7200 4. The automatically generated man page for command line options had a minor
7201 bug that caused no ill effects; however, a more serious problem was that
7202 the procedure for building the man page automatically didn't always
7203 operate. Consequently, release 4.20 contains an out-of-date version. This
7204 shouldn't happen again.
7206 5. When building Exim with embedded Perl support, the script that builds the
7207 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7208 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7210 6. The freeze_tell option was not being used for messages that were frozen on
7211 arrival, either by an ACL or by local_scan().
7213 7. Added the smtp_incomplete_transaction log selector.
7215 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7216 was accepting AUTH without a new EHLO.
7218 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7219 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7222 10. Logging of TCP/IP connections (when configured) now happens in the main
7223 daemon process instead of the child process, so that the TCP/IP connection
7224 count is more accurate (but it can never be perfect).
7226 11. The use of "drop" in a nested ACL was not being handled correctly in the
7227 outer ACL. Now, if condition failure induced by the nested "drop" causes
7228 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7229 or "require"), the connection is dropped.
7231 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7232 that yield "discard" can now be used with an "accept" or a "discard" verb,
7233 but an error is generated for any others (because I can't see a useful way
7234 to define what should happen).
7236 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7237 are now processed in the same way as lines in the Exim configuration file.
7238 In particular, continuation lines are supported.
7240 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7242 15. Added -ti meaning -t -i.
7244 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7245 domains, and warn by logging if others are found.
7247 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7248 returned from GIFCONF when Exim is trying to find the list of interfaces on
7249 a host. The code in os.c has been modified to copy each ifreq to an aligned
7250 structure in all cases.
7252 Also, in some cases, the returned ifreq's were being copied to a 'struct
7253 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7254 means the last couple of bytes of an IPv6 address could be chopped if the
7255 ifreq contained only a normal sockaddr (14 bytes storage).
7257 18. Named domain lists were not supported in the hosts_treat_as_local option.
7258 An entry such as +xxxx was not recognized, and was treated as a literal
7261 19. Ensure that header lines added by a DATA ACL are included in the reject log
7262 if the ACL subsequently rejects the message.
7264 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7265 MD5 (which is deprecated).
7267 21. When testing a filter file using -bf, Exim was writing a message when it
7268 took the sender from a "From " line in the message, but it was not doing so
7269 when it took $return_path from a Return-Path: header line. It now does.
7271 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7272 with a valid header line field name (a series of printing characters
7273 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7275 23. Changed "disc" in the source to "disk" to conform to the documentation and
7276 the book and for uniformity.
7278 24. Ignore Sendmail's -Ooption=value command line item.
7280 25. When execve() failed while trying to run a command in a pipe transport,
7281 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7282 could be confused with a return value of 69 from the command itself. This
7283 has been changed to 127, the value the shell returns if it is asked to run
7284 a non-existent command. The wording for the related log line suggests a
7285 non-existent command as the problem.
7287 26. If received_header_text expands to an empty string, do not add a Received:
7288 header line to the message. (Well, it adds a token one on the spool, but
7289 marks it "old" so that it doesn't get used or transmitted.)
7291 27. Installed eximstats 1.28 (addition of -nt option).
7293 28. There was no check for failure on the call to getsockname() in the daemon
7294 code. This can fail if there is a shortage of resources on the system, with
7295 ENOMEM, for example. A temporary error is now given on failure.
7297 29. Contrary to the C standard, it seems that in some environments, the
7298 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7299 program. Exim now does this explicitly; it affects the formatting of
7300 timestamps using strftime().
7302 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7303 complaints. I've now initialized all the variables, to avoid this.
7305 32. Header lines added by a system filter were not being "seen" during
7306 transport-time rewrites.
7308 33. The info_callback() function passed to OpenSSL is set up with type void
7309 (*)(SSL *, int, int), as described somewhere. However, when calling the
7310 function (actually a macro) that sets it up, the type void(*)() is
7311 expected. I've put in a cast to prevent warnings from picky compilers.
7313 34. If a DNS black list lookup found a CNAME record, but there were no A
7314 records associated with the domain it pointed at, Exim crashed.
7316 35. If a DNS black list lookup returned more than one A record, Exim ignored
7317 all but the first. It now scans all returned addresses if a particular IP
7318 value is being sought. In this situation, the contents of the
7319 $dnslist_value variable are a list of all the addresses, separated by a
7322 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7323 to accept a host name and all its aliases if the forward lookup for any of
7324 them yielded the IP address of the incoming connection. Now it accepts only
7325 those names whose forward lookup yields the correct IP address. Any other
7326 names are discarded. This closes a loophole whereby a rogue DNS
7327 administrator could create reverse DNS records to break through a
7328 wildcarded host restriction in an ACL.
7330 37. If a user filter or a system filter that ran in a subprocess used any of
7331 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7332 the wrong values were passed to the pipe command ($thisaddress had the
7333 value of $0, $0 had the value of $1, etc). This bug was introduced by
7334 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7336 38. Improved the line breaking for long SMTP error messages from ACLs.
7337 Previously, if there was no break point between 40 and 75 characters, Exim
7338 left the rest of the message alone. Two changes have been made: (a) I've
7339 reduced the minimum length to 35 characters; (b) if it can't find a break
7340 point between 35 and 75 characters, it looks ahead and uses the first one
7341 that it finds. This may give the occasional overlong line, but at least the
7342 remaining text gets split now.
7344 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7345 file descriptors might be low, and that setting 1000 would always raise it.
7346 It turns out that in some environments, the limit is already over 1000 and
7347 that lowering it causes trouble. So now Exim takes care not to decrease it.
7349 40. When delivering a message, the value of $return_path is set to $sender_
7350 address at the start of routing (routers may change the value). By an
7351 oversight, this default was not being set up when an address was tested by
7352 -bt or -bv, which affected the outcome if any router or filter referred to
7355 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7356 log only when "message" or "log_message" are set. However, if one of the
7357 conditions was an address verification, or a call to a nested ACL, the
7358 messages generated by the underlying test were being passed through. This
7359 no longer happens. The underlying message is available in $acl_verify_
7360 message for both "message" and "log_message" expansions, so it can be
7361 passed through if needed.
7363 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7364 new expansion $bh_ to give the encoded byte string without charset
7365 translation. Translation happens only if iconv() is available; HAVE_ICONV
7366 indicates this at build time. HEADERS_CHARSET gives the charset to
7367 translate to; headers_charset can change it in the configuration, and
7368 "headers charset" can change it in an individual filter file.
7370 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7371 that creates RFC 2047 encoded "words" labels them as that charset instead
7372 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7373 expansion operator; (ii) when Exim creates a From: line for a local
7374 message; (iii) when a header line is rewritten to include a "phrase" part.
7376 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7377 buggy, causing it to skip the first lines of messages whose message ID
7378 ended in 'D'. This would not have bitten before Exim release 4.14, because
7379 message IDs were unlikely to end in 'D' before then. The effect was to have
7380 incorrect size information for certain domains.
7382 45. #include "config.h" was missing at the start of the crypt16.c module. This
7383 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7386 46. If there was a timeout during a "random" callout check, Exim treated it as
7387 a failure of the random address, and carried on sending RSET and the real
7388 address. If the delay was just some slowness somewhere, the response to the
7389 original RCPT would be taken as a response to RSET and so on, causing
7390 mayhem of various kinds.
7392 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7393 when I implemented it. It didn't allow for the fact that some option values
7394 may legitimately be negative (e.g. size_addition), and it didn't even do
7395 the right test for positive values.
7397 48. Domain names in DNS records are case-independent. Exim always looks them up
7398 in lower case. Some resolvers return domain names in exactly the case they
7399 appear in the zone file, that is, they may contain uppercase letters. Not
7400 all resolvers do this - some return always lower case. Exim was treating a
7401 change of case by a resolver as a change of domain, similar to a widening
7402 of a domain abbreviation. This triggered its re-routing code and so it was
7403 trying to route what was effectively the same domain again. This normally
7404 caused routing to fail (because the router wouldn't handle the domain
7405 twice). Now Exim checks for this case specially, and just changes the
7406 casing of the domain that it ultimately uses when it transmits the message
7409 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7412 50. If a filter generated a file delivery with a non-absolute name (possible if
7413 no home directory exists for the router), the forbid_file option was not
7416 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7417 the existing equality matching.
7419 52. Exim was using ints instead of ino_t variables in some places where it was
7420 dealing with inode numbers.
7422 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7423 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7424 called TMPDIR, and if it finds it is different, it changes its value.
7426 54. The smtp_printf() function is now made available to local_scan() so
7427 additional output lines can be written before returning. There is also an
7428 smtp_fflush() function to enable the detection of a dropped connection.
7429 The variables smtp_input and smtp_batched_input are exported to
7432 55. Changed the default runtime configuration: the message "Unknown user"
7433 has been removed from the ACL, and instead placed on the localuser router,
7434 using the cannot_route_message feature. This means that any verification
7435 failures that generate their own messages won't get overridden. Similarly,
7436 the "Unrouteable address" message that was in the ACL for unverifiable
7437 relay addresses has also been removed.
7439 56. Added hosts_avoid_esmtp to the smtp transport.
7441 57. The exicyclog script was not checking for the esoteric option
7442 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7443 will work only if exicyclog is run under the appropriate euid.
7445 58. Following a discussion on the list, the rules by which Exim recognises line
7446 endings on incoming messages have been changed. The -dropcr and drop_cr
7447 options are now no-ops, retained only for backwards compatibility. The
7448 following line terminators are recognized: LF CRLF CR. However, special
7449 processing applies to CR:
7451 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7452 nor a local message in the state where . is a terminator.
7454 (ii) If a bare CR is encountered in a header line, an extra space is added
7455 after the line terminator so as not to end the header. The reasoning
7456 behind this is that bare CRs in header lines are most likely either
7457 to be mistakes, or people trying to play silly games.
7459 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7460 was being incorrectly given as 18 bytes larger than it should have been.
7461 This is a VOB (very old bug).
7463 60. This may never have affected anything current, but just in case it has:
7464 When the local host is found other than at the start of a list of hosts,
7465 the local host, those with the same MX, and any that follow, are discarded.
7466 When the list in question was part of a longer list of hosts, the following
7467 hosts (not currently being processed) were also being discarded. This no
7468 longer happens. I'm not sure if this situation could ever has previously
7471 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7474 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7475 header. This is something that is recommended in a new Internet Draft, and
7476 is something that is documented as being done by Sendmail. There are two
7477 possible values. For messages generated by the autoreply transport, Exim
7480 Auto-Submitted: auto-replied
7482 whereas for all other generated messages (e.g. bounces) it adds
7484 Auto-Submitted: auto-generated
7486 63. The "personal" condition in filters now includes a test for the
7487 Auto-Submitted: header. If it contains the string "auto-" the message it
7488 not considered personal.
7490 64. Added rcpt_include_affixes as a generic transport option.
7492 65. Added queue_only_override (default true).
7494 66. Added the syslog_duplication option.
7496 67. If what should have been the first header line of a message consisted of
7497 a space followed by a colon, Exim was mis-interpreting it as a header line.
7498 It isn't of course - it is syntactically invalid and should therefore be
7499 treated as the start of the message body. The misbehaviour could have
7500 caused a number of strange effects, including loss of data in subsequent
7501 header lines, and spool format errors.
7503 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7504 client host had authenticated. This control can now be exercised by an ACL
7505 for more flexibility.
7507 69. By default, callouts do not happen when testing with -bh. There is now a
7508 variant, -bhc, which does actually run the callout code, including
7509 consulting and updating the callout cache.
7511 70. Added support for saslauthd authentication, courtesy of Alexander
7514 71. If statvfs() failed on the spool or log directories while checking their
7515 size for availability, Exim confusingly gave the error "space shortage".
7516 Furthermore, in debugging mode it crashed with a floating point exception.
7517 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7518 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7519 serious problem, Exim now writes to the main and panic logs when this
7520 happens, with details of the failure. It then refuses to accept the
7521 incoming message, giving the message "spool directory problem" or "log
7522 directory problem" with a 421 code for SMTP messages.
7524 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7525 0, 1, and 2 exist, because some OS complain for execs without them (see
7526 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7527 descriptors. However, the code omitted to check that the open succeeded,
7528 causing mysterious errors if for some reason the permissions on /dev/null
7529 got screwed. Now Exim writes a message to the main and panic logs, and
7530 bombs out if it can't open /dev/null.
7532 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7533 interact so that it is all more flexible. It is supposed to remain
7534 backwards compatible. Also added extra_local_interfaces.
7536 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7537 to bomb out with an assertion failure - to the client this appears as a
7538 connection drop. This problem occurs in the part of the code that was taken
7539 from the Samba project. Fortunately, the assertion is in a very simple
7540 function, so I have fixed this by reproducing the function inline in the
7541 one place where it is called, and arranging for authentication to fail
7542 instead of killing the process with assert().
7544 75. The SPA client code was not working when the server requested OEM rather
7545 than Unicode encoding.
7547 76. Added code to make require_files with a specific uid setting more usable in
7548 the case where statting the file as root fails - usually a non-root-mounted
7549 NFS file system. When this happens and the failure is EACCES, Exim now
7550 forks a subprocess and does the per-uid checking as the relevant uid.
7552 77. Added process_log_path.
7554 78. If log_file_path was not explicitly set, a setting of check_log_space or
7555 check_log_inodes was ignored.
7557 79. If a space check for the spool or log partitions fails, the incident is now
7558 logged. Of course, in the latter case the data may get lost...
7560 80. Added the %p formatting code to string_format() so that it can be used to
7561 print addresses in debug_print(). Adjusted all the address printing in the
7562 debugging in store.c to use %p rather than %d.
7564 81. There was a concern that a line of code in smtp_in.c could overflow a
7565 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7566 initially expressed, the concern was not well-founded, because trailing
7567 spaces are removed early. However, if the trailing spaces were followed by
7568 a NULL, they did not get removed, so the overflow was possible. Two fixes
7571 (a) I re-wrote the offending code in a cleaner fashion.
7572 (b) If an incoming SMTP command contains a NULL character, it is rejected
7575 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7576 runs initgroups(), so that if the Exim user is in any additional groups,
7577 they will be used during message reception.
7583 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob