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.
224 JH/01 Remove code calling the customisable local_scan function, unless a new
225 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
227 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
228 non-signal-safe functions being used.
230 JH/03 Bug 2269: When presented with a received message having a stupidly large
231 number of DKIM-Signature headers, disable DKIM verification to avoid
232 a resource-consumption attack. The limit is set at twenty.
234 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
235 report of oldest_pass in ${authres } in consequence, and separate out
236 some descriptions of reasons for verification fail.
238 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
239 files in the spool were present and unlocked. A queue-runner could spot
240 them, resulting in a duplicate delivery. Fix that by doing the unlock
241 after the unlink. Investigation by Tim Stewart. Take the opportunity to
242 add more error-checking on spoolfile handling while that code is being
245 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
246 No known attacks, no CVE, this is defensive hardening.
248 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
249 a queue-runner could start a delivery while other operations were ongoing.
250 Cutthrough delivery was a common victim, resulting in duplicate delivery.
251 Found and investigated by Tim Stewart. Fix by using the open message data
252 file handle rather than opening another, and not locally closing it (which
253 releases a lock) for that case, while creating the temporary .eml format
254 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
256 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
257 $sender_verify_failure/$recipient_verify_failure to "random".
259 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
262 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
263 Previously this would segfault.
265 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
268 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
269 like zero, since the resolver should be doing this for us, But we need one
270 as a CNAME but no MX presence gets the CNAME returned; we need to check
271 that doesn't point to an MX to declare it "no MX returned" rather than
272 "error, loop". A new main option is added so the older capability of
273 following some limited number of chain links is maintained.
275 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
277 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
278 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
279 now-deprecated earlier definitions used only specified the range up to TLS
280 1.2 (in the older-version library docs).
282 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
284 JH/15 Rework TLS client-side context management. Stop using a global, and
285 explicitly pass a context around. This enables future use of TLS for
286 connections to service-daemons (eg. malware scanning) while a client smtp
287 connection is using TLS; with cutthrough connections this is quite likely.
289 JH/16 Fix ARC verification to do AS checks in reverse order.
291 JH/17 Support a "tls" option on the ${readsocket } expansion item.
293 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
294 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
295 Previously the "utf8" would be re-prepended for every additional message.
297 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
298 Previously thery were accepted, resulting in issues when attempting to
299 forward messages to a non-supporting MTA.
301 PP/02 Let -n work with printing macros too, not just options.
303 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
304 one parent address was copied, and bogus data was used at delivery-logging
305 time. Either a crash (after delivery) or bogus log data could result.
306 Discovery and analysis by Tim Stewart.
308 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
309 Previously if the string ended mid-character, we did not insert the
310 promised '?' replacement.
312 PP/04 Documentation: current string operators work on bytes, not codepoints.
314 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
315 should pack well giving a smaller memory footprint so better caching and
316 therefore performance. Group the declarations where this can't be done so
317 that the byte-sized flag variables are not interspersed among pointer
318 variables, giving a better chance of good packing by the compiler.
320 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
321 non-null, to avoid issues with sites running BATV. Previously reports were
322 sent with an empty envelope sender so looked like bounces.
324 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
325 The ignore_error flag wasn't being returned from the filter subprocess so
326 was not set for later routers. Investigation and fix by Matthias Kurz.
328 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
329 and a msg:complete for the whole, when a message is manually removed using
330 -Mrm. Developement by Matthias Kurz, hacked on by JH.
332 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
333 a "Gnu special" function, asprintf() in the DB utility binary builds; I
334 hope that is portable enough.
336 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
337 requiring a known-CA anchor certificate; make it now rely entirely on the
338 TLSA as an anchor. Checking the name on the leaf cert against the name
339 on the A-record for the host is still done for TA (but not for EE mode).
341 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
342 included in delivery lines for non-proxied connections, when compiled with
343 SUPPORT_SOCKS and running with proxy logging enabled.
345 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
346 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
347 move the existing event to fire before the normal logging of message
348 failure so that custom logging is bracketed by normal logging.
350 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
351 msg:fail:internal event. Developement by Matthias Kurz.
353 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
354 far too small for todays use of crypto signatures stored there. Go all
355 the way to the max DNS message size of 64kB, even though this might be
356 overmuch for IOT constrained device use.
358 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
359 copy a string over itself. The library routine is documented as not
360 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
362 JH/32 For main options check_spool_space and check_inode_space, where the
363 platform supports 64b integers, support more than the previous 2^31 kB
364 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
365 the previous G, M, k.
367 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
368 $authenticated_fail_id variable on authentication failure. Previously
371 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
372 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
373 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
374 GNUTLS_SEC_PARAM_MEDIUM.
376 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
377 side. Previously we would continue as if no SNI had been received.
379 JH/36 Harden the handling of string-lists. When a list consisted of a sole
380 "<" character, which should be a list-separator specification, we walked
381 off past the nul-terimation.
383 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
384 causes) even when the retry time is not yet met. Previously they were
385 not, meaning that when (say) an account was over-quota and temp-rejecting,
386 and multiple senders' messages were queued, only one sender would get
387 notified on each configured delay_warning cycle.
389 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
391 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
392 cipher-suites, an error can be left on the stack even for a succeeding
393 accept; this results in impossible error messages when a later operation
396 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
397 return error codes indicating retry. Under TLS1.3 this becomes required.
399 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
400 it only wrote the new authenticators, resulting in a lack of tracking of
401 peer changes of ESMTP extensions until the next cache flush.
403 JH/41 Fix the loop reading a message header line to check for integer overflow,
404 and more-often against header_maxsize. Previously a crafted message could
405 induce a crash of the recive process; now the message is cleanly rejected.
407 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
408 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
414 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
415 When redis_servers is set to a list of > 1 element, and the Redis servers
416 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
417 case of MOVED into a DEFER case instead, thus moving the query onto the
418 next server in the list. For a cluster of N elements, all N servers must
419 be defined in redis_servers.
421 GF/02 Catch and remove uninitialized value warning in exiqsumm
422 Check for existence of @ARGV before looking at $ARGV[0]
424 JH/01 Replace the store_release() internal interface with store_newblock(),
425 which internalises the check required to safely use the old one, plus
426 the allocate and data copy operations duplicated in both (!) of the
427 extant use locations.
429 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
430 modifier. This matches the restriction on the commandline.
432 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
433 Previously only the last row was returned.
435 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
436 we assumed that tags in the header were well-formed, and parsed the
437 element content after inspecting only the first char of the tag.
438 Assumptions at that stage could crash the receive process on malformed
441 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
442 While running the DKIM ACL we operate on the Permanent memory pool so that
443 variables created with "set" persist to the DATA ACL. Also (at any time)
444 DNS lookups that fail create cache records using the Permanent pool. But
445 expansions release any allocations made on the current pool - so a dnsdb
446 lookup expansion done in the DKIM ACL releases the memory used for the
447 DNS negative-cache, and bad things result. Solution is to switch to the
448 Main pool for expansions.
449 While we're in that code, add checks on the DNS cache during store_reset,
450 active in the testsuite.
451 Problem spotted, and debugging aided, by Wolfgang Breyha.
453 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
454 When none of the hosts presented to a transport match an already-open
455 connection, close it and proceed with the list. Previously we would
456 queue the message. Spotted by Lena with Yahoo, probably involving
459 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
460 Previously a spurious "250 OK id=" response was appended to the proper
463 JH/08 The "support for" informational output now, which built with Content
464 Scanning support, has a line for the malware scanner interfaces compiled
465 in. Interface can be individually included or not at build time.
467 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
468 by the template makefile "src/EDITME". The "STREAM" support for an older
469 ClamAV interface method is removed.
471 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
472 rows affected is given instead).
474 JH/11 The runtime Berkeley DB library version is now additionally output by
475 "exim -d -bV". Previously only the compile-time version was shown.
477 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
478 SMTP connection. Previously, when one had more receipients than the
479 first, an abortive onward connection was made. Move to full support for
480 multiple onward connections in sequence, handling cutthrough connection
481 for all multi-message initiating connections.
483 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
484 routers. Previously, a multi-recipient message would fail to match the
485 onward-connection opened for the first recipient, and cause its closure.
487 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
488 a timeout on read on a GnuTLS initiating connection, resulting in the
489 initiating connection being dropped. This mattered most when the callout
490 was marked defer_ok. Fix to keep the two timeout-detection methods
493 JH/15 Relax results from ACL control request to enable cutthrough, in
494 unsupported situations, from error to silently (except under debug)
495 ignoring. This covers use with PRDR, frozen messages, queue-only and
498 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
500 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
501 metadata, resulting in a crash in free().
503 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
504 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
505 Broken also in d185889f4, with init system revamp.
507 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
508 Previously we abruptly closed the connection after reading a malware-
509 found indication; now we go on to read the "scan ok" response line,
512 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
513 ACL. Previously, a crash would result.
515 JH/19 Speed up macro lookups during configuration file read, by skipping non-
516 macro text after a replacement (previously it was only once per line) and
517 by skipping builtin macros when searching for an uppercase lead character.
519 JH/20 DANE support moved from Experimental to mainline. The Makefile control
520 for the build is renamed.
522 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
523 was allocated for every new TLS startup, meaning one per message. Fix
524 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
526 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
527 reported the original. Fix to report (as far as possible) the ACL
528 result replacing the original.
530 JH/23 Fix memory leak during multi-message connections using STARTTLS under
531 OpenSSL. Certificate information is loaded for every new TLS startup,
532 and the resources needed to be freed.
534 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
536 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
539 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
540 DATA response info to the (existing) per-recipient response info for
541 the "C=" log element. It can have useful tracking info from the
542 destination system. Patch from Simon Arlott.
544 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
545 length value. Previously this would segfault.
547 HS/02 Support Avast multiline protoocol, this allows passing flags to
548 newer versions of the scanner.
550 JH/28 Ensure that variables possibly set during message acceptance are marked
551 dead before release of memory in the daemon loop. This stops complaints
552 about them when the debug_store option is enabled. Discovered specifically
553 for sender_rate_period, but applies to a whole set of variables.
554 Do the same for the queue-runner and queue-list loops, for variables set
555 from spool message files. Do the same for the SMTP per-message loop, for
556 certain variables indirectly set in ACL operations.
558 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
559 as a multi-recipient message from a mailinglist manager). The coding had
560 an arbitrary cutoff number of characters while checking for more input;
561 enforced by writing a NUL into the buffer. This corrupted long / fast
562 input. The problem was exposed more widely when more pipelineing of SMTP
563 responses was introduced, and one Exim system was feeding another.
564 The symptom is log complaints of SMTP syntax error (NUL chars) on the
565 receiving system, and refused recipients seen by the sending system
566 (propating to people being dropped from mailing lists).
567 Discovered and pinpointed by David Carter.
569 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
570 replaced by the ${authresults } expansion.
572 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
574 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
575 allows proper process termination in container environments.
577 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
578 Previously the "final dot" had a newline after it; ensure it is CR,LF.
580 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
581 and "err_perm", deprecated since 4.83 when the RFC-defined words
582 "temperror" and "permerror" were introduced.
584 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
585 transport-filters or DKIM-signing. The restriction was lost in the
586 consolidation of verify-callout and delivery SMTP handling.
587 Extend the restriction to also cover ARC-signing.
589 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
590 in defer=pass mode supply a 450 to the initiator. Previously the message
593 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
594 tls_require_ciphers is used as before.
596 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
597 "pass_unscanned". Only tmpfails from the scanner are written to
598 the paniclog, as they may require admin intervention (permission
599 denied, license issues). Other scanner errors (like decompression
600 bombs) do not cause a paniclog entry.
602 JH/36 Fix reinitialisation of DKIM logging variable between messages.
603 Previously it was possible to log spurious information in receive log
606 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
607 triggered odd behaviour from Outlook Express clients.
609 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
612 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
613 since the IETF WG has not yet settled on that versus the original
614 "bare" representation.
616 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
617 Previously the millisecond value corrupted the output.
618 Fix also for syslog_pid=no and log_selector +pid, for which the pid
619 corrupted the output.
625 JH/01 Rework error string handling in TLS interface so that the caller in
626 more cases is responsible for logging. This permits library-sourced
627 string to be attached to addresses during delivery, and collapses
628 pairs of long lines into single ones.
630 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
631 during configuration. Wildcards are allowed and expanded.
633 JH/02 Rework error string handling in DKIM to pass more info back to callers.
634 This permits better logging.
636 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
637 do not close it down and have the child transport start it up again on
638 the passed-on TCP connection. Instead, proxy the child (and any
639 subsequent ones) for TLS via a unix-domain socket channel. Logging is
640 affected: the continued delivery log lines do not have any DNSSEC, TLS
641 Certificate or OCSP information. TLS cipher information is still logged.
643 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
644 identical IP addresses on different listening ports. Will also affect
647 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
648 add noisy ifdef guards to special-case this sillyness.
649 Patch from Bernd Kuhls.
651 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
652 than 255 are no longer allowed.
654 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
655 Disable the session-cache too, which might reduce our load. Since we
656 currrectly use a new context for every connection, both as server and
657 client, there is no benefit for these.
658 GnuTLS appears to not support tickets server-side by default (we don't
659 call gnutls_session_ticket_enable_server()) but client side is enabled
660 by default on recent versions (3.1.3 +) unless the PFS priority string
663 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
664 <https://reproducible-builds.org/specs/source-date-epoch/>.
666 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
667 the check for any unsuccessful recipients did not notice the limit, and
668 erroneously found still-pending ones.
670 JH/08 Pipeline CHUNKING command and data together, on kernels that support
671 MSG_MORE. Only in-clear (not on TLS connections).
673 JH/09 Avoid using a temporary file during transport using dkim. Unless a
674 transport-filter is involved we can buffer the headers in memory for
675 creating the signature, and read the spool data file once for the
676 signature and again for transmission.
678 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
679 4.77 as the kernel support then wasn't solid, having issues in 64bit
680 mode. Now, it's been long enough. Add support for FreeBSD also.
682 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
683 case where the routing stage had gathered several addresses to send to
684 a host before calling the transport for the first, we previously failed
685 to close down TLS in the old transport process before passing the TCP
686 connection to the new process. The new one sent a STARTTLS command
687 which naturally failed, giving a failed delivery and bloating the retry
688 database. Investigation and fix prototype from Wolfgang Breyha.
690 JH/12 Fix check on SMTP command input synchronisation. Previously there were
691 false-negatives in the check that the sender had not preempted a response
692 or prompt from Exim (running as a server), due to that code's lack of
693 awareness of the SMTP input buffering.
695 PP/04 Add commandline_checks_require_admin option.
696 Exim drops privileges sanely, various checks such as -be aren't a
697 security problem, as long as you trust local users with access to their
698 own account. When invoked by services which pass untrusted data to
699 Exim, this might be an issue. Set this option in main configuration
700 AND make fixes to the calling application, such as using `--` to stop
703 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
704 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
705 responses to those, into a single TLS record each way (this usually means
706 a single packet). As a side issue, smtp_enforce_sync now works on TLS
709 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
710 affects you only if you're dancing at the edge of the param size limits.
711 If you are, and this message makes sense to you, then: raise the
712 configured limit or use OpenSSL 1.1. Nothing we can do for older
715 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
716 cmdline element to get the documented default one. Previously it was
719 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
720 are unsuable for non-message-specific reasons - eg. connection timeout,
721 banner-time rejection.
723 JH/16 Fix logging of delivery remote port, when specified by router, under
726 PP/06 Repair manualroute's ability to take options in any order, even if one
727 is the name of a transport.
730 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
732 JH/17 Change the list-building routines interface to use the expanding-string
733 triplet model, for better allocation and copying behaviour.
735 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
736 Previously it was constructed the first time a possibly-matching string
737 was met in the configuration file input during startup; now it is done
740 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
741 compatible one, to avoid the (poorly documented) possibility of a config
742 file in the working directory redirecting the DB files, possibly correpting
743 some existing file. CVE-2017-10140 assigned for BDB.
745 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
746 cache-hot. Previously, although the result was properly cached, the
747 initial verify call returned a defer.
749 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
750 the main verify for receipient in uncached-mode.
752 JH/22 Retire historical build files to an "unsupported" subdir. These are
753 defined as "ones for which we have no current evidence of testing".
755 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
756 if present. Previously it was ignored.
758 JH/24 Start using specified-initialisers in C structure init coding. This is
759 a C99 feature (it's 2017, so now considered safe).
761 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
762 if was a fixed-sized field and bitmask ops via macros; it is now more
765 PP/07 GitHub PR 56: Apply MariaDB build fix.
766 Patch provided by Jaroslav Škarvada.
768 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
769 during Coverity cleanups [4.87 JH/47]
770 Diagnosis and fix provided by Michael Fischer v. Mollard.
772 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
773 the right size to place the terminating semicolon on its own folded
774 line, the header hash was calculated to an incorrect value thanks to
775 the (relaxed) space the fold became.
777 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
778 and confused the parent.
780 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
781 which could crash as a result. This could lead to undeliverable messages.
783 JH/28 Logging: "next input sent too soon" now shows where input was truncated
786 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
787 matters on fast-turnover and PID-randomising systems, which were getting
788 out-of-order delivery.
790 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
791 a possibly-overlapping copy. The symptom was that "Remote host closed
792 connection in response to HELO" was logged instead of the actual 4xx
795 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
796 Previously only that bufferd was discarded, resulting in SYMTP command
799 JH/32 DKIM: when a message has multiple signatures matching an identity given
800 in dkim_verify_signers, run the dkim acl once for each. Previously only
801 one run was done. Bug 2189.
803 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
804 "panic the current process" to "deliberately defer". The panic log is
805 still written with the problem list name; the mail and reject logs now
806 get a temp-reject line for the message that was being handled, saying
807 something like "domains check lookup or other defer". The SMTP 451
808 message is still "Temporary local problem".
810 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
811 A crafted sequence of BDAT commands could result in in-use memory beeing
812 freed. CVE-2017-16943.
814 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
815 from SMTP input. Previously it was always done; now only done for DATA
816 and not BDAT commands. CVE-2017-16944.
818 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
819 to the message (such as an overlong header line). Previously this was
820 not done and we did not exit BDAT mode. Followon from the previous item
821 though a different problem.
827 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
828 than -2003 did; needs libidn2 in addition to libidn.
830 JH/02 The path option on a pipe transport is now expanded before use.
832 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
833 Patch provided by "Björn", documentation fix added too.
835 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
836 missing a wire-to-host endian conversion.
838 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
839 close after a BDAT command line could be taken as a following command,
840 giving a synch failure. Fix by only checking for synch immediately
841 before acknowledging the chunk.
843 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
844 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
845 macro. Patches provided by Josh Soref.
847 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
848 Previously we did not; the RFC seems ambiguous and VRFY is not listed
849 by IANA as a service extension. However, John Klensin suggests that we
852 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
853 the dkim code may be unix-mode line endings rather than smtp wire-format
854 CRLF, so prepend a CR to any bare LF.
856 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
857 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
859 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
860 the first were themselves being wrongly included in the feed into dkim
861 processing; with most chunk sizes in use this resulted in an incorrect
862 body hash calculated value.
864 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
865 DKIM signature block, for verification. Although advised against by
866 standards it is specifically not ruled illegal.
868 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
870 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
871 missing a body hash (the bh= tag).
873 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
874 It seems that HAProxy sends the Proxy Protocol information in clear and
875 only then does a TLS startup, so do the same.
877 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
878 TCP connections (such as for Spamd) unless the daemon successfully set
879 Fast Open mode on its listening sockets. This fixes breakage seen on
880 too-old kernels or those not configured for Fast Open, at the cost of
881 requiring both directions being enabled for TFO, and TFO never being used
882 by non-daemon-related Exim processes.
884 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
885 endings, at least on the first header line. Try to canonify any that get
886 past that check, despite the cost.
888 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
889 now limited to an arbitrary five deep, while parsing addresses with the
890 strip_excess_angle_brackets option enabled.
892 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
893 instead leave the unprompted TLS handshake in socket buffer for the
894 TLS library to consume.
896 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
898 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
900 JH/16 Drop variables when they go out of scope. Memory management drops a whole
901 region in one operation, for speed, and this leaves assigned pointers
902 dangling. Add checks run only under the testsuite which checks all
903 variables at a store-reset and panics on a dangling pointer; add code
904 explicitly nulling out all the variables discovered. Fixes one known
905 bug: a transport crash, where a dangling pointer for $sending_ip_address
906 originally assigned in a verify callout, is re-used.
908 PP/06 Drop '.' from @INC in various Perl scripts.
910 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
912 PP/08 Reduce a number of compilation warnings under clang; building with
913 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
914 should be warning-free.
916 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
918 HS/01 Fix portability problems introduced by PP/08 for platforms where
919 realloc(NULL) is not equivalent to malloc() [SunOS et al].
921 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
922 chunk. This allows us to accept broken chunked messages. We need a more
923 general solution here.
925 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
926 already-broken messages in the queue.
928 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
930 JH/19 Fix reference counting bug in routing-generated-address tracking.
936 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
937 supports it and a size is available (ie. the sending peer gave us one).
939 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
940 years of being deprecated). The replacements are the ACLs
941 acl_smtp_mime and acl_not_smtp_mime.
943 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
944 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
945 or even an in-clear connection were permitted. Now, if the host lookup
946 was dnssec and dane was requested then the host is only used if the
947 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
948 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
949 if one fails this test.
950 This means that a poorly-configured remote DNS will make it incommunicado;
951 but it protects against a DNS-interception attack on it.
953 JH/04 Bug 1810: make continued-use of an open smtp transport connection
954 non-noisy when a race steals the message being considered.
956 JH/05 If main configuration option tls_certificate is unset, generate a
957 self-signed certificate for inbound TLS connections.
959 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
960 in rewrites and routers.
962 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
963 and logged a warning sing 4.83; now they are a configuration file error.
965 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
966 (lacking @domain). Apply the same qualification processing as RCPT.
968 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
970 JH/10 Support ${sha256:} applied to a string (as well as the previous
973 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
974 a cutthrough deliver is pending, as we always want to make a connection.
975 This also avoids re-routing the message when later placing the cutthrough
976 connection after a verify cache hit.
977 Do not update it with the verify result either.
979 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
980 when routing results in more than one destination address.
982 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
983 signing (which inhibits the cutthrough capability). Previously only
984 the presence of an option was tested; now an expansion evaluating as
985 empty is permissible (obviously it should depend only on data available
986 when the cutthrough connection is made).
988 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
989 the relevant preceding SMTP command did not note the pipelining mode.
991 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
992 Previously they were not counted.
994 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
995 as one having no matching records. Previously we deferred the message
996 that needed the lookup.
998 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
999 distinguished as "(=".
1001 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1002 for missing MX records. Previously it only worked for missing A records.
1004 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1006 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1007 after the data-go-ahead and data-ack. Patch from Jason Betts.
1009 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1010 even for a "none" policy. Patch from Tony Meyer.
1012 JH/22 Fix continued use of a connection for further deliveries. If a port was
1013 specified by a router, it must also match for the delivery to be
1016 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1017 When one of the recipients of a message was unsuitable for the connection
1018 (has no matching addresses), we lost track of needing to mark it
1019 deferred. As a result mail would be lost.
1021 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1023 JH/25 Decoding ACL controls is now done using a binary search; the source code
1024 takes up less space and should be simpler to maintain. Merge the ACL
1025 condition decode tables also, with similar effect.
1027 JH/26 Fix problem with one_time used on a redirect router which returned the
1028 parent address unchanged. A retry would see the parent address marked as
1029 delivered, so not attempt the (identical) child. As a result mail would
1032 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1033 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1034 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1037 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1038 Main config options check_{log,spool}_{inodes,space} are now
1039 100 inodes, 10MB unless set otherwise in the configuration.
1041 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1042 Previously it only applied to the main-section connection policy
1045 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1047 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1048 by me. Added RFC7919 DH primes as an alternative.
1050 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1051 are not in the system include path.
1053 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1054 GnuTLS, when a session startup failed (eg because the client disconnected)
1055 Exim did stdio operations after fclose. This was exposed by a recent
1056 change which nulled out the file handle after the fclose.
1058 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1059 signed directly by the cert-signing cert, rather than an intermediate
1060 OCSP-signing cert. This is the model used by LetsEncrypt.
1062 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1064 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1065 an incoming connection.
1067 HS/02 Bug 1802: Do not half-close the connection after sending a request
1070 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1071 fallback to "prime256v1".
1073 JH/34 SECURITY: Use proper copy of DATA command in error message.
1074 Could leak key material. Remotely exploitable. CVE-2016-9963.
1080 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1081 and 3.4.4 - once the server is enabled to respond to an OCSP request
1082 it does even when not requested, resulting in a stapling non-aware
1083 client dropping the TLS connection.
1085 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1086 support variable-length bit vectors. No functional change.
1088 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1089 The I= interface field on outgoing lines is now after the H= remote
1090 host field, same as incoming lines. There is a separate
1091 outgoing_interface log selector which allows you to disable the
1094 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1095 If not running log_selector +smtp_connection the mainlog would be held
1096 open indefinitely after a "too many connections" event, including to a
1097 deleted file after a log rotate. Leave the per net connection logging
1098 leaving it open for efficiency as that will be quickly detected by the
1099 check on the next write.
1101 HS/01 Bug 1671: Fix post transport crash.
1102 Processing the wait-<transport> messages could crash the delivery
1103 process if the message IDs didn't exist for some reason. When
1104 using 'split_spool_directory=yes' the construction of the spool
1105 file name failed already, exposing the same netto behaviour.
1107 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1108 mime_regex ACL conditions.
1110 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1111 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1112 to HELO, local diagnostic string.
1114 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1115 log line to debug. Even when configured with a tls authenticator many
1116 client connections are expected to not authenticate in this way, so
1117 an authenticate fail is not an error.
1119 HS/02 Add the Exim version string to the process info. This way exiwhat
1120 gives some more detail about the running daemon.
1122 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1123 matter for fast-change records such as DNSBLs.
1125 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1126 retry record, even if constant. There may be multiple transports with
1127 different interface settings and the retry behaviour needs to be kept
1130 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1132 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1134 JH/10 Bug 840: fix log_defer_output option of pipe transport
1136 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1137 hosts_randomize. This matters a lot when combined with mua_wrapper.
1139 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1140 ${quote_pgsql:<string>} operator.
1142 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1144 JH/14 Change header file name for freeradius-client. Relevant if compiling
1145 with Radius support; from the Gentoo tree and checked under Fedora.
1147 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1149 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1150 option transport_filter, meaning no filtering.
1152 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1154 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1155 defaults to "*" (all hosts). The variable is now available when not built
1156 with TLS, default unset, mainly to enable keeping the testsuite sane.
1157 If a server certificate is not supplied (via tls_certificate) an error is
1158 logged, and clients will find TLS connections fail on startup. Presumably
1159 they will retry in-clear.
1160 Packagers of Exim are strongly encouraged to create a server certificate
1161 at installation time.
1163 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1164 with the $config_file variable.
1166 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1167 in transport context, after the attempt, and per-recipient. The latter type
1168 is per host attempted. The event data is the error message, and the errno
1169 information encodes the lookup type (A vs. MX) used for the (first) host,
1170 and the trailing two digits of the smtp 4xx response.
1172 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1173 to write to mainlog (or rejectlog, paniclog) in the window between file
1174 creation and permissions/ownership being changed. Particularly affects
1175 installations where exicyclog is run as root, rather than exim user;
1176 result is that the running daemon panics and dies.
1178 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1180 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1181 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1182 "pri" and "weight". Note that the previous implicit priority given by the
1183 list order is no longer honoured.
1185 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1186 for DKIM processing.
1188 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1189 by defining SUPPORT_SOCKS.
1191 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1192 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1193 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1194 variables are renamed to proxy_{local,external}_{address,port}.
1196 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1197 for a build by defining SUPPORT_I18N
1199 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1200 of the query string, and make ${quote_redis:} do that quoting.
1202 JH/29 Move Events support from Experimental to mainline, enabled by default
1203 and removable for a build by defining DISABLE_EVENT.
1205 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1207 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1208 cached by the daemon.
1210 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1211 by defining LOOKUP_REDIS. The libhiredis library is required.
1213 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1214 keys are given for lookup.
1216 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1217 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1218 only supported when built with TLS support. The PolarSSL SHA routines
1219 are still used when the TLS library is too old for convenient support.
1221 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1222 openssl_options), for security. OpenSSL forces this from version 1.1.0
1223 server-side so match that on older versions.
1225 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1226 allocation for $value could be released as the expansion processing
1227 concluded, but leaving the global pointer active for it.
1229 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1230 and to use the domains and local_parts ACL conditions.
1232 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1233 incorrectly not doubled on cutthrough transmission, hence seen as a
1234 body-termination at the receiving system - resulting in truncated mails.
1235 Commonly the sender saw a TCP-level error, and retransmitted the message
1236 via the normal store-and-forward channel. This could result in duplicates
1237 received - but deduplicating mailstores were liable to retain only the
1238 initial truncated version.
1240 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1242 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1244 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1245 we're in there, support oversigning also; bug 1309.
1247 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1249 HS/04 Add support for keep_environment and add_environment options.
1251 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1252 either intentional arithmetic overflow during PRNG, or testing config-
1255 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1256 delivery resulted in actual delivery. Cancel cutthrough before DATA
1259 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1260 rejects a recipient: pass the reject to the originator.
1262 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1263 Many were false-positives and ignorable, but it's worth fixing the
1266 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1267 for the new environment-manipulation done at startup. Move the routines
1268 from being local to tls.c to being global via the os.c file.
1270 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1271 an extract embedded as result-arg for a map, the first arg for extract
1272 is unavailable so we cannot tell if this is a numbered or keyed
1273 extraction. Accept either.
1279 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1282 JH/02 The smtp transport option "multi_domain" is now expanded.
1284 JH/03 The smtp transport now requests PRDR by default, if the server offers
1287 JH/04 Certificate name checking on server certificates, when exim is a client,
1288 is now done by default. The transport option tls_verify_cert_hostnames
1289 can be used to disable this per-host. The build option
1290 EXPERIMENTAL_CERTNAMES is withdrawn.
1292 JH/05 The value of the tls_verify_certificates smtp transport and main options
1293 default to the word "system" to access the system default CA bundle.
1294 For GnuTLS, only version 3.0.20 or later.
1296 JH/06 Verification of the server certificate for a TLS connection is now tried
1297 (but not required) by default. The verification status is now logged by
1298 default, for both outbound TLS and client-certificate supplying inbound
1301 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1304 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1305 Status Notification (bounce) messages are now MIME format per RFC 3464.
1306 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1307 under the control of the dsn_advertise_hosts option, and routers may
1308 have a dsn_lasthop option.
1310 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1311 default, modifiable by a malware= option. The list separator for
1312 the options can now be changed in the usual way. Bug 68.
1314 JH/10 The smtp_receive_timeout main option is now expanded before use.
1316 JH/11 The incoming_interface log option now also enables logging of the
1317 local interface on delivery outgoing connections.
1319 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1320 if the interface and destination host and port all match.
1322 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1325 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1326 Patch from Andrew Lewis.
1328 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1329 now supports optional time-restrictions, weighting, and priority
1330 modifiers per server. Patch originally by <rommer@active.by>.
1332 JH/16 The spamd_address main option now supports a mixed list of local
1333 and remote servers. Remote servers can be IPv6 addresses, and
1334 specify a port-range.
1336 JH/17 Bug 68: The spamd_address main option now supports an optional
1337 timeout value per server.
1339 JH/18 Bug 1581: Router and transport options headers_add/remove can
1340 now have the list separator specified.
1342 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1345 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1348 JH/21 Support for the A6 type of dns record is withdrawn.
1350 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1351 rather than the verbs used.
1353 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1354 from 255 to 1024 chars.
1356 JH/24 Verification callouts now attempt to use TLS by default.
1358 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1359 are generic router options now. The defaults didn't change.
1361 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1362 Original patch from Alexander Shikoff, worked over by JH.
1364 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1365 files if ZCAT_COMMAND is not executable.
1367 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1369 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1371 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1372 Normally benign, it bites when the pair was led to by a CNAME;
1373 modern usage is to not canonicalize the domain to a CNAME target
1374 (and we were inconsistent anyway for A-only vs AAAA+A).
1376 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1378 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1379 when evaluating $sender_host_dnssec.
1381 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1382 $sender_helo_dnssec variable.
1384 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1386 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1388 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1390 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1391 documented as working, but never had. Support all but $spam_report.
1393 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1394 added for tls authenticator.
1396 HS/03 Add perl_taintmode main config option
1402 TL/01 When running the test suite, the README says that variables such as
1403 no_msglog_check are global and can be placed anywhere in a specific
1404 test's script, however it was observed that placement needed to be near
1405 the beginning for it to behave that way. Changed the runtest perl
1406 script to read through the entire script once to detect and set these
1407 variables, reset to the beginning of the script, and then run through
1408 the script parsing/test process like normal.
1410 TL/02 The BSD's have an arc4random API. One of the functions to induce
1411 adding randomness was arc4random_stir(), but it has been removed in
1412 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1413 function when detected.
1415 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1416 cause callback expansion.
1418 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1419 syntax errors in an expansion can be treated as a string instead of
1420 logging or causing an error, due to the internal use of bool_lax
1421 instead of bool when processing it.
1423 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1424 server certificates when making smtp deliveries.
1426 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1428 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1430 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1432 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1433 Merged patch from Sebastian Wiedenroth.
1435 JH/05 Fix results-pipe from transport process. Several recipients, combined
1436 with certificate use, exposed issues where response data items split
1437 over buffer boundaries were not parsed properly. This eventually
1438 resulted in duplicates being sent. This issue only became common enough
1439 to notice due to the introduction of connection certificate information,
1440 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1442 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1443 size buffer was used, resulting in syntax errors when an expansion
1446 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1447 version 3.3.6 or later.
1449 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1450 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1451 both become "event_action", the variables become $event_name, $event_data
1452 and $event_defer_errno. There is a new variable $verify_mode, usable in
1453 routers, transports and related events. The tls:cert event is now also
1454 raised for inbound connections, if the main configuration event_action
1457 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1458 early OCSP support, but no stapling (appears to be less than 1.0.0).
1460 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1461 server certificate names available under the smtp transport option
1462 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1465 JH/10 Time-related extraction expansions from certificates now use the main
1466 option "timezone" setting for output formatting, and are consistent
1467 between OpenSSL and GnuTLS compilations. Bug 1541.
1469 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1470 encoded parameter in the incoming message. Bug 1558.
1472 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1473 include certificate info, eximon was claiming there were spoolfile
1476 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1478 JH/14 Log delivery-related information more consistently, using the sequence
1479 "H=<name> [<ip>]" wherever possible.
1481 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1482 are problematic for Debian distribution, omit them from the release
1485 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1487 JH/16 Fix string representation of time values on 64bit time_t architectures.
1490 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1491 output list separator was used.
1496 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1497 checkers that were complaining about end of non-void function with no
1500 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1501 This was a regression introduced in 4.83 by another bugfix.
1503 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1505 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1506 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1512 TF/01 Correctly close the server side of TLS when forking for delivery.
1514 When a message was received over SMTP with TLS, Exim failed to clear up
1515 the incoming connection properly after forking off the child process to
1516 deliver the message. In some situations the subsequent outgoing
1517 delivery connection happened to have the same fd number as the incoming
1518 connection previously had. Exim would try to use TLS and fail, logging
1519 a "Bad file descriptor" error.
1521 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1522 utilities have not been installed.
1524 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1525 temporary space as the ACL may create new global variables.
1527 TL/01 LDAP support uses per connection or global context settings, depending
1528 upon the detected version of the libraries at build time.
1530 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1531 to extract and use the src ip:port in logging and expansions as if it
1532 were a direct connection from the outside internet. PPv2 support was
1533 updated based on HAProxy spec change in May 2014.
1535 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1537 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1538 Properly escape header and check for NULL return.
1540 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1543 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1545 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1546 characters in header names, implemented as a verify condition.
1547 Contributed by Michael Fischer v. Mollard.
1549 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1550 results permerror and temperror. Previous values are deprecated but
1551 still accepted. In a future release, err_perm and err_temp will be
1552 completely removed, which will be a backward incompatibility if the
1553 ACL tests for either of these two old results. Patch contributed by
1554 user bes-internal on the mailing list.
1556 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1558 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1559 selectors, in both main and reject logs.
1561 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1564 JH/07 Add malware type "sock" for talking to simple daemon.
1566 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1568 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1569 routers/transports under cutthrough routing.
1571 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1572 numbers. Touch up "bool" conditional to keep the same definition.
1574 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1576 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1578 JH/12 Expand items in router/transport headers_add or headers_remove lists
1579 individually rather than the list as a whole. Bug 1452.
1581 Required for reasonable handling of multiple headers_ options when
1582 they may be empty; requires that headers_remove items with embedded
1583 colons must have them doubled (or the list-separator changed).
1585 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1586 view the policy declared in the DMARC record. Currently, $dmarc_status
1587 is a combined value of both the record presence and the result of the
1590 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1592 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1593 dnslookup router and the smtp transport (applying to the forward
1596 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1597 of ldap servers used for a specific lookup. Patch provided by Heiko
1600 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1601 New variable $lookup_dnssec_authenticated for observability.
1603 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1604 Patch submitted by Lars Timman.
1606 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1608 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1609 Requires trusted mode and valid format message id, aborts otherwise.
1610 Patch contributed by Heiko Schlichting.
1612 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1613 certextract with support for various fields. Bug 1358.
1615 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1616 is requested by default, modifiable by smtp transport option
1619 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1620 operate on certificate variables to give certificate fingerprints
1621 Also new ${sha256:cert_variable}.
1623 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1625 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1626 Christian Aistleitner.
1628 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1630 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1631 file. Patch from Wolfgang Breyha.
1633 JH/25 Expand the coverage of the delivery $host and $host_address to
1634 client authenticators run in verify callout. Bug 1476.
1636 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1637 align with daemon_smtp_ports. Bug 72.
1639 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1640 support and error reporting did not work properly.
1642 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1643 and is readable. Patch from Andrew Colin Kissa.
1645 TL/14 Enhance documentation of ${run expansion and how it parses the
1646 commandline after expansion, particularly in the case when an
1647 unquoted variable expansion results in an empty value.
1649 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1651 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1652 by renaming away from it. Addresses GH issue 15, reported by
1655 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1657 TL/15 SECURITY: prevent double expansion in math comparison functions
1658 (can expand unsanitized data). Not remotely exploitable.
1665 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1667 PP/02 Make -n do something, by making it not do something.
1668 When combined with -bP, the name of an option is not output.
1670 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1673 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1674 $sender_host_name and config options to manage this, and basic check
1677 PP/05 DSCP support for outbound connections and control modifier for inbound.
1679 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1680 (Only plugin which currently uses this is kerberos4, which nobody should
1681 be using, but we should make it available and other future plugins might
1682 conceivably use it, even though it would break NAT; stuff *should* be
1683 using channel bindings instead).
1685 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1686 name; added for Sendmail compatibility; requires admin caller.
1687 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1688 just ignore it); requires trusted caller.
1689 Also parse but ignore: -Ac -Am -X<logfile>
1692 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1694 TL/02 Add +smtp_confirmation as a default logging option.
1696 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1697 Patch by Magnus Holmgren from 2007-02-20.
1699 TL/04 Bugzilla 1281 - Spec typo.
1700 Bugzilla 1283 - Spec typo.
1701 Bugzilla 1290 - Spec grammar fixes.
1703 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1705 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1707 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1708 mailing list by by Dmitry Isaikin.
1710 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1712 JH/02 Support "G" suffix to numbers in ${if comparisons.
1714 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1716 NM/01 Bugzilla 1197 - Spec typo
1717 Bugzilla 1196 - Spec examples corrections
1719 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1721 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1722 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1725 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1726 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1728 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1729 "acl {{name}{arg}...}", and optional args on acl condition
1732 JH/05 Permit multiple router/transport headers_add/remove lines.
1734 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1736 JH/07 Avoid using a waiting database for a single-message-only transport.
1737 Performance patch from Paul Fisher. Bugzilla 1262.
1739 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1742 JH/09 Add $headers_added variable, with content from use of ACL modifier
1743 add_header (but not yet added to the message). Bugzilla 199.
1745 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1746 Pulled from Bugzilla 817 by Wolfgang Breyha.
1748 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1750 (nb: this is the same fix as in Exim 4.80.1)
1752 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1755 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1758 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1760 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1761 advertise SMTP AUTH mechanism to us, instead of a generic
1762 protocol violation error. Also, make Exim more robust to bad
1763 data from the Dovecot auth socket.
1765 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1767 When a queue runner is handling a message, Exim first routes the
1768 recipient addresses, during which it prunes them based on the retry
1769 hints database. After that it attempts to deliver the message to
1770 any remaining recipients. It then updates the hints database using
1773 So if a recipient address works intermittently, it can get repeatedly
1774 deferred at routing time. The retry hints record remains fresh so the
1775 address never reaches the final cutoff time.
1777 This is a fairly common occurrence when a user is bumping up against
1778 their storage quota. Exim had some logic in its local delivery code
1779 to deal with this. However it did not apply to per-recipient defers
1780 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1782 This change adds a proper retry rule check during routing so that the
1783 final cutoff time is checked against the message's age. We only do
1784 this check if there is an address retry record and there is not a
1785 domain retry record; this implies that previous attempts to handle
1786 the address had the retry_use_local_parts option turned on. We use
1787 this as an approximation for the destination being like a local
1788 delivery, as in LMTP.
1790 I suspect this new check makes the old local delivery cutoff check
1791 redundant, but I have not verified this so I left the code in place.
1793 TF/02 Correct gecos expansion when From: is a prefix of the username.
1795 Test 0254 submits a message to Exim with the header
1799 When I ran the test suite under the user fanf2, Exim expanded
1800 the header to contain my full name, whereas it should have added
1801 a Resent-Sender: header. It erroneously treats any prefix of the
1802 username as equal to the username.
1804 This change corrects that bug.
1806 GF/01 DCC debug and logging tidyup
1807 Error conditions log to paniclog rather than rejectlog.
1808 Debug lines prefixed by "DCC: " to remove any ambiguity.
1810 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1812 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1813 Bug spotted by Jeremy Harris; was flawed since initial commit.
1814 Would have resulted in OCSP responses post-SNI triggering an Exim
1815 NULL dereference and crash.
1817 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1819 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1820 Bug detection, analysis and fix by Samuel Thibault.
1821 Bugzilla 1331, Debian bug #698092.
1823 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1825 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1826 Server implementation by Todd Lyons, client by JH.
1827 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1828 config variable "prdr_enable" controls whether the server
1829 advertises the facility. If the client requests PRDR a new
1830 acl_data_smtp_prdr ACL is called once for each recipient, after
1831 the body content is received and before the acl_smtp_data ACL.
1832 The client is controlled by both of: a hosts_try_prdr option
1833 on the smtp transport, and the server advertisement.
1834 Default client logging of deliveries and rejections involving
1835 PRDR are flagged with the string "PRDR".
1837 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1838 fclose(). Diagnosis by Todd Lyons.
1840 PP/17 Update configure.default to handle IPv6 localhost better.
1841 Patch by Alain Williams (plus minor tweaks).
1844 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1845 This is now consistent with GnuTLS, and is now documented: the
1846 previous undocumented portable approach to treating the option as
1847 unset was to force an expansion failure. That still works, and
1848 an empty string is now equivalent.
1850 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1851 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1852 not performing validation itself.
1854 PP/20 Added force_command boolean option to pipe transport.
1855 Patch from Nick Koston, of cPanel Inc.
1857 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1860 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1862 PP/21 Fix eximon continuous updating with timestamped log-files.
1863 Broken in a format-string cleanup in 4.80, missed when I repaired the
1864 other false fix of the same issue.
1865 Report and fix from Heiko Schlichting.
1868 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1869 Report from Prashanth Katuri.
1871 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1872 It's SecureTransport, so affects any MacOS clients which use the
1873 system-integrated TLS libraries, including email clients.
1875 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1876 using a MIME ACL for non-SMTP local injection.
1877 Report and assistance in diagnosis by Warren Baker.
1879 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1881 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1883 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1884 last id that failed so it may be referenced in subsequent ACL's.
1886 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1889 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1890 ldap library initialization, allowing self-signed CA's to be
1891 used. Also properly sets require_cert option later in code by
1892 using NULL (global ldap config) instead of ldap handle (per
1893 session). Bug diagnosis and testing by alxgomz.
1895 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1896 the src/util/ subdirectory.
1898 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1899 renamed to Transport Post Delivery Action by Jeremy Harris, as
1902 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1903 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1904 redis_servers = needs to be configured which will be used by the redis
1905 lookup. Patch from Warren Baker, of The Packet Hub.
1907 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1909 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1910 hostname or reverse DNS when processing a host list. Used suggestions
1911 from multiple comments on this bug.
1913 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1915 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1916 Missed a few lines, added it to make the runtest require no keyboard
1919 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1920 contains upper case chars. Make router use caseful_local_part.
1922 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1923 support when GnuTLS has been built with p11-kit.
1929 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1931 This, or similar/improved, will also be change PP/11 of 4.82.
1937 PP/01 Handle short writes when writing local log-files.
1938 In practice, only affects FreeBSD (8 onwards).
1939 Bugzilla 1053, with thanks to Dmitry Isaikin.
1941 NM/01 Bugzilla 949 - Documentation tweak
1943 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1946 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1948 PP/02 Implemented gsasl authenticator.
1950 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1952 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1953 `pkg-config foo` for cflags/libs.
1955 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1956 with rest of GSASL and with heimdal_gssapi.
1958 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1959 `pkg-config foo` for cflags/libs for the TLS implementation.
1961 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1962 properties get this fed in as external SSF. A number of robustness
1963 and debugging improvements to the cyrus_sasl authenticator.
1965 PP/08 cyrus_sasl server now expands the server_realm option.
1967 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1968 Patch by Jeremy Harris.
1970 PP/10 Added dbmjz lookup type.
1972 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1974 PP/12 MAIL args handles TAB as well as SP, for better interop with
1975 non-compliant senders.
1976 Analysis and variant patch by Todd Lyons.
1978 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1979 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1980 Patch from Dirk Mueller <dmueller@suse.com>
1982 PP/13 tls_peerdn now print-escaped for spool files.
1983 Observed some $tls_peerdn in wild which contained \n, which resulted
1984 in spool file corruption.
1986 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1987 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1988 or write after TLS renegotiation, which otherwise led to messages
1991 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1992 as a tracking header (ie: a signed header comes before the signature).
1993 Patch from Wolfgang Breyha.
1995 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1996 comma-sep list; embedded commas doubled.
1998 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2000 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2002 Report and patch from Dmitry Banschikov.
2004 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2005 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2006 failures. We appear to now support TLS1.1+ with Exim.
2008 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2009 lets Exim select keys and certificates based upon TLS SNI from client.
2010 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2011 before an outbound SMTP session. New log_selector, +tls_sni.
2013 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2014 NULL dereference. Report and patch from Alun Jones.
2016 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2017 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2018 Not seeing resolver debug output on NetBSD, but suspect this is a
2019 resolver implementation change.
2021 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2022 Left warnings. Added "eximon gdb" invocation mode.
2024 PP/21 Defaulting "accept_8bitmime" to true, not false.
2026 PP/22 Added -bw for inetd wait mode support.
2028 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2029 locate the relevant includes and libraries. Made this the default.
2031 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2032 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2034 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2035 This may cause build issues on older platforms.
2037 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2038 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2039 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2040 Added SNI support via GnuTLS too.
2041 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2043 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2045 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2046 Applied second patch from Janne, implementing suggestion to default
2047 multiple-strings-in-record handling to match SPF spec.
2049 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2051 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2052 read-only, out of scope).
2053 Patch from Wolfgang Breyha, report from Stuart Northfield.
2055 PP/29 Fix three issues highlighted by clang analyser static analysis.
2056 Only crash-plausible issue would require the Cambridge-specific
2057 iplookup router and a misconfiguration.
2058 Report from Marcin Mirosław.
2060 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2062 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2063 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2064 As part of this, removing so much warning spew let me fix some minor
2065 real issues in debug logging.
2067 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2068 assignment on my part. Fixed.
2070 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2071 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2072 Janne Snabb (who went above and beyond: thank you).
2074 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2075 string otherwise requires a connection and a bunch more work and it's
2076 relatively easy to get wrong. Should also expose TLS library linkage
2079 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2080 64-bit ${eval} (JH/03).
2082 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2083 GNU libc to support some of the 64-bit stuff, should not lead to
2084 conflicts. Defined before os.h is pulled in, so if a given platform
2085 needs to override this, it can.
2087 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2088 protection layer was required, which is not implemented.
2089 Bugzilla 1254, patch from Wolfgang Breyha.
2091 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2092 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2093 tls_dhparam take prime identifiers. Also unbreak combination of
2094 OpenSSL+DH_params+TLSSNI.
2096 PP/39 Disable SSLv2 by default in OpenSSL support.
2102 PP/01 Solaris build fix for Oracle's LDAP libraries.
2103 Bugzilla 1109, patch from Stephen Usher.
2105 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2107 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2110 TF/02 Fix a couple more cases where we did not log the error message
2111 when unlink() failed. See also change 4.74-TF/03.
2113 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2114 lock up or crash if it happened to be inside a call to libc when it
2115 got a SIGUSR1 from exiwhat.
2117 The SIGUSR1 handler appends the current process status to the process
2118 log which is later printed by exiwhat. It used to use the general
2119 purpose logging code to do this, but several functions it calls are
2120 not safe for signals.
2122 The new output code in the SIGUSR1 handler is specific to the process
2123 log, and simple enough that it's easy to inspect for signal safety.
2124 Removing some special cases also simplifies the general logging code.
2125 Removing the spurious timestamps from the process log simplifies
2128 TF/04 Improved ratelimit ACL condition.
2130 The /noupdate option has been deprecated in favour of /readonly which
2131 has clearer semantics. The /leaky, /strict, and /readonly update modes
2132 are mutually exclusive. The update mode is no longer included in the
2133 database key; it just determines when the database is updated. (This
2134 means that when you upgrade Exim will forget old rate measurements.)
2136 Exim now checks that the per_* options are used with an update mode that
2137 makes sense for the current ACL. For example, when Exim is processing a
2138 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2139 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2140 must specify per_mail/readonly. If you omit the update mode it defaults to
2141 /leaky where that makes sense (as before) or /readonly where required.
2143 The /noupdate option is now undocumented but still supported for
2144 backwards compatibility. It is equivalent to /readonly except that in
2145 ACLs where /readonly is required you may specify /leaky/noupdate or
2146 /strict/noupdate which are treated the same as /readonly.
2148 A useful new feature is the /count= option. This is a generalization
2149 of the per_byte option, so that you can measure the throughput of other
2150 aggregate values. For example, the per_byte option is now equivalent
2151 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2153 The per_rcpt option has been generalized using the /count= mechanism
2154 (though it's more complicated than the per_byte equivalence). When it is
2155 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2156 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2157 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2158 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2159 Note that using per_rcpt with a non-readonly update mode in more than
2160 one ACL will cause the recipients to be double-counted. (The per_mail
2161 and per_byte options don't have this problem.)
2163 The handling of very low rates has changed slightly. If the computed rate
2164 is less than the event's count (usually one) then this event is the first
2165 after a long gap. In this case the rate is set to the same as this event's
2166 count, so that the first message of a spam run is counted properly.
2168 The major new feature is a mechanism for counting the rate of unique
2169 events. The new per_addr option counts the number of different
2170 recipients that someone has sent messages to in the last time period. It
2171 behaves like per_rcpt if all the recipient addresses are different, but
2172 duplicate recipient addresses do not increase the measured rate. Like
2173 the /count= option this is a general mechanism, so the per_addr option
2174 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2175 example, measure the rate that a client uses different sender addresses
2176 with the options per_mail/unique=$sender_address. There are further
2177 details in the main documentation.
2179 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2181 TF/06 Removed a few PCRE remnants.
2183 TF/07 Automatically extract Exim's version number from tags in the git
2184 repository when doing development or release builds.
2186 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2187 Bugzilla 879. Patch from Paul Fisher.
2189 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2190 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2193 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2195 PP/05 Variable $av_failed, true if the AV scanner deferred.
2196 Bugzilla 1078. Patch from John Horne.
2198 PP/06 Stop make process more reliably on build failure.
2199 Bugzilla 1087. Patch from Heiko Schlittermann.
2201 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2202 Bugzilla 1089. Patch from Heiko Schlittermann.
2204 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2205 Bugzilla 1131. Patch from Holger Weiß.
2207 PP/09 Handle IPv6 addresses with SPF.
2208 Bugzilla 860. Patch from Wolfgang Breyha.
2210 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2212 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2215 PP/11 match_* no longer expand right-hand-side by default.
2216 New compile-time build option, EXPAND_LISTMATCH_RHS.
2217 New expansion conditions, "inlist", "inlisti".
2219 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2221 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2223 PP/14 fix log_write() format string regression from TF/03.
2224 Bugzilla 1152. Patch from Dmitry Isaikin.
2230 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2232 PP/02 Harmonised TLS library version reporting; only show if debugging.
2233 Layout now matches that introduced for other libraries in 4.74 PP/03.
2235 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2237 PP/04 New "dns_use_edns0" global option.
2239 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2242 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2243 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2245 TK/01 Updated PolarSSL code to 0.14.2.
2246 Bugzilla 1097. Patch from Andreas Metzler.
2248 PP/07 Catch divide-by-zero in ${eval:...}.
2249 Fixes bugzilla 1102.
2251 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2254 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2255 format-string attack -- SECURITY: remote arbitrary code execution.
2257 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2258 time unintentionally subject to list matching rules, letting the header
2259 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2260 arbitrary string expansion). This allowed for information disclosure.
2262 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2263 INT_MIN/-1 -- value coerced to INT_MAX.
2269 NM/01 Workaround for PCRE version dependency in version reporting
2272 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2273 This fixes portability to compilers other than gcc, notably
2274 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2276 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2277 makefiles for portability to HP-UX and POSIX correctness.
2279 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2280 Also via indented variable definition in the Makefile.
2281 (Debugging by Oliver Heesakkers).
2283 PP/02 Restore caching of spamd results with expanded spamd_address.
2284 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2286 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2287 Improves build reliability. Fix from: Frank Elsner
2289 NM/02 Fix wide character breakage in the rfc2047 coding
2290 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2292 NM/03 Allow underscore in dnslist lookups
2293 Fixes bug 1026. Patch from Graeme Fowler
2295 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2296 Code patches from Adam Ciarcinski of NetBSD.
2298 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2301 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2302 is logged, to avoid truncation. Patch from John Horne.
2304 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2305 Patch from Jakob Hirsch.
2307 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2308 SQL string expansion failure details.
2309 Patch from Andrey Oktyabrski.
2311 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2312 Patch from Simon Arlott.
2314 PP/09 New lookups functionality failed to compile on old gcc which rejects
2315 extern declarations in function scope.
2316 Patch from Oliver Fleischmann
2318 PP/10 Use sig_atomic_t for flags set from signal handlers.
2319 Check getgroups() return and improve debugging.
2320 Fixed developed for diagnosis in bug 927 (which turned out to be
2323 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2324 Patch from Mark Zealey.
2326 PP/12 Bugzilla 1056: Improved spamd server selection.
2327 Patch from Mark Zealey.
2329 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2330 Based on patch from Heiko Schlittermann.
2332 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2333 Patch from Uwe Doering, sign-off by Michael Haardt.
2335 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2336 variable declaration deep within a block. Bug and patch from
2339 PP/15 lookups-Makefile IRIX compatibility coercion.
2341 PP/16 Make DISABLE_DKIM build knob functional.
2343 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2344 Patch by Simon Arlott
2346 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2347 variable argument macros. Our copy now differs from upstream.
2353 TF/01 Failure to get a lock on a hints database can have serious
2354 consequences so log it to the panic log.
2356 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2357 controlled using the smtp_confirmation log selector.
2359 TF/03 Include the error message when we fail to unlink a spool file.
2361 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2362 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2363 for maintaining out-of-tree patches for some time.
2365 PP/01 Bugzilla 139: Documentation and portability issues.
2366 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2367 Handle per-OS dynamic-module compilation flags.
2369 PP/02 Let /dev/null have normal permissions.
2370 The 4.73 fixes were a little too stringent and complained about the
2371 permissions on /dev/null. Exempt it from some checks.
2372 Reported by Andreas M. Kirchwitz.
2374 PP/03 Report version information for many libraries, including
2375 Exim version information for dynamically loaded libraries. Created
2376 version.h, now support a version extension string for distributors
2377 who patch heavily. Dynamic module ABI change.
2379 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2380 privilege escalation vulnerability whereby the Exim run-time user
2381 can cause root to append content of the attacker's choosing to
2384 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2387 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2388 If dropping privileges for untrusted macros, we disabled normal logging
2389 on the basis that it would fail; for the Exim run-time user, this is not
2390 the case, and it resulted in successful deliveries going unlogged.
2391 Fixed. Reported by Andreas Metzler.
2397 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2398 only prepend for the Resent-* case. Fixes regression introduced in
2399 Exim 4.70 by NM/22 for Bugzilla 607.
2401 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2402 increasing numbers of administrators be bitten by this.
2404 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2406 PP/03 Bugzilla 994: added openssl_options main configuration option.
2408 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2410 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2412 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2414 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2415 without a peer certificate, leading to a segfault because of an
2416 assumption that peers always have certificates. Be a little more
2417 paranoid. Problem reported by Martin Tscholak.
2419 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2420 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2421 NB: ClamAV planning to remove STREAM in "middle of 2010".
2422 CL also introduces -bmalware, various -d+acl logging additions and
2423 more caution in buffer sizes.
2425 PP/09 Implemented reverse_ip expansion operator.
2427 PP/10 Bugzilla 937: provide a "debug" ACL control.
2429 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2431 PP/12 Bugzilla 973: Implement --version.
2433 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2435 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2437 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2439 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2440 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2441 ignore trailing whitespace.
2443 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2445 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2448 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2449 Notification from Dr Andrew Aitchison.
2451 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2452 ExtendedDetectionInfo response format.
2453 Notification from John Horne.
2455 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2458 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2459 XSL and documented dependency on system catalogs, with examples of how
2462 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2465 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2466 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2467 configuration file which is writeable by the Exim user or group.
2469 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2470 of configuration files to cover files specified with the -C option if
2471 they are going to be used with root privileges, not just the default
2474 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2475 option (effectively making it always true).
2477 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2478 files to be used while preserving root privileges.
2480 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2481 that rogue child processes cannot use them.
2483 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2484 run-time user, instead of root.
2486 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2487 Exim run-time user without dropping privileges.
2489 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2490 result string, instead of calling string_vformat() twice with the same
2493 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2494 for other users. Others should always drop root privileges if they use
2495 -C on the command line, even for a whitelisted configure file.
2497 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2499 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2505 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2506 $data_path, and $header_path variables; fixed documentation bugs and
2509 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2510 exipick to access non-standard spools, including the "frozen" queue
2513 NM/01 Bugzilla 965: Support mysql stored procedures.
2514 Patch from Alain Williams
2516 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2518 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2519 Patch from Andreas Metzler
2521 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2522 Patch from Kirill Miazine
2524 NM/05 Bugzilla 671: Added umask to procmail example.
2526 JJ/03 installed exipick 20100323.0, fixing doc bug
2528 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2529 directory. Notification and patch from Dan Rosenberg.
2531 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2533 TK/02 Improve log output when DKIM signing operation fails.
2535 MH/01 Treat the transport option dkim_domain as a colon separated
2536 list, not as a single string, and sign the message with each element,
2537 omitting multiple occurences of the same signer.
2539 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2540 Bugzilla 985, 986. Patch by Simon Arlott
2542 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2543 Patch by Simon Arlott
2545 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2546 MBX locking. Notification from Dan Rosenberg.
2552 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2554 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2556 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2558 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2560 NM/04 Bugzilla 915: Flag broken perl installation during build.
2566 TK/01 Added patch by Johannes Berg that expands the main option
2567 "spamd_address" if it starts with a dollar sign.
2569 TK/02 Write list of recipients to X-Envelope-Sender header when building
2570 the mbox-format spool file for content scanning (suggested by Jakob
2573 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2574 (http://www.dcc-servers.net/) support via dccifd. Activated by
2575 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2577 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2578 by Mark Daniel Reidel <mr@df.eu>.
2580 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2581 When building exim an external PCRE library is now needed -
2582 PCRE is a system library on the majority of modern systems.
2583 See entry on PCRE_LIBS in EDITME file.
2585 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2586 conversation. Added nologin parameter to request.
2587 Patch contributed by Kirill Miazine.
2589 TF/01 Do not log submission mode rewrites if they do not change the address.
2591 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2593 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2594 log files in place. Contributed by Roberto Lima.
2596 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2598 TF/03 Bugzilla 615: When checking the local_parts router precondition
2599 after a local_part_suffix or local_part_prefix option, Exim now
2600 does not use the address's named list lookup cache, since this
2601 contains cached lookups for the whole local part.
2603 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2604 Robert Millan. Documentation is in experimental-spec.txt.
2606 TF/04 Bugzilla 668: Fix parallel build (make -j).
2608 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2610 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2611 Patch provided by Jan Srzednicki.
2613 TF/05 Leading white space used to be stripped from $spam_report which
2614 wrecked the formatting. Now it is preserved.
2616 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2617 that they are available at delivery time.
2619 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2621 TF/08 TLS error reporting now respects the incoming_interface and
2622 incoming_port log selectors.
2624 TF/09 Produce a more useful error message if an SMTP transport's hosts
2625 setting expands to an empty string.
2627 NM/06 Bugzilla 744: EXPN did not work under TLS.
2628 Patch provided by Phil Pennock.
2630 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2631 Patch provided by Richard Godbee.
2633 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2634 acl_smtp_notquit, added index entry.
2636 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2637 Patch provided by Eugene Bujak.
2639 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2640 accept(). Patch provided by Maxim Dounin.
2642 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2643 Patch provided by Phil Pennock.
2645 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2647 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2648 Patch provided by Brad "anomie" Jorsch.
2650 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2651 Patch provided by Dean Brooks.
2653 TK/05 Add native DKIM support (does not depend on external libraries).
2655 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2656 Patch provided by Graeme Fowler.
2658 NM/16 Bugzilla 851: Documentation example syntax fix.
2660 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2662 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2665 NM/19 Bugzilla 745: TLS version reporting.
2666 Patch provided by Phil Pennock.
2668 NM/20 Bugzilla 167: bool: condition support.
2669 Patch provided by Phil Pennock.
2671 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2672 clients. Patch provided by Phil Pennock.
2674 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2675 Patch provided by Brad "anomie" Jorsch.
2677 NM/23 Bugzilla 687: Fix misparses in eximstats.
2678 Patch provided by Heiko Schlittermann.
2680 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2681 Patch provided by Heiko Schlittermann.
2683 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2684 plus update to original patch.
2686 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2688 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2689 Patch provided by David Brownlee.
2691 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2693 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2695 NM/30 Bugzilla 888: TLS documentation bugfixes.
2697 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2699 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2700 Unlike the original bugzilla I have changed all shell scripts in src tree.
2702 NM/33 Bugzilla 898: Transport filter timeout fix.
2703 Patch by Todd Rinaldo.
2705 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2706 Patch by Serge Demonchaux.
2708 NM/35 Bugzilla 39: Base64 decode bug fixes.
2709 Patch by Jakob Hirsch.
2711 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2713 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2715 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2717 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2723 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2724 ALT-N's libdkim that I have put here:
2725 http://duncanthrax.net/exim-experimental/
2727 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2728 were called 'true' and it seems that C99 defines that as a reserved
2729 keyword to be used with 'bool' variable types. That means you could
2730 not include C99-style headers which use bools without triggering
2731 build errors in sieve.c.
2733 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2734 as mailq or other aliases. Changed the --help handling significantly
2735 to do whats expected. exim_usage() emits usage/help information.
2737 SC/01 Added the -bylocaldomain option to eximstats.
2739 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2741 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2743 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2749 PH/01 Another patch from the Sieve maintainer.
2751 PH/02 When an IPv6 address is converted to a string for single-key lookup
2752 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2753 dots are used instead of colons so that keys in lsearch files need not
2754 contain colons. This was done some time before quoting was made available
2755 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2756 (notated using the quote facility) so as to distinguish them from IPv4
2757 keys. This meant that lookups for IP addresses in host lists did not work
2758 for iplsearch lookups.
2760 This has been fixed by arranging for IPv6 addresses to be expressed with
2761 colons if the lookup type is iplsearch. This is not incompatible, because
2762 previously such lookups could never work.
2764 The situation is now rather anomalous, since one *can* have colons in
2765 ordinary lsearch keys. However, making the change in all cases is
2766 incompatible and would probably break a number of configurations.
2768 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2771 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2772 conversion specification without a maximum field width, thereby enabling
2773 a rogue spamd server to cause a buffer overflow. While nobody in their
2774 right mind would setup Exim to query an untrusted spamd server, an
2775 attacker that gains access to a server running spamd could potentially
2776 exploit this vulnerability to run arbitrary code as the Exim user.
2778 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2779 $primary_hostname instead of what libspf2 thinks the hosts name is.
2781 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2782 a directory entry by the name of the lookup key. Previously, if a
2783 symlink pointed to a non-existing file or a file in a directory that
2784 Exim lacked permissions to read, a lookup for a key matching that
2785 symlink would fail. Now it is enough that a matching directory entry
2786 exists, symlink or not. (Bugzilla 503.)
2788 PH/03 The body_linecount and body_zerocount variables are now exported in the
2791 PH/04 Added the $dnslist_matched variable.
2793 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2794 This means they are set thereafter only if the connection becomes
2797 PH/06 Added the client_condition to authenticators so that some can be skipped
2798 by clients under certain conditions.
2800 PH/07 The error message for a badly-placed control=no_multiline_responses left
2801 "_responses" off the end of the name.
2803 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2805 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2806 (without spaces) instead of just copying the configuration text.
2808 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2810 PH/11 Added $max_received_linelength.
2812 PH/12 Added +ignore_defer and +include_defer to host lists.
2814 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2815 way in which PCRE > 7.0 is built.
2817 PH/14 Implemented queue_only_load_latch.
2819 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2820 MAIL command. The effect was to mangle the value on 64-bit systems.
2822 PH/16 Another patch from the Sieve maintainer.
2824 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2826 PH/18 If a system quota error occurred while trying to create the file for
2827 a maildir delivery, the message "Mailbox is full" was not appended to the
2828 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2829 only to a quota excession during the actual writing of the file.
2831 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2832 characters?) which causes problems in log lines. The DN values are now
2833 passed through string_printing() before being added to log lines.
2835 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2836 and InterBase are left for another time.)
2838 PH/21 Added message_body_newlines option.
2840 PH/22 Guard against possible overflow in moan_check_errorcopy().
2842 PH/23 POSIX allows open() to be a macro; guard against that.
2844 PH/24 If the recipient of an error message contained an @ in the local part
2845 (suitably quoted, of course), incorrect values were put in $domain and
2846 $local_part during the evaluation of errors_copy.
2852 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2853 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2856 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2857 issue a MAIL command.
2859 PH/02 In an ACL statement such as
2861 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2863 if a client was not listed at all, or was listed with a value other than
2864 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2865 the condition was not true (as it should be), so access was not denied.
2866 The bug was that the ! inversion was incorrectly passed on to the second
2867 item. This has been fixed.
2869 PH/03 Added additional dnslists conditions == and =& which are different from
2870 = and & when the dns lookup returns more than one IP address.
2872 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2873 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2875 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2876 FSYNC, which compiles an option called disable_fsync that allows for
2877 bypassing fsync(). The documentation is heavily laced with warnings.
2879 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2881 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2882 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2883 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2884 including adding "make clean"; (3) Added -fPIC when compiling the test
2885 dynamically loaded module, to get rid of a warning.
2887 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2888 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2889 = 0s. The bug is otherwise harmless.
2891 PH/07 There was a bug in the dovecot authenticator such that the value of
2892 $auth1 could be overwritten, and so not correctly preserved, after a
2893 successful authentication. This usually meant that the value preserved by
2894 the server_setid option was incorrect.
2896 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2898 PH/09 Installed PCRE release 7.0.
2900 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2901 run for batched SMTP input. It is now run at the start of every message
2902 in the batch. While fixing this I discovered that the process information
2903 (output by running exiwhat) was not always getting set for -bs and -bS
2904 input. This is fixed, and it now also says "batched" for BSMTP.
2906 PH/11 Added control=no_pipelining.
2908 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2909 patch, slightly modified), and move the expansion of helo_data till after
2910 the connection is made in the smtp transport (so it can use these
2913 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2915 PH/14 Added log_selector = +pid.
2917 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2919 PH/16 Add ${if forany and ${if forall.
2921 PH/17 Added dsn_from option to vary the From: line in DSNs.
2923 PH/18 Flush SMTP output before performing a callout, unless control =
2924 no_callout_flush is set.
2926 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2927 was true (the default) a successful delivery failed to delete the retry
2928 item, thus causing premature timeout of the address. The bug is now
2931 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2933 PH/21 Long custom messages for fakedefer and fakereject are now split up
2934 into multiline responses in the same way that messages for "deny" and
2935 other ACL rejections are.
2937 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2938 with slight modification.
2940 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2941 draft, changing the syntax and factoring some duplicate code".
2943 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2944 for deliveries of the second and subsequent messages over the same SMTP
2947 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2948 ${reduce, with only minor "tidies".
2950 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2952 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2953 expansion side effects.
2955 PH/27 When a message times out after an over-quota error from an Exim-imposed
2956 quota, the bounce message says "mailbox is full". This message was not
2957 being given when it was a system quota that was exceeded. It now should
2960 MH/03 Made $recipients available in local_scan(). local_scan() already has
2961 better access to the recipient list through recipients_list[], but
2962 $recipients can be useful in postmaster-provided expansion strings.
2964 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2965 in the case of a MAIL command with additional options following the
2966 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2967 were accidentally chopped off.
2969 PH/29 SMTP synchronization checks are implemented when a command is read -
2970 there is a check that no more input is waiting when there shouldn't be
2971 any. However, for some commands, a delay in an ACL can mean that it is
2972 some time before the response is written. In this time, more input might
2973 arrive, invalidly. So now there are extra checks after an ACL has run for
2974 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2975 pipelining has not been advertised.
2977 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2979 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2980 returned in the SMTP response when smtp_return_error_details was false.
2981 This has been fixed.
2983 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2984 instead of the C I/O that was originally supplied, because problems were
2985 reported on Solaris.
2987 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2988 Exim which did not show up earlier: it was assuming that a call to
2989 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2990 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2991 was a macro that became an assignment, so it seemed to work. This has
2992 changed to a proper function call with a void return, hence the compile
2993 error. Exim's code has been fixed.
2995 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2998 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3000 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3002 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3003 because it turns out to be too big or there is a timeout) there is no
3004 "Completed" line in the log. When some messages of this type were
3005 selected by exigrep, they were listed as "not completed". Others were
3006 picked up by some special patterns. I have improved the selection
3007 criteria to be more general.
3009 PH/38 The host_find_failed option in the manualroute router can now be set
3010 to "ignore", to completely ignore a host whose IP address cannot be
3011 found. If all hosts are ignored, the behaviour is controlled by the new
3012 host_all_ignored option.
3014 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3015 homing or because of multiple MX records with /mx) generated more than
3016 one IP address, and the following item turned out to be the local host,
3017 all the secondary addresses of the first item were incorrectly removed
3018 from the list, along with the local host and any following hosts (which
3019 is what is supposed to happen).
3021 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3022 whoever called Exim into the -H file. In the case of the daemon it was
3023 behaving confusingly. When first started, it used values for whoever
3024 started the daemon, but after a SIGHUP it used the Exim user (because it
3025 calls itself on a restart). I have changed the code so that it now always
3028 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3029 message are rejected with the same error (e.g. no authentication or bad
3030 sender address), and a DATA command is nevertheless sent (as can happen
3031 with PIPELINING or a stupid MUA), the error message that was given to the
3032 RCPT commands is included in the rejection of the DATA command. This is
3033 intended to be helpful for MUAs that show only the final error to their
3036 PH/42 Another patch from the Sieve maintainer.
3038 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3039 Eximstats - Fixed some broken HTML links and added missing column headers
3041 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3042 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3044 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3046 PH/43 Yet another patch from the Sieve maintainer.
3048 PH/44 I found a way to check for a TCP/IP connection going away before sending
3049 the response to the final '.' that terminates a message, but only in the
3050 case where the client has not sent further data following the '.'
3051 (unfortunately, this is allowed). However, in many cases there won't be
3052 any further data because there won't be any more messages to send. A call
3053 to select() can be used: if it shows that the input is "ready", there is
3054 either input waiting, or the socket has been closed. An attempt to read
3055 the next input character can distinguish the two cases. Previously, Exim
3056 would have sent an OK response which the client would never have see.
3057 This could lead to message repetition. This fix should cure that, at
3058 least in a lot of common cases.
3060 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3061 advertised in response to EHLO.
3067 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3068 fixed by 4.65/MH/01 (is this a record?) are fixed:
3070 (i) An empty string was always treated as zero by the numeric comparison
3071 operators. This behaviour has been restored.
3073 (ii) It is documented that the numeric comparison operators always treat
3074 their arguments as decimal numbers. This was broken in that numbers
3075 starting with 0 were being interpreted as octal.
3077 While fixing these problems I realized that there was another issue that
3078 hadn't been noticed. Values of message_size_limit (both the global option
3079 and the transport option) were treated as octal if they started with 0.
3080 The documentation was vague. These values are now always treated as
3081 decimal, and I will make that clear in the documentation.
3087 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3088 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3091 MH/01 Don't check that the operands of numeric comparison operators are
3092 integers when their expansion is in "skipping" mode (fixes bug
3093 introduced by 4.64-PH/07).
3095 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3096 child addresses, Exim now panics and dies. Previously, because the count
3097 is held in a short int, deliveries were likely to be lost. As such a
3098 large number of recipients for a single message is ridiculous
3099 (performance will be very, very poor), I have chosen to impose a limit
3100 rather than extend the field.
3106 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3107 leftover -K file (the existence of which was triggered by #402).
3108 While we were at it, introduced process PID as part of the -K
3109 filename. This should rule out race conditions when creating
3112 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3113 processing considerably. Previous code took too long for large mails,
3114 triggering a timeout which in turn triggers #401.
3116 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3117 in the DK code in transports.c. sendfile() is not really portable,
3118 hence the _LINUX specificness.
3120 TF/01 In the add_headers option to the mail command in an Exim filter,
3121 there was a bug that Exim would claim a syntax error in any
3122 header after the first one which had an odd number of characters
3125 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3126 callout verification, Exim cached a "reject" for the entire domain. This
3127 is correct for most verifications, but it is not correct for a recipient
3128 verification with use_sender or use_postmaster set, because in that case
3129 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3130 case of MAIL FROM:<> rejection from other early rejections (e.g.
3131 rejection of HELO). When verifying a recipient using a non-null MAIL
3132 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3133 Whatever the result of the callout, the value of the domain cache is
3134 left unchanged (for any other kind of callout, getting as far as trying
3135 RCPT means that the domain itself is ok).
3137 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3140 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3141 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3142 session, the daemon ignores ECONNECT errors and logs others; it now
3143 ignores EPIPE as well.
3145 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3146 (quoted-printable decoding).
3148 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3149 later the small subsequent patch to fix an introduced bug.
3151 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3153 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3155 PH/08 An error is now given if message_size_limit is specified negative.
3157 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3158 to be given (somewhat) arbitrary names.
3160 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3163 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3164 miscellaneous code fixes
3166 PH/10 Added the log_reject_target ACL modifier to specify where to log
3169 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3170 hostname. This is wrong, because it relates to the incoming message (and
3171 probably the interface on which it is arriving) and not to the outgoing
3172 callout (which could be using a different interface). This has been
3173 changed to use the value of the helo_data option from the smtp transport
3174 instead - this is what is used when a message is actually being sent. If
3175 there is no remote transport (possible with a router that sets up host
3176 addresses), $smtp_active_hostname is used.
3178 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3179 tweaks were necessary in order to get it to work (see also 21 below):
3180 (a) The code assumed that strncpy() returns a negative number on buffer
3181 overflow, which isn't the case. Replaced with Exim's string_format()
3183 (b) There were several signed/unsigned issues. I just did the minimum
3184 hacking in of casts. There is scope for a larger refactoring.
3185 (c) The code used strcasecmp() which is not a standard C function.
3186 Replaced with Exim's strcmpic() function.
3187 (d) The code set only $1; it now sets $auth1 as well.
3188 (e) A simple test gave the error "authentication client didn't specify
3189 service in request". It would seem that Dovecot has changed its
3190 interface. Fortunately there's a specification; I followed it and
3191 changed what the client sends and it appears to be working now.
3193 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3196 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3197 address A is aliased to B and C, where B exists and C does not. Without
3198 -v the output is "A verified" because verification stops after a
3199 successful redirection if more than one address is generated. However,
3200 with -v the child addresses are also verified. Exim was outputting "A
3201 failed to verify" and then showing the successful verification for C,
3202 with its parentage. It now outputs "B failed to verify", showing B's
3203 parentage before showing the successful verification of C.
3205 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3206 look up a TXT record in a specific list after matching in a combined
3209 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3210 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3211 they consult the DNS. I had assumed they would set it the way they
3212 wanted; and indeed my experiments on Linux seem to show that in some
3213 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3214 To be on the safe side, however, I have now made the interface to
3215 host_find_byname() similar to host_find_bydns(), with an argument
3216 containing the DNS resolver options. The host_find_byname() function now
3217 sets these options at its start, just as host_find_bydns() does. The smtp
3218 transport options dns_qualify_single and dns_search_parents are passed to
3219 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3220 of host_find_byname() use the default settings of RES_DEFNAMES
3221 (qualify_single) but not RES_DNSRCH (search_parents).
3223 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3224 spool_read_header() do less string testing, by means of a preliminary
3225 switch on the second character of optional "-foo" lines. (This is
3226 overdue, caused by the large number of possibilities that now exist.
3227 Originally there were few.) While I was there, I also converted the
3228 str(n)cmp tests so they don't re-test the leading "-" and the first
3229 character, in the hope this might squeeze out yet more improvement.
3231 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3232 flag allowing group syntax was set by the header_syntax check but not
3233 turned off, possible causing trouble later; (2) The flag was not being
3234 set at all for the header_verify test, causing "group"-style headers to
3235 be rejected. I have now set it in this case, and also caused header_
3236 verify to ignore an empty address taken from a group. While doing this, I
3237 came across some other cases where the code for allowing group syntax
3238 while scanning a header line wasn't quite right (mostly, not resetting
3239 the flag correctly in the right place). These bugs could have caused
3240 trouble for malformed header lines. I hope it is now all correct.
3242 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3243 with the "reply" argument non-NULL. The code, however (which originally
3244 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3245 but it didn't always do it. This confused somebody who was copying the
3246 code for some other use. I have removed all the tests.
3248 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3249 feature that was used to support insecure browsers during the U.S. crypto
3250 embargo. It requires special client support, and Exim is probably the
3251 only MTA that supported it -- and would never use it because real RSA is
3252 always available. This code has been removed, because it had the bad
3253 effect of slowing Exim down by computing (never used) parameters for the
3254 RSA_EXPORT functionality.
3256 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3257 authenticator to fail if there's a tab character in the incoming data
3258 (there should never be unless someone is messing about, as it's supposed
3259 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3260 if the connection is using TLS or if the remote IP is the same as the
3261 local IP, and the "valid-client-cert option" if a client certificate has
3264 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3265 authenticators. This can be used for authorization after authentication
3266 succeeds. (In the case of plaintext, it servers for both authentication
3269 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3270 if any retry times were supplied.
3272 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3273 connection, where there is no client IP address to check. In this
3274 situation, the verify now always succeeds.
3276 PH/25 Applied John Jetmore's -Mset patch.
3278 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3280 PH/27 In a string expansion for a processed (not raw) header when multiple
3281 headers of the same name were present, leading whitespace was being
3282 removed from all of them, but trailing whitespace was being removed only
3283 from the last one. Now trailing whitespace is removed from each header
3284 before concatenation. Completely empty headers in a concatenation (as
3285 before) are ignored.
3287 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3288 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3290 PH/29 [Removed. This was a change that I later backed out, and forgot to
3291 correct the ChangeLog entry (that I had efficiently created) before
3292 committing the later change.]
3294 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3295 address errors (4xx response to RCPT) over the same connection as other
3296 messages routed to the same hosts. Such deliveries are always "forced",
3297 so retry times are not inspected. This resulted in far too many retries
3298 for the affected addresses. The effect occurred only when there were more
3299 hosts than the hosts_max_try setting in the smtp transport when it had
3300 the 4xx errors. Those hosts that it had tried were not added to the list
3301 of hosts for which the message was waiting, so if all were tried, there
3302 was no problem. Two fixes have been applied:
3304 (i) If there are any address or message errors in an SMTP delivery, none
3305 of the hosts (tried or untried) are now added to the list of hosts
3306 for which the message is waiting, so the message should not be a
3307 candidate for sending over the same connection that was used for a
3308 successful delivery of some other message. This seems entirely
3309 reasonable: after all the message is NOT "waiting for some host".
3310 This is so "obvious" that I'm not sure why it wasn't done
3311 previously. Hope I haven't missed anything, but it can't do any
3312 harm, as the worst effect is to miss an optimization.
3314 (ii) If, despite (i), such a delivery is accidentally attempted, the
3315 routing retry time is respected, so at least it doesn't keep
3316 hammering the server.
3318 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3319 in ${readsocket because some servers need this prod.
3321 PH/32 Added some extra debug output when updating a wait-xxx database.
3323 PH/33 The hint "could be header name not terminated by colon", which has been
3324 given for certain expansion errors for a long time, was not being given
3325 for the ${if def:h_colon_omitted{... case.
3327 PH/34 The spec says: "With one important exception, whenever a domain list is
3328 being scanned, $domain contains the subject domain." There was at least
3329 one case where this was not true.
3331 PH/35 The error "getsockname() failed: connection reset by peer" was being
3332 written to the panic log as well as the main log, but it isn't really
3333 panic-worthy as it just means the connection died rather early on. I have
3334 removed the panic log writing for the ECONNRESET error when getsockname()
3337 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3338 runs only) independently of the message's sender address. This meant
3339 that, if the 4xx error was in fact related to the sender, a different
3340 message to the same recipient with a different sender could confuse
3341 things. In particular, this can happen when sending to a greylisting
3342 server, but other circumstances could also provoke similar problems.
3343 I have changed the default so that the retry time for these errors is now
3344 based a combination of the sender and recipient addresses. This change
3345 can be overridden by setting address_retry_include_sender=false in the
3348 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3349 remote server are returned as part of bounce messages. This was not
3350 happening for LMTP over a pipe (the lmtp transport), but now it is the
3351 same for both kinds of LMTP.
3353 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3354 in header lines that were in fact CNAMEs. This is no longer the case.
3356 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3357 and queue runs started by the daemon processed all messages. This has
3358 been fixed so that -R and -S can now usefully be given with -q<time>.
3360 PH/40 Import PCRE release 6.7 (fixes some bugs).
3362 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3364 PH/42 Give an error if -q is specified more than once.
3366 PH/43 Renamed the variables $interface_address and $interface_port as
3367 $received_ip_address and $received_port, to make it clear that these
3368 values apply to message reception, and not to the outgoing interface when
3369 a message is delivered. (The old names remain recognized, of course.)
3371 PH/44 There was no timeout on the connect() call when using a Unix domain
3372 socket in the ${readsocket expansion. There now is.
3374 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3375 be meaningful with "accept".
3377 SC/01 Eximstats V1.43
3378 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3380 SC/02 Eximstats V1.44
3381 Use a glob alias rather than an array ref in the generated
3382 parser. This improves both readability and performance.
3384 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3385 Collect SpamAssassin and rejection statistics.
3386 Don't display local sender or destination tables unless
3387 there is data to show.
3388 Added average volumes into the top table text output.
3390 SC/04 Eximstats V1.46
3391 Collect data on the number of addresses (recipients)
3392 as well as the number of messages.
3394 SC/05 Eximstats V1.47
3395 Added 'Message too big' to the list of mail rejection
3396 reasons (thanks to Marco Gaiarin).
3398 SC/06 Eximstats V1.48
3399 Mainlog lines which have GMT offsets and are too short to
3400 have a flag are now skipped.
3402 SC/07 Eximstats V1.49 (Alain Williams)
3403 Added the -emptyok flag.
3405 SC/08 Eximstats V1.50
3406 Fixes for obtaining the IP address from reject messages.
3408 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3409 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3410 whitespace changes from 4.64-PH/27
3412 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3415 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3416 are found, allow negative numbers in numeric criteria)
3418 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3420 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3421 to match changes made in 4.64-PH/43
3423 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3425 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3426 "domain" is actually an IP address, and force a failure. This locks out
3427 those revolvers/nameservers that support "A-for-A" lookups, in
3428 contravention of the specifications.
3430 PH/48 When a host name was looked up from an IP address, and the subsequent
3431 forward lookup of the name timed out, the host name was left in
3432 $sender_host_name, contrary to the specification.
3434 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3435 restricted to single-key lookups, Exim was not diagnosing an error if
3436 * or *@ was used with a query-style lookup.
3438 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3440 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3441 long ago, but noone interested enough thought of it. Let's just say that
3442 the "1.1" means that there are some new functions that weren't there at
3443 some point in the past.
3445 PH/51 Error processing for expansion failure of helo_data from an smtp
3446 transport during callout processing was broken.
3448 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3449 tested/used via the -bh/-bhc/-bs options.
3451 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3452 bug, fixed in subsequent PCRE releases).
3454 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3455 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3457 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3463 SC/01 Use a glob alias rather than an array ref in eximstats generated
3464 parser. This improves both readability and performance.
3466 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3467 Don't display local sender or destination tables in eximstats unless
3468 there is data to show.
3469 Added average volumes into the eximstats top table text output.
3471 SC/03 Collect data on the number of addresses (recipients) as well
3472 as the number of messages in eximstats.
3474 TF/01 Correct an error in the documentation for the redirect router. Exim
3475 does (usually) call initgroups() when daemonizing.
3477 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3478 with consistent privilege compared to when running as a daemon.
3480 TF/03 Note in the spec that $authenticated_id is not set for local
3481 submissions from trusted users.
3483 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3484 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3486 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3487 by adding some example configuration directives to the default
3488 configuration file. A little bit of work is required to uncomment the
3489 directives and define how usernames and passwords are checked, but
3490 there is now a framework to start from.
3492 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3493 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3494 without this. I don't know how relevant this is to other LDAP libraries.
3496 PH/02 Add the verb name to the "unknown ACL verb" error.
3498 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3500 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3502 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3503 directory not expanded when it should be if an expanded home directory
3504 was set for the address (which is overridden by the transport).
3506 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3509 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3510 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3511 because it is too late at that time, and has no effect.
3513 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3514 security issue with \' (bugzilla #107). I could not use the
3515 PQescapeStringConn() function, because it needs a PGconn value as one of
3518 PH/08 When testing addresses using -bt, indicate those final addresses that
3519 are duplicates that would not cause an additional delivery. At least one
3520 person was confused, thinking that -bt output corresponded to deliveries.
3521 (Suppressing duplicates isn't a good idea as you lose the information
3522 about possibly different redirections that led to the duplicates.)
3524 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3525 systems where poll() doesn't work, in particular OS X.
3527 PH/10 Added more information to debugging output for retry time not reached.
3529 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3530 operations in malware.c.
3532 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3535 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3536 syslog_duplication set false, log lines that would normally be written
3537 both the the main log and to the reject log were not written to syslog at
3540 PH/14 In the default configuration, change the use of "message" in ACL warn
3541 statements to "add_header".
3543 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3544 not followed by a command (e.g. "seen endif").
3546 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3547 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3550 PH/17 Added extra conditions to the default value of delay_warning_condition
3554 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3555 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3556 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3559 The Auto-Submitted: and various List- headers are standardised, whereas I
3560 don't think Precedence: ever was.
3562 PH/18 Refactored debugging code in route_finduser() to show more information,
3563 in particular, the error code if getpwnam() issues one.
3565 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3566 This is apparently needed in addition to the PH/07 change above to avoid
3567 any possible encoding problems.
3569 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3570 but not after initializing Perl.
3572 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3573 output them only if debugging. By default they are written stderr,
3574 apparently, which is not desirable.
3576 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3579 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3582 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3584 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3585 authenticated or an ident call has been made. Suppress the default
3586 values for $authenticated_id and $authenticated_sender (but permit -oMai
3587 and -oMas) when testing with -bh.
3589 PH/24 Re-jigged the order of the tests in the default configuration so that the
3590 tests for valid domains and recipients precede the DNS black list and CSA
3591 tests, on the grounds that those ones are more expensive.
3593 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3594 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3595 This bug exists in every version of Exim that I still have, right back to
3598 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3599 However, an attempt to turn on case-sensitivity in a regex key by
3600 including (?-i) didn't work because the subject string was already
3601 lowercased, and the effects were non-intuitive. It turns out that a
3602 one-line patch can be used to allow (?-i) to work as expected.
3608 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3609 other effects) broke the use of negated acl sub-conditions.
3611 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3614 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3615 "Deny" causes Exim to reject the incoming connection with a 554 error.
3616 Unfortunately, if there is a major crisis, such as a disk failure,
3617 tcp-wrappers gives "deny", whereas what one would like would be some
3618 kind of temporary error. A kludge has been added to help with this.
3619 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3620 554 error is used if errno is still zero or contains ENOENT (which occurs
3621 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3624 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3626 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3627 errors. Otherwise a message that provokes a temporary error (when other
3628 messages do not) can cause a whole host to time out.
3630 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3631 addresses were routed directly to files or pipes from a redirect router.
3632 File deliveries just didn't batch; pipe deliveries might have suffered
3635 PH/06 A failure to get a lock for a hints database would erroneously always say
3636 "Failed to get write lock", even when it was really a read lock.
3638 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3639 of 0600. This has been changed to use the value of the lockfile_mode
3640 option (which defaults to 0600).
3642 PH/08 Applied small patch from the Sieve maintainer.
3644 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3645 folder from quota calculations, a direct delivery into this folder messed
3646 up the contents of the maildirsize file. This was because the regex was
3647 used only to exclude .Trash (or whatever) when the size of the mailbox
3648 was calculated. There was no check that a delivery was happening into an
3649 excluded directory. This bug has been fixed by ignoring all quota
3650 processing for deliveries into excluded directories.
3652 PH/10 Added the maildirfolder_create_regex option to appendfile.
3658 PH/01 The code for finding all the local interface addresses on a FreeBSD
3659 system running IPv6 was broken. This may well have applied to all BSD
3660 systems, as well as to others that have similar system calls. The broken
3661 code found IPv4 interfaces correctly, but gave incorrect values for the
3662 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3663 that it would not match correctly against @[] and not recognize the IPv6
3666 PH/02 The ipliteral router was not recognizing addresses of the form user@
3667 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3669 PH/03 Added disable_ipv6.
3671 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3672 decoded form, because it is most often used to construct To: headers
3673 lines in autoreplies, and the decoded form may well be syntactically
3674 invalid. However, $reply_address has leading white space removed, and all
3675 newlines turned into spaces so that the autoreply transport does not
3678 PH/05 If group was specified without a user on a router, and no group or user
3679 was specified on a transport, the group from the router was ignored.
3681 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3682 for visible compile-time settings that can be used to change these
3683 numbers, for those that want even more. Backwards compatibility with old
3684 spool files has been maintained. However, going back to a previous Exim
3685 release will lost any variables that are in spool files.
3687 PH/07 Two small changes when running in the test harness: increase delay when
3688 passing a TCP/IP connection to a new process, in case the original
3689 process has to generate a bounce, and remove special handling of
3690 127.0.0.2 (sic), which is no longer necessary.
3692 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3693 be the same on different OS.
3695 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3698 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3699 whether --show-vars was specified or not
3701 JJ/02 exipick: Added support for new ACL variable spool format introduced
3704 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3705 syntactically invalid From: or Reply-to: line, and a filter used this to
3706 generate an autoreply, and therefore failed to obtain an address for the
3707 autoreply, Exim could try to deliver to a non-existent relative file
3708 name, causing unrelated and misleading errors. What now happens is that
3709 it logs this as a hard delivery error, but does not attempt to create a
3712 PH/11 The exinext utility has a -C option for testing purposes, but although
3713 the given file was scanned by exinext itself; it wasn't being passed on
3714 when Exim was called.
3716 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3717 an end-of-file indication when reading a command response.
3719 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3720 compiled. In many other places in Exim, IPv6 addresses are always
3721 recognized, so I have changed this. It also means that IPv4 domain
3722 literals of the form [IPV4:n.n.n.n] are now always recognized.
3724 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3725 used if the router is not running as root, for example, when verifying at
3726 ACL time, or when using -bh. The debugging output from this situation was
3727 non-existent - all you got was a failure to exec. I have made two
3730 (a) Failures to set uid/gid, the current directory, or a process leader
3731 in a subprocess such as that created by queryprogram now generate
3732 suitable debugging output when -d is set.
3734 (b) The queryprogram router detects when it is not running as root,
3735 outputs suitable debugging information if -d is set, and then runs
3736 the subprocess without attempting to change uid/gid.
3738 PH/15 Minor change to Makefile for building test_host (undocumented testing
3741 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3742 additional section of a DNS packet that returns MX or SRV records.
3743 Instead, it always explicitly searches for A/AAAA records. This avoids
3744 major problems that occur when a DNS server includes only records of one
3745 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3746 fixed another bug: if SRV records were looked up and the corresponding
3747 address records were *not* found in the additional section, the port
3748 values from the SRV records were lost.
3750 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3751 using the correct key (the original address) when searching the retry
3752 rules in order to find which one to use for generating the retry hint.
3754 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3755 adding the default one. Similarly, if it contains a Reply-To: header, the
3756 errors_reply_to option, if set, is not used.
3758 PH/19 When calculating a retry time, Exim used to measure the "time since
3759 failure" by looking at the "first failed" field in the retry record. Now
3760 it does not use this if it is later than than the arrival time of the
3761 message. Instead it uses the arrival time. This makes for better
3762 behaviour in cases where some deliveries succeed, thus re-setting the
3763 "first failed" field. An example is a quota failure for a huge message
3764 when small messages continue to be delivered. Without this change, the
3765 "time since failure" will always be short, possible causing more frequent
3766 delivery attempts for the huge message than are intended.
3767 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3769 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3770 $1, $2, $3) because the numerical variables can be reset during some
3771 expansion items (e.g. "match"), thereby losing the authentication data.
3773 PH/21 Make -bV show the size of off_t variables so that the test suite can
3774 decide whether to run tests for quotas > 2G.
3776 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3777 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3778 filecount value is greater than 2G or if a quota value is greater than 2G
3779 on a system where the size of off_t is not greater than 4, a panic error
3782 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3783 never match. The debug and -bh output now contains an explicit error
3784 message indicating a malformed IPv4 address or mask.
3786 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3787 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3788 PH/23 above applies.
3790 PH/25 Do not write to syslog when running in the test harness. The only
3791 occasion when this arises is a failure to open the main or panic logs
3792 (for which there is an explicit test).
3794 PH/26 Added the /no_tell option to "control=freeze".
3796 PH/27 If a host name lookup failed very early in a connection, for example, if
3797 the IP address matched host_lookup and the reverse lookup yielded a name
3798 that did not have a forward lookup, an error message of the form "no IP
3799 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3800 could be logged. Now it outputs the IP address instead of "NULL".
3802 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3803 allows the sender and the authenticated sender to be set when
3804 submitting a message from within Exim. Since child_open_exim() is
3805 documented for local_scan(), the new function should be too.
3807 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3808 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3809 results in an empty string is now treated as unset.
3811 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3813 PH/31 Added sender_verify_fail logging option.
3815 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3816 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3817 tidied the source and removed it altogether.
3819 PH/33 When a queue run was abandoned because the load average was too high, a
3820 log line was always written; now it is written only if the queue_run log
3821 selector is set. In addition, the log line for abandonment now contains
3822 information about the queue run such as the pid. This is always present
3823 in "start" and "stop" lines but was omitted from the "abandon" line.
3825 PH/34 Omit spaces between a header name and the colon in the error message that
3826 is given when verify = headers_syntax fails (if there are lots of them,
3827 the message gets confusing).
3829 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3830 names, as there are now some PTR records that contain slashes. This check
3831 is only to protect against broken name servers that fall over on strange
3832 characters, so the fact that it applies to all lookups doesn't matter.
3834 PH/36 Now that the new test suite is complete, we can remove some of the
3835 special code in Exim that was needed for the old test suite. For example,
3836 sorting DNS records because real resolvers return them in an arbitrary
3837 order. The new test suite's fake resolver always returns records in the
3840 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3841 bounces) except when queue_only is set, to avoid logging races between
3842 the different processes.
3844 PH/38 Panic-die if .include specifies a non-absolute path.
3846 PH/39 A tweak to the "H" retry rule from its user.
3848 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3849 a label. They prevented compilation on older perls.
3851 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3852 a warning to be raised on newish perls.
3854 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3855 on queue. Changes to match documented behaviour of showing count of
3856 messages matching specified criteria.
3858 PH/40 Changed the default ident timeout from 30s to 5s.
3860 PH/41 Added support for the use of login_cap features, on those BSD systems
3861 that have them, for controlling the resources used by pipe deliveries.
3863 PH/42 The content-scanning code uses fopen() to create files in which to put
3864 message data. Previously it was not paying any attention to the mode of
3865 the files. Exim runs with umask(0) because the rest of the code creates
3866 files with open(), and sets the required mode explicitly. Thus, these
3867 files were ending up world-writeable. This was not a big issue, because,
3868 being within the spool directory, they were not world-accessible. I have
3869 created a function called modefopen, which takes an additional mode
3870 argument. It sets umask(777), creates the file, chmods it to the required
3871 mode, then resets the umask. All the relevant calls to fopen() in the
3872 content scanning code have been changed to use this function.
3874 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3875 to 24 hours. This avoids potential overflow problems when processing G
3876 and H retry rules. I suspect nobody ever tinkers with this value.
3878 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3880 PH/45 When the plaintext authenticator is running as a client, the server's
3881 challenges are checked to ensure they are valid base64 strings. By
3882 default, the authentication attempt is cancelled if an invalid string is
3883 received. Setting client_ignore_invalid_base64 true ignores these errors.
3884 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3885 they are received. Thus, the responses can be made to depend on the
3886 challenges. If an invalid string is ignored, an empty string is placed in
3889 PH/46 Messages that are created by the autoreply transport now contains a
3890 References: header, in accordance with RFCs 2822 and 3834.
3892 PH/47 Added authenticated_sender_force to the smtp transport.
3894 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3896 PH/49 Installed latest patch from the Sieve maintainer.
3898 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3899 was also set, the appendfile transport was unnecessarily scanning a
3900 directory of message files (e.g. for maildir delivery) to find the count
3901 of files (along with the size), even though it did not need this
3902 information. It now does the scan only if it needs to find either the
3903 size of the count of files.
3905 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3907 PH/52 Two bugs concerned with error handling when the smtp transport is
3910 (i) Exim was not creating retry information for temporary errors given
3911 for individual recipients after the DATA command when the smtp transport
3912 was used in LMTP mode. This meant that they could be retried too
3913 frequently, and not timed out correctly.
3915 (ii) Exim was setting the flag that allows error details to be returned
3916 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3917 recipients that were returned after the DATA command.
3919 PH/53 This is related to PH/52, but is more general: for any failing address,
3920 when detailed error information was permitted to be returned to the
3921 sender, but the error was temporary, then after the final timeout, only
3922 "retry timeout exceeded" was returned. Now it returns the full error as
3923 well as "retry timeout exceeded".
3925 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3926 do this, and (what is worse) MTAs that accept it.
3928 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3929 will now be deprecated.
3931 PH/56 New os.c-cygwin from the Cygwin maintainer.
3933 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3934 formats (previously only available in exim formats via -bpr, -bpru,
3935 and -bpra. Now also available in native and exiqgrep formats)
3937 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3938 with very large, slow to parse queues
3940 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3942 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3944 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3945 responses to authentication challenges, though it was showing the
3946 challenges; (ii) I've removed the CR characters from the debug output for
3949 PH/58 Allow for the insertion of a newline as well as a space when a string
3950 is turned into more than one encoded-word during RFC 2047 encoding. The
3951 Sieve code now uses this.
3953 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3954 data_4xx, lost_connection, tls_required.
3956 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3957 message was being sent as an SMTP response.
3959 PH/61 Add -l and -k options to exicyclog.
3961 PH/62 When verifying, if an address was redirected to one new address, so that
3962 verification continued, and the new address failed or deferred after
3963 having set something in $address_data, the value of $address_data was not
3964 passed back to the ACL. This was different to the case when no
3965 redirection occurred. The value is now passed back in both cases.
3967 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3968 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3969 use login_cap.h, so on its own it isn't the distinguishing feature. The
3970 new name refers directly to the setclassresources() function.
3972 PH/65 Added configuration files for NetBSD3.
3974 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3976 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3977 is preferred over IPv4.
3979 PH/68 The bounce_return_message and bounce_return_body options were not being
3980 honoured for bounces generated during the reception of non-SMTP messages.
3981 In particular, this applied to messages rejected by the ACL. This bug has
3982 been fixed. However, if bounce_return_message is true and bounce_return_
3983 body is false, the headers that are returned for a non-SMTP message
3984 include only those that have been read before the error was detected.
3985 (In the case of an ACL rejection, they have all been read.)
3987 PH/69 The HTML version of the specification is now built in a directory called
3988 spec_html instead of spec.html, because the latter looks like a path with
3989 a MIME-type, and this confuses some software.
3991 PH/70 Catch two compiler warnings in sieve.c.
3993 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3994 function verify_get_ident() calls ip_connect() to connect a socket, but
3995 if the "connect()" function timed out, ip_connect() used to close the
3996 socket. However, verify_get_ident() also closes the socket later, and in
3997 between Exim writes to the log, which may get opened at this point. When
3998 the socket was closed in ip_connect(), the log could get the same file
3999 descriptor number as the socket. This naturally causes chaos. The fix is
4000 not to close the socket in ip_connect(); the socket should be closed by
4001 the function that creates it. There was only one place in the code where
4002 this was missing, in the iplookup router, which I don't think anybody now
4003 uses, but I've fixed it anyway.
4005 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4006 well as to direct DNS lookups. Otherwise the handling of names in host
4007 lists is inconsistent and therefore confusing.
4013 PH/01 Two changes to the default runtime configuration:
4015 (1) Move the checks for relay_from_hosts and authenticated clients from
4016 after to before the (commented out) DNS black list checks.
4018 (2) Add control=submission to the relay_from_hosts and authenticated
4019 clients checks, on the grounds that messages accepted by these
4020 statements are most likely to be submissions.
4022 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4024 (1) Generate an error if the third argument for the ${prvs expansion is
4027 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4030 (3) Reset the variables that are obtained from the first argument of
4031 ${prvscheck and used in the second argument before leaving the code,
4032 because their memory is reclaimed, so using them afterwards may do
4035 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4036 one (it's much easier than Tom thought :-).
4038 (5) Because of (4), we can now allow for the use of $prvscheck_result
4039 inside the third argument.
4041 PH/03 For some reason, the default setting of PATH when running a command from
4042 a pipe transport was just "/usr/bin". I have changed it to
4045 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4046 anything to be listed in the output from -bV.
4048 PH/05 When a filter generated an autoreply, the entire To: header line was
4049 quoted in the delivery log line, like this:
4051 => >A.N.Other <ano@some.domain> <original@ddress> ...
4053 This has been changed so that it extracts the operative address. There
4054 may be more than one such address. If so, they are comma-separated, like
4057 => >ano@some.domain,ona@other.domain <original@ddress> ...
4059 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4060 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4061 not being looked up in the rDNS to get a host name, Exim was showing the
4062 IP address twice in Received: lines, even though the IP addresses were
4063 identical. For example:
4065 Received: from [1.2.3.4] (helo=[1.2.3.4])
4067 However, if the real host name was known, it was omitting the HELO data
4068 if it matched the actual IP address. This has been tidied up so that it
4069 doesn't show the same IP address twice.
4071 PH/07 When both +timestamp and +memory debugging was on, the value given by
4072 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4073 called by the debug printing, thereby overwriting the timestamp buffer.
4074 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4076 PH/08 When the original message was included in an autoreply transport, it
4077 always said "this is a copy of the message, including all the headers",
4078 even if body_only or headers_only was set. It now gives an appropriate
4081 PH/09 Applied a patch from the Sieve maintainer which:
4083 o fixes some comments
4084 o adds the (disabled) notify extension core
4085 o adds some debug output for the result of if/elsif tests
4086 o points to the current vacation draft in the documentation
4087 and documents the missing references header update
4091 o fixes a bug in processing the envelope test (when testing
4092 multiple envelope elements, the last element determined the
4095 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4096 Electronic Mail") by including:
4098 Auto-submitted: auto-generated
4100 in the messages that it generates (bounce messages and others, such as
4101 warnings). In the case of bounce messages for non-SMTP messages, there was
4102 also a typo: it was using "Auto_submitted" (underscore instead of
4103 hyphen). Since every message generated by Exim is necessarily in response
4104 to another message, thes have all been changed to:
4106 Auto-Submitted: auto-replied
4108 in accordance with these statements in the RFC:
4110 The auto-replied keyword:
4112 - SHOULD be used on messages sent in direct response to another
4113 message by an automatic process,
4115 - MUST NOT be used on manually-generated messages,
4117 - MAY be used on Delivery Status Notifications (DSNs) and Message
4118 Disposition Notifications (MDNs),
4120 - MUST NOT be used on messages generated by automatic or periodic
4121 processes, except for messages which are automatic responses to
4124 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4125 to the default Received: header definition.
4127 PH/12 Added log selector acl_warn_skipped (default on).
4129 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4130 variables because (a) they are in the wrong storage pool and (b) even if
4131 they were copied, it wouldn't work properly because of the caching.
4133 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4134 checking when decoding. Apparently there are clients that generate
4135 overlong encoded strings. Why am I not surprised?
4137 PH/15 If the first argument of "${if match_address" was not empty, but did not
4138 contain an "@" character, Exim crashed. Now it writes a panic log message
4139 and treats the condition as false.
4141 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4143 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4144 extension "envelope-auth". The code is finished and in agreement with
4145 other implementations, but there is no documentation so far and in fact,
4146 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4147 not changing the active code.
4149 Print executed "if" and "elsif" statements when debugging is used. This
4150 helps a great deal to understand what a filter does.
4152 Document more things not specified clearly in RFC3028. I had all this
4153 sorted out, when out of a sudden new issues came to my mind. Oops."
4155 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4158 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4159 canonical form (as documented). However, after a host name lookup from
4160 the IP address, check_host() was doing a simple string comparison with
4161 addresses acquired from the DNS when checking that the found name did
4162 have the original IP as one of its addresses. Since any found IPv6
4163 addresses are likely to be in abbreviated form, the comparison could
4164 fail. Luckily, there already exists a function for doing the comparison
4165 by converting both addresses to binary, so now that is used instead of
4166 the text comparison.
4168 PH/20 There was another similar case to PH/19, when a complete host name was
4169 given in a host list; looking up its IP address could give an abbreviated
4170 form, whereas the current host's name might or might not be abbreviated.
4171 The same fix has been applied.
4177 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4178 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4181 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4182 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4184 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4186 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4187 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4188 Also, for "Auto-Submitted", treat anything other than "no" as
4189 non-personal, in accordance with RFC 3834. (Previously it treated
4190 anything starting "auto-" as non-personal.)
4192 TF/01 The control=submission/name=... option had a problem with syntax
4193 errors if the name included a slash character. The /name= option
4194 now slurps the rest of the string, so it can include any characters
4195 but it must come last in the list of options (after /sender_retain
4198 PH/05 Some modifications to the interface to the fake nameserver for the new
4206 TK/01 Added the "success_on_redirect" address verification option. See
4207 NewStuff for rationale and an example.
4209 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4211 PH/02 Patch to exigrep to allow it to work on syslog lines.
4213 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4214 fread() to skip over the body file's header line, because in Cygwin the
4215 header line is locked and is inaccessible.
4217 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4218 co-exist for some time) to make it clear that it is the Exim ID that is
4219 referenced, not the Message-ID: header line.
4221 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4222 string_format() function, because snprintf() does not exist on all
4225 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4226 ${stat: expansion item.
4228 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4229 protocol synchronization error", to keep the pedants happy.
4231 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4232 well as for IRIX systems, when gcc is being used. See the host.c source
4235 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4237 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4240 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4241 transport and to the smtp transport in LMTP mode.
4243 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4245 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4246 run a filter in a subprocess. This could lead to confusion in subsequent
4247 lookups in the parent process. There should also be a search_tidyup() at
4248 the end of the subprocess.
4250 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4251 only if the host matched helo_try_verify_hosts, which caused the
4252 verification to occur when the EHLO/HELO command was issued. The ACL just
4253 tested the remembered result. Now, if a previous verification attempt has
4254 not happened, "verify = helo" does it there and then.
4256 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4258 TK/03 Fix log output including CR from clamd.
4260 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4261 exist provoked a memory error which could cause a segfault.
4263 PH/15 Installed PCRE 6.2
4265 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4267 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4268 of the problem. Specifically, suggested +O2 rather than +O1 for the
4271 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4273 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4274 record, the port was not correctly logged when the outgoing_port log
4275 selector was set (it logged the transort's default port).
4277 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4278 fallback_hosts, and "hosts" in the smtp transport.
4280 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4281 host errors such as "Connection refused".
4283 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4284 authentication with radiusclient 0.4.9:
4286 - Error returned from rc_read_config was caught wrongly
4287 - Username/password not passed on to radius server due to wrong length.
4289 The presumption is that some radiusclient API changes for 4.51/PH/17
4290 were not taken care of correctly. The code is still untested by me (my
4291 Linux distribution still has 0.3.2 of radiusclient), but it was
4292 contributed by a Radius user.
4294 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4295 expanding the "port" option of the smtp transport.
4297 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4298 while reading a MIME header. Thanks to Tom Hughes for a patch.
4300 PH/24 Include config.h inside local_scan.h so that configuration settings are
4303 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4304 that in an ACL for RCPT (for example), you can examine exactly what was
4307 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4308 commands, but it was not correctly comparing the address with the actual
4309 client host address. Thus, it would show the EHLO address in Received:
4310 header lines when this was not necessary.
4312 PH/27 Added the % operator to ${eval:}.
4314 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4315 it should be ignoring failures (because with -C, for example, it has lost
4316 privilege). It wasn't ignoring creation failures other than "already
4319 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4322 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4323 because an input line was too long, either on its own, or by virtue of
4324 too many continuations, the temporary file was not being removed, and the
4325 return code was incorrect.
4327 PH/31 Missing "BOOL" in function definition in filtertest.c.
4329 PH/32 Applied Sieve patches from the maintainer.
4331 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4333 PH/33 Added "verify = not_blind".
4335 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4336 Local/Makefile (with some defaults set). These are used in built scripts
4337 such as exicyclog, but they have never been used in the exim_install
4338 script (though there are many overriding facilities there). I have
4339 arranged that the exim_install script now takes note of these two
4342 PH/35 Installed configuration files for Dragonfly.
4344 PH/36 When a locally submitted message by a trusted user did not contain a
4345 From: header, and the sender address was obtained from -f or from an SMTP
4346 MAIL command, and the trusted user did not use -F to supply a sender
4347 name, $originator_name was incorrectly used when constructing a From:
4348 header. Furthermore, $originator_name was used for submission mode
4349 messages from external hosts without From: headers in a similar way,
4350 which is clearly wrong.
4352 PH/37 Added control=suppress_local_fixups.
4354 PH/38 When log_selector = +received_sender was set, and the addition of the
4355 sender made the log line's construction buffer exactly full, or one byte
4356 less than full, an overflow happened when the terminating "\n" was
4359 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4360 when the result of a list match is failure because a DNS lookup failed.
4362 PH/40 RM_COMMAND is now used in the building process.
4364 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4365 the "build-* directories that it finds.
4367 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4368 address in a domain literal was a prefix of an interface address.
4370 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4371 when verifying a sender address, unless rewrite_headers is false.
4373 PH/44 Wrote a long comment about why errors_to addresses are verified as
4374 recipients, not senders.
4376 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4377 the ratelimit ACL was added.
4379 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4381 PH/46 Added extra information about PostgreSQL errors to the error string.
4383 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4384 avoiding the need to install special zones in a real server. This is
4385 backwards compatible; if it can't find the fake resolver, it drops back.
4386 Thus, both old and new test suites can be run.
4388 TF/02 Added util/ratelimit.pl
4390 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4391 clock is set back in time.
4393 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4394 Candler <B.Candler@pobox.com>.
4396 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4397 OK for addresses that are the result of redirections.
4399 PH/48 A number of further additions for the benefit of the new test suite,
4400 including a fake gethostbyname() that interfaces to the fake DNS resolver
4403 TF/06 The fix for widen_domains has also been applied to qualify_single and
4404 search_parents which are the other dnslookup options that can cause
4407 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4410 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4412 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4413 service" if it encountered an MX record with an empty target hostname.
4414 The message is now "an MX or SRV record indicated no SMTP service".
4416 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4417 if the DNS of the sending site is misconfigured. This is quite a
4418 common situation. This change restores the behaviour of treating a
4419 helo verification defer as a failure.
4421 PH/51 If self=fail was set on a router, the bounce message did not include the
4422 actual error message.
4428 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4430 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4431 command itself ended in error, the underlying message about the transport
4432 filter timeout was being overwritten with the pipe command error. Now the
4433 underlying error message should be appended to the second error message.
4435 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4437 PH/02 Reduce the amount of output that "make" produces by default. Full output
4438 can still be requested.
4440 PH/03 The warning log line about a condition test deferring for a "warn" verb
4441 was being output only once per connection, rather than after each
4442 occurrence (because it was using the same function as for successful
4443 "warn" verbs). This seems wrong, so I have changed it.
4445 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4446 it should not have, which might have caused a crash in the right
4447 circumstances, but probably never did.
4449 PH/04 Installed a modified version of Tony Finch's patch to make submission
4450 mode fix the return path as well as the Sender: header line, and to
4451 add a /name= option so that you can make the user's friendly name appear
4454 TF/03 Added the control = fakedefer ACL modifier.
4456 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4457 Mark Lowes for thorough testing.
4459 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4461 TK/03 Merged latest SRS patch from Miles Wilton.
4463 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4464 with the definition in sysexits.h (which is #included earlier).
4465 Fortunately, Exim does not actually use EX_OK. The code used to try to
4466 preserve the sysexits.h value, by assuming that macro definitions were
4467 scanned for macro replacements. I have been disabused of this notion,
4468 so now the code just undefines EX_OK before #including unistd.h.
4470 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4471 in the smtp transport. When a block could not be written in a single
4472 write() function, the timeout was being re-applied to each part-write.
4473 This seems wrong - if the receiver was accepting one byte at a time it
4474 would take for ever. The timeout is now adjusted when this happens. It
4475 doesn't have to be particularly precise.
4477 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4478 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4480 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4481 without a domain if the check to <postmaster@domain> fails.
4483 SC/01 Eximstats: added -xls and the ability to specify output files
4484 (patch written by Frank Heydlauf).
4486 SC/02 Eximstats: use FileHandles for outputting results.
4488 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4490 SC/04 Eximstats: fixed display of large numbers with -nvr option
4492 SC/05 Eximstats: fixed merging of reports with empty tables.
4494 SC/06 Eximstats: added the -include_original_destination flag
4496 SC/07 Eximstats: removed tabs and trailing whitespace.
4498 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4500 TK/06 MBOX spool code: Add real "From " MBOX separator line
4501 so the .eml file is really in mbox format (even though
4502 most programs do not really care). Patch from Alex Miller.
4504 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4505 The latter is generated from $received_to and is only set if the
4506 message has one envelope recipient. SA can use these headers,
4507 obviously out-of-the-box. Patch from Alex Miller.
4509 PH/08 The ${def test on a variable was returning false if the variable's
4510 value was "0", contrary to what the specification has always said!
4511 The result should be true unless the variable is empty.
4513 PH/09 The syntax error of a character other than { following "${if
4514 def:variable_name" (after optional whitespace) was not being diagnosed.
4515 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4516 accidental colon was present, for example, could give incorrect results.
4518 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4519 result is used (not including appendfile, where other changes are about
4522 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4523 This involved changing a lot of size variables from int to off_t. It
4524 should work with maildirs and everything.
4526 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4527 spamd dying while we are connected to it.
4529 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4532 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4533 function for BDB 4.3.
4535 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4537 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4538 the sources for PCRE have been re-arranged and more files are now
4541 PH/15 The code I had for printing potentially long long variables in PH/11
4542 above was not the best (it lost precision). The length of off_t variables
4543 is now inspected at build time, and an appropriate printing format (%ld
4544 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4545 to be "long long int" or "long int". This is needed for the internal
4546 formatting function string_vformat().
4548 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4549 the configuration file to be ":syslog", then the script "guesses" where
4550 the logs files are, rather than using the compiled in default. In our
4551 case the guess is not the same as the compiled default, so the script
4552 suddenly stopped working when I started to use syslog. The patch checks
4553 to see if log_file_path is "". If so, it attempts to read it from exim
4554 with no configuration file to get the compiled in version, before it
4555 falls back to the previous guessing code."
4557 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4558 implementing BATV in an Exim configuration. See NewStuff for the gory
4561 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4562 Makefile that are specific to HP-UX.
4564 PH/18 If the "use_postmaster" option was set for a recipient callout together
4565 with the "random" option, the postmaster address was used as the MAIL
4566 FROM address for the random test, but not for the subsequent recipient
4567 test. It is now used for both.
4569 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4570 patch removes a few documentation additions to RFC 3028, because the
4571 latest draft now contains them. It adds the new en;ascii-case comparator
4572 and a new error check for 8bit text in MIME parts. Comparator and
4573 require names are now matched exactly. I enabled the subaddress
4574 extension, but it is not well tested yet (read: it works for me)."
4576 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4577 rework some of the code of TK/09 above to avoid the hardwired use of
4578 "%lld" and "long long". Replaced the call to snprintf() with a call to
4581 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4582 records point to non-existent hosts", "retry timeout exceeded", and
4583 "retry time not reached for any host after a long failure period".
4585 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4586 experimental DomainKeys support:
4588 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4589 (2) On an error such as an illegally used "control", the wrong name for
4590 the control was given.
4592 These problems did NOT occur unless DomainKeys support was compiled.
4594 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4596 PH/24 Added ${if match_ip condition.
4598 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4599 databases so that it will be absolutely obvious if a crash occurs in the
4600 DB library. This is a regular occurrence (often caused by mis-matched
4603 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4604 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4605 picked up on a user's system that detects such things. There doesn't seem
4606 to be a gcc warning option for this - only an attribute that has to be
4607 put on the function's prototype. It seems that in Fedora Core 4 they have
4608 set this on a number of new functions. No doubt there will be more in due
4611 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4612 specify a transport. However, if an address that was verified by such a
4613 router was the subject of a callout, Exim crashed because it tried to
4614 read the rcpt_include_affixes from the non-existent transport. Now it
4615 just assumes that the setting of that option is false. This bug was
4616 introduced by 4.51/PH/31.
4618 PH/28 Changed -d+all to exclude +memory, because that information is very
4619 rarely of interest, but it makes the output a lot bigger. People tend to
4620 do -d+all out of habit.
4622 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4623 code in os-type was giving problems when libc.so lives in lib64, like on
4626 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4627 aren't the modern standard, and it seems that some systems' include files
4628 don't always have them. Exim was already checking for some of the newer
4629 ones like T_AAAA, and defining it itself. I've added checks for all the
4630 record types that Exim uses.
4632 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4633 not automatically generating a new one, as it is supposed to. This
4634 prevented TLS from working. If the file did exist, but contained invalid
4635 data, a new version was generated, as expected. It was only the case of a
4636 non-existent file that was broken.
4638 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4639 with a change in libdomainkeys > 0.64.
4641 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4642 from DNS. If the selector record carries the flag, it now has
4643 precedence over the domain-wide flag.
4645 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4647 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4648 the use of a port name that isn't defined in /etc/services) occurred, the
4649 message was deferred as in a normal delivery, and thus remained on the
4650 spool, instead of being failed because of the mua_wrapper setting. This
4651 is now fixed, and I tidied up some of the mua_wrapper messages at the
4654 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4655 the messages in a hash of arrays rather than using individual hashes.
4656 This is a bit cleaner and results in dramatic memory savings, albeit
4657 at a slight CPU cost.
4659 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4660 as requested by Marc Sherman.
4662 SC/10 Eximstats: added histograms for user specified patterns as requested
4665 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4667 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4668 fopen() in the content-scanning modules that did not already have it.
4674 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4675 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4677 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4679 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4681 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4682 to test Sieve filters that use "vacation".
4684 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4685 that changes the way the GnuTLS parameters are stored in the cache file.
4686 The new format can be generated externally. For backward compatibility,
4687 if the data in the cache doesn't make sense, Exim assumes it has read an
4688 old-format file, and it generates new data and writes a new file. This
4689 means that you can't go back to an older release without removing the
4692 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4693 work if there are any delivery delays because "one_time" forces the
4694 parent to be marked "delivered", so its unseen clone is never tried
4695 again. For this reason, Exim now forbids the simultaneous setting of
4698 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4699 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4700 entry if you want to know the details. The fix, however, neglected to
4701 consider the case when local delivery batching is involved. The test for
4702 "previously delivered" was not happening when checking to see if an
4703 address could be batched with a previous (undelivered) one; under
4704 certain circumstances this could lead to multiple deliveries to the same
4707 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4708 in its include files, and this causes problems building Exim.
4710 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4711 header_syntax) but Exim was just ignoring anything given after a slash.
4712 In particular, this caused confusion with an attempt to use "verify =
4713 reverse_host_lookup/defer_ok". An error is now given when options are
4714 supplied for verify items that do not have them. (Maybe reverse_host_
4715 lookup should have a defer_ok option, but that's a different point.)
4717 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4718 defined by RFC 821) to 2048, because there were problems with some AUTH
4719 commands, and RFC 1869 says the size should be increased for extended
4720 SMTP commands that take arguments.
4722 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4725 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4726 "unknown" error; now it says that the functionality isn't in the binary.
4728 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4729 an address' error message when a string expansion fails (syntax or
4730 whatever). Otherwise the password may appear in the log. Following change
4731 PH/42 below, there is no longer a chance of it appearing in a bounce
4734 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4736 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4737 of its addresses was ever tried. (Bugzilla bug #2.)
4739 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4740 the result incorrectly in the debug output. (It correctly added a newline
4741 to what was transported.)
4743 TF/01 Added $received_time.
4745 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4746 commented out examples of how to interface to a virus scanner and to
4747 SpamAssassin. Also added commented examples of av_scanner and
4748 spamd_address settings.
4750 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4751 and controls are allowed in which ACLs. There were a couple of minor
4752 errors. Some of the entries in the conditions table (which is a table of
4753 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4754 negation of where the condition IS allowed.
4756 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4758 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4759 header file does not have a version number, so I've had to invent a new
4760 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4761 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4762 radiusclient), but it was contributed by a Radius user.
4764 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4765 files or directories, for OpenSSL.
4767 PH/19 When an Exim process that is running as root has to create an Exim log
4768 file, it does so in a subprocess that runs as exim:exim so as to get the
4769 ownership right at creation (otherwise, other Exim processes might see
4770 the file with the wrong ownership). There was no test for failure of this
4771 fork() call, which would lead to the process getting stuck as it waited
4772 for a non-existent subprocess. Forks do occasionally fail when resources
4773 run out. I reviewed all the other calls to fork(); they all seem to check
4776 PH/20 When checking for unexpected SMTP input at connect time (before writing
4777 the banner), Exim was not dealing correctly with a non-positive return
4778 from the read() function. If the client had disconnected by this time,
4779 the result was a log entry for a synchronization error with an empty
4780 string after "input=" when read() returned zero. If read() returned -1
4781 (an event I could not check), uninitialized data bytes were printed.
4782 There were reports of junk text (parts of files, etc) appearing after
4785 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4787 PH/22 Added support for macro redefinition, and (re)definition in between
4788 driver and ACL definitions.
4790 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4791 forgetting to use the resulting value; it was using the unexpanded value.
4793 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4794 hadn't been configured. The fix is from Juergen Kreileder, who
4795 understands it better than I do:
4797 "Here's what I see happening with three configured cyrus_sasl
4798 authenticators configured (plain, login, cram-md5):
4800 On startup auth_cyrus_sasl_init() gets called for each of these.
4801 This means three calls to sasl_listmech() without a specified mech_list.
4802 => SASL tests which mechs of all available mechs actually work
4803 => three warnings about OTP not working
4804 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4806 With the patch, sasl_listmech() also gets called three times. But now
4807 SASL's mech_list option is set to the server_mech specified in the the
4808 authenticator. Or in other words, the answer from sasl_listmech()
4809 gets limited to just the mech you're testing for (which is different
4811 => the return list contains just 'plain' or 'login', 'cram-md5' or
4812 nothing depending on the value of ob->server_mech.
4814 I've just tested the patch: Authentication still works fine,
4815 unavailable mechs specified in the exim configuration are still
4816 caught, and the auth.log warnings about OTP are gone."
4818 PH/25 When debugging is enabled, the contents of the command line are added
4819 to the debugging output, even when log_selector=+arguments is not
4822 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4823 answer is "GNU", and only if the return is "GNU/something" is the answer
4826 PH/27 $acl_verify_message is now set immediately after the failure of a
4827 verification in an ACL, and so is available in subsequent modifiers. In
4828 particular, the message can be preserved by coding like this:
4830 warn !verify = sender
4831 set acl_m0 = $acl_verify_message
4833 Previously, $acl_verify_message was set only while expanding "message"
4834 and "log_message" when a very denied access.
4836 PH/28 Modified OS/os.c-Linux with
4838 -#ifndef OS_LOAD_AVERAGE
4839 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4841 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4842 nomenclature these days.)
4844 PH/29 Installed patch from the Sieve maintainer that adds the options
4845 sieve_useraddress and sieve_subaddress to the redirect router.
4847 PH/30 In these circumstances:
4848 . Two addresses routed to the same list of hosts;
4849 . First host does not offer TLS;
4850 . First host accepts first address;
4851 . First host gives temporary error to second address;
4852 . Second host offers TLS and a TLS session is established;
4853 . Second host accepts second address.
4854 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4855 and peerdn, if requested) that were in fact used only for the second
4858 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4859 attention to any local part prefix or suffix that was matched by the
4860 router that accepted the address. It now behaves in the same way as it
4861 does for delivery: the affixes are removed from the local part unless
4862 rcpt_include_affixes is set on the transport.
4864 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4865 timeout during the DATA phase of an incoming message.
4867 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4868 applied a patch sanctioned by the Sieve maintainer.
4870 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4871 the uid or gid is negative. A case of a negative gid caused this to be
4872 noticed. The fix allows for either to be negative.
4874 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4875 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4878 PH/36 The change PH/12 above was broken. Fixed it.
4880 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4881 the grounds that routing the same address twice would always produce the
4882 same answer. This might have been true once, but it is certainly no
4883 longer true now. Routing a child address may depend on the previous
4884 routing that produced that child. Some complicated redirection strategies
4885 went wrong when messages had multiple recipients, and made Exim's
4886 behaviour dependent on the order in which the addresses were given.
4888 I have moved the duplicate checking until after the routing is complete.
4889 Exim scans the addresses that are assigned to local and remote
4890 transports, and removes any duplicates. This means that more work will be
4891 done, as duplicates will always all be routed, but duplicates are
4892 presumably rare, so I don't expect this is of any significance.
4894 For deliveries to pipes, files, and autoreplies, the duplicate checking
4895 still happens during the routing process, since they are not going to be
4898 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4899 It corrects a timeout issue with spamd. This is Ian's comment: "The
4900 background is that sometimes spamd either never reads data from a
4901 connection it has accepted, or it never writes response data. The exiscan
4902 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4903 blindly assumes that writes won't block so it may never time out."
4905 PH/39 Allow G after quota size as well as K and M.
4907 PH/40 The value set for $authenticated_id in an authenticator may not contain
4908 binary zeroes or newlines because the value is written to log lines and
4909 to spool files. There was no check on this. Now the value is run through
4910 the string_printing() function so that such characters are converted to
4911 printable escape sequences.
4913 PH/41 $message_linecount is a new variable that contains the total number of
4914 lines in the message. Compare $body_linecount, which is the count for the
4917 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4918 bounce and delay warning messages, except in certain special cases, which
4921 (a) An SMTP error message from a remote host;
4922 (b) A message specified in a :fail: redirection;
4923 (c) A message specified in a "fail" command in a system filter;
4924 (d) A message specified in a FAIL return from the queryprogram router;
4925 (e) A message specified by the cannot_route_message router option.
4927 In these cases only, Exim does include the error details in bounce and
4928 warning messages. There are also a few cases where bland messages such
4929 as "unrouteable address" or "local delivery error" are given.
4931 PH/43 $value is now also set for the "else" part of a ${run expansion.
4933 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4934 being worked on, but at least Exim now implements the latest version to
4937 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4938 process to complete was treated as a delivery failure, a timeout while
4939 writing the message to the pipe was logged, but erroneously treated as a
4940 successful delivery. Such timeouts include transport filter timeouts. For
4941 consistency with the overall process timeout, these timeouts are now
4942 treated as errors, giving rise to delivery failures by default. However,
4943 there is now a new Boolean option for the pipe transport called
4944 timeout_defer, which, if set TRUE, converts the failures into defers for
4945 both kinds of timeout. A transport filter timeout is now identified in
4948 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4949 systems where "make" and "gmake" are different, calling "gmake" at top
4950 level broke things. I've arranged for the value of $(MAKE) to be passed
4951 from the Makefile to this script so that it can call the same version of
4955 A note about Exim versions 4.44 and 4.50
4956 ----------------------------------------
4958 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4959 changes of various kinds. As a consequence, a big documentation update was
4960 needed. This delayed the release for rather longer than seemed good, especially
4961 in the light of a couple of (minor) security issues. Therefore, the changes
4962 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4963 release. So 4.44 and 4.50 are in effect two different branches that both start
4966 I have left the 4.50 change log unchanged; it contains all the changes since
4967 4.43. The change log for 4.44 is below; many of its items are identical to
4968 those for 4.50. This seems to be the most sensible way to preserve the
4969 historical information.
4975 1. Minor wording change to the doc/README.SIEVE file.
4977 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4978 computation of the current number of files was incorrect.
4980 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4981 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4984 4. Give more explanation in the error message when the command for a transport
4985 filter fails to execute.
4987 5. There are several places where Exim runs a non-Exim command in a
4988 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4989 was being done only for the command run by the queryprogram router. It is
4990 now done for all such subprocesses. The other cases are: ${run, transport
4991 filters, and the commands run by the lmtp and pipe transports.
4993 6. Added CONFIGURE_GROUP build-time option.
4995 7. Some older OS have a limit of 256 on the maximum number of file
4996 descriptors. Exim was using setrlimit() to set 1000 as a large value
4997 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4998 systems. I've change it so that if it can't get 1000, it tries for 256.
5000 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5001 was an oversight, and furthermore, ever since the addition of extra
5002 controls (e.g. 4.43/32), the checks on when to allow different forms of
5003 "control" were broken. There should now be diagnostics for all cases when a
5004 control that does not make sense is encountered.
5006 9. Added the /retain_sender option to "control=submission".
5008 10. $recipients is now available in the predata ACL (oversight).
5010 11. Tidy the search cache before the fork to do a delivery from a message
5011 received from the command line. Otherwise the child will trigger a lookup
5012 failure and thereby defer the delivery if it tries to use (for example) a
5013 cached ldap connection that the parent has called unbind on.
5015 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5016 of $address_data from the recipient verification was clobbered by the
5017 sender verification.
5019 13. The value of address_data from a sender verification is now available in
5020 $sender_address_data in subsequent conditions in the ACL statement.
5022 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5024 15. Added a new option "connect=<time>" to callout options, to set a different
5027 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5028 was its contents. (It was OK if the option was not defined at all.)
5030 17. A "Completed" log line is now written for messages that are removed from
5031 the spool by the -Mrm option.
5033 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5034 information about exactly what failed.
5036 19. Added -dd to debug only the daemon process.
5038 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5039 handles timeouts, both on the server side and network timeouts. Renamed the
5040 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5042 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5043 It is now set to "smtps".
5045 22. $host_address is now set to the target address during the checking of
5046 ignore_target_hosts.
5048 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5049 being passed; this would have caused $sender_host_name to have been used if
5050 matching the list had actually called for a host name (not very likely,
5051 since this list is usually IP addresses). A host name is now passed as
5054 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5055 code that specifies a non-restarting handler (typically sigaction() in
5056 modern systems) in an attempt to fix a rare and obscure crash bug.
5058 25. Narrowed the window for a race in the daemon that could cause it to ignore
5059 SIGCHLD signals. This is not a major problem, because they are used only to
5060 wake it up if nothing else does.
5062 26. A malformed maildirsize file could cause Exim to calculate negative values
5063 for the mailbox size or file count. Odd effects could occur as a result.
5064 The maildirsize information is now recalculated if the size or filecount
5067 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5068 support for a long time. Removed HAVE_SYS_VFS_H.
5070 28. Installed the latest version of exipick from John Jetmore.
5072 29. In an address list, if the pattern was not a regular expression, an empty
5073 subject address (from a bounce message) matched only if the pattern was an
5074 empty string. Non-empty patterns were not even tested. This was the wrong
5075 because it is perfectly reasonable to use an empty address as part of a
5076 database query. An empty address is now tested by patterns that are
5077 lookups. However, all the other forms of pattern expect the subject to
5078 contain a local part and a domain, and therefore, for them, an empty
5079 address still always fails if the pattern is not itself empty.
5081 30. Exim went into a mad DNS loop when attempting to do a callout where the
5082 host was specified on an smtp transport, and looking it up yielded more
5083 than one IP address.
5085 31. Re-factored the code for checking spool and log partition space into a
5086 function that finds that data and another that does the check. The former
5087 is then used to implement four new variables: $spool_space, $log_space,
5088 $spool_inodes, and $log_inodes.
5090 32. The RFC2047 encoding function was originally intended for short strings
5091 such as real names; it was not keeping to the 75-character limit for
5092 encoded words that the RFC imposes. It now respects the limit, and
5093 generates multiple encoded words if necessary. To be on the safe side, I
5094 have increased the buffer size for the ${rfc2047: expansion operator from
5097 33. It is now permitted to omit both strings after an "if" condition; if the
5098 condition is true, the result is "true". As before, when the second string
5099 is omitted, a false condition yields an empty string. This makes it less
5100 cumbersome to write custom ACL and router conditions.
5102 34. Failure to deliver a bounce message always caused it to be frozen, even if
5103 there was an errors_to setting on the router. The errors_to setting is now
5106 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5107 canonical form (fully expanded) before being placed in
5108 $sender_host_address.
5110 36. The table in the code that translates DNS record types into text (T_A to
5111 "A" for instance) was missing entries for NS and CNAME. It is just possible
5112 that this could have caused confusion if both these types were looked up
5113 for the same domain, because the text type is used as part of Exim's
5114 per-process caching. But the chance of anyone hitting this buglet seems
5117 37. The dnsdb lookup has been extended in a number of ways.
5119 (1) There is a new type, "zns", which walks up the domain tree until it
5120 finds some nameserver records. It should be used with care.
5122 (2) There is a new type, "mxh", which is like "mx" except that it returns
5123 just the host names, not the priorities.
5125 (3) It is now possible to give a list of domains (or IP addresses) to be
5126 looked up. The behaviour when one of the lookups defers can be
5127 controlled by a keyword.
5129 (4) It is now possible to specify the separator character for use when
5130 multiple records are returned.
5132 38. The dnslists ACL condition has been extended: it is now possible to supply
5133 a list of IP addresses and/or domains to be looked up in a particular DNS
5136 39. Added log_selector=+queue_time_overall.
5138 40. When running the queue in the test harness, wait just a tad after forking a
5139 delivery process, to get repeatability of debugging output.
5141 41. Include certificate and key file names in error message when GnuTLS fails
5142 to set them up, because the GnuTLS error message doesn't include the name
5143 of the failing file when there is a problem reading it.
5145 42. Allow both -bf and -bF in the same test run.
5147 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5149 44. The "Exiscan patch" is now merged into the mainline Exim source.
5151 45. Sometimes the final signoff response after QUIT could fail to get
5152 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5153 < 0 before doing a fflush(). This bug looks as though it goes back to the
5154 introduction of TLS in release 3.20, but "sometimes" must have been rare
5155 because the tests only now provoked it.
5157 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5158 (this can affect the format of dates).
5160 47. exim_tidydb, when checking for the continued existence of a message for
5161 which it has found a message-specific retry record, was not finding
5162 messages that were in split spool directories. Consequently, it was
5163 deleting retry records that should have stayed in existence.
5165 48. Steve fixed some bugs in eximstats.
5167 49. The SPA authentication driver was not abandoning authentication and moving
5168 on to the next authenticator when an expansion was forced to fail,
5169 contradicting the general specification for all authenticators. Instead it
5170 was generating a temporary error. It now behaves as specified.
5172 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5173 (the order specifies the preference for clients). The order is now AES256,
5174 AES128, 3DES, ARCFOUR128.
5176 51. Small patch to Sieve code - explicitly set From: when generating an
5179 52. Exim crashed if a remote delivery caused a very long error message to be
5180 recorded - for instance if somebody sent an entire SpamAssassin report back
5181 as a large number of 550 error lines. This bug was coincidentally fixed by
5182 increasing the size of one of Exim's internal buffers (big_buffer) that
5183 happened as part of the Exiscan merge. However, to be on the safe side, I
5184 have made the code more robust (and fixed the comments that describe what
5187 53. Now that there can be additional text after "Completed" in log lines (if
5188 the queue_time_overall log selector is set), a one-byte patch to exigrep
5189 was needed to allow it to recognize "Completed" as not the last thing in
5192 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5193 patch that reportedly fixes this has been added. I am not expert enough to
5194 create a test for it. This is what the patch creator wrote:
5196 "I found a little strange behaviour of ldap code when working with
5197 Windows 2003 AD Domain, where users was placed in more than one
5198 Organization Units. When I tried to give exim partial DN, the exit code
5199 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5200 But simultaneously result of request was absolutely normal ldap result,
5201 so I produce this patch..."
5203 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5204 REFERENCE, so I have modified the code to exclude the patch when that macro
5207 55. Some experimental protocols are using DNS PTR records for new purposes. The
5208 keys for these records are domain names, not reversed IP addresses. The
5209 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5210 leaves it alone. Component reversal etc. now happens only for IP addresses.
5213 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5215 57. Double the size of the debug message buffer (to 2048) so that more of very
5216 long debug lines gets shown.
5218 58. The exicyclog utility now does better if the number of log files to keep
5219 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5221 59. Two changes related to the smtp_active_hostname option:
5223 (1) $smtp_active_hostname is now available as a variable.
5224 (2) The default for smtp_banner uses $smtp_active_hostname instead
5225 of $primary_hostname.
5227 60. The host_aton() function is supposed to be passed a string that is known
5228 to be a valid IP address. However, in the case of IPv6 addresses, it was
5229 not checking this. This is a hostage to fortune. Exim now panics and dies
5230 if the condition is not met. A case was found where this could be provoked
5231 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5232 components; fortuitously, this particular loophole had already been fixed
5233 by change 4.50/55 above.
5235 If there are any other similar loopholes, the new check in host_aton()
5236 itself should stop them being exploited. The report I received stated that
5237 data on the command line could provoke the exploit when Exim was running as
5238 exim, but did not say which command line option was involved. All I could
5239 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5240 running as the user.
5243 61. There was a buffer overflow vulnerability in the SPA authentication code
5244 (which came originally from the Samba project). I have added a test to the
5245 spa_base64_to_bits() function which I hope fixes it.
5248 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5249 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5251 63. The daemon start-up calls getloadavg() while still root for those OS that
5252 need the first call to be done as root, but it missed one case: when
5253 deliver_queue_load_max is set with deliver_drop_privilege. This is
5254 necessary for the benefit of the queue runner, because there is no re-exec
5255 when deliver_drop_privilege is set.
5257 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5258 This has been fixed.
5260 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5261 was in use, was not putting the data itself into the right store pool;
5262 consequently, it could be overwritten for a subsequent message in the same
5263 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5266 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5268 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5269 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5270 the function were treating the return as a boolean value, which happened to
5271 work because 0=false and not-0=true, but is not correct code.
5273 68. The host_aton() function was not handling scoped IPv6 addresses (those
5274 with, for example, "%eth0" on the end) correctly.
5276 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5277 negated items (that is, ~something) in unsigned ints. Some compilers
5278 apparently mutter when there is no cast.
5280 70. If an address verification called from an ACL failed, and did not produce a
5281 user-specific message (i.e. there was only a "system" message), nothing was
5282 put in $acl_verify_message. In this situation, it now puts the system
5285 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5286 session; change 4.31/43 added the unwanted input to the log line - except
5287 that it did not do this in the start of session case. It now does.
5289 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5290 This is wrong and can cause the other end to generate a synchronization
5291 error if it is another Exim or anything else that does the synchronization
5292 check. A QUIT command is no longer sent after a timeout.
5294 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5295 during host lookups.
5297 74. The defer_ok option of callout verification was not working if it was used
5298 when verifying addresses in header lines, that is, for this case:
5300 verify = header_sender/callout=defer_ok
5302 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5303 those file descriptors could be used for SMTP connections. If anything
5304 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5305 could be sent to the SMTP client, causing chaos. The daemon now opens
5306 stdin, stdout, and stderr to /dev/null when it puts itself into the
5309 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5310 log by default. The user can override this with suitable Perl magic.
5312 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5313 the log message when discard triggers, was not working for the DATA ACL or
5314 for the non-SMTP ACL.
5316 78. Error message wording change in sieve.c.
5318 79. If smtp_accept_max_per_host was set, the number of connections could be
5319 restricted to fewer than expected, because the daemon was trying to set up
5320 a new connection before checking whether the processes handling previous
5321 connections had finished. The check for completed processes is now done
5322 earlier. On busy systems, this bug wouldn't be noticed because something
5323 else would have woken the daemon, and it would have reaped the completed
5326 80. If a message was submitted locally by a user whose login name contained one
5327 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5328 It caused a spool format error. I have fixed the spool reading code. A
5329 related problem was that the "from" clause in the Received: line became
5330 illegal because of the space(s). It is now covered by ${quote_local_part.
5332 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5335 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5337 83. Patch from Sieve maintainer to fix unterminated string problem in
5338 "vacation" handling.
5340 84. Some minor changes to the Linux configuration files to help with other
5341 OS variants using glibc.
5343 85. One more patch for Sieve to update vacation handling to latest spec.
5346 ----------------------------------------------------
5347 See the note above about the 4.44 and 4.50 releases.
5348 ----------------------------------------------------
5354 1. Change 4.43/35 introduced a bug that caused file counts to be
5355 incorrectly computed when quota_filecount was set in an appendfile
5358 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5359 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5362 3. Give more explanation in the error message when the command for a transport
5363 filter fails to execute.
5365 4. There are several places where Exim runs a non-Exim command in a
5366 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5367 was being done only for the command run by the queryprogram router. It is
5368 now done for all such subprocesses. The other cases are: ${run, transport
5369 filters, and the commands run by the lmtp and pipe transports.
5371 5. Some older OS have a limit of 256 on the maximum number of file
5372 descriptors. Exim was using setrlimit() to set 1000 as a large value
5373 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5374 systems. I've change it so that if it can't get 1000, it tries for 256.
5376 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5377 was an oversight, and furthermore, ever since the addition of extra
5378 controls (e.g. 4.43/32), the checks on when to allow different forms of
5379 "control" were broken. There should now be diagnostics for all cases when a
5380 control that does not make sense is encountered.
5382 7. $recipients is now available in the predata ACL (oversight).
5384 8. Tidy the search cache before the fork to do a delivery from a message
5385 received from the command line. Otherwise the child will trigger a lookup
5386 failure and thereby defer the delivery if it tries to use (for example) a
5387 cached ldap connection that the parent has called unbind on.
5389 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5390 of $address_data from the recipient verification was clobbered by the
5391 sender verification.
5393 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5394 was its contents. (It was OK if the option was not defined at all.)
5396 11. A "Completed" log line is now written for messages that are removed from
5397 the spool by the -Mrm option.
5399 12. $host_address is now set to the target address during the checking of
5400 ignore_target_hosts.
5402 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5403 being passed; this would have caused $sender_host_name to have been used if
5404 matching the list had actually called for a host name (not very likely,
5405 since this list is usually IP addresses). A host name is now passed as
5408 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5409 code that specifies a non-restarting handler (typically sigaction() in
5410 modern systems) in an attempt to fix a rare and obscure crash bug.
5412 15. Narrowed the window for a race in the daemon that could cause it to ignore
5413 SIGCHLD signals. This is not a major problem, because they are used only to
5414 wake it up if nothing else does.
5416 16. A malformed maildirsize file could cause Exim to calculate negative values
5417 for the mailbox size or file count. Odd effects could occur as a result.
5418 The maildirsize information is now recalculated if the size or filecount
5421 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5422 support for a long time. Removed HAVE_SYS_VFS_H.
5424 18. Updated exipick to current release from John Jetmore.
5426 19. Allow an empty sender to be matched against a lookup in an address list.
5427 Previously the only cases considered were a regular expression, or an
5430 20. Exim went into a mad DNS lookup loop when doing a callout where the
5431 host was specified on the transport, if the DNS lookup yielded more than
5434 21. The RFC2047 encoding function was originally intended for short strings
5435 such as real names; it was not keeping to the 75-character limit for
5436 encoded words that the RFC imposes. It now respects the limit, and
5437 generates multiple encoded words if necessary. To be on the safe side, I
5438 have increased the buffer size for the ${rfc2047: expansion operator from
5441 22. Failure to deliver a bounce message always caused it to be frozen, even if
5442 there was an errors_to setting on the router. The errors_to setting is now
5445 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5446 canonical form (fully expanded) before being placed in
5447 $sender_host_address.
5449 24. Updated eximstats to version 1.33
5451 25. Include certificate and key file names in error message when GnuTLS fails
5452 to set them up, because the GnuTLS error message doesn't include the name
5453 of the failing file when there is a problem reading it.
5455 26. Expand error message when OpenSSL has problems setting up cert/key files.
5458 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5459 (this can affect the format of dates).
5461 28. exim_tidydb, when checking for the continued existence of a message for
5462 which it has found a message-specific retry record, was not finding
5463 messages that were in split spool directories. Consequently, it was
5464 deleting retry records that should have stayed in existence.
5466 29. eximstats updated to version 1.35
5467 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5468 1.35 - bugfix such that pie charts by volume are generated correctly
5470 30. The SPA authentication driver was not abandoning authentication and moving
5471 on to the next authenticator when an expansion was forced to fail,
5472 contradicting the general specification for all authenticators. Instead it
5473 was generating a temporary error. It now behaves as specified.
5475 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5476 (the order specifies the preference for clients). The order is now AES256,
5477 AES128, 3DES, ARCFOUR128.
5479 31. Small patch to Sieve code - explicitly set From: when generating an
5482 32. Exim crashed if a remote delivery caused a very long error message to be
5483 recorded - for instance if somebody sent an entire SpamAssassin report back
5484 as a large number of 550 error lines. This bug was coincidentally fixed by
5485 increasing the size of one of Exim's internal buffers (big_buffer) that
5486 happened as part of the Exiscan merge. However, to be on the safe side, I
5487 have made the code more robust (and fixed the comments that describe what
5490 33. Some experimental protocols are using DNS PTR records for new purposes. The
5491 keys for these records are domain names, not reversed IP addresses. The
5492 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5493 leaves it alone. Component reversal etc. now happens only for IP addresses.
5496 34. The host_aton() function is supposed to be passed a string that is known
5497 to be a valid IP address. However, in the case of IPv6 addresses, it was
5498 not checking this. This is a hostage to fortune. Exim now panics and dies
5499 if the condition is not met. A case was found where this could be provoked
5500 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5501 components; fortuitously, this particular loophole had already been fixed
5502 by change 4.50/55 or 4.44/33 above.
5504 If there are any other similar loopholes, the new check in host_aton()
5505 itself should stop them being exploited. The report I received stated that
5506 data on the command line could provoke the exploit when Exim was running as
5507 exim, but did not say which command line option was involved. All I could
5508 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5509 running as the user.
5512 35. There was a buffer overflow vulnerability in the SPA authentication code
5513 (which came originally from the Samba project). I have added a test to the
5514 spa_base64_to_bits() function which I hope fixes it.
5517 36. The daemon start-up calls getloadavg() while still root for those OS that
5518 need the first call to be done as root, but it missed one case: when
5519 deliver_queue_load_max is set with deliver_drop_privilege. This is
5520 necessary for the benefit of the queue runner, because there is no re-exec
5521 when deliver_drop_privilege is set.
5523 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5524 was in use, was not putting the data itself into the right store pool;
5525 consequently, it could be overwritten for a subsequent message in the same
5526 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5529 38. Sometimes the final signoff response after QUIT could fail to get
5530 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5531 < 0 before doing a fflush(). This bug looks as though it goes back to the
5532 introduction of TLS in release 3.20, but "sometimes" must have been rare
5533 because the tests only now provoked it.
5539 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5540 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5541 Now it returns an integer. A number of calls were still expecting a T/F
5542 return. Fortuitously, in all cases, the tests worked in OK situations,
5543 which is the norm. However, things would have gone wrong on any write
5544 failures on the smtp file descriptor. This function is used when sending
5545 messages over SMTP and also when doing verify callouts.
5547 2. When Exim is called to do synchronous delivery of a locally submitted
5548 message (the -odf or -odi options), it no longer closes stderr before doing
5551 3. Implemented the mua_wrapper option.
5553 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5555 5. Implemented the functions header_remove(), header_testname(),
5556 header_add_at_position(), and receive_remove_recipient(), and exported them
5559 6. If an ACL "warn" statement specified the addition of headers, Exim already
5560 inserted X-ACL-Warn: at the start if there was no header name. However, it
5561 was not making this test for the second and subsequent header lines if
5562 there were newlines in the string. This meant that an invalid header could
5563 be inserted if Exim was badly configured.
5565 7. Allow an ACL "warn" statement to add header lines at the start or after all
5566 the Received: headers, as well as at the end.
5568 8. Added the rcpt_4xx retry error code.
5570 9. Added postmaster_mailfrom=xxx to callout verification option.
5572 10. Added mailfrom=xxxx to the callout verification option, for verify=
5575 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5576 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5578 12. Inserted some casts to stop certain compilers warning when using pointer
5579 differences as field lengths or precisions in printf-type calls (mostly
5580 affecting debugging statements).
5582 13. Added optional readline() support for -be (dynamically loaded).
5584 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5585 same clock tick as a message's arrival, so that its received time was the
5586 same as the "first fail" time on the retry record, and that message
5587 remained on the queue past the ultimate address timeout, every queue runner
5588 would try a delivery (because it was past the ultimate address timeout) but
5589 after another failure, the ultimate address timeout, which should have then
5590 bounced the address, did not kick in. This was a "< instead of <=" error;
5591 in most cases the first failure would have been in the next clock tick
5592 after the received time, and all would be well.
5594 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5595 being recognized when the domain list was tested by the match_domain
5596 condition in an expansion string.
5598 16. Added the ${str2b64: operator.
5600 17. Exim was always calling setrlimit() to set a large limit for the number of
5601 processes, without checking whether the existing limit was already
5602 adequate. (It did check for the limit on file descriptors.) Furthermore,
5603 errors from getrlimit() and setrlimit() were being ignored. Now they are
5604 logged to the main and panic logs, but Exim does carry on, to try to do its
5605 job under whatever limits there are.
5607 18. Imported PCRE 5.0.
5609 19. Trivial typo in log message " temporarily refused connection" (the leading
5612 20. If the log selector return_path_on_delivery was set and an address was
5613 redirected to /dev/null, the delivery process crashed because it assumed
5614 that a return path would always be set for a "successful" delivery. In this
5615 case, the whole delivery is bypassed as an optimization, and therefore no
5618 21. Internal re-arrangement: the function for sending a challenge and reading
5619 a response while authentication was assuming a zero-terminated challenge
5620 string. It's now changed to take a pointer and a length, to allow for
5621 binary data in such strings.
5623 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5625 23. Exim was not respecting finduser_retries when seeking the login of the
5626 uid under which it was called; it was always trying 10 times. (The default
5627 setting of finduser_retries is zero.) Also, it was sleeping after the final
5628 failure, which is pointless.
5630 24. Implemented tls_on_connect_ports.
5632 25. Implemented acl_smtp_predata.
5634 26. If the domain in control=submission is set empty, Exim assumes that the
5635 authenticated id is a complete email address when it generates From: or
5636 Sender: header lines.
5638 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5639 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5640 chown and chgrp in /bin and hostname in /usr/bin.
5642 28. Exim was keeping the "process log" file open after each use, just as it
5643 does for the main log. This opens the possibility of it remaining open for
5644 long periods when the USR1 signal hits a daemon. Occasional processlog
5645 errors were reported, that could have been caused by this. Anyway, it seems
5646 much more sensible not to leave this file open at all, so that is what now
5649 29. The long-running daemon process does not normally write to the log once it
5650 has entered its main loop, and it closes the log before doing so. This is
5651 so that log files can straightforwardly be renamed and moved. However,
5652 there are a couple of unusual error situations where the daemon does write
5653 log entries, and I had neglected to close the log afterwards.
5655 30. The text of an SMTP error response that was received during a remote
5656 delivery was being truncated at 512 bytes. This is too short for some of
5657 the long messages that one sometimes sees. I've increased the limit to
5660 31. It is now possible to make retry rules that apply only when a message has a
5661 specific sender, in particular, an empty sender.
5663 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5664 it possible to be selective about when SMTP synchronization is enforced.
5666 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5668 32. Implemented hosts_connection_nolog.
5670 33. Added an ACL for QUIT.
5672 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5675 35. Added mailbox_size and mailbox_filecount to appendfile.
5677 36. Added control = no_multiline_responses to ACLs.
5679 37. There was a bug in the logic of the code that waits for the clock to tick
5680 in the case where the clock went backwards by a substantial amount such
5681 that the microsecond fraction of "now" was more than the microsecond
5682 fraction of "then" (but the whole seconds number was less).
5684 38. Added support for the libradius Radius client library this is found on
5685 FreeBSD (previously only the radiusclient library was supported).
5691 1. When certain lookups returned multiple values in the form name=value, the
5692 quoting of the values was not always being done properly. Specifically:
5693 (a) If the value started with a double quote, but contained no whitespace,
5695 (b) If the value contained whitespace other than a space character (i.e.
5696 tabs or newlines or carriage returns) it was not quoted.
5697 This fix has been applied to the mysql and pgsql lookups by writing a
5698 separate quoting function and calling it from the lookup code. The fix
5699 should probably also be applied to nisplus, ibase and oracle lookups, but
5700 since I cannot test any of those, I have not disturbed their existing code.
5702 2. A hit in the callout cache for a specific address caused a log line with no
5703 reason for rejecting RCPT. Now it says "Previous (cached) callout
5704 verification failure".
5706 3. There was an off-by-one bug in the queryprogram router. An over-long
5707 return line was truncated at 256 instead of 255 characters, thereby
5708 overflowing its buffer with the terminating zero. As well as fixing this, I
5709 have increased the buffer size to 1024 (and made a note to document this).
5711 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5712 when Exim is waiting for an SMTP response from a remote server, Exim
5713 restarts its select() call on the socket, thereby resetting its timeout.
5714 This is not a problem when such interrupts are rare. Somebody set up a cron
5715 job to run exiwhat every 2 minutes, which is less than the normal select()
5716 timeout (5 or 10 minutes). This meant that the select() timeout never
5717 kicked in because it was always reset. I have fixed this by comparing the
5718 time when an interrupt arrives with the time at the start of the first call
5719 to select(). If more time than the timeout has elapsed, the interrupt is
5720 treated as a timeout.
5722 5. Some internal re-factoring in preparation for the addition of Sieve
5723 extensions (by MH). In particular, the "personal" test is moved to a
5724 separate function, and given an option for scanning Cc: and Bcc: (which is
5725 not set for Exim filters).
5727 6. When Exim created an email address using the login of the caller as the
5728 local part (e.g. when creating a From: or Sender: header line), it was not
5729 quoting the local part when it contained special characters such as @.
5731 7. Installed new OpenBSD configuration files.
5733 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5734 try to make them clearer.
5736 9. Callout options, other than the timeout value, were being ignored when
5737 verifying sender addresses in header lines. For example, when using
5739 verify = header_sender/callout=no_cache
5741 the cache was (incorrectly) being used.
5743 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5744 only the Cygwin environment.
5746 11. When return_path_on_delivery was set as a log selector, if different remote
5747 addresses in the same message used different return paths and parallel
5748 remote delivery occurred, the wrong values would sometimes be logged.
5749 (Whenever a remote delivery process finished, the return path value from
5750 the most recently started remote delivery process was logged.)
5752 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5753 lines when AUTH and/or TLS are in use. This is the "received protocol"
5754 field. Exim used to use "asmtp" for authenticated SMTP, without any
5755 indication (in the protocol name) for TLS use. Now it follows the RFC and
5756 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5757 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5758 names appear in log lines as well as in Received: header lines.
5760 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5761 extensions, and comparison tests, and to fix some bugs.
5763 14. Changes to the "personal" filter test:
5765 (1) The test was buggy in that it was just doing the equivalent of
5766 "contains" tests on header lines. For example, if a user's address was
5767 anne@some.where, the "personal" test would incorrectly be true for
5769 To: susanne@some.where
5771 This test is now done by extracting each address from the header in turn,
5772 and checking the entire address. Other tests that are part of "personal"
5773 are now done using regular expressions (for example, to check local parts
5774 of addresses in From: header lines).
5776 (2) The list of non-personal local parts in From: addresses has been
5777 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5778 taken from the Sieve specification recommendations.
5780 (3) If the message contains any header line starting with "List-" it is
5781 treated as non-personal.
5783 (4) The test for "circular" in the Subject: header line has been removed
5784 because it now seems ill-conceived.
5786 15. Minor typos in src/EDITME comments corrected.
5788 16. Installed latest exipick from John Jetmore.
5790 17. If headers_add on a router specified a text string that was too long for
5791 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5792 of string_sprintf() is now avoided.
5794 18. $message_body_size was not set (it was always zero) when running the DATA
5795 ACL and the local_scan() function.
5797 19. For the "mail" command in an Exim filter, no default was being set for
5798 the once_repeat time, causing a random time value to be used if "once" was
5799 specified. (If the value happened to be <= 0, no repeat happened.) The
5800 default is now 0s, meaning "never repeat". The "vacation" command was OK
5801 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5802 (I found it when inspecting the code).
5804 20. There is now an overall timeout for performing a callout verification. It
5805 defaults to 4 times the callout timeout, which applies to individual SMTP
5806 commands during the callout. The overall timeout applies when there is more
5807 than one host that can be tried. The timeout is checked before trying the
5808 next host. This prevents very long delays if there are a large number of
5809 hosts and all are timing out (e.g. when the network connections are timing
5810 out). The value of the overall timeout can be changed by specifying an
5811 additional sub-option for "callout", called "maxwait". For example:
5813 verify = sender/callout=5s,maxwait=20s
5815 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5816 to the end before writing, but this should make it even safer).
5818 22. Exim was forgetting that it had advertised PIPELINING for the second and
5819 subsequent messages on an SMTP connection. It was also not resetting its
5820 memory on STARTTLS and an internal HELO.
5822 23. When Exim logs an SMTP synchronization error within a session, it now
5823 records whether PIPELINING has been advertised or not.
5825 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5826 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5827 rather than long int.
5829 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5831 26. Added the never_mail option to autoreply.
5837 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5838 crash if the getsockname() call failed; this can happen if a connection is
5839 closed very soon after it is established. The problem was simply in the
5840 order in which certain operations were done, causing Exim to try to write
5841 to the SMTP stream before it had set up the file descriptor. The bug has
5842 been fixed by making things happen in the correct order.
5848 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5849 before the connection was closed, thus losing the rejection response.
5851 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5852 some early Solaris releases, but causes trouble in current releases where
5853 socklen_t is defined.
5855 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5858 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5861 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5862 that releasing the top of it at the end releases what was used for sub-
5863 expansions (unless the block got too big). However, discard this block if
5864 the first thing is a variable or header, so that we can use its block when
5865 it is dynamic (useful for very large $message_headers, for example).
5867 6. Lookups now cache *every* query, not just the most recent. A new, separate
5868 store pool is used for this. It can be recovered when all lookup caches are
5869 flushed. Lookups now release memory at the end of their result strings.
5870 This has involved some general refactoring of the lookup sources.
5872 7. Some code has been added to the store_xxx() functions to reduce the amount
5873 of flapping under certain conditions.
5875 8. log_incoming_interface used to affect only the <= reception log lines. Now
5876 it causes the local interface and port to be added to several more SMTP log
5877 lines, for example "SMTP connection from", and rejection lines.
5879 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5881 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5883 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5884 could be overwritten at the end of the current message (or the start of a
5885 new message if it was set in a HELO ACL). The value is now preserved for
5886 the duration of the SMTP connection.
5888 12. If a transport had a headers_rewrite setting, and a matching header line
5889 contained an unqualified address, that address was qualified, even if it
5890 did not match any rewriting rules. The underlying bug was that the values
5891 of the flags that permit the existence of unqualified sender and recipient
5892 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5893 non-local messages, and by -bnq for local messages) were not being
5894 preserved with the message after it was received.
5896 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5897 "next input=" as part of the text comprising the host identity instead of
5898 the correct text. The code was using the same buffer for two different
5899 strings. However, depending on which order the printing function evaluated
5900 its arguments, the bug did not always show up. Under Linux, for example, my
5901 test suite worked just fine.
5903 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5904 doesn't work with some older versions of Perl. It has been changed to "my",
5905 which in any case is probably the better facility to use.
5907 15. A really picky compiler found some instances of statements for creating
5908 error messages that either had too many or two few arguments for the format
5911 16. The size of the buffer for calls to the DNS resolver has been increased
5912 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5913 for addresses that have a lot of PTR records. This alleviates a problem; it
5914 does not fully solve it.
5916 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5917 buffer now truncates the list and logs the incident, which is the same
5918 action as happens when Exim is looking up a host name and its aliases.
5919 Previously in this situation something unpredictable would happen;
5920 sometimes it was "internal error: store_reset failed".
5922 18. If a server dropped the connection unexpectedly when an Exim client was
5923 using GnuTLS and trying to read a response, the client delivery process
5924 crashed while trying to generate an error log message.
5926 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5927 string, for example:
5929 warn message = H1: something\nH2: something
5931 the text was added as a single header line from Exim's point of view
5932 though it ended up OK in the delivered message. However, searching for the
5933 second and subsequent header lines using $h_h2: did not work. This has been
5934 fixed. Similarly, if a system filter added multiple headers in this way,
5935 the routers could not see them.
5937 20. Expanded the error message when iplsearch is called with an invalid key to
5938 suggest using net-iplsearch in a host list.
5940 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5941 ACLs are no longer actually imposed (and a message to that effect is
5944 22. If a "gecos" field in a passwd entry contained escaped characters, in
5945 particular, if it contained a \" sequence, Exim got it wrong when building
5946 a From: or a Sender: header from that name. A second bug also caused
5947 incorrect handling when an unquoted " was present following a character
5948 that needed quoting.
5950 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5951 was not being matched caselessly.
5953 24. Arranged for all hyphens in the exim.8 source to be escaped with
5956 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5957 properly. Recipient callout cache records were still being keyed to include
5958 the sender, even when use_sender was set false. This led to far more
5959 callouts that were necessary. The sender is no longer included in the key
5960 when use_sender is false.
5962 26. Added "control = submission" modifier to ACLs.
5964 27. Added the ${base62d: operator to decode base 62 numbers.
5966 28. dnsdb lookups can now access SRV records.
5968 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5969 the configuration file.
5971 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5972 (-v) mode. This makes the output for a verbose queue run more intelligible.
5974 31. Added a use_postmaster feature to recipient callouts.
5976 32. Added the $body_zerocount variable, containing the number of binary zero
5977 bytes in the message body.
5979 33. The time of last modification of the "new" subdirectory is now used as the
5980 "mailbox time last read" when there is a quota error for a maildir
5983 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5985 35. Added +ignore_unknown as a special item in host lists.
5987 36. Code for decoding IPv6 addresses in host lists is now included, even if
5988 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5989 address was recognized as an IP address, but was then not correctly decoded
5990 into binary, causing unexpected and incorrect effects when compared with
5997 1. Very minor rewording of debugging text in manualroute to say "list of
5998 hosts" instead of "hostlist".
6000 2. If verify=header_syntax was set, and a header line with an unqualified
6001 address (no domain) and a large number of spaces between the end of the
6002 name and the colon was received, the reception process suffered a buffer
6003 overflow, and (when I tested it) crashed. This was caused by some obsolete
6004 code that should have been removed. The fix is to remove it!
6006 3. When running in the test harness, delay a bit after writing a bounce
6007 message to get a bit more predictability in the log output.
6009 4. Added a call to search_tidyup() just before forking a reception process. In
6010 theory, someone could use a lookup in the expansion of smtp_accept_max_
6011 per_host which, without the tidyup, could leave open a database connection.
6013 5. Added the variables $recipient_data and $sender_data which get set from a
6014 lookup success in an ACL "recipients" or "senders" condition, or a router
6015 "senders" option, similar to $domain_data and $local_part_data.
6017 6. Moved the writing of debug_print from before to after the "senders" test
6020 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6021 problems for message scanning, either using a data ACL, or using
6022 local_scan() because the Received: header was not generated till after they
6023 were called (in order to set the time as the time of reception completion).
6024 I have revised the way this works. The header is now generated after the
6025 body is received, but before the ACL or local_scan() are called. After they
6026 are run, the timestamp in the header is updated.
6032 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6033 before starting a queue runner without re-exec. This happened only when
6034 deliver_drop_privilege was set or when the Exim user was set to root. The
6035 effect of the bug was that timeouts during subsequent deliveries caused
6036 crashes instead of being properly handled. The handler is now left at its
6037 default (and expected) setting.
6039 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6040 message, again when deliver_drop_privilege is set or Exim is run as root.
6041 The bug described in (1) was not present in this case, but the tidying up
6042 of the other signals was missing. I have made the two cases consistent.
6044 3. The ignore_target_hosts setting on a manualroute router was being ignored
6045 for hosts that were looked up using the /MX notation.
6047 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6050 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6051 operated on the sender address. After changing the $sender_address to <>
6052 for the sender address verify, Exim was re-instated it as the original
6053 (before rewriting) address, but remembering that it had rewritten it, so it
6054 wasn't rewriting it again. This bug also had the effect of breaking the
6055 sender address verification caching when the sender address was rewritten.
6057 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6058 This has been changed so that if the ip literal address matches
6059 ignore_target_hosts, the router declines.
6061 7. Added expansion conditions match_domain, match_address, and match_local_
6062 part (NOT match_host).
6064 8. The placeholder for the Received: header didn't have a length field set.
6066 9. Added code to Exim itself and to exim_lock to test for a specific race
6067 condition that could lead to file corruption when using MBX delivery. The
6068 issue is with the lockfile that is created in /tmp. If this file is removed
6069 after a process has opened it but before that process has acquired a lock,
6070 there is the potential for a second process to recreate the file and also
6071 acquire a lock. This could lead to two Exim processes writing to the file
6072 at the same time. The added code performs the same test as UW imapd; it
6073 checks after acquiring the lock that its file descriptor still refers to
6074 the same named file.
6076 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6077 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6080 11. Added the smtp_active_hostname option. If used, this will typically be made
6081 to depend on the incoming interface address. Because $interface_address is
6082 not set up until the daemon has forked a reception process, error responses
6083 that can happen earlier (such as "too many connections") no longer contain
6086 12. If an expansion in a condition on a "warn" statement fails because a lookup
6087 defers, the "warn" statement is abandoned, and the next ACL statement is
6088 processed. Previously this caused the whole ACL to be aborted.
6090 13. Added the iplsearch lookup type.
6092 14. Added ident_timeout as a log selector.
6094 15. Added tls_certificate_verified as a log selector.
6096 16. Added a global option tls_require_ciphers (compare the smtp transport
6097 option of the same name). This controls incoming TLS connections.
6099 17. I finally figured out how to make tls_require_ciphers do a similar thing
6100 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6101 before starting the TLS session.
6103 18. Tabs are now shown as \t in -bP output.
6105 19. If the log selector return_path_on_delivery was set, Exim crashed when
6106 bouncing a message because it had too many Received: header lines.
6108 20. If two routers both had headers_remove settings, and the first one included
6109 a superfluous trailing colon, the final name in the first list and the
6110 first name in the second list were incorrectly joined into one item (with a
6111 colon in the middle).
6117 1. Added -C and -D options to the exinext utility, mainly to make it easier
6118 to include in the automated testing, but these could be helpful when
6119 multiple configurations are in use.
6121 2. The exinext utility was not formatting the output nicely when there was
6122 an alternate port involved in the retry record key, nor when there was a
6123 message id as well (for retries that were specific to a specific message
6124 and a specific host). It was also confused by IPv6 addresses, because of
6125 the additional colons they contain. I have fixed the IPv4 problem, and
6126 patched it up to do a reasonable job for IPv6.
6128 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6129 delivery, the log line now contains "pipelined" if PIPELINING was used.
6131 4. An SMTP transport process used to panic and die if the bind() call to set
6132 an explicit outgoing interface failed. This has been changed; it is now
6133 treated in the same way as a connect() failure.
6135 5. A reference to $sender_host_name in the part of a conditional expansion
6136 that was being skipped was still causing a DNS lookup. This no longer
6139 6. The def: expansion condition was not recognizing references to header lines
6140 that used bh_ and bheader_.
6142 7. Added the _cache feature to named lists.
6144 8. The code for checking quota_filecount in the appendfile transport was
6145 allowing one more file than it should have been.
6147 9. For compatibility with Sendmail, the command line option
6155 and sets the incoming protocol and host name (for trusted callers). The
6156 host name and its colon can be omitted when only the protocol is to be set.
6157 Note the Exim already has two private options, -pd and -ps, that refer to
6158 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6159 "s", but I don't think that's a major issue.
6161 10. A number of refactoring changes to the code, none of which should affect
6164 (a) The number of logging options was getting close to filling up the
6165 32-bit word that was used as a bit map. I have split them into two classes:
6166 those that are passed in the argument to log_write(), and those that are
6167 only ever tested independently outside of that function. These are now in
6168 separate 32-bit words, so there is plenty of room for expansion again.
6169 There is no change in the user interface or the logging behaviour.
6171 (b) When building, for example, log lines, the code previously used a
6172 macro that called string_cat() twice, in order to add two strings. This is
6173 not really sufficiently general. Furthermore, there was one instance where
6174 it was actually wrong because one of the argument was used twice, and in
6175 one call a function was used. (As it happened, calling the function twice
6176 did not affect the overall behaviour.) The macro has been replaced by a
6177 function that can join an arbitrary number of extra strings onto a growing
6180 (c) The code for expansion conditions now uses a table and a binary chop
6181 instead of a serial search (which was left over from when there were very
6182 few conditions). Also, it now recognizes conditions like "pam" even when
6183 the relevant support is not compiled in: a suitably worded error message is
6184 given if an attempt is made to use such a condition.
6186 11. Added ${time_interval:xxxxx}.
6188 12. A bug was causing one of the ddress fields not to be passed back correctly
6189 from remote delivery subprocesses. The field in question was not being
6190 subsequently used, so this caused to problems in practice.
6192 13. Added new log selectors queue_time and deliver_time.
6194 14. Might have fixed a bug in maildirsizefile handling that threw up
6195 "unexpected character" debug warnings, and recalculated the data
6196 unnecessarily. In any case, I expanded the warning message to give more
6199 15. Added the message "Restricted characters in address" to the statements in
6200 the default ACL that block characters like @ and % in local parts.
6202 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6203 Three changes have been made:
6205 (a) There was a serious bug; a negative response to MAIL caused the whole
6206 recipient domain to be cached as invalid, thereby blocking all messages
6207 to all local parts at the same domain, from all senders. This bug has
6208 been fixed. The domain is no longer cached after a negative response to
6209 MAIL if the sender used is not empty.
6211 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6214 (c) A new callout option, "use_sender" has been added for people who want
6215 the modified behaviour.
6221 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6224 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6225 indeed breaks things for older releases.
6227 3. Added additional logging to the case where there is a problem reading data
6228 from a filter that is running in a subprocess using a pipe, in order to
6229 try to track down a specific problem.
6231 4. Testing facility fudge: when running in the test harness and attempting
6232 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6233 getting "No route to host". Convert this to a timeout.
6235 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6238 6. Some OS don't have socklen_t but use size_t instead. This affects the
6239 fifth argument of getsockopt() amongst other things. This is now
6240 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6241 can be set for individual OS. I have set it for SunOS5, OSF1, and
6242 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6243 some earlier ones do not.
6245 7. Change 4.30/15 was not doing the test caselessly.
6247 8. The standard form for an IPv6 address literal was being rejected by address
6248 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6249 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6250 this, as well as the form without the "IPv6" on the front (but only when
6251 address literals are enabled, of course).
6253 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6255 10. Exim crashed if a message with an empty sender address specified by -f
6256 encountered a router with an errors_to setting. This could be provoked only
6257 by a command such as
6261 where an empty string was supplied; "<>" did not hit this bug.
6263 11. Installed PCRE release 4.5.
6265 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6266 remained set. It is now erased.
6268 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6269 times from message ids (which are base 36 rather than the normal 62).
6271 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6272 were being counted as actual protocol errors, and logged if the log
6273 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6274 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6275 and DATA following a set of rejected RCPTs do not count as protocol errors.
6276 In other words, Exim assumes they were pipelined, though this may not
6277 actually be the case. Of course, in all cases the client gets an
6278 appropriate error code.
6280 15. If a lookup fails in an ACL condition, a message about the failure may
6281 be available; it is used if testing the ACL cannot continue, because most
6282 such messages specify what the cause of the deferral is. However, some
6283 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6284 that caused an old message to be retained and used if a later statement
6285 caused a defer, replacing the real cause of the deferral.
6287 16. If an IP address had so many PTR records that the DNS lookup buffer
6288 was not large enough to hold them, Exim could crash while trying to process
6289 the truncated data. It now detects and logs this case.
6291 17. Further to 4.21/58, another change has been made: if (and only if) the
6292 first line of a message (the first header line) ends with CRLF, a bare LF
6293 in a subsequent header line has a space inserted after it, so as not to
6294 terminate the header.
6296 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6297 unnecessarily, used atoi() instead of strtol(), and didn't check the
6298 termination when getting file sizes from file names by regex.
6300 19. Completely re-implemented the support for maildirsize files, in the light
6301 of a number of problems with the previous contributed implementation
6302 (4.30/29). In particular:
6304 . If the quota is zero, the maildirsize file is maintained, but no quota is
6307 . If the maildir directory does not exist, it is created before any attempt
6308 to write a maildirsize file.
6310 . The quota value in the file is just a cache; if the quota is changed in
6311 the transport, the new value overrides.
6313 . A regular expression is available for excluding directories from the
6316 20. The autoreply transport checks the characters in options that define the
6317 message's headers; it allows continued headers, but it was checking with
6318 isspace() after an embedded newline instead of explicitly looking for a
6321 21. If all the "regular" hosts to which an address was routed had passed their
6322 expiry times, and had not reached their retry times, the address was
6323 bounced, even if fallback hosts were defined. Now Exim should go on to try
6326 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6327 equivalent code in the SMTP transport. Some hosts send humungous responses
6328 to HELO/EHLO, more than 1024 it seems.
6330 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6331 gives compiler warnings in some environments. I've now done it "properly",
6334 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6335 (because of problems with the built-in one) was declared to return uschar *
6336 instead of char *, causing compiler failure.
6338 25. Fixed a file descriptor leak when processing alias/forward files.
6340 26. Fixed a minor format string issue in dbfn.c.
6342 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6344 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6345 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6348 29. When a sender address is verified, it is cached, to save repeating the test
6349 when there is more than one recipient in a message. However, when the
6350 verification involves a callout, it is possible for different callout
6351 options to be set for different recipients. It is too complicated to keep
6352 track of this in the cache, so now Exim always runs a verification when a
6353 callout is required, relying on the callout cache for the optimization.
6354 The overhead is duplication of the address routing, but this should not be
6357 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6358 to be verified with callout=postmaster, and the main callout worked but the
6359 postmaster check failed, the verification correctly failed. However, if a
6360 subsequent RCPT command asked for sender verification *without* the
6361 postmaster check, incorrect caching caused this verification also to fail,
6364 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6365 it was not caching the DNS options (qualify_single, search_parents) that
6366 were used when the lookup failed. A subsequent lookup with different
6367 options therefore always gave the same answer, though there were cases
6368 where it should not have. (Example: a "domains = !$mx_any" option on a
6369 dnslookup router: the "domains" option is always processed without any
6370 widening, but the router might have qualify_single set.) Now Exim uses the
6371 cached value only when the same options are set.
6373 32. Added John Jetmore's "exipick" utility to the distribution.
6375 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6376 than a timeout (e.g. a certificate is required, and is not provided), an
6377 Exim server now closes the connection immediately. Previously it waited for
6378 the client to close - but if the client is SSL, it seems that they each
6379 wait for each other, leading to a delay before one of them times out.
6381 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6382 maintained 0.8.x compatibility because I don't think many are using it, and
6383 it is clearly obsolete.
6385 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6388 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6389 client certificate was expired. A simple patch fixes this, though I don't
6390 understand the full logic of why the verify callback is called multiple
6393 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6394 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6395 which causes problems with some clients (such as the Certicom SSL Plus
6396 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6397 disables the coutermeasure allowing Eudora to connect."
6399 38. Exim was not checking that a write() to a log file succeeded. This could
6400 lead to Bad Things if a log got too big, in particular if it hit a file
6401 size limit. Exim now panics and dies if it cannot write to a log file, just
6402 as it does if it cannot open a log file.
6404 39. Modified OS/Makefile-Linux so that it now contains
6406 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6408 The two -D definitions ensure that Exim is compiled with large file
6409 support, which makes it possible to handle log files that are bigger than
6412 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6413 instance) a domain was checked against a named list that involved a lookup,
6414 causing $domain_data to be set, then another domain was checked against the
6415 same list, then the first domain was re-checked, the value of $domain_data
6416 after the final check could be wrong. In particular, if the second check
6417 failed, it could be set empty. This bug probably also applied to
6420 41. The strip_trailing_dot option was not being applied to the address given
6421 with the -f command-line option.
6423 42. The code for reading a message's header from the spool was incrementing
6424 $received_count, but never initializing it. This meant that the value was
6425 incorrect (doubled) while delivering a message in the same process in which
6426 it was received. In the most common configuration of Exim, this never
6427 happens - a fresh exec is done - but it can happen when
6428 deliver_drop_privilege is set.
6430 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6431 it now includes up to 150 characters of the unexpected data in the log
6434 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6435 and building data strings. The size of both of these buffers was 10 000
6436 bytes - far larger than anybody would *ever* want, thought I. Needless to
6437 say, somebody hit the limit. I have increased the maximum line length to
6438 20 000 and the maximum data length of concatenated lines to 100 000. I have
6439 also fixed two bugs, because there was no checking on these buffers. Tsk,
6440 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6441 buffer is too small.
6443 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6444 lsearch lookups. Now it does.
6446 46. When parsing a route_list item in a manualroute router, a fixed-length
6447 buffer was used for the list of hosts. I made this 1024 bytes long,
6448 thinking that nobody would ever have a list of hosts that long. Wrong.
6449 Somebody had a whole pile of complicated expansion conditions, and the
6450 string was silently truncated, leading to an expansion error. It turns out
6451 that it is easier to change to an unlimited length (owing to other changes
6452 that have happened since this code was originally written) than to build
6453 structure for giving a limitation error. The length of the item that
6454 expands into the list of hosts is now unlimited.
6456 47. The lsearch lookup could not handle data where the length of text line was
6457 more than 4095 characters. Such lines were truncated, leading to shortened
6458 data being returned. It should now handle lines of any length.
6460 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6461 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6464 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6465 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6466 no longer happens in comment lines. A list of replacements is now placed
6467 at the head of all of the source files, except those whose only change is
6468 to replace PERL_COMMAND in the very first #! line.
6470 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6471 messages on the queue, with a bottom-up merge sort, using code contributed
6472 by Michael Haardt. This should make operations like -bp somewhat faster on
6473 large queues. It won't affect queue runners, except when queue_run_in_order
6476 51. Installed eximstats 1.31 in the distribution.
6478 52. Added support for SRV lookups to the dnslookup router.
6480 53. If an ACL referred to $message_body or $message_body_end, the value was not
6481 reset for any messages that followed in the same SMTP session.
6483 54. The store-handling optimization for building very long strings was not
6484 differentiating between the different store pools. I don't think this
6485 actually made any difference in practice, but I've tidied it.
6487 55. While running the routers to verify a sender address, $sender_address
6488 was still set to the sender address. This is wrong, because when routing to
6489 send a bounce to the sender, it would be empty. Therefore, I have changed
6490 it so that, while verifying a sender address, $sender_address is set to <>.
6491 (There is no change to what happens when verifying a recipient address.)
6493 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6494 target A or AAAA records (if not already returned) without resetting the
6495 qualify_single or search_parents options of the DNS resolver. These are
6496 inappropriate in this case because the targets of MX and SRV records must
6497 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6498 target that, when qualified, matched something in the local domain. These
6499 two options are now turned off when doing these lookups.
6501 57. It seems that at least some releases of Reiserfs (which does not have the
6502 concept of a fixed number of inodes) returns zero and not -1 for the
6503 number of available inodes. This interacted badly with check_spool_inodes,
6504 which assumed that -1 was the "no such thing" setting. What I have done is
6505 to check that the total number of inodes is greater than zero before doing
6506 the test of how many are available.
6508 58. When a "warn" ACL statement has a log_message modifier, the message is
6509 remembered, and not repeated. This is to avoid a lot of repetition when a
6510 message has many recipients that cause the same warning to be written.
6511 However, Exim was preserving the list of already written lines for an
6512 entire SMTP session, which doesn't seem right. The memory is now reset if a
6513 new message is started.
6515 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6516 debugging output unless log_rewrite was also set.
6518 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6519 of (int)(handle) when we know that handle contains (void *)(-1).
6521 61. The Exim daemon panic-logs an error return when it closes the incoming
6522 connection. However "connection reset by peer" seems to be common, and
6523 isn't really an error worthy of noting specially, so that particular error
6526 62. When Exim is trying to find all the local interfaces, it used to panic and
6527 die if the ioctl to get the interface flags failed. However, it seems that
6528 on at least one OS (Solaris 9) it is possible to have an interface that is
6529 included in the list of interfaces, but for which you get a failure error
6530 for this call. This happens when the interface is not "plumbed" into a
6531 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6532 failure of the "get flags" call assumes that the interface is down.
6534 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6535 makes life easier for people who are doing arithmetic on fields extracted
6536 from dates, where you often get leading zeros that should not be
6537 interpreted as octal.
6539 64. Added qualify_domain to the redirect router, to override the global
6542 65. If a pathologically long header line contained very many addresses (the
6543 report of this problem mentioned 10 000) and each of them was rewritten,
6544 Exim could use up a very large amount of memory. (It kept on making new
6545 copies of the header line as it rewrote, and never released the old ones.)
6546 At the expense of a bit more processing, the header rewriting function has
6547 been changed so that it no longer eats memory in this way.
6549 66. The generation of the Received: header has been moved from the time that a
6550 message starts to be received, to the time that it finishes. The timestamp
6551 in the Received: header should now be very close to that of the <= log
6552 line. There are two side-effects of this change:
6554 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6555 logged header lines no longer include the local Received: line, because
6556 it has not yet been created. The same applies to a copy of the message
6557 that is returned to a non-SMTP sender when a message is rejected.
6559 (b) When a filter file is tested using -bf, no additional Received: header
6560 is added to the test message. After some thought, I decided that this
6563 This change does not affect the value of $received_for. It is still set
6564 after address rewriting, but before local_scan() is called.
6566 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6568 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6569 gave an unhelpful panic error message, and a defer error. I have managed to
6570 change this behaviour so that it now rejects any supplied certificate,
6571 which seems right, as the list of acceptable certificates is empty.
6573 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6574 gave an unhelpful defer error. I have not managed to make this reject any
6575 supplied certificates, but the error message it gives is "no certificate
6576 supplied", which is not helpful.
6578 70. exigrep's output now also includes lines that are not associated with any
6579 message, but which match the given pattern. Implemented by a patch from
6580 Martin Sluka, which also tidied up the Perl a bit.
6582 71. Recipient callout verification, like sender verification, was using <> in
6583 the MAIL FROM command. This isn't really the right thing, since the actual
6584 sender may affect whether the remote host accepts the recipient or not. I
6585 have changed it to use the actual sender in the callout; this means that
6586 the cache record is now keyed on a recipient/sender pair, not just the
6587 recipient address. There doesn't seem to be a real danger of callout loops,
6588 since a callout by the remote host to check the sender would use <>.
6589 [SEE ABOVE: changed after hitting problems.]
6591 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6592 temporary errors. However, in the case of such a code being given after
6593 the end of a data transmission (i.e. after ".") Exim was failing to write
6594 a retry record for the message. (Yes, there was some broken host that was
6595 actually sending 8xx at this point.)
6597 73. An unknown lookup type in a host list could cause Exim to panic-die when
6598 the list was checked. (An example that provoked this was putting <; in the
6599 middle of a list instead of at the start.) If this happened during a DATA
6600 ACL check, a -D file could be left lying around. This kind of configuration
6601 error no longer causes Exim to die; instead it causes a defer error. The
6602 incident is still logged to the main and panic logs.
6604 74. Buglet left over from Exim 3 conversion. The message "too many messages
6605 in one connection" was written to the rejectlog but not the mainlog, except
6606 when address rewriting (yes!) was being logged.
6608 75. Added write_rejectlog option.
6610 76. When a system filter was run not as root (that is, when system_filter_user
6611 was set), the values of the $n variables were not being returned to the
6612 main process; thus, they were not subsequently available in the $sn
6615 77. Added +return_path_on_delivery log selector.
6617 78. A connection timeout was being treated differently from recipients deferred
6618 when testing hosts_max_try with a message that was older than the host's
6619 retry timeout. (The host should not be counted, thus allowing all hosts to
6620 be tried at least once before bouncing.) This may have been the cause of an
6621 occasionally reported bug whereby a message would remain on the queue
6622 longer than the retry timeout, but would be bounced if a delivery was
6623 forced. I say "may" because I never totally pinned down the problem;
6624 setting up timeout/retry tests is difficult. See also the next item.
6626 79. The ultimate address timeout was not being applied to errors that involved
6627 a combination of host plus message (for example, a timeout on a MAIL
6628 command). When an address resolved to a number of possible hosts, and they
6629 were not all tried for each delivery (e.g. because of hosts_max_try), a
6630 message could remain on the queue longer than the retry timeout.
6632 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6635 81. Fixed an obscure SMTP outgoing bug which required at least the following
6636 conditions: (a) there was another message waiting for the same server;
6637 (b) the server returned 5xx to all RCPT commands in the first message so
6638 that the message was not completed; (c) the server dropped the connection
6639 or gave a negative response to the RSET that Exim sends to abort the
6640 transaction. The observed case was a dropped connection after DATA that had
6641 been sent in pipelining mode. That is, the server had advertised PIPELINING
6642 but was not implementing it correctly. The effect of the bug was incorrect
6643 behaviour, such as trying another host, and this could lead to a crash.
6649 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6650 and daemon.c were passed as pointers to ints; they should have been
6651 pointers to socklen_t variables (which are typically unsigned ints).
6653 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6656 3. Fixed a really odd bug that affected only the testing scheme; patching a
6657 certain fixed string in the binary changed the value of another string that
6658 happened to be identical to the end of the original first string.
6660 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6661 name", it returns that address as the IP address. On some operating
6662 systems (e.g. Solaris), it also passes back the IP address string as the
6663 "host name". However, on others (e.g. Linux), it passes back an empty
6664 string. Exim wasn't checking for this, and was changing the host name to an
6665 empty string, assuming it had been canonicalized.
6667 5. Although rare, it is permitted to have more than one PTR record for a given
6668 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6669 all the names associated with an address, because they do in Solaris.
6670 However, it seems that they do not in Linux for data that comes from the
6671 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6672 I found this out when I moved to a new Linux workstation and tried to run
6673 the Exim test suite.
6675 To get round this problem I have changed the code so that it now does its
6676 own call to the DNS to look up PTR records when searching for a host name.
6677 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6678 addresses that are only in /etc/hosts are still found.
6680 This behaviour is, however, controlled by an option called host_lookup_
6681 order, which defaults to "bydns:byaddr". If people want to use the other
6682 order, or indeed, just use one or the other means of lookup, they can
6683 specify it in this variable.
6685 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6686 some operating systems, this comes back from gethostbyaddr() as an empty
6687 string, and this is what Exim used to test for. However, it seems that in
6688 other systems, "." is yielded. Exim now tests for this case too.
6690 7. The values of check_spool_space and check_log_space are now held internally
6691 as a number of kilobytes instead of an absolute number of bytes. If a
6692 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6693 kilobyte. This means that much larger values can be stored.
6695 8. Exim monitor: an attempt to get the action menu when not actually pointing
6696 at a message produces an empty menu entitled "No message selected". This
6697 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6698 no entries in it ("Shell widget menu has zero width and/or height"). So I
6699 have added a single, blank menu entry in this case.
6701 9. Added ${quote_local_part.
6703 10. MIME decoding is now applied to the contents of Subject: header lines when
6706 11. Now that a reference to $sender_host_address automatically causes a reverse
6707 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6708 host lookup before query-style lookups in lists that might use this
6709 variable. This has therefore been abolished, and the "net-" prefix is no
6710 longer necessary for query-style lookups.
6712 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6713 have been a typo for LFLAGS, so it has been changed.
6715 13. The install script calls Exim with "-C /dev/null" in order to find the
6716 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6717 to be output. However, since Exim outputs its version number before the
6718 error, it didn't break the script. It just looked ugly. I fixed this by
6719 always allowing "-C /dev/null" if the caller is root.
6721 14. Ignore overlarge ACL variable number when reading spool file - insurance
6722 against a later release with more variables having written the file.
6724 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6725 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6726 this, as well as the form without the "IPv6" on the front.
6728 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6729 OS/Makefile-Darwin file.
6731 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6732 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6734 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6735 variable. However, some OS use ints for inodes. I've added cast to long int
6736 to get rid of the compiler warning.
6738 19. I had forgotten to lock out "/../" in configuration file names when
6739 ALT_CONFIG_PREFIX was set.
6741 20. Routers used for verification do not need to specify transports. However,
6742 if such a router generated a host list, and callout was configured, Exim
6743 crashed, because it could not find a port number from the (non-existent)
6744 transport. It now assumes port 25 in this circumstance.
6746 21. Added the -t option to exigrep.
6748 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6749 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6750 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6751 them is set, LOOKUP_LSEARCH is forced.
6753 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6755 24. Added sender and host information to the "rejected by local_scan()" log
6756 line; previously there was no indication of these.
6758 25. Added .include_if_exists.
6760 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6761 Linux. It turns out that not all file systems support this. Apparently some
6762 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6763 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6764 not supported on the file descriptor, is now ignored when Exim is trying to
6765 sync a directory. This applies only to Linux.
6767 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6769 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6770 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6773 29. Added support for maildirsize files from supplied patch (modified a bit).
6775 30. The use of :fail: followed by an empty string could lead Exim to respond to
6776 sender verification failures with (e.g.):
6778 550 Verification failed for <xxx>
6779 550 Sender verify failed
6781 where the first response line was missing the '-' that indicates it is not
6782 the final line of the response.
6784 31. The loop for finding the name of the user that called Exim had a hardwired
6785 limit of 10; it now uses the value of finduser_retries, which is used for
6786 all other user lookups.
6788 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6791 33. Exim was neglecting to zero errno before one call of strtol() when
6792 expanding a string and expecting an integer value. On some systems this
6793 resulted in spurious "integer overflow" errors. Also, it was casting the
6794 result into an int without checking.
6796 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6797 did not work. The code looks as if it has *never* worked, though it appears
6798 to have been documented since at least release 1.62. I have made it work.
6800 35. The "timeout_DNS" error in retry rules, also documented since at least
6801 1.62, also never worked. As it isn't clear exactly what this means, and
6802 clearly it isn't a major issue, I have abolished the feature by treating it
6803 as "timeout", and writing a warning to the main and panic logs.
6805 36. The display of retry rules for -brt wasn't always showing the error code
6808 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6809 timeout_connect_A, timeout_connect_MX.
6811 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6812 to the empty sender.
6814 39. The daemon was not analysing the content of -oX till after it had closed
6815 stderr and disconnected from the controlling terminal. This meant that any
6816 syntax errors were only noted on the panic log, and the return code from
6817 the command was 0. By re-arranging the code a little, I've made the
6818 decoding happen first, so such errors now appear on stderr, and the return
6819 code is 1. However, the actual setting up of the sockets still happens in
6820 the disconnected process, so errors there are still only recorded on the
6823 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6824 connections (as happens on some IP stacks) was logged at start up time as
6825 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6826 it from "IPv6 and IPv4", which means that two separate sockets are being
6829 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6830 says whether AF_INET or AF_INET6 was passed as an argument.
6832 42. Exiwhat output was messed up when time zones were included in log
6835 43. Exiwhat now gives more information about the daemon's listening ports,
6836 and whether -tls-on-connect was used.
6838 44. The "port" option of the smtp transport is now expanded.
6840 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6841 silently ignored. Now an error message is written to the main and panic
6844 46. There's a new ACL modifier called "logwrite" which writes to a log file
6845 as soon as it is encountered.
6847 47. Added $local_user_uid and $local_user_gid at routing time.
6849 48. Exim crashed when trying to verify a sender address that was being
6852 49. Exim was recognizing only a space character after ".include". It now also
6853 recognizes a tab character.
6855 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6856 extracting the relevant information from the specification. The man page no
6857 longer contains scrambled data for the -d option, and I've added a section
6858 at the front about calling Exim under different names.
6860 51. Added "extra_headers" argument to the "mail" command in filter files.
6862 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6865 53. Installed eximstats 1.29.
6867 54. Added transport_filter_timeout as a generic transport option.
6869 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6870 Cc: header lines. This was required by RFC 822, but it not required by RFC
6873 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6874 incoming messages that did not have them. Now it does so only if the
6875 message originates locally, that is, if there is no associated remote host
6876 address. When Resent- header lines are present, this applies to the Resent-
6877 lines rather than the non-Resent- lines.
6879 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6880 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6882 58. Messages for configuration errors now include the name of the main
6883 configuration file - useful now that there may be more than one file in a
6884 list (.included file names were always shown).
6886 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6887 for those rare installations that do not start the daemon as root or run it
6888 setuid root. I've cut out the call to initgroups() if the daemon is not
6891 60. The Exim user and group can now be bound into the binary as text strings
6892 that are looked up at the start of Exim's processing.
6894 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6896 62. Added $mailstore_basename variable.
6898 63. Installed patch to sieve.c from Michael Haardt.
6900 64. When Exim failed to open the panic log after failing to open the main log,
6901 the original message it was trying to log was written to stderr and debug
6902 output, but if they were not available (the usual case in production), it
6903 was lost. Now it is written to syslog before the two lines that record the
6904 failures to open the logs.
6906 65. Users' Exim filters run in subprocesses under the user's uid. It is
6907 possible for a "deliver" command or an alias in a "personal" command to
6908 provoke an address rewrite. If logging of address rewriting is configured,
6909 this fails because the process is not running as root or exim. There may be
6910 a better way of dealing with this, but for the moment (because 4.30 needs
6911 to be released), I have disabled address rewrite logging when running a
6912 filter in a non-root, non-exim process.
6918 1. The buildconfig auxiliary program wasn't quoting the value set for
6919 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6920 not defined. This bug was masked in 4.22 by the effect that was fixed in
6923 2. Some messages that were rejected after a message id was allocated were
6924 shown as "incomplete" by exigrep. It no longer does this for messages that
6925 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6927 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6928 have allow_domain_literals set, the ID did not get logged in the <= line.
6929 Domain literals are now always recognized in Message-ID: header lines.
6931 4. The first argument for a ${extract expansion item is the key name or field
6932 number. Leading and trailing spaces in this item were not being ignored,
6933 causing some misleading effects.
6935 5. When deliver_drop_privilege was set, single queue runner processes started
6936 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6937 same command in the process it spins off) were not dropping privilege.
6939 6. When the daemon running as "exim" started a queue runner, it always
6940 re-executed Exim in the spun-off process. This is a waste of effort when
6941 deliver_drop_privilege is set. The new process now just calls the
6942 queue-runner function directly.
6948 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6951 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6952 ignored. Though the use of -oP was forcing the writing of a pid file, it
6953 was always written to the default place.
6955 3. If the message "no IP address found for host xxxx" is generated during
6956 incoming verification, it is now followed by identification of the incoming
6957 connection (so you can more easily find what provoked it).
6959 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6961 5. Added some features to "harden" Exim a bit more against certain attacks:
6963 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6964 be put in Local/Makefile. This is like the never_users runtime option,
6965 but it cannot be overridden. The default setting is "root".
6967 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6968 prefix string with which any file named in a -C command line option
6971 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6972 is retained for -C and -D only if the caller of Exim is root. Without
6973 it, the exim user may also use -C and -D and retain privilege.
6975 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6976 command line option is disabled.
6978 6. Macro names set by the -D option must start with an upper case letter, just
6979 like macro names defined in the configuration file.
6981 7. Added "dereference=" facility to LDAP.
6983 8. Two instances of the typo "uknown" in the source files are fixed.
6985 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6986 the Configure-Makefile script screwed up while processing it.
6988 10. Incorporated PCRE 4.4.
6990 11. The SMTP synchronization check was not operating right at the start of an
6991 SMTP session. For example, it could not catch a HELO sent before the client
6992 waited for the greeting. There is now a check for outstanding input at the
6993 point when the greeting is written. Because of the duplex, asynchronous
6994 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6995 way, but not yet received, when the check is performed.
6997 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6998 on TCP/IP sockets, because this apparently causes some broken clients to
7001 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7002 unchanged) from the Cygwin maintainer.
7004 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7005 was supported instead of testing for mbx. Effectively a typo.
7007 15. The spa authenticator server code was not checking that the input it
7008 received was valid base64.
7010 16. The debug output line for the "set" modifier in ACLs was not showing the
7011 name of the variable that was being set.
7013 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7015 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7016 lookup on its own. Something else was needed to trigger the lookup. For
7017 example, a match in host_lookup or the need for a host name in a host list.
7018 Now, if $sender_host_name is referenced and the host name has not yet been
7019 looked up, a lookup is performed. If the lookup fails, the variable remains
7020 empty, and $host_lookup_failed is set to "1".
7022 19. Added "eqi" as a case-independent comparison operator.
7024 20. The saslauthd authentication condition could segfault if neither service
7025 nor realm was specified.
7027 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7028 error message that was logged was misleading, and incoming SMTP
7029 connections were dropped. The message is now more accurate, and temporary
7030 errors are given to SMTP connections.
7032 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7033 (except RSET) with a 421 error, until QUIT is received. However, it was
7034 failing to send a response to QUIT.
7036 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7037 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7038 ACL is now run after the helo_try_verify_hosts code.
7040 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7041 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7042 case-independent, so other case variants are also recognized). Apparently
7043 some systems use these upper case variants.
7045 25. If more than two messages were waiting for the same host, and a transport
7046 filter was specified for the transport, Exim sent two messages over the
7047 same TCP/IP connection, and then failed with "socket operation on non-
7048 socket" when it tried to send the third.
7050 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7052 27. The extern definition of crypt16() in expand.c was not being excluded when
7053 the OS had its own crypt16() function.
7055 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7056 as a preferred synonym for return_size_limit, both as an option and as an
7059 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7061 30. Changed the default configuration ACL to relax the local part checking rule
7062 for addresses that are not in any local domains. For these addresses,
7063 slashes and pipe symbols are allowed within local parts, but the sequence
7064 /../ is explicitly forbidden.
7066 31. SPA server authentication was not clearing the challenge buffer before
7069 32. log_message in a "warn" ACL statement was writing to the reject log as
7070 well as to the main log, which contradicts the documentation and doesn't
7071 seem right (because no rejection is happening). So I have stopped it.
7073 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7074 However, I am unable to do any testing of this.
7076 34. Fixed an infelicity in the appendfile transport. When checking directories
7077 for a mailbox, to see if any needed to be created, it was accidentally
7078 using path names with one or more superfluous leading slashes; tracing
7079 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7081 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7082 contents are added to the log line that is written for every discarded
7083 recipient. (Previously a log_message setting was ignored.)
7085 36. The ${quote: operator now quotes the string if it is empty.
7087 37. The install script runs exim in order to find its version number. If for
7088 some reason other than non-existence or emptiness, which it checks, it
7089 could not run './exim', it was installing it with an empty version number,
7090 i.e. as "exim-". This error state is now caught, and the installation is
7093 38. An argument was missing from the function that creates an error message
7094 when Exim fails to connect to the socket for saslauthd authentication.
7095 This could cause Exim to crash, or give a corrupted message.
7097 39. Added isip, isip4, and isip6 to ${if conditions.
7099 40. The ACL variables $acl_xx are now saved with the message, and can be
7100 accessed later in routers, transports, and filters.
7102 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7103 strings in the file are not string-expanded.
7105 42. If a MAIL command specified a SIZE value that was too large to fit into an
7106 int variable, the check against message_size_limit failed. Such values are
7107 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7108 day this will have to be increased, but I don't think I want to be around
7109 when emails are that large.
7116 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7117 iconv() is not standard in FreeBSD.
7119 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7120 IPv6 enabled. The observed symptom was a segmentation fault on return from
7121 the function os_common_find_running_interfaces() in src/os.c.
7123 3. In the check_special_case() function in daemon.c I had used "errno" as an
7124 argument name, which causes warnings on some systems. This was basically a
7125 typo, since it was named "eno" in the comments!
7127 4. The code that waits for the clock to tick (at a resolution of some fraction
7128 of a second) so as to ensure message-id uniqueness was always waiting for
7129 at least one whole tick, when it could have waited for less. [This is
7130 almost certainly not relevant at current processor speeds, where it is
7131 unlikely to ever wait at all. But we try to future-proof.]
7133 5. The function that sleeps for a time interval that includes fractions of a
7134 second contained a race. It did not block SIGALRM between setting the
7135 timer, and suspending (a couple of lines later). If the interval was short
7136 and the sigsuspend() was delayed until after it had expired, the suspension
7137 never ended. On busy systems this could lead to processes getting stuck for
7140 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7141 process, before it forks any delivery processes. The open lookup caching
7142 mechanism meant that the open file or database connection was passed into
7143 the delivery process. The problem was that delivery processes always tidy
7144 up cached lookup data. This could cause a problem for the next delivery
7145 process started by the queue runner, because the external queue runner
7146 process does not know about the closure. So the next delivery process
7147 still has data in the lookup cache. In the case of a file lookup, there was
7148 no problem because closing a file descriptor in a subprocess doesn't affect
7149 the parent. However, if the lookup was caching a connection to a database,
7150 the connection was closed, and the second delivery process was likely to
7151 see errors such as "PGSQL: query failed: server closed the connection
7152 unexpectedly". The problem has been fixed by closing all cached lookups
7153 in a queue runner before running a delivery process.
7155 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7156 seem to have the "const" qualifier which it has on other OS. I've
7159 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7160 *of the same type* (client or server) with the same public name that an
7161 error should be diagnosed.
7163 9. When Exim looked up a host name for an IP address, but failed to find the
7164 original IP address when looking up the host name (a safety check), it
7165 output the message "<ip address> does not match any IP for NULL", which was
7166 confusing, to say the least. The bug was that the host name should have
7167 appeared instead of "NULL".
7169 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7170 user that is built into the binary, and the -C or -D options is used, root
7171 privilege is dropped before the configuration file is read. In addition,
7172 logging is switched to stderr instead of the normal log files. If the
7173 configuration then re-defines the Exim user, the unprivileged environment
7174 is probably not what is expected, so Exim logs a panic warning message (but
7177 However, if deliver_drop_privilege is set, the unprivileged state may well
7178 be exactly what is intended, so the warning has been cut out in that case,
7179 and Exim is allowed to try to write to its normal log files.
7185 1. smtp_return_error_details was not giving details for temporary sender
7186 or receiver verification errors.
7188 2. Diagnose a configuration error if two authenticators have the same public
7191 3. Exim used not to create the message log file for a message until the first
7192 delivery attempt. This could be confusing when incoming messages were held
7193 for policy or load reasons. The message log file is now created at the time
7194 the message is received, and an initial "Received" line is written to it.
7196 4. The automatically generated man page for command line options had a minor
7197 bug that caused no ill effects; however, a more serious problem was that
7198 the procedure for building the man page automatically didn't always
7199 operate. Consequently, release 4.20 contains an out-of-date version. This
7200 shouldn't happen again.
7202 5. When building Exim with embedded Perl support, the script that builds the
7203 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7204 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7206 6. The freeze_tell option was not being used for messages that were frozen on
7207 arrival, either by an ACL or by local_scan().
7209 7. Added the smtp_incomplete_transaction log selector.
7211 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7212 was accepting AUTH without a new EHLO.
7214 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7215 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7218 10. Logging of TCP/IP connections (when configured) now happens in the main
7219 daemon process instead of the child process, so that the TCP/IP connection
7220 count is more accurate (but it can never be perfect).
7222 11. The use of "drop" in a nested ACL was not being handled correctly in the
7223 outer ACL. Now, if condition failure induced by the nested "drop" causes
7224 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7225 or "require"), the connection is dropped.
7227 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7228 that yield "discard" can now be used with an "accept" or a "discard" verb,
7229 but an error is generated for any others (because I can't see a useful way
7230 to define what should happen).
7232 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7233 are now processed in the same way as lines in the Exim configuration file.
7234 In particular, continuation lines are supported.
7236 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7238 15. Added -ti meaning -t -i.
7240 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7241 domains, and warn by logging if others are found.
7243 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7244 returned from GIFCONF when Exim is trying to find the list of interfaces on
7245 a host. The code in os.c has been modified to copy each ifreq to an aligned
7246 structure in all cases.
7248 Also, in some cases, the returned ifreq's were being copied to a 'struct
7249 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7250 means the last couple of bytes of an IPv6 address could be chopped if the
7251 ifreq contained only a normal sockaddr (14 bytes storage).
7253 18. Named domain lists were not supported in the hosts_treat_as_local option.
7254 An entry such as +xxxx was not recognized, and was treated as a literal
7257 19. Ensure that header lines added by a DATA ACL are included in the reject log
7258 if the ACL subsequently rejects the message.
7260 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7261 MD5 (which is deprecated).
7263 21. When testing a filter file using -bf, Exim was writing a message when it
7264 took the sender from a "From " line in the message, but it was not doing so
7265 when it took $return_path from a Return-Path: header line. It now does.
7267 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7268 with a valid header line field name (a series of printing characters
7269 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7271 23. Changed "disc" in the source to "disk" to conform to the documentation and
7272 the book and for uniformity.
7274 24. Ignore Sendmail's -Ooption=value command line item.
7276 25. When execve() failed while trying to run a command in a pipe transport,
7277 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7278 could be confused with a return value of 69 from the command itself. This
7279 has been changed to 127, the value the shell returns if it is asked to run
7280 a non-existent command. The wording for the related log line suggests a
7281 non-existent command as the problem.
7283 26. If received_header_text expands to an empty string, do not add a Received:
7284 header line to the message. (Well, it adds a token one on the spool, but
7285 marks it "old" so that it doesn't get used or transmitted.)
7287 27. Installed eximstats 1.28 (addition of -nt option).
7289 28. There was no check for failure on the call to getsockname() in the daemon
7290 code. This can fail if there is a shortage of resources on the system, with
7291 ENOMEM, for example. A temporary error is now given on failure.
7293 29. Contrary to the C standard, it seems that in some environments, the
7294 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7295 program. Exim now does this explicitly; it affects the formatting of
7296 timestamps using strftime().
7298 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7299 complaints. I've now initialized all the variables, to avoid this.
7301 32. Header lines added by a system filter were not being "seen" during
7302 transport-time rewrites.
7304 33. The info_callback() function passed to OpenSSL is set up with type void
7305 (*)(SSL *, int, int), as described somewhere. However, when calling the
7306 function (actually a macro) that sets it up, the type void(*)() is
7307 expected. I've put in a cast to prevent warnings from picky compilers.
7309 34. If a DNS black list lookup found a CNAME record, but there were no A
7310 records associated with the domain it pointed at, Exim crashed.
7312 35. If a DNS black list lookup returned more than one A record, Exim ignored
7313 all but the first. It now scans all returned addresses if a particular IP
7314 value is being sought. In this situation, the contents of the
7315 $dnslist_value variable are a list of all the addresses, separated by a
7318 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7319 to accept a host name and all its aliases if the forward lookup for any of
7320 them yielded the IP address of the incoming connection. Now it accepts only
7321 those names whose forward lookup yields the correct IP address. Any other
7322 names are discarded. This closes a loophole whereby a rogue DNS
7323 administrator could create reverse DNS records to break through a
7324 wildcarded host restriction in an ACL.
7326 37. If a user filter or a system filter that ran in a subprocess used any of
7327 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7328 the wrong values were passed to the pipe command ($thisaddress had the
7329 value of $0, $0 had the value of $1, etc). This bug was introduced by
7330 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7332 38. Improved the line breaking for long SMTP error messages from ACLs.
7333 Previously, if there was no break point between 40 and 75 characters, Exim
7334 left the rest of the message alone. Two changes have been made: (a) I've
7335 reduced the minimum length to 35 characters; (b) if it can't find a break
7336 point between 35 and 75 characters, it looks ahead and uses the first one
7337 that it finds. This may give the occasional overlong line, but at least the
7338 remaining text gets split now.
7340 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7341 file descriptors might be low, and that setting 1000 would always raise it.
7342 It turns out that in some environments, the limit is already over 1000 and
7343 that lowering it causes trouble. So now Exim takes care not to decrease it.
7345 40. When delivering a message, the value of $return_path is set to $sender_
7346 address at the start of routing (routers may change the value). By an
7347 oversight, this default was not being set up when an address was tested by
7348 -bt or -bv, which affected the outcome if any router or filter referred to
7351 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7352 log only when "message" or "log_message" are set. However, if one of the
7353 conditions was an address verification, or a call to a nested ACL, the
7354 messages generated by the underlying test were being passed through. This
7355 no longer happens. The underlying message is available in $acl_verify_
7356 message for both "message" and "log_message" expansions, so it can be
7357 passed through if needed.
7359 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7360 new expansion $bh_ to give the encoded byte string without charset
7361 translation. Translation happens only if iconv() is available; HAVE_ICONV
7362 indicates this at build time. HEADERS_CHARSET gives the charset to
7363 translate to; headers_charset can change it in the configuration, and
7364 "headers charset" can change it in an individual filter file.
7366 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7367 that creates RFC 2047 encoded "words" labels them as that charset instead
7368 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7369 expansion operator; (ii) when Exim creates a From: line for a local
7370 message; (iii) when a header line is rewritten to include a "phrase" part.
7372 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7373 buggy, causing it to skip the first lines of messages whose message ID
7374 ended in 'D'. This would not have bitten before Exim release 4.14, because
7375 message IDs were unlikely to end in 'D' before then. The effect was to have
7376 incorrect size information for certain domains.
7378 45. #include "config.h" was missing at the start of the crypt16.c module. This
7379 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7382 46. If there was a timeout during a "random" callout check, Exim treated it as
7383 a failure of the random address, and carried on sending RSET and the real
7384 address. If the delay was just some slowness somewhere, the response to the
7385 original RCPT would be taken as a response to RSET and so on, causing
7386 mayhem of various kinds.
7388 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7389 when I implemented it. It didn't allow for the fact that some option values
7390 may legitimately be negative (e.g. size_addition), and it didn't even do
7391 the right test for positive values.
7393 48. Domain names in DNS records are case-independent. Exim always looks them up
7394 in lower case. Some resolvers return domain names in exactly the case they
7395 appear in the zone file, that is, they may contain uppercase letters. Not
7396 all resolvers do this - some return always lower case. Exim was treating a
7397 change of case by a resolver as a change of domain, similar to a widening
7398 of a domain abbreviation. This triggered its re-routing code and so it was
7399 trying to route what was effectively the same domain again. This normally
7400 caused routing to fail (because the router wouldn't handle the domain
7401 twice). Now Exim checks for this case specially, and just changes the
7402 casing of the domain that it ultimately uses when it transmits the message
7405 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7408 50. If a filter generated a file delivery with a non-absolute name (possible if
7409 no home directory exists for the router), the forbid_file option was not
7412 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7413 the existing equality matching.
7415 52. Exim was using ints instead of ino_t variables in some places where it was
7416 dealing with inode numbers.
7418 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7419 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7420 called TMPDIR, and if it finds it is different, it changes its value.
7422 54. The smtp_printf() function is now made available to local_scan() so
7423 additional output lines can be written before returning. There is also an
7424 smtp_fflush() function to enable the detection of a dropped connection.
7425 The variables smtp_input and smtp_batched_input are exported to
7428 55. Changed the default runtime configuration: the message "Unknown user"
7429 has been removed from the ACL, and instead placed on the localuser router,
7430 using the cannot_route_message feature. This means that any verification
7431 failures that generate their own messages won't get overridden. Similarly,
7432 the "Unrouteable address" message that was in the ACL for unverifiable
7433 relay addresses has also been removed.
7435 56. Added hosts_avoid_esmtp to the smtp transport.
7437 57. The exicyclog script was not checking for the esoteric option
7438 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7439 will work only if exicyclog is run under the appropriate euid.
7441 58. Following a discussion on the list, the rules by which Exim recognises line
7442 endings on incoming messages have been changed. The -dropcr and drop_cr
7443 options are now no-ops, retained only for backwards compatibility. The
7444 following line terminators are recognized: LF CRLF CR. However, special
7445 processing applies to CR:
7447 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7448 nor a local message in the state where . is a terminator.
7450 (ii) If a bare CR is encountered in a header line, an extra space is added
7451 after the line terminator so as not to end the header. The reasoning
7452 behind this is that bare CRs in header lines are most likely either
7453 to be mistakes, or people trying to play silly games.
7455 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7456 was being incorrectly given as 18 bytes larger than it should have been.
7457 This is a VOB (very old bug).
7459 60. This may never have affected anything current, but just in case it has:
7460 When the local host is found other than at the start of a list of hosts,
7461 the local host, those with the same MX, and any that follow, are discarded.
7462 When the list in question was part of a longer list of hosts, the following
7463 hosts (not currently being processed) were also being discarded. This no
7464 longer happens. I'm not sure if this situation could ever has previously
7467 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7470 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7471 header. This is something that is recommended in a new Internet Draft, and
7472 is something that is documented as being done by Sendmail. There are two
7473 possible values. For messages generated by the autoreply transport, Exim
7476 Auto-Submitted: auto-replied
7478 whereas for all other generated messages (e.g. bounces) it adds
7480 Auto-Submitted: auto-generated
7482 63. The "personal" condition in filters now includes a test for the
7483 Auto-Submitted: header. If it contains the string "auto-" the message it
7484 not considered personal.
7486 64. Added rcpt_include_affixes as a generic transport option.
7488 65. Added queue_only_override (default true).
7490 66. Added the syslog_duplication option.
7492 67. If what should have been the first header line of a message consisted of
7493 a space followed by a colon, Exim was mis-interpreting it as a header line.
7494 It isn't of course - it is syntactically invalid and should therefore be
7495 treated as the start of the message body. The misbehaviour could have
7496 caused a number of strange effects, including loss of data in subsequent
7497 header lines, and spool format errors.
7499 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7500 client host had authenticated. This control can now be exercised by an ACL
7501 for more flexibility.
7503 69. By default, callouts do not happen when testing with -bh. There is now a
7504 variant, -bhc, which does actually run the callout code, including
7505 consulting and updating the callout cache.
7507 70. Added support for saslauthd authentication, courtesy of Alexander
7510 71. If statvfs() failed on the spool or log directories while checking their
7511 size for availability, Exim confusingly gave the error "space shortage".
7512 Furthermore, in debugging mode it crashed with a floating point exception.
7513 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7514 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7515 serious problem, Exim now writes to the main and panic logs when this
7516 happens, with details of the failure. It then refuses to accept the
7517 incoming message, giving the message "spool directory problem" or "log
7518 directory problem" with a 421 code for SMTP messages.
7520 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7521 0, 1, and 2 exist, because some OS complain for execs without them (see
7522 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7523 descriptors. However, the code omitted to check that the open succeeded,
7524 causing mysterious errors if for some reason the permissions on /dev/null
7525 got screwed. Now Exim writes a message to the main and panic logs, and
7526 bombs out if it can't open /dev/null.
7528 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7529 interact so that it is all more flexible. It is supposed to remain
7530 backwards compatible. Also added extra_local_interfaces.
7532 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7533 to bomb out with an assertion failure - to the client this appears as a
7534 connection drop. This problem occurs in the part of the code that was taken
7535 from the Samba project. Fortunately, the assertion is in a very simple
7536 function, so I have fixed this by reproducing the function inline in the
7537 one place where it is called, and arranging for authentication to fail
7538 instead of killing the process with assert().
7540 75. The SPA client code was not working when the server requested OEM rather
7541 than Unicode encoding.
7543 76. Added code to make require_files with a specific uid setting more usable in
7544 the case where statting the file as root fails - usually a non-root-mounted
7545 NFS file system. When this happens and the failure is EACCES, Exim now
7546 forks a subprocess and does the per-uid checking as the relevant uid.
7548 77. Added process_log_path.
7550 78. If log_file_path was not explicitly set, a setting of check_log_space or
7551 check_log_inodes was ignored.
7553 79. If a space check for the spool or log partitions fails, the incident is now
7554 logged. Of course, in the latter case the data may get lost...
7556 80. Added the %p formatting code to string_format() so that it can be used to
7557 print addresses in debug_print(). Adjusted all the address printing in the
7558 debugging in store.c to use %p rather than %d.
7560 81. There was a concern that a line of code in smtp_in.c could overflow a
7561 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7562 initially expressed, the concern was not well-founded, because trailing
7563 spaces are removed early. However, if the trailing spaces were followed by
7564 a NULL, they did not get removed, so the overflow was possible. Two fixes
7567 (a) I re-wrote the offending code in a cleaner fashion.
7568 (b) If an incoming SMTP command contains a NULL character, it is rejected
7571 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7572 runs initgroups(), so that if the Exim user is in any additional groups,
7573 they will be used during message reception.
7579 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob