1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Avoid costly startup code when not strictly needed. This reduces time
12 for some exim process initialisations. It does mean that the logging
13 of TLS configuration problems is only done for the daemon startup.
15 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
17 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
18 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
25 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
26 side, to decode a TLS 1.3 packet capture.
28 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
29 Previously the default library behaviour applied, sending two, each in
32 JH/03 Debug output for ACL now gives the config file name and line number for
35 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
37 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
39 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
40 buffer overrun for (non-chunking) other transports.
42 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
43 TLS1.3, means that a server rejecting a client certificate is not visible
44 to the client until the first read of encrypted data (typically the
45 response to EHLO). Add detection for that case and treat it as a failed
46 TLS connection attempt, so that the normal retry-in-clear can work (if
49 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
50 and/or domain. Found and fixed by Jason Betts.
52 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
53 configuration). If a CNAME target was not a wellformed name pattern, a
56 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
57 the OS reports them interleaved with other addresses.
59 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
60 used both for input and for a verify callout, both encrypted, SMTP
61 responses being sent by the server could be lost. This resulted in
62 dropped connections and sometimes bounces generated by a peer sending
65 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
66 string. Previously it was possible to cause undefined behaviour in a
67 library routine (usually a crash). Found by "zerons".
69 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
72 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
73 API was removed, so update to use the newer ones.
75 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
76 any timeout set, is taking a long time. Previously we would hang on to a
77 rotated logfile "forever" if the input was arriving with long gaps
78 (a previous attempt to fix addressed lack, for a long time, of initial
81 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
82 shared (NFS) environment. The length of the tempfile name is now
83 4 + 16 ("hdr.$message_exim_id") which might break on file
84 systems which restrict the file name length to lower values.
87 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
88 shared (NFS) environment.
90 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
91 did for all versions <4.90). Notably -M, -m, --invert, -I may be
94 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
95 on some platforms for bit 31.
97 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
98 to changes apparently associated with TLS1.3 handling some of the APIs
99 previously used were either nonfunctional or inappropriate. Strings
100 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
101 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
102 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
103 This affects log line X= elements, the $tls_{in,out}_cipher variables,
104 and the use of specific cipher names in the encrypted= ACL condition.
106 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
108 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
109 verification result was not updated unless hosts_require_ocsp applied.
111 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
112 queue_list_requires_admin set to false, non-admin users were denied the
115 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
116 directory-of-certs mode. Previously they were advertised despite the
119 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
120 A single TCP connection by a client will now hold a TLS connection open
121 for multiple message deliveries, by default. Previoud the default was to
124 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
125 default. If built with the facility, DANE will be used. The facility
126 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
128 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
129 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
130 must be defined and you must still, unless you define DISABLE_TLS, manage
131 the the include-dir and library-file requirements that go with that
132 choice. Non-TLS builds are still supported.
134 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
137 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
138 default. If the platform supports and has the facility enabled, it will
139 be requested on all coneections.
141 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
142 controlled by the build-time option SUPPORT_PIPE_CONNECT.
144 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
146 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
147 success-DSN messages. Previously the From: header was always the default
148 one for these; the option was ignored.
150 JH/28 Fix the timeout on smtp response to apply to the whole response.
151 Previously it was reset for every read, so a teergrubing peer sending
152 single bytes within the time limit could extend the connection for a
153 long time. Credit to Qualsys Security Advisory Team for the discovery.
155 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
156 delivery address, which leaked information of the results of local
157 forwarding. Change to the original envelope recipient address, per
160 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
161 requested. Previously not bounce was generated and a log entry of
162 error ignored was made.
164 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
166 JH/32 Introduce a general tainting mechanism for values read from the input
167 channel, and values derived from them. Refuse to expand any tainted
168 values, to catch one form of exploit.
170 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
171 was unused and the unexpanded text used for the test. Found and
172 fixed by Ruben Jenster.
174 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
175 an attempt to use a TLS library read routine dereffed a nul pointer,
178 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
179 them in our smtp response.
181 JH/36 Have the general router option retry_use_local_part default to true when
182 any of the restrictive preconditions are set (to anything). Previously it
183 was only for check_local user. The change removes one item of manual
184 configuration which is required for proper retries when a remote router
185 handles a subset of addresses for a domain.
187 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
188 link count into consideration.
190 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
191 caused the extension of big_buffer, the following lines were ignored.
193 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
194 accordance with RFC 2308. Previously there was no expiry, so a longlived
195 receive process (eg. due to ACL delays) versus a short SOA value could
198 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
200 JH/39 Promote DMARC support to mainline.
202 JH/40 Bug 2452: Add a References: header to DSNs.
204 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
205 parameters. The relevant library call is documented as "Deprecated: This
206 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
207 3.6.0, DH parameters are negotiated following RFC7919."
213 JH/01 Remove code calling the customisable local_scan function, unless a new
214 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
216 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
217 non-signal-safe functions being used.
219 JH/03 Bug 2269: When presented with a received message having a stupidly large
220 number of DKIM-Signature headers, disable DKIM verification to avoid
221 a resource-consumption attack. The limit is set at twenty.
223 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
224 report of oldest_pass in ${authres } in consequence, and separate out
225 some descriptions of reasons for verification fail.
227 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
228 files in the spool were present and unlocked. A queue-runner could spot
229 them, resulting in a duplicate delivery. Fix that by doing the unlock
230 after the unlink. Investigation by Tim Stewart. Take the opportunity to
231 add more error-checking on spoolfile handling while that code is being
234 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
235 No known attacks, no CVE, this is defensive hardening.
237 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
238 a queue-runner could start a delivery while other operations were ongoing.
239 Cutthrough delivery was a common victim, resulting in duplicate delivery.
240 Found and investigated by Tim Stewart. Fix by using the open message data
241 file handle rather than opening another, and not locally closing it (which
242 releases a lock) for that case, while creating the temporary .eml format
243 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
245 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
246 $sender_verify_failure/$recipient_verify_failure to "random".
248 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
251 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
252 Previously this would segfault.
254 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
257 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
258 like zero, since the resolver should be doing this for us, But we need one
259 as a CNAME but no MX presence gets the CNAME returned; we need to check
260 that doesn't point to an MX to declare it "no MX returned" rather than
261 "error, loop". A new main option is added so the older capability of
262 following some limited number of chain links is maintained.
264 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
266 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
267 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
268 now-deprecated earlier definitions used only specified the range up to TLS
269 1.2 (in the older-version library docs).
271 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
273 JH/15 Rework TLS client-side context management. Stop using a global, and
274 explicitly pass a context around. This enables future use of TLS for
275 connections to service-daemons (eg. malware scanning) while a client smtp
276 connection is using TLS; with cutthrough connections this is quite likely.
278 JH/16 Fix ARC verification to do AS checks in reverse order.
280 JH/17 Support a "tls" option on the ${readsocket } expansion item.
282 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
283 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
284 Previously the "utf8" would be re-prepended for every additional message.
286 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
287 Previously thery were accepted, resulting in issues when attempting to
288 forward messages to a non-supporting MTA.
290 PP/02 Let -n work with printing macros too, not just options.
292 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
293 one parent address was copied, and bogus data was used at delivery-logging
294 time. Either a crash (after delivery) or bogus log data could result.
295 Discovery and analysis by Tim Stewart.
297 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
298 Previously if the string ended mid-character, we did not insert the
299 promised '?' replacement.
301 PP/04 Documentation: current string operators work on bytes, not codepoints.
303 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
304 should pack well giving a smaller memory footprint so better caching and
305 therefore performance. Group the declarations where this can't be done so
306 that the byte-sized flag variables are not interspersed among pointer
307 variables, giving a better chance of good packing by the compiler.
309 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
310 non-null, to avoid issues with sites running BATV. Previously reports were
311 sent with an empty envelope sender so looked like bounces.
313 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
314 The ignore_error flag wasn't being returned from the filter subprocess so
315 was not set for later routers. Investigation and fix by Matthias Kurz.
317 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
318 and a msg:complete for the whole, when a message is manually removed using
319 -Mrm. Developement by Matthias Kurz, hacked on by JH.
321 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
322 a "Gnu special" function, asprintf() in the DB utility binary builds; I
323 hope that is portable enough.
325 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
326 requiring a known-CA anchor certificate; make it now rely entirely on the
327 TLSA as an anchor. Checking the name on the leaf cert against the name
328 on the A-record for the host is still done for TA (but not for EE mode).
330 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
331 included in delivery lines for non-proxied connections, when compiled with
332 SUPPORT_SOCKS and running with proxy logging enabled.
334 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
335 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
336 move the existing event to fire before the normal logging of message
337 failure so that custom logging is bracketed by normal logging.
339 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
340 msg:fail:internal event. Developement by Matthias Kurz.
342 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
343 far too small for todays use of crypto signatures stored there. Go all
344 the way to the max DNS message size of 64kB, even though this might be
345 overmuch for IOT constrained device use.
347 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
348 copy a string over itself. The library routine is documented as not
349 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
351 JH/32 For main options check_spool_space and check_inode_space, where the
352 platform supports 64b integers, support more than the previous 2^31 kB
353 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
354 the previous G, M, k.
356 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
357 $authenticated_fail_id variable on authentication failure. Previously
360 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
361 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
362 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
363 GNUTLS_SEC_PARAM_MEDIUM.
365 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
366 side. Previously we would continue as if no SNI had been received.
368 JH/36 Harden the handling of string-lists. When a list consisted of a sole
369 "<" character, which should be a list-separator specification, we walked
370 off past the nul-terimation.
372 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
373 causes) even when the retry time is not yet met. Previously they were
374 not, meaning that when (say) an account was over-quota and temp-rejecting,
375 and multiple senders' messages were queued, only one sender would get
376 notified on each configured delay_warning cycle.
378 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
380 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
381 cipher-suites, an error can be left on the stack even for a succeeding
382 accept; this results in impossible error messages when a later operation
385 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
386 return error codes indicating retry. Under TLS1.3 this becomes required.
388 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
389 it only wrote the new authenticators, resulting in a lack of tracking of
390 peer changes of ESMTP extensions until the next cache flush.
392 JH/41 Fix the loop reading a message header line to check for integer overflow,
393 and more-often against header_maxsize. Previously a crafted message could
394 induce a crash of the recive process; now the message is cleanly rejected.
396 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
397 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
403 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
404 When redis_servers is set to a list of > 1 element, and the Redis servers
405 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
406 case of MOVED into a DEFER case instead, thus moving the query onto the
407 next server in the list. For a cluster of N elements, all N servers must
408 be defined in redis_servers.
410 GF/02 Catch and remove uninitialized value warning in exiqsumm
411 Check for existence of @ARGV before looking at $ARGV[0]
413 JH/01 Replace the store_release() internal interface with store_newblock(),
414 which internalises the check required to safely use the old one, plus
415 the allocate and data copy operations duplicated in both (!) of the
416 extant use locations.
418 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
419 modifier. This matches the restriction on the commandline.
421 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
422 Previously only the last row was returned.
424 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
425 we assumed that tags in the header were well-formed, and parsed the
426 element content after inspecting only the first char of the tag.
427 Assumptions at that stage could crash the receive process on malformed
430 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
431 While running the DKIM ACL we operate on the Permanent memory pool so that
432 variables created with "set" persist to the DATA ACL. Also (at any time)
433 DNS lookups that fail create cache records using the Permanent pool. But
434 expansions release any allocations made on the current pool - so a dnsdb
435 lookup expansion done in the DKIM ACL releases the memory used for the
436 DNS negative-cache, and bad things result. Solution is to switch to the
437 Main pool for expansions.
438 While we're in that code, add checks on the DNS cache during store_reset,
439 active in the testsuite.
440 Problem spotted, and debugging aided, by Wolfgang Breyha.
442 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
443 When none of the hosts presented to a transport match an already-open
444 connection, close it and proceed with the list. Previously we would
445 queue the message. Spotted by Lena with Yahoo, probably involving
448 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
449 Previously a spurious "250 OK id=" response was appended to the proper
452 JH/08 The "support for" informational output now, which built with Content
453 Scanning support, has a line for the malware scanner interfaces compiled
454 in. Interface can be individually included or not at build time.
456 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
457 by the template makefile "src/EDITME". The "STREAM" support for an older
458 ClamAV interface method is removed.
460 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
461 rows affected is given instead).
463 JH/11 The runtime Berkeley DB library version is now additionally output by
464 "exim -d -bV". Previously only the compile-time version was shown.
466 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
467 SMTP connection. Previously, when one had more receipients than the
468 first, an abortive onward connection was made. Move to full support for
469 multiple onward connections in sequence, handling cutthrough connection
470 for all multi-message initiating connections.
472 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
473 routers. Previously, a multi-recipient message would fail to match the
474 onward-connection opened for the first recipient, and cause its closure.
476 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
477 a timeout on read on a GnuTLS initiating connection, resulting in the
478 initiating connection being dropped. This mattered most when the callout
479 was marked defer_ok. Fix to keep the two timeout-detection methods
482 JH/15 Relax results from ACL control request to enable cutthrough, in
483 unsupported situations, from error to silently (except under debug)
484 ignoring. This covers use with PRDR, frozen messages, queue-only and
487 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
489 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
490 metadata, resulting in a crash in free().
492 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
493 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
494 Broken also in d185889f4, with init system revamp.
496 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
497 Previously we abruptly closed the connection after reading a malware-
498 found indication; now we go on to read the "scan ok" response line,
501 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
502 ACL. Previously, a crash would result.
504 JH/19 Speed up macro lookups during configuration file read, by skipping non-
505 macro text after a replacement (previously it was only once per line) and
506 by skipping builtin macros when searching for an uppercase lead character.
508 JH/20 DANE support moved from Experimental to mainline. The Makefile control
509 for the build is renamed.
511 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
512 was allocated for every new TLS startup, meaning one per message. Fix
513 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
515 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
516 reported the original. Fix to report (as far as possible) the ACL
517 result replacing the original.
519 JH/23 Fix memory leak during multi-message connections using STARTTLS under
520 OpenSSL. Certificate information is loaded for every new TLS startup,
521 and the resources needed to be freed.
523 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
525 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
528 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
529 DATA response info to the (existing) per-recipient response info for
530 the "C=" log element. It can have useful tracking info from the
531 destination system. Patch from Simon Arlott.
533 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
534 length value. Previously this would segfault.
536 HS/02 Support Avast multiline protoocol, this allows passing flags to
537 newer versions of the scanner.
539 JH/28 Ensure that variables possibly set during message acceptance are marked
540 dead before release of memory in the daemon loop. This stops complaints
541 about them when the debug_store option is enabled. Discovered specifically
542 for sender_rate_period, but applies to a whole set of variables.
543 Do the same for the queue-runner and queue-list loops, for variables set
544 from spool message files. Do the same for the SMTP per-message loop, for
545 certain variables indirectly set in ACL operations.
547 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
548 as a multi-recipient message from a mailinglist manager). The coding had
549 an arbitrary cutoff number of characters while checking for more input;
550 enforced by writing a NUL into the buffer. This corrupted long / fast
551 input. The problem was exposed more widely when more pipelineing of SMTP
552 responses was introduced, and one Exim system was feeding another.
553 The symptom is log complaints of SMTP syntax error (NUL chars) on the
554 receiving system, and refused recipients seen by the sending system
555 (propating to people being dropped from mailing lists).
556 Discovered and pinpointed by David Carter.
558 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
559 replaced by the ${authresults } expansion.
561 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
563 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
564 allows proper process termination in container environments.
566 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
567 Previously the "final dot" had a newline after it; ensure it is CR,LF.
569 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
570 and "err_perm", deprecated since 4.83 when the RFC-defined words
571 "temperror" and "permerror" were introduced.
573 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
574 transport-filters or DKIM-signing. The restriction was lost in the
575 consolidation of verify-callout and delivery SMTP handling.
576 Extend the restriction to also cover ARC-signing.
578 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
579 in defer=pass mode supply a 450 to the initiator. Previously the message
582 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
583 tls_require_ciphers is used as before.
585 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
586 "pass_unscanned". Only tmpfails from the scanner are written to
587 the paniclog, as they may require admin intervention (permission
588 denied, license issues). Other scanner errors (like decompression
589 bombs) do not cause a paniclog entry.
591 JH/36 Fix reinitialisation of DKIM logging variable between messages.
592 Previously it was possible to log spurious information in receive log
595 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
596 triggered odd behaviour from Outlook Express clients.
598 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
601 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
602 since the IETF WG has not yet settled on that versus the original
603 "bare" representation.
605 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
606 Previously the millisecond value corrupted the output.
607 Fix also for syslog_pid=no and log_selector +pid, for which the pid
608 corrupted the output.
614 JH/01 Rework error string handling in TLS interface so that the caller in
615 more cases is responsible for logging. This permits library-sourced
616 string to be attached to addresses during delivery, and collapses
617 pairs of long lines into single ones.
619 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
620 during configuration. Wildcards are allowed and expanded.
622 JH/02 Rework error string handling in DKIM to pass more info back to callers.
623 This permits better logging.
625 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
626 do not close it down and have the child transport start it up again on
627 the passed-on TCP connection. Instead, proxy the child (and any
628 subsequent ones) for TLS via a unix-domain socket channel. Logging is
629 affected: the continued delivery log lines do not have any DNSSEC, TLS
630 Certificate or OCSP information. TLS cipher information is still logged.
632 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
633 identical IP addresses on different listening ports. Will also affect
636 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
637 add noisy ifdef guards to special-case this sillyness.
638 Patch from Bernd Kuhls.
640 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
641 than 255 are no longer allowed.
643 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
644 Disable the session-cache too, which might reduce our load. Since we
645 currrectly use a new context for every connection, both as server and
646 client, there is no benefit for these.
647 GnuTLS appears to not support tickets server-side by default (we don't
648 call gnutls_session_ticket_enable_server()) but client side is enabled
649 by default on recent versions (3.1.3 +) unless the PFS priority string
652 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
653 <https://reproducible-builds.org/specs/source-date-epoch/>.
655 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
656 the check for any unsuccessful recipients did not notice the limit, and
657 erroneously found still-pending ones.
659 JH/08 Pipeline CHUNKING command and data together, on kernels that support
660 MSG_MORE. Only in-clear (not on TLS connections).
662 JH/09 Avoid using a temporary file during transport using dkim. Unless a
663 transport-filter is involved we can buffer the headers in memory for
664 creating the signature, and read the spool data file once for the
665 signature and again for transmission.
667 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
668 4.77 as the kernel support then wasn't solid, having issues in 64bit
669 mode. Now, it's been long enough. Add support for FreeBSD also.
671 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
672 case where the routing stage had gathered several addresses to send to
673 a host before calling the transport for the first, we previously failed
674 to close down TLS in the old transport process before passing the TCP
675 connection to the new process. The new one sent a STARTTLS command
676 which naturally failed, giving a failed delivery and bloating the retry
677 database. Investigation and fix prototype from Wolfgang Breyha.
679 JH/12 Fix check on SMTP command input synchronisation. Previously there were
680 false-negatives in the check that the sender had not preempted a response
681 or prompt from Exim (running as a server), due to that code's lack of
682 awareness of the SMTP input buffering.
684 PP/04 Add commandline_checks_require_admin option.
685 Exim drops privileges sanely, various checks such as -be aren't a
686 security problem, as long as you trust local users with access to their
687 own account. When invoked by services which pass untrusted data to
688 Exim, this might be an issue. Set this option in main configuration
689 AND make fixes to the calling application, such as using `--` to stop
692 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
693 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
694 responses to those, into a single TLS record each way (this usually means
695 a single packet). As a side issue, smtp_enforce_sync now works on TLS
698 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
699 affects you only if you're dancing at the edge of the param size limits.
700 If you are, and this message makes sense to you, then: raise the
701 configured limit or use OpenSSL 1.1. Nothing we can do for older
704 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
705 cmdline element to get the documented default one. Previously it was
708 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
709 are unsuable for non-message-specific reasons - eg. connection timeout,
710 banner-time rejection.
712 JH/16 Fix logging of delivery remote port, when specified by router, under
715 PP/06 Repair manualroute's ability to take options in any order, even if one
716 is the name of a transport.
719 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
721 JH/17 Change the list-building routines interface to use the expanding-string
722 triplet model, for better allocation and copying behaviour.
724 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
725 Previously it was constructed the first time a possibly-matching string
726 was met in the configuration file input during startup; now it is done
729 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
730 compatible one, to avoid the (poorly documented) possibility of a config
731 file in the working directory redirecting the DB files, possibly correpting
732 some existing file. CVE-2017-10140 assigned for BDB.
734 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
735 cache-hot. Previously, although the result was properly cached, the
736 initial verify call returned a defer.
738 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
739 the main verify for receipient in uncached-mode.
741 JH/22 Retire historical build files to an "unsupported" subdir. These are
742 defined as "ones for which we have no current evidence of testing".
744 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
745 if present. Previously it was ignored.
747 JH/24 Start using specified-initialisers in C structure init coding. This is
748 a C99 feature (it's 2017, so now considered safe).
750 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
751 if was a fixed-sized field and bitmask ops via macros; it is now more
754 PP/07 GitHub PR 56: Apply MariaDB build fix.
755 Patch provided by Jaroslav Škarvada.
757 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
758 during Coverity cleanups [4.87 JH/47]
759 Diagnosis and fix provided by Michael Fischer v. Mollard.
761 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
762 the right size to place the terminating semicolon on its own folded
763 line, the header hash was calculated to an incorrect value thanks to
764 the (relaxed) space the fold became.
766 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
767 and confused the parent.
769 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
770 which could crash as a result. This could lead to undeliverable messages.
772 JH/28 Logging: "next input sent too soon" now shows where input was truncated
775 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
776 matters on fast-turnover and PID-randomising systems, which were getting
777 out-of-order delivery.
779 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
780 a possibly-overlapping copy. The symptom was that "Remote host closed
781 connection in response to HELO" was logged instead of the actual 4xx
784 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
785 Previously only that bufferd was discarded, resulting in SYMTP command
788 JH/32 DKIM: when a message has multiple signatures matching an identity given
789 in dkim_verify_signers, run the dkim acl once for each. Previously only
790 one run was done. Bug 2189.
792 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
793 "panic the current process" to "deliberately defer". The panic log is
794 still written with the problem list name; the mail and reject logs now
795 get a temp-reject line for the message that was being handled, saying
796 something like "domains check lookup or other defer". The SMTP 451
797 message is still "Temporary local problem".
799 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
800 A crafted sequence of BDAT commands could result in in-use memory beeing
801 freed. CVE-2017-16943.
803 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
804 from SMTP input. Previously it was always done; now only done for DATA
805 and not BDAT commands. CVE-2017-16944.
807 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
808 to the message (such as an overlong header line). Previously this was
809 not done and we did not exit BDAT mode. Followon from the previous item
810 though a different problem.
816 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
817 than -2003 did; needs libidn2 in addition to libidn.
819 JH/02 The path option on a pipe transport is now expanded before use.
821 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
822 Patch provided by "Björn", documentation fix added too.
824 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
825 missing a wire-to-host endian conversion.
827 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
828 close after a BDAT command line could be taken as a following command,
829 giving a synch failure. Fix by only checking for synch immediately
830 before acknowledging the chunk.
832 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
833 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
834 macro. Patches provided by Josh Soref.
836 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
837 Previously we did not; the RFC seems ambiguous and VRFY is not listed
838 by IANA as a service extension. However, John Klensin suggests that we
841 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
842 the dkim code may be unix-mode line endings rather than smtp wire-format
843 CRLF, so prepend a CR to any bare LF.
845 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
846 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
848 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
849 the first were themselves being wrongly included in the feed into dkim
850 processing; with most chunk sizes in use this resulted in an incorrect
851 body hash calculated value.
853 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
854 DKIM signature block, for verification. Although advised against by
855 standards it is specifically not ruled illegal.
857 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
859 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
860 missing a body hash (the bh= tag).
862 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
863 It seems that HAProxy sends the Proxy Protocol information in clear and
864 only then does a TLS startup, so do the same.
866 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
867 TCP connections (such as for Spamd) unless the daemon successfully set
868 Fast Open mode on its listening sockets. This fixes breakage seen on
869 too-old kernels or those not configured for Fast Open, at the cost of
870 requiring both directions being enabled for TFO, and TFO never being used
871 by non-daemon-related Exim processes.
873 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
874 endings, at least on the first header line. Try to canonify any that get
875 past that check, despite the cost.
877 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
878 now limited to an arbitrary five deep, while parsing addresses with the
879 strip_excess_angle_brackets option enabled.
881 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
882 instead leave the unprompted TLS handshake in socket buffer for the
883 TLS library to consume.
885 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
887 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
889 JH/16 Drop variables when they go out of scope. Memory management drops a whole
890 region in one operation, for speed, and this leaves assigned pointers
891 dangling. Add checks run only under the testsuite which checks all
892 variables at a store-reset and panics on a dangling pointer; add code
893 explicitly nulling out all the variables discovered. Fixes one known
894 bug: a transport crash, where a dangling pointer for $sending_ip_address
895 originally assigned in a verify callout, is re-used.
897 PP/06 Drop '.' from @INC in various Perl scripts.
899 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
901 PP/08 Reduce a number of compilation warnings under clang; building with
902 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
903 should be warning-free.
905 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
907 HS/01 Fix portability problems introduced by PP/08 for platforms where
908 realloc(NULL) is not equivalent to malloc() [SunOS et al].
910 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
911 chunk. This allows us to accept broken chunked messages. We need a more
912 general solution here.
914 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
915 already-broken messages in the queue.
917 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
919 JH/19 Fix reference counting bug in routing-generated-address tracking.
925 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
926 supports it and a size is available (ie. the sending peer gave us one).
928 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
929 years of being deprecated). The replacements are the ACLs
930 acl_smtp_mime and acl_not_smtp_mime.
932 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
933 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
934 or even an in-clear connection were permitted. Now, if the host lookup
935 was dnssec and dane was requested then the host is only used if the
936 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
937 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
938 if one fails this test.
939 This means that a poorly-configured remote DNS will make it incommunicado;
940 but it protects against a DNS-interception attack on it.
942 JH/04 Bug 1810: make continued-use of an open smtp transport connection
943 non-noisy when a race steals the message being considered.
945 JH/05 If main configuration option tls_certificate is unset, generate a
946 self-signed certificate for inbound TLS connections.
948 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
949 in rewrites and routers.
951 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
952 and logged a warning sing 4.83; now they are a configuration file error.
954 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
955 (lacking @domain). Apply the same qualification processing as RCPT.
957 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
959 JH/10 Support ${sha256:} applied to a string (as well as the previous
962 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
963 a cutthrough deliver is pending, as we always want to make a connection.
964 This also avoids re-routing the message when later placing the cutthrough
965 connection after a verify cache hit.
966 Do not update it with the verify result either.
968 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
969 when routing results in more than one destination address.
971 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
972 signing (which inhibits the cutthrough capability). Previously only
973 the presence of an option was tested; now an expansion evaluating as
974 empty is permissible (obviously it should depend only on data available
975 when the cutthrough connection is made).
977 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
978 the relevant preceding SMTP command did not note the pipelining mode.
980 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
981 Previously they were not counted.
983 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
984 as one having no matching records. Previously we deferred the message
985 that needed the lookup.
987 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
988 distinguished as "(=".
990 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
991 for missing MX records. Previously it only worked for missing A records.
993 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
995 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
996 after the data-go-ahead and data-ack. Patch from Jason Betts.
998 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
999 even for a "none" policy. Patch from Tony Meyer.
1001 JH/22 Fix continued use of a connection for further deliveries. If a port was
1002 specified by a router, it must also match for the delivery to be
1005 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1006 When one of the recipients of a message was unsuitable for the connection
1007 (has no matching addresses), we lost track of needing to mark it
1008 deferred. As a result mail would be lost.
1010 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1012 JH/25 Decoding ACL controls is now done using a binary search; the source code
1013 takes up less space and should be simpler to maintain. Merge the ACL
1014 condition decode tables also, with similar effect.
1016 JH/26 Fix problem with one_time used on a redirect router which returned the
1017 parent address unchanged. A retry would see the parent address marked as
1018 delivered, so not attempt the (identical) child. As a result mail would
1021 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1022 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1023 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1026 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1027 Main config options check_{log,spool}_{inodes,space} are now
1028 100 inodes, 10MB unless set otherwise in the configuration.
1030 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1031 Previously it only applied to the main-section connection policy
1034 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1036 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1037 by me. Added RFC7919 DH primes as an alternative.
1039 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1040 are not in the system include path.
1042 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1043 GnuTLS, when a session startup failed (eg because the client disconnected)
1044 Exim did stdio operations after fclose. This was exposed by a recent
1045 change which nulled out the file handle after the fclose.
1047 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1048 signed directly by the cert-signing cert, rather than an intermediate
1049 OCSP-signing cert. This is the model used by LetsEncrypt.
1051 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1053 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1054 an incoming connection.
1056 HS/02 Bug 1802: Do not half-close the connection after sending a request
1059 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1060 fallback to "prime256v1".
1062 JH/34 SECURITY: Use proper copy of DATA command in error message.
1063 Could leak key material. Remotely exploitable. CVE-2016-9963.
1069 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1070 and 3.4.4 - once the server is enabled to respond to an OCSP request
1071 it does even when not requested, resulting in a stapling non-aware
1072 client dropping the TLS connection.
1074 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1075 support variable-length bit vectors. No functional change.
1077 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1078 The I= interface field on outgoing lines is now after the H= remote
1079 host field, same as incoming lines. There is a separate
1080 outgoing_interface log selector which allows you to disable the
1083 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1084 If not running log_selector +smtp_connection the mainlog would be held
1085 open indefinitely after a "too many connections" event, including to a
1086 deleted file after a log rotate. Leave the per net connection logging
1087 leaving it open for efficiency as that will be quickly detected by the
1088 check on the next write.
1090 HS/01 Bug 1671: Fix post transport crash.
1091 Processing the wait-<transport> messages could crash the delivery
1092 process if the message IDs didn't exist for some reason. When
1093 using 'split_spool_directory=yes' the construction of the spool
1094 file name failed already, exposing the same netto behaviour.
1096 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1097 mime_regex ACL conditions.
1099 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1100 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1101 to HELO, local diagnostic string.
1103 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1104 log line to debug. Even when configured with a tls authenticator many
1105 client connections are expected to not authenticate in this way, so
1106 an authenticate fail is not an error.
1108 HS/02 Add the Exim version string to the process info. This way exiwhat
1109 gives some more detail about the running daemon.
1111 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1112 matter for fast-change records such as DNSBLs.
1114 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1115 retry record, even if constant. There may be multiple transports with
1116 different interface settings and the retry behaviour needs to be kept
1119 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1121 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1123 JH/10 Bug 840: fix log_defer_output option of pipe transport
1125 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1126 hosts_randomize. This matters a lot when combined with mua_wrapper.
1128 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1129 ${quote_pgsql:<string>} operator.
1131 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1133 JH/14 Change header file name for freeradius-client. Relevant if compiling
1134 with Radius support; from the Gentoo tree and checked under Fedora.
1136 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1138 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1139 option transport_filter, meaning no filtering.
1141 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1143 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1144 defaults to "*" (all hosts). The variable is now available when not built
1145 with TLS, default unset, mainly to enable keeping the testsuite sane.
1146 If a server certificate is not supplied (via tls_certificate) an error is
1147 logged, and clients will find TLS connections fail on startup. Presumably
1148 they will retry in-clear.
1149 Packagers of Exim are strongly encouraged to create a server certificate
1150 at installation time.
1152 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1153 with the $config_file variable.
1155 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1156 in transport context, after the attempt, and per-recipient. The latter type
1157 is per host attempted. The event data is the error message, and the errno
1158 information encodes the lookup type (A vs. MX) used for the (first) host,
1159 and the trailing two digits of the smtp 4xx response.
1161 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1162 to write to mainlog (or rejectlog, paniclog) in the window between file
1163 creation and permissions/ownership being changed. Particularly affects
1164 installations where exicyclog is run as root, rather than exim user;
1165 result is that the running daemon panics and dies.
1167 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1169 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1170 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1171 "pri" and "weight". Note that the previous implicit priority given by the
1172 list order is no longer honoured.
1174 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1175 for DKIM processing.
1177 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1178 by defining SUPPORT_SOCKS.
1180 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1181 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1182 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1183 variables are renamed to proxy_{local,external}_{address,port}.
1185 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1186 for a build by defining SUPPORT_I18N
1188 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1189 of the query string, and make ${quote_redis:} do that quoting.
1191 JH/29 Move Events support from Experimental to mainline, enabled by default
1192 and removable for a build by defining DISABLE_EVENT.
1194 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1196 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1197 cached by the daemon.
1199 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1200 by defining LOOKUP_REDIS. The libhiredis library is required.
1202 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1203 keys are given for lookup.
1205 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1206 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1207 only supported when built with TLS support. The PolarSSL SHA routines
1208 are still used when the TLS library is too old for convenient support.
1210 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1211 openssl_options), for security. OpenSSL forces this from version 1.1.0
1212 server-side so match that on older versions.
1214 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1215 allocation for $value could be released as the expansion processing
1216 concluded, but leaving the global pointer active for it.
1218 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1219 and to use the domains and local_parts ACL conditions.
1221 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1222 incorrectly not doubled on cutthrough transmission, hence seen as a
1223 body-termination at the receiving system - resulting in truncated mails.
1224 Commonly the sender saw a TCP-level error, and retransmitted the message
1225 via the normal store-and-forward channel. This could result in duplicates
1226 received - but deduplicating mailstores were liable to retain only the
1227 initial truncated version.
1229 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1231 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1233 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1234 we're in there, support oversigning also; bug 1309.
1236 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1238 HS/04 Add support for keep_environment and add_environment options.
1240 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1241 either intentional arithmetic overflow during PRNG, or testing config-
1244 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1245 delivery resulted in actual delivery. Cancel cutthrough before DATA
1248 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1249 rejects a recipient: pass the reject to the originator.
1251 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1252 Many were false-positives and ignorable, but it's worth fixing the
1255 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1256 for the new environment-manipulation done at startup. Move the routines
1257 from being local to tls.c to being global via the os.c file.
1259 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1260 an extract embedded as result-arg for a map, the first arg for extract
1261 is unavailable so we cannot tell if this is a numbered or keyed
1262 extraction. Accept either.
1268 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1271 JH/02 The smtp transport option "multi_domain" is now expanded.
1273 JH/03 The smtp transport now requests PRDR by default, if the server offers
1276 JH/04 Certificate name checking on server certificates, when exim is a client,
1277 is now done by default. The transport option tls_verify_cert_hostnames
1278 can be used to disable this per-host. The build option
1279 EXPERIMENTAL_CERTNAMES is withdrawn.
1281 JH/05 The value of the tls_verify_certificates smtp transport and main options
1282 default to the word "system" to access the system default CA bundle.
1283 For GnuTLS, only version 3.0.20 or later.
1285 JH/06 Verification of the server certificate for a TLS connection is now tried
1286 (but not required) by default. The verification status is now logged by
1287 default, for both outbound TLS and client-certificate supplying inbound
1290 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1293 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1294 Status Notification (bounce) messages are now MIME format per RFC 3464.
1295 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1296 under the control of the dsn_advertise_hosts option, and routers may
1297 have a dsn_lasthop option.
1299 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1300 default, modifiable by a malware= option. The list separator for
1301 the options can now be changed in the usual way. Bug 68.
1303 JH/10 The smtp_receive_timeout main option is now expanded before use.
1305 JH/11 The incoming_interface log option now also enables logging of the
1306 local interface on delivery outgoing connections.
1308 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1309 if the interface and destination host and port all match.
1311 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1314 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1315 Patch from Andrew Lewis.
1317 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1318 now supports optional time-restrictions, weighting, and priority
1319 modifiers per server. Patch originally by <rommer@active.by>.
1321 JH/16 The spamd_address main option now supports a mixed list of local
1322 and remote servers. Remote servers can be IPv6 addresses, and
1323 specify a port-range.
1325 JH/17 Bug 68: The spamd_address main option now supports an optional
1326 timeout value per server.
1328 JH/18 Bug 1581: Router and transport options headers_add/remove can
1329 now have the list separator specified.
1331 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1334 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1337 JH/21 Support for the A6 type of dns record is withdrawn.
1339 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1340 rather than the verbs used.
1342 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1343 from 255 to 1024 chars.
1345 JH/24 Verification callouts now attempt to use TLS by default.
1347 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1348 are generic router options now. The defaults didn't change.
1350 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1351 Original patch from Alexander Shikoff, worked over by JH.
1353 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1354 files if ZCAT_COMMAND is not executable.
1356 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1358 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1360 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1361 Normally benign, it bites when the pair was led to by a CNAME;
1362 modern usage is to not canonicalize the domain to a CNAME target
1363 (and we were inconsistent anyway for A-only vs AAAA+A).
1365 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1367 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1368 when evaluating $sender_host_dnssec.
1370 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1371 $sender_helo_dnssec variable.
1373 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1375 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1377 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1379 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1380 documented as working, but never had. Support all but $spam_report.
1382 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1383 added for tls authenticator.
1385 HS/03 Add perl_taintmode main config option
1391 TL/01 When running the test suite, the README says that variables such as
1392 no_msglog_check are global and can be placed anywhere in a specific
1393 test's script, however it was observed that placement needed to be near
1394 the beginning for it to behave that way. Changed the runtest perl
1395 script to read through the entire script once to detect and set these
1396 variables, reset to the beginning of the script, and then run through
1397 the script parsing/test process like normal.
1399 TL/02 The BSD's have an arc4random API. One of the functions to induce
1400 adding randomness was arc4random_stir(), but it has been removed in
1401 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1402 function when detected.
1404 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1405 cause callback expansion.
1407 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1408 syntax errors in an expansion can be treated as a string instead of
1409 logging or causing an error, due to the internal use of bool_lax
1410 instead of bool when processing it.
1412 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1413 server certificates when making smtp deliveries.
1415 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1417 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1419 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1421 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1422 Merged patch from Sebastian Wiedenroth.
1424 JH/05 Fix results-pipe from transport process. Several recipients, combined
1425 with certificate use, exposed issues where response data items split
1426 over buffer boundaries were not parsed properly. This eventually
1427 resulted in duplicates being sent. This issue only became common enough
1428 to notice due to the introduction of connection certificate information,
1429 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1431 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1432 size buffer was used, resulting in syntax errors when an expansion
1435 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1436 version 3.3.6 or later.
1438 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1439 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1440 both become "event_action", the variables become $event_name, $event_data
1441 and $event_defer_errno. There is a new variable $verify_mode, usable in
1442 routers, transports and related events. The tls:cert event is now also
1443 raised for inbound connections, if the main configuration event_action
1446 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1447 early OCSP support, but no stapling (appears to be less than 1.0.0).
1449 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1450 server certificate names available under the smtp transport option
1451 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1454 JH/10 Time-related extraction expansions from certificates now use the main
1455 option "timezone" setting for output formatting, and are consistent
1456 between OpenSSL and GnuTLS compilations. Bug 1541.
1458 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1459 encoded parameter in the incoming message. Bug 1558.
1461 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1462 include certificate info, eximon was claiming there were spoolfile
1465 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1467 JH/14 Log delivery-related information more consistently, using the sequence
1468 "H=<name> [<ip>]" wherever possible.
1470 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1471 are problematic for Debian distribution, omit them from the release
1474 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1476 JH/16 Fix string representation of time values on 64bit time_t architectures.
1479 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1480 output list separator was used.
1485 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1486 checkers that were complaining about end of non-void function with no
1489 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1490 This was a regression introduced in 4.83 by another bugfix.
1492 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1494 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1495 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1501 TF/01 Correctly close the server side of TLS when forking for delivery.
1503 When a message was received over SMTP with TLS, Exim failed to clear up
1504 the incoming connection properly after forking off the child process to
1505 deliver the message. In some situations the subsequent outgoing
1506 delivery connection happened to have the same fd number as the incoming
1507 connection previously had. Exim would try to use TLS and fail, logging
1508 a "Bad file descriptor" error.
1510 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1511 utilities have not been installed.
1513 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1514 temporary space as the ACL may create new global variables.
1516 TL/01 LDAP support uses per connection or global context settings, depending
1517 upon the detected version of the libraries at build time.
1519 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1520 to extract and use the src ip:port in logging and expansions as if it
1521 were a direct connection from the outside internet. PPv2 support was
1522 updated based on HAProxy spec change in May 2014.
1524 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1526 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1527 Properly escape header and check for NULL return.
1529 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1532 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1534 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1535 characters in header names, implemented as a verify condition.
1536 Contributed by Michael Fischer v. Mollard.
1538 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1539 results permerror and temperror. Previous values are deprecated but
1540 still accepted. In a future release, err_perm and err_temp will be
1541 completely removed, which will be a backward incompatibility if the
1542 ACL tests for either of these two old results. Patch contributed by
1543 user bes-internal on the mailing list.
1545 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1547 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1548 selectors, in both main and reject logs.
1550 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1553 JH/07 Add malware type "sock" for talking to simple daemon.
1555 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1557 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1558 routers/transports under cutthrough routing.
1560 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1561 numbers. Touch up "bool" conditional to keep the same definition.
1563 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1565 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1567 JH/12 Expand items in router/transport headers_add or headers_remove lists
1568 individually rather than the list as a whole. Bug 1452.
1570 Required for reasonable handling of multiple headers_ options when
1571 they may be empty; requires that headers_remove items with embedded
1572 colons must have them doubled (or the list-separator changed).
1574 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1575 view the policy declared in the DMARC record. Currently, $dmarc_status
1576 is a combined value of both the record presence and the result of the
1579 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1581 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1582 dnslookup router and the smtp transport (applying to the forward
1585 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1586 of ldap servers used for a specific lookup. Patch provided by Heiko
1589 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1590 New variable $lookup_dnssec_authenticated for observability.
1592 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1593 Patch submitted by Lars Timman.
1595 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1597 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1598 Requires trusted mode and valid format message id, aborts otherwise.
1599 Patch contributed by Heiko Schlichting.
1601 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1602 certextract with support for various fields. Bug 1358.
1604 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1605 is requested by default, modifiable by smtp transport option
1608 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1609 operate on certificate variables to give certificate fingerprints
1610 Also new ${sha256:cert_variable}.
1612 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1614 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1615 Christian Aistleitner.
1617 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1619 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1620 file. Patch from Wolfgang Breyha.
1622 JH/25 Expand the coverage of the delivery $host and $host_address to
1623 client authenticators run in verify callout. Bug 1476.
1625 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1626 align with daemon_smtp_ports. Bug 72.
1628 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1629 support and error reporting did not work properly.
1631 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1632 and is readable. Patch from Andrew Colin Kissa.
1634 TL/14 Enhance documentation of ${run expansion and how it parses the
1635 commandline after expansion, particularly in the case when an
1636 unquoted variable expansion results in an empty value.
1638 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1640 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1641 by renaming away from it. Addresses GH issue 15, reported by
1644 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1646 TL/15 SECURITY: prevent double expansion in math comparison functions
1647 (can expand unsanitized data). Not remotely exploitable.
1654 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1656 PP/02 Make -n do something, by making it not do something.
1657 When combined with -bP, the name of an option is not output.
1659 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1662 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1663 $sender_host_name and config options to manage this, and basic check
1666 PP/05 DSCP support for outbound connections and control modifier for inbound.
1668 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1669 (Only plugin which currently uses this is kerberos4, which nobody should
1670 be using, but we should make it available and other future plugins might
1671 conceivably use it, even though it would break NAT; stuff *should* be
1672 using channel bindings instead).
1674 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1675 name; added for Sendmail compatibility; requires admin caller.
1676 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1677 just ignore it); requires trusted caller.
1678 Also parse but ignore: -Ac -Am -X<logfile>
1681 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1683 TL/02 Add +smtp_confirmation as a default logging option.
1685 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1686 Patch by Magnus Holmgren from 2007-02-20.
1688 TL/04 Bugzilla 1281 - Spec typo.
1689 Bugzilla 1283 - Spec typo.
1690 Bugzilla 1290 - Spec grammar fixes.
1692 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1694 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1696 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1697 mailing list by by Dmitry Isaikin.
1699 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1701 JH/02 Support "G" suffix to numbers in ${if comparisons.
1703 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1705 NM/01 Bugzilla 1197 - Spec typo
1706 Bugzilla 1196 - Spec examples corrections
1708 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1710 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1711 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1714 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1715 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1717 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1718 "acl {{name}{arg}...}", and optional args on acl condition
1721 JH/05 Permit multiple router/transport headers_add/remove lines.
1723 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1725 JH/07 Avoid using a waiting database for a single-message-only transport.
1726 Performance patch from Paul Fisher. Bugzilla 1262.
1728 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1731 JH/09 Add $headers_added variable, with content from use of ACL modifier
1732 add_header (but not yet added to the message). Bugzilla 199.
1734 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1735 Pulled from Bugzilla 817 by Wolfgang Breyha.
1737 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1739 (nb: this is the same fix as in Exim 4.80.1)
1741 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1744 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1747 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1749 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1750 advertise SMTP AUTH mechanism to us, instead of a generic
1751 protocol violation error. Also, make Exim more robust to bad
1752 data from the Dovecot auth socket.
1754 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1756 When a queue runner is handling a message, Exim first routes the
1757 recipient addresses, during which it prunes them based on the retry
1758 hints database. After that it attempts to deliver the message to
1759 any remaining recipients. It then updates the hints database using
1762 So if a recipient address works intermittently, it can get repeatedly
1763 deferred at routing time. The retry hints record remains fresh so the
1764 address never reaches the final cutoff time.
1766 This is a fairly common occurrence when a user is bumping up against
1767 their storage quota. Exim had some logic in its local delivery code
1768 to deal with this. However it did not apply to per-recipient defers
1769 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1771 This change adds a proper retry rule check during routing so that the
1772 final cutoff time is checked against the message's age. We only do
1773 this check if there is an address retry record and there is not a
1774 domain retry record; this implies that previous attempts to handle
1775 the address had the retry_use_local_parts option turned on. We use
1776 this as an approximation for the destination being like a local
1777 delivery, as in LMTP.
1779 I suspect this new check makes the old local delivery cutoff check
1780 redundant, but I have not verified this so I left the code in place.
1782 TF/02 Correct gecos expansion when From: is a prefix of the username.
1784 Test 0254 submits a message to Exim with the header
1788 When I ran the test suite under the user fanf2, Exim expanded
1789 the header to contain my full name, whereas it should have added
1790 a Resent-Sender: header. It erroneously treats any prefix of the
1791 username as equal to the username.
1793 This change corrects that bug.
1795 GF/01 DCC debug and logging tidyup
1796 Error conditions log to paniclog rather than rejectlog.
1797 Debug lines prefixed by "DCC: " to remove any ambiguity.
1799 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1801 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1802 Bug spotted by Jeremy Harris; was flawed since initial commit.
1803 Would have resulted in OCSP responses post-SNI triggering an Exim
1804 NULL dereference and crash.
1806 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1808 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1809 Bug detection, analysis and fix by Samuel Thibault.
1810 Bugzilla 1331, Debian bug #698092.
1812 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1814 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1815 Server implementation by Todd Lyons, client by JH.
1816 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1817 config variable "prdr_enable" controls whether the server
1818 advertises the facility. If the client requests PRDR a new
1819 acl_data_smtp_prdr ACL is called once for each recipient, after
1820 the body content is received and before the acl_smtp_data ACL.
1821 The client is controlled by both of: a hosts_try_prdr option
1822 on the smtp transport, and the server advertisement.
1823 Default client logging of deliveries and rejections involving
1824 PRDR are flagged with the string "PRDR".
1826 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1827 fclose(). Diagnosis by Todd Lyons.
1829 PP/17 Update configure.default to handle IPv6 localhost better.
1830 Patch by Alain Williams (plus minor tweaks).
1833 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1834 This is now consistent with GnuTLS, and is now documented: the
1835 previous undocumented portable approach to treating the option as
1836 unset was to force an expansion failure. That still works, and
1837 an empty string is now equivalent.
1839 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1840 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1841 not performing validation itself.
1843 PP/20 Added force_command boolean option to pipe transport.
1844 Patch from Nick Koston, of cPanel Inc.
1846 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1849 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1851 PP/21 Fix eximon continuous updating with timestamped log-files.
1852 Broken in a format-string cleanup in 4.80, missed when I repaired the
1853 other false fix of the same issue.
1854 Report and fix from Heiko Schlichting.
1857 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1858 Report from Prashanth Katuri.
1860 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1861 It's SecureTransport, so affects any MacOS clients which use the
1862 system-integrated TLS libraries, including email clients.
1864 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1865 using a MIME ACL for non-SMTP local injection.
1866 Report and assistance in diagnosis by Warren Baker.
1868 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1870 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1872 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1873 last id that failed so it may be referenced in subsequent ACL's.
1875 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1878 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1879 ldap library initialization, allowing self-signed CA's to be
1880 used. Also properly sets require_cert option later in code by
1881 using NULL (global ldap config) instead of ldap handle (per
1882 session). Bug diagnosis and testing by alxgomz.
1884 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1885 the src/util/ subdirectory.
1887 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1888 renamed to Transport Post Delivery Action by Jeremy Harris, as
1891 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1892 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1893 redis_servers = needs to be configured which will be used by the redis
1894 lookup. Patch from Warren Baker, of The Packet Hub.
1896 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1898 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1899 hostname or reverse DNS when processing a host list. Used suggestions
1900 from multiple comments on this bug.
1902 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1904 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1905 Missed a few lines, added it to make the runtest require no keyboard
1908 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1909 contains upper case chars. Make router use caseful_local_part.
1911 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1912 support when GnuTLS has been built with p11-kit.
1918 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1920 This, or similar/improved, will also be change PP/11 of 4.82.
1926 PP/01 Handle short writes when writing local log-files.
1927 In practice, only affects FreeBSD (8 onwards).
1928 Bugzilla 1053, with thanks to Dmitry Isaikin.
1930 NM/01 Bugzilla 949 - Documentation tweak
1932 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1935 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1937 PP/02 Implemented gsasl authenticator.
1939 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1941 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1942 `pkg-config foo` for cflags/libs.
1944 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1945 with rest of GSASL and with heimdal_gssapi.
1947 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1948 `pkg-config foo` for cflags/libs for the TLS implementation.
1950 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1951 properties get this fed in as external SSF. A number of robustness
1952 and debugging improvements to the cyrus_sasl authenticator.
1954 PP/08 cyrus_sasl server now expands the server_realm option.
1956 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1957 Patch by Jeremy Harris.
1959 PP/10 Added dbmjz lookup type.
1961 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1963 PP/12 MAIL args handles TAB as well as SP, for better interop with
1964 non-compliant senders.
1965 Analysis and variant patch by Todd Lyons.
1967 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1968 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1969 Patch from Dirk Mueller <dmueller@suse.com>
1971 PP/13 tls_peerdn now print-escaped for spool files.
1972 Observed some $tls_peerdn in wild which contained \n, which resulted
1973 in spool file corruption.
1975 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1976 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1977 or write after TLS renegotiation, which otherwise led to messages
1980 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1981 as a tracking header (ie: a signed header comes before the signature).
1982 Patch from Wolfgang Breyha.
1984 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1985 comma-sep list; embedded commas doubled.
1987 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1989 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1991 Report and patch from Dmitry Banschikov.
1993 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1994 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1995 failures. We appear to now support TLS1.1+ with Exim.
1997 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1998 lets Exim select keys and certificates based upon TLS SNI from client.
1999 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2000 before an outbound SMTP session. New log_selector, +tls_sni.
2002 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2003 NULL dereference. Report and patch from Alun Jones.
2005 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2006 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2007 Not seeing resolver debug output on NetBSD, but suspect this is a
2008 resolver implementation change.
2010 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2011 Left warnings. Added "eximon gdb" invocation mode.
2013 PP/21 Defaulting "accept_8bitmime" to true, not false.
2015 PP/22 Added -bw for inetd wait mode support.
2017 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2018 locate the relevant includes and libraries. Made this the default.
2020 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2021 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2023 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2024 This may cause build issues on older platforms.
2026 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2027 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2028 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2029 Added SNI support via GnuTLS too.
2030 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2032 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2034 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2035 Applied second patch from Janne, implementing suggestion to default
2036 multiple-strings-in-record handling to match SPF spec.
2038 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2040 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2041 read-only, out of scope).
2042 Patch from Wolfgang Breyha, report from Stuart Northfield.
2044 PP/29 Fix three issues highlighted by clang analyser static analysis.
2045 Only crash-plausible issue would require the Cambridge-specific
2046 iplookup router and a misconfiguration.
2047 Report from Marcin Mirosław.
2049 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2051 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2052 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2053 As part of this, removing so much warning spew let me fix some minor
2054 real issues in debug logging.
2056 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2057 assignment on my part. Fixed.
2059 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2060 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2061 Janne Snabb (who went above and beyond: thank you).
2063 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2064 string otherwise requires a connection and a bunch more work and it's
2065 relatively easy to get wrong. Should also expose TLS library linkage
2068 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2069 64-bit ${eval} (JH/03).
2071 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2072 GNU libc to support some of the 64-bit stuff, should not lead to
2073 conflicts. Defined before os.h is pulled in, so if a given platform
2074 needs to override this, it can.
2076 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2077 protection layer was required, which is not implemented.
2078 Bugzilla 1254, patch from Wolfgang Breyha.
2080 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2081 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2082 tls_dhparam take prime identifiers. Also unbreak combination of
2083 OpenSSL+DH_params+TLSSNI.
2085 PP/39 Disable SSLv2 by default in OpenSSL support.
2091 PP/01 Solaris build fix for Oracle's LDAP libraries.
2092 Bugzilla 1109, patch from Stephen Usher.
2094 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2096 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2099 TF/02 Fix a couple more cases where we did not log the error message
2100 when unlink() failed. See also change 4.74-TF/03.
2102 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2103 lock up or crash if it happened to be inside a call to libc when it
2104 got a SIGUSR1 from exiwhat.
2106 The SIGUSR1 handler appends the current process status to the process
2107 log which is later printed by exiwhat. It used to use the general
2108 purpose logging code to do this, but several functions it calls are
2109 not safe for signals.
2111 The new output code in the SIGUSR1 handler is specific to the process
2112 log, and simple enough that it's easy to inspect for signal safety.
2113 Removing some special cases also simplifies the general logging code.
2114 Removing the spurious timestamps from the process log simplifies
2117 TF/04 Improved ratelimit ACL condition.
2119 The /noupdate option has been deprecated in favour of /readonly which
2120 has clearer semantics. The /leaky, /strict, and /readonly update modes
2121 are mutually exclusive. The update mode is no longer included in the
2122 database key; it just determines when the database is updated. (This
2123 means that when you upgrade Exim will forget old rate measurements.)
2125 Exim now checks that the per_* options are used with an update mode that
2126 makes sense for the current ACL. For example, when Exim is processing a
2127 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2128 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2129 must specify per_mail/readonly. If you omit the update mode it defaults to
2130 /leaky where that makes sense (as before) or /readonly where required.
2132 The /noupdate option is now undocumented but still supported for
2133 backwards compatibility. It is equivalent to /readonly except that in
2134 ACLs where /readonly is required you may specify /leaky/noupdate or
2135 /strict/noupdate which are treated the same as /readonly.
2137 A useful new feature is the /count= option. This is a generalization
2138 of the per_byte option, so that you can measure the throughput of other
2139 aggregate values. For example, the per_byte option is now equivalent
2140 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2142 The per_rcpt option has been generalized using the /count= mechanism
2143 (though it's more complicated than the per_byte equivalence). When it is
2144 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2145 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2146 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2147 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2148 Note that using per_rcpt with a non-readonly update mode in more than
2149 one ACL will cause the recipients to be double-counted. (The per_mail
2150 and per_byte options don't have this problem.)
2152 The handling of very low rates has changed slightly. If the computed rate
2153 is less than the event's count (usually one) then this event is the first
2154 after a long gap. In this case the rate is set to the same as this event's
2155 count, so that the first message of a spam run is counted properly.
2157 The major new feature is a mechanism for counting the rate of unique
2158 events. The new per_addr option counts the number of different
2159 recipients that someone has sent messages to in the last time period. It
2160 behaves like per_rcpt if all the recipient addresses are different, but
2161 duplicate recipient addresses do not increase the measured rate. Like
2162 the /count= option this is a general mechanism, so the per_addr option
2163 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2164 example, measure the rate that a client uses different sender addresses
2165 with the options per_mail/unique=$sender_address. There are further
2166 details in the main documentation.
2168 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2170 TF/06 Removed a few PCRE remnants.
2172 TF/07 Automatically extract Exim's version number from tags in the git
2173 repository when doing development or release builds.
2175 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2176 Bugzilla 879. Patch from Paul Fisher.
2178 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2179 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2182 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2184 PP/05 Variable $av_failed, true if the AV scanner deferred.
2185 Bugzilla 1078. Patch from John Horne.
2187 PP/06 Stop make process more reliably on build failure.
2188 Bugzilla 1087. Patch from Heiko Schlittermann.
2190 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2191 Bugzilla 1089. Patch from Heiko Schlittermann.
2193 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2194 Bugzilla 1131. Patch from Holger Weiß.
2196 PP/09 Handle IPv6 addresses with SPF.
2197 Bugzilla 860. Patch from Wolfgang Breyha.
2199 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2201 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2204 PP/11 match_* no longer expand right-hand-side by default.
2205 New compile-time build option, EXPAND_LISTMATCH_RHS.
2206 New expansion conditions, "inlist", "inlisti".
2208 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2210 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2212 PP/14 fix log_write() format string regression from TF/03.
2213 Bugzilla 1152. Patch from Dmitry Isaikin.
2219 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2221 PP/02 Harmonised TLS library version reporting; only show if debugging.
2222 Layout now matches that introduced for other libraries in 4.74 PP/03.
2224 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2226 PP/04 New "dns_use_edns0" global option.
2228 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2231 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2232 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2234 TK/01 Updated PolarSSL code to 0.14.2.
2235 Bugzilla 1097. Patch from Andreas Metzler.
2237 PP/07 Catch divide-by-zero in ${eval:...}.
2238 Fixes bugzilla 1102.
2240 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2243 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2244 format-string attack -- SECURITY: remote arbitrary code execution.
2246 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2247 time unintentionally subject to list matching rules, letting the header
2248 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2249 arbitrary string expansion). This allowed for information disclosure.
2251 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2252 INT_MIN/-1 -- value coerced to INT_MAX.
2258 NM/01 Workaround for PCRE version dependency in version reporting
2261 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2262 This fixes portability to compilers other than gcc, notably
2263 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2265 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2266 makefiles for portability to HP-UX and POSIX correctness.
2268 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2269 Also via indented variable definition in the Makefile.
2270 (Debugging by Oliver Heesakkers).
2272 PP/02 Restore caching of spamd results with expanded spamd_address.
2273 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2275 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2276 Improves build reliability. Fix from: Frank Elsner
2278 NM/02 Fix wide character breakage in the rfc2047 coding
2279 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2281 NM/03 Allow underscore in dnslist lookups
2282 Fixes bug 1026. Patch from Graeme Fowler
2284 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2285 Code patches from Adam Ciarcinski of NetBSD.
2287 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2290 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2291 is logged, to avoid truncation. Patch from John Horne.
2293 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2294 Patch from Jakob Hirsch.
2296 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2297 SQL string expansion failure details.
2298 Patch from Andrey Oktyabrski.
2300 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2301 Patch from Simon Arlott.
2303 PP/09 New lookups functionality failed to compile on old gcc which rejects
2304 extern declarations in function scope.
2305 Patch from Oliver Fleischmann
2307 PP/10 Use sig_atomic_t for flags set from signal handlers.
2308 Check getgroups() return and improve debugging.
2309 Fixed developed for diagnosis in bug 927 (which turned out to be
2312 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2313 Patch from Mark Zealey.
2315 PP/12 Bugzilla 1056: Improved spamd server selection.
2316 Patch from Mark Zealey.
2318 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2319 Based on patch from Heiko Schlittermann.
2321 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2322 Patch from Uwe Doering, sign-off by Michael Haardt.
2324 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2325 variable declaration deep within a block. Bug and patch from
2328 PP/15 lookups-Makefile IRIX compatibility coercion.
2330 PP/16 Make DISABLE_DKIM build knob functional.
2332 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2333 Patch by Simon Arlott
2335 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2336 variable argument macros. Our copy now differs from upstream.
2342 TF/01 Failure to get a lock on a hints database can have serious
2343 consequences so log it to the panic log.
2345 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2346 controlled using the smtp_confirmation log selector.
2348 TF/03 Include the error message when we fail to unlink a spool file.
2350 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2351 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2352 for maintaining out-of-tree patches for some time.
2354 PP/01 Bugzilla 139: Documentation and portability issues.
2355 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2356 Handle per-OS dynamic-module compilation flags.
2358 PP/02 Let /dev/null have normal permissions.
2359 The 4.73 fixes were a little too stringent and complained about the
2360 permissions on /dev/null. Exempt it from some checks.
2361 Reported by Andreas M. Kirchwitz.
2363 PP/03 Report version information for many libraries, including
2364 Exim version information for dynamically loaded libraries. Created
2365 version.h, now support a version extension string for distributors
2366 who patch heavily. Dynamic module ABI change.
2368 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2369 privilege escalation vulnerability whereby the Exim run-time user
2370 can cause root to append content of the attacker's choosing to
2373 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2376 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2377 If dropping privileges for untrusted macros, we disabled normal logging
2378 on the basis that it would fail; for the Exim run-time user, this is not
2379 the case, and it resulted in successful deliveries going unlogged.
2380 Fixed. Reported by Andreas Metzler.
2386 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2387 only prepend for the Resent-* case. Fixes regression introduced in
2388 Exim 4.70 by NM/22 for Bugzilla 607.
2390 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2391 increasing numbers of administrators be bitten by this.
2393 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2395 PP/03 Bugzilla 994: added openssl_options main configuration option.
2397 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2399 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2401 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2403 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2404 without a peer certificate, leading to a segfault because of an
2405 assumption that peers always have certificates. Be a little more
2406 paranoid. Problem reported by Martin Tscholak.
2408 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2409 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2410 NB: ClamAV planning to remove STREAM in "middle of 2010".
2411 CL also introduces -bmalware, various -d+acl logging additions and
2412 more caution in buffer sizes.
2414 PP/09 Implemented reverse_ip expansion operator.
2416 PP/10 Bugzilla 937: provide a "debug" ACL control.
2418 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2420 PP/12 Bugzilla 973: Implement --version.
2422 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2424 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2426 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2428 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2429 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2430 ignore trailing whitespace.
2432 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2434 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2437 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2438 Notification from Dr Andrew Aitchison.
2440 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2441 ExtendedDetectionInfo response format.
2442 Notification from John Horne.
2444 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2447 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2448 XSL and documented dependency on system catalogs, with examples of how
2451 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2454 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2455 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2456 configuration file which is writeable by the Exim user or group.
2458 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2459 of configuration files to cover files specified with the -C option if
2460 they are going to be used with root privileges, not just the default
2463 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2464 option (effectively making it always true).
2466 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2467 files to be used while preserving root privileges.
2469 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2470 that rogue child processes cannot use them.
2472 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2473 run-time user, instead of root.
2475 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2476 Exim run-time user without dropping privileges.
2478 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2479 result string, instead of calling string_vformat() twice with the same
2482 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2483 for other users. Others should always drop root privileges if they use
2484 -C on the command line, even for a whitelisted configure file.
2486 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2488 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2494 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2495 $data_path, and $header_path variables; fixed documentation bugs and
2498 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2499 exipick to access non-standard spools, including the "frozen" queue
2502 NM/01 Bugzilla 965: Support mysql stored procedures.
2503 Patch from Alain Williams
2505 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2507 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2508 Patch from Andreas Metzler
2510 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2511 Patch from Kirill Miazine
2513 NM/05 Bugzilla 671: Added umask to procmail example.
2515 JJ/03 installed exipick 20100323.0, fixing doc bug
2517 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2518 directory. Notification and patch from Dan Rosenberg.
2520 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2522 TK/02 Improve log output when DKIM signing operation fails.
2524 MH/01 Treat the transport option dkim_domain as a colon separated
2525 list, not as a single string, and sign the message with each element,
2526 omitting multiple occurences of the same signer.
2528 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2529 Bugzilla 985, 986. Patch by Simon Arlott
2531 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2532 Patch by Simon Arlott
2534 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2535 MBX locking. Notification from Dan Rosenberg.
2541 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2543 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2545 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2547 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2549 NM/04 Bugzilla 915: Flag broken perl installation during build.
2555 TK/01 Added patch by Johannes Berg that expands the main option
2556 "spamd_address" if it starts with a dollar sign.
2558 TK/02 Write list of recipients to X-Envelope-Sender header when building
2559 the mbox-format spool file for content scanning (suggested by Jakob
2562 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2563 (http://www.dcc-servers.net/) support via dccifd. Activated by
2564 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2566 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2567 by Mark Daniel Reidel <mr@df.eu>.
2569 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2570 When building exim an external PCRE library is now needed -
2571 PCRE is a system library on the majority of modern systems.
2572 See entry on PCRE_LIBS in EDITME file.
2574 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2575 conversation. Added nologin parameter to request.
2576 Patch contributed by Kirill Miazine.
2578 TF/01 Do not log submission mode rewrites if they do not change the address.
2580 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2582 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2583 log files in place. Contributed by Roberto Lima.
2585 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2587 TF/03 Bugzilla 615: When checking the local_parts router precondition
2588 after a local_part_suffix or local_part_prefix option, Exim now
2589 does not use the address's named list lookup cache, since this
2590 contains cached lookups for the whole local part.
2592 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2593 Robert Millan. Documentation is in experimental-spec.txt.
2595 TF/04 Bugzilla 668: Fix parallel build (make -j).
2597 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2599 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2600 Patch provided by Jan Srzednicki.
2602 TF/05 Leading white space used to be stripped from $spam_report which
2603 wrecked the formatting. Now it is preserved.
2605 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2606 that they are available at delivery time.
2608 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2610 TF/08 TLS error reporting now respects the incoming_interface and
2611 incoming_port log selectors.
2613 TF/09 Produce a more useful error message if an SMTP transport's hosts
2614 setting expands to an empty string.
2616 NM/06 Bugzilla 744: EXPN did not work under TLS.
2617 Patch provided by Phil Pennock.
2619 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2620 Patch provided by Richard Godbee.
2622 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2623 acl_smtp_notquit, added index entry.
2625 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2626 Patch provided by Eugene Bujak.
2628 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2629 accept(). Patch provided by Maxim Dounin.
2631 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2632 Patch provided by Phil Pennock.
2634 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2636 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2637 Patch provided by Brad "anomie" Jorsch.
2639 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2640 Patch provided by Dean Brooks.
2642 TK/05 Add native DKIM support (does not depend on external libraries).
2644 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2645 Patch provided by Graeme Fowler.
2647 NM/16 Bugzilla 851: Documentation example syntax fix.
2649 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2651 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2654 NM/19 Bugzilla 745: TLS version reporting.
2655 Patch provided by Phil Pennock.
2657 NM/20 Bugzilla 167: bool: condition support.
2658 Patch provided by Phil Pennock.
2660 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2661 clients. Patch provided by Phil Pennock.
2663 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2664 Patch provided by Brad "anomie" Jorsch.
2666 NM/23 Bugzilla 687: Fix misparses in eximstats.
2667 Patch provided by Heiko Schlittermann.
2669 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2670 Patch provided by Heiko Schlittermann.
2672 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2673 plus update to original patch.
2675 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2677 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2678 Patch provided by David Brownlee.
2680 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2682 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2684 NM/30 Bugzilla 888: TLS documentation bugfixes.
2686 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2688 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2689 Unlike the original bugzilla I have changed all shell scripts in src tree.
2691 NM/33 Bugzilla 898: Transport filter timeout fix.
2692 Patch by Todd Rinaldo.
2694 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2695 Patch by Serge Demonchaux.
2697 NM/35 Bugzilla 39: Base64 decode bug fixes.
2698 Patch by Jakob Hirsch.
2700 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2702 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2704 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2706 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2712 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2713 ALT-N's libdkim that I have put here:
2714 http://duncanthrax.net/exim-experimental/
2716 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2717 were called 'true' and it seems that C99 defines that as a reserved
2718 keyword to be used with 'bool' variable types. That means you could
2719 not include C99-style headers which use bools without triggering
2720 build errors in sieve.c.
2722 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2723 as mailq or other aliases. Changed the --help handling significantly
2724 to do whats expected. exim_usage() emits usage/help information.
2726 SC/01 Added the -bylocaldomain option to eximstats.
2728 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2730 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2732 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2738 PH/01 Another patch from the Sieve maintainer.
2740 PH/02 When an IPv6 address is converted to a string for single-key lookup
2741 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2742 dots are used instead of colons so that keys in lsearch files need not
2743 contain colons. This was done some time before quoting was made available
2744 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2745 (notated using the quote facility) so as to distinguish them from IPv4
2746 keys. This meant that lookups for IP addresses in host lists did not work
2747 for iplsearch lookups.
2749 This has been fixed by arranging for IPv6 addresses to be expressed with
2750 colons if the lookup type is iplsearch. This is not incompatible, because
2751 previously such lookups could never work.
2753 The situation is now rather anomalous, since one *can* have colons in
2754 ordinary lsearch keys. However, making the change in all cases is
2755 incompatible and would probably break a number of configurations.
2757 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2760 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2761 conversion specification without a maximum field width, thereby enabling
2762 a rogue spamd server to cause a buffer overflow. While nobody in their
2763 right mind would setup Exim to query an untrusted spamd server, an
2764 attacker that gains access to a server running spamd could potentially
2765 exploit this vulnerability to run arbitrary code as the Exim user.
2767 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2768 $primary_hostname instead of what libspf2 thinks the hosts name is.
2770 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2771 a directory entry by the name of the lookup key. Previously, if a
2772 symlink pointed to a non-existing file or a file in a directory that
2773 Exim lacked permissions to read, a lookup for a key matching that
2774 symlink would fail. Now it is enough that a matching directory entry
2775 exists, symlink or not. (Bugzilla 503.)
2777 PH/03 The body_linecount and body_zerocount variables are now exported in the
2780 PH/04 Added the $dnslist_matched variable.
2782 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2783 This means they are set thereafter only if the connection becomes
2786 PH/06 Added the client_condition to authenticators so that some can be skipped
2787 by clients under certain conditions.
2789 PH/07 The error message for a badly-placed control=no_multiline_responses left
2790 "_responses" off the end of the name.
2792 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2794 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2795 (without spaces) instead of just copying the configuration text.
2797 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2799 PH/11 Added $max_received_linelength.
2801 PH/12 Added +ignore_defer and +include_defer to host lists.
2803 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2804 way in which PCRE > 7.0 is built.
2806 PH/14 Implemented queue_only_load_latch.
2808 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2809 MAIL command. The effect was to mangle the value on 64-bit systems.
2811 PH/16 Another patch from the Sieve maintainer.
2813 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2815 PH/18 If a system quota error occurred while trying to create the file for
2816 a maildir delivery, the message "Mailbox is full" was not appended to the
2817 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2818 only to a quota excession during the actual writing of the file.
2820 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2821 characters?) which causes problems in log lines. The DN values are now
2822 passed through string_printing() before being added to log lines.
2824 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2825 and InterBase are left for another time.)
2827 PH/21 Added message_body_newlines option.
2829 PH/22 Guard against possible overflow in moan_check_errorcopy().
2831 PH/23 POSIX allows open() to be a macro; guard against that.
2833 PH/24 If the recipient of an error message contained an @ in the local part
2834 (suitably quoted, of course), incorrect values were put in $domain and
2835 $local_part during the evaluation of errors_copy.
2841 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2842 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2845 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2846 issue a MAIL command.
2848 PH/02 In an ACL statement such as
2850 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2852 if a client was not listed at all, or was listed with a value other than
2853 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2854 the condition was not true (as it should be), so access was not denied.
2855 The bug was that the ! inversion was incorrectly passed on to the second
2856 item. This has been fixed.
2858 PH/03 Added additional dnslists conditions == and =& which are different from
2859 = and & when the dns lookup returns more than one IP address.
2861 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2862 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2864 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2865 FSYNC, which compiles an option called disable_fsync that allows for
2866 bypassing fsync(). The documentation is heavily laced with warnings.
2868 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2870 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2871 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2872 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2873 including adding "make clean"; (3) Added -fPIC when compiling the test
2874 dynamically loaded module, to get rid of a warning.
2876 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2877 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2878 = 0s. The bug is otherwise harmless.
2880 PH/07 There was a bug in the dovecot authenticator such that the value of
2881 $auth1 could be overwritten, and so not correctly preserved, after a
2882 successful authentication. This usually meant that the value preserved by
2883 the server_setid option was incorrect.
2885 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2887 PH/09 Installed PCRE release 7.0.
2889 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2890 run for batched SMTP input. It is now run at the start of every message
2891 in the batch. While fixing this I discovered that the process information
2892 (output by running exiwhat) was not always getting set for -bs and -bS
2893 input. This is fixed, and it now also says "batched" for BSMTP.
2895 PH/11 Added control=no_pipelining.
2897 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2898 patch, slightly modified), and move the expansion of helo_data till after
2899 the connection is made in the smtp transport (so it can use these
2902 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2904 PH/14 Added log_selector = +pid.
2906 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2908 PH/16 Add ${if forany and ${if forall.
2910 PH/17 Added dsn_from option to vary the From: line in DSNs.
2912 PH/18 Flush SMTP output before performing a callout, unless control =
2913 no_callout_flush is set.
2915 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2916 was true (the default) a successful delivery failed to delete the retry
2917 item, thus causing premature timeout of the address. The bug is now
2920 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2922 PH/21 Long custom messages for fakedefer and fakereject are now split up
2923 into multiline responses in the same way that messages for "deny" and
2924 other ACL rejections are.
2926 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2927 with slight modification.
2929 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2930 draft, changing the syntax and factoring some duplicate code".
2932 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2933 for deliveries of the second and subsequent messages over the same SMTP
2936 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2937 ${reduce, with only minor "tidies".
2939 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2941 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2942 expansion side effects.
2944 PH/27 When a message times out after an over-quota error from an Exim-imposed
2945 quota, the bounce message says "mailbox is full". This message was not
2946 being given when it was a system quota that was exceeded. It now should
2949 MH/03 Made $recipients available in local_scan(). local_scan() already has
2950 better access to the recipient list through recipients_list[], but
2951 $recipients can be useful in postmaster-provided expansion strings.
2953 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2954 in the case of a MAIL command with additional options following the
2955 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2956 were accidentally chopped off.
2958 PH/29 SMTP synchronization checks are implemented when a command is read -
2959 there is a check that no more input is waiting when there shouldn't be
2960 any. However, for some commands, a delay in an ACL can mean that it is
2961 some time before the response is written. In this time, more input might
2962 arrive, invalidly. So now there are extra checks after an ACL has run for
2963 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2964 pipelining has not been advertised.
2966 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2968 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2969 returned in the SMTP response when smtp_return_error_details was false.
2970 This has been fixed.
2972 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2973 instead of the C I/O that was originally supplied, because problems were
2974 reported on Solaris.
2976 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2977 Exim which did not show up earlier: it was assuming that a call to
2978 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2979 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2980 was a macro that became an assignment, so it seemed to work. This has
2981 changed to a proper function call with a void return, hence the compile
2982 error. Exim's code has been fixed.
2984 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2987 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2989 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2991 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2992 because it turns out to be too big or there is a timeout) there is no
2993 "Completed" line in the log. When some messages of this type were
2994 selected by exigrep, they were listed as "not completed". Others were
2995 picked up by some special patterns. I have improved the selection
2996 criteria to be more general.
2998 PH/38 The host_find_failed option in the manualroute router can now be set
2999 to "ignore", to completely ignore a host whose IP address cannot be
3000 found. If all hosts are ignored, the behaviour is controlled by the new
3001 host_all_ignored option.
3003 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3004 homing or because of multiple MX records with /mx) generated more than
3005 one IP address, and the following item turned out to be the local host,
3006 all the secondary addresses of the first item were incorrectly removed
3007 from the list, along with the local host and any following hosts (which
3008 is what is supposed to happen).
3010 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3011 whoever called Exim into the -H file. In the case of the daemon it was
3012 behaving confusingly. When first started, it used values for whoever
3013 started the daemon, but after a SIGHUP it used the Exim user (because it
3014 calls itself on a restart). I have changed the code so that it now always
3017 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3018 message are rejected with the same error (e.g. no authentication or bad
3019 sender address), and a DATA command is nevertheless sent (as can happen
3020 with PIPELINING or a stupid MUA), the error message that was given to the
3021 RCPT commands is included in the rejection of the DATA command. This is
3022 intended to be helpful for MUAs that show only the final error to their
3025 PH/42 Another patch from the Sieve maintainer.
3027 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3028 Eximstats - Fixed some broken HTML links and added missing column headers
3030 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3031 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3033 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3035 PH/43 Yet another patch from the Sieve maintainer.
3037 PH/44 I found a way to check for a TCP/IP connection going away before sending
3038 the response to the final '.' that terminates a message, but only in the
3039 case where the client has not sent further data following the '.'
3040 (unfortunately, this is allowed). However, in many cases there won't be
3041 any further data because there won't be any more messages to send. A call
3042 to select() can be used: if it shows that the input is "ready", there is
3043 either input waiting, or the socket has been closed. An attempt to read
3044 the next input character can distinguish the two cases. Previously, Exim
3045 would have sent an OK response which the client would never have see.
3046 This could lead to message repetition. This fix should cure that, at
3047 least in a lot of common cases.
3049 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3050 advertised in response to EHLO.
3056 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3057 fixed by 4.65/MH/01 (is this a record?) are fixed:
3059 (i) An empty string was always treated as zero by the numeric comparison
3060 operators. This behaviour has been restored.
3062 (ii) It is documented that the numeric comparison operators always treat
3063 their arguments as decimal numbers. This was broken in that numbers
3064 starting with 0 were being interpreted as octal.
3066 While fixing these problems I realized that there was another issue that
3067 hadn't been noticed. Values of message_size_limit (both the global option
3068 and the transport option) were treated as octal if they started with 0.
3069 The documentation was vague. These values are now always treated as
3070 decimal, and I will make that clear in the documentation.
3076 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3077 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3080 MH/01 Don't check that the operands of numeric comparison operators are
3081 integers when their expansion is in "skipping" mode (fixes bug
3082 introduced by 4.64-PH/07).
3084 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3085 child addresses, Exim now panics and dies. Previously, because the count
3086 is held in a short int, deliveries were likely to be lost. As such a
3087 large number of recipients for a single message is ridiculous
3088 (performance will be very, very poor), I have chosen to impose a limit
3089 rather than extend the field.
3095 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3096 leftover -K file (the existence of which was triggered by #402).
3097 While we were at it, introduced process PID as part of the -K
3098 filename. This should rule out race conditions when creating
3101 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3102 processing considerably. Previous code took too long for large mails,
3103 triggering a timeout which in turn triggers #401.
3105 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3106 in the DK code in transports.c. sendfile() is not really portable,
3107 hence the _LINUX specificness.
3109 TF/01 In the add_headers option to the mail command in an Exim filter,
3110 there was a bug that Exim would claim a syntax error in any
3111 header after the first one which had an odd number of characters
3114 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3115 callout verification, Exim cached a "reject" for the entire domain. This
3116 is correct for most verifications, but it is not correct for a recipient
3117 verification with use_sender or use_postmaster set, because in that case
3118 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3119 case of MAIL FROM:<> rejection from other early rejections (e.g.
3120 rejection of HELO). When verifying a recipient using a non-null MAIL
3121 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3122 Whatever the result of the callout, the value of the domain cache is
3123 left unchanged (for any other kind of callout, getting as far as trying
3124 RCPT means that the domain itself is ok).
3126 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3129 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3130 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3131 session, the daemon ignores ECONNECT errors and logs others; it now
3132 ignores EPIPE as well.
3134 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3135 (quoted-printable decoding).
3137 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3138 later the small subsequent patch to fix an introduced bug.
3140 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3142 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3144 PH/08 An error is now given if message_size_limit is specified negative.
3146 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3147 to be given (somewhat) arbitrary names.
3149 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3152 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3153 miscellaneous code fixes
3155 PH/10 Added the log_reject_target ACL modifier to specify where to log
3158 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3159 hostname. This is wrong, because it relates to the incoming message (and
3160 probably the interface on which it is arriving) and not to the outgoing
3161 callout (which could be using a different interface). This has been
3162 changed to use the value of the helo_data option from the smtp transport
3163 instead - this is what is used when a message is actually being sent. If
3164 there is no remote transport (possible with a router that sets up host
3165 addresses), $smtp_active_hostname is used.
3167 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3168 tweaks were necessary in order to get it to work (see also 21 below):
3169 (a) The code assumed that strncpy() returns a negative number on buffer
3170 overflow, which isn't the case. Replaced with Exim's string_format()
3172 (b) There were several signed/unsigned issues. I just did the minimum
3173 hacking in of casts. There is scope for a larger refactoring.
3174 (c) The code used strcasecmp() which is not a standard C function.
3175 Replaced with Exim's strcmpic() function.
3176 (d) The code set only $1; it now sets $auth1 as well.
3177 (e) A simple test gave the error "authentication client didn't specify
3178 service in request". It would seem that Dovecot has changed its
3179 interface. Fortunately there's a specification; I followed it and
3180 changed what the client sends and it appears to be working now.
3182 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3185 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3186 address A is aliased to B and C, where B exists and C does not. Without
3187 -v the output is "A verified" because verification stops after a
3188 successful redirection if more than one address is generated. However,
3189 with -v the child addresses are also verified. Exim was outputting "A
3190 failed to verify" and then showing the successful verification for C,
3191 with its parentage. It now outputs "B failed to verify", showing B's
3192 parentage before showing the successful verification of C.
3194 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3195 look up a TXT record in a specific list after matching in a combined
3198 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3199 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3200 they consult the DNS. I had assumed they would set it the way they
3201 wanted; and indeed my experiments on Linux seem to show that in some
3202 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3203 To be on the safe side, however, I have now made the interface to
3204 host_find_byname() similar to host_find_bydns(), with an argument
3205 containing the DNS resolver options. The host_find_byname() function now
3206 sets these options at its start, just as host_find_bydns() does. The smtp
3207 transport options dns_qualify_single and dns_search_parents are passed to
3208 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3209 of host_find_byname() use the default settings of RES_DEFNAMES
3210 (qualify_single) but not RES_DNSRCH (search_parents).
3212 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3213 spool_read_header() do less string testing, by means of a preliminary
3214 switch on the second character of optional "-foo" lines. (This is
3215 overdue, caused by the large number of possibilities that now exist.
3216 Originally there were few.) While I was there, I also converted the
3217 str(n)cmp tests so they don't re-test the leading "-" and the first
3218 character, in the hope this might squeeze out yet more improvement.
3220 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3221 flag allowing group syntax was set by the header_syntax check but not
3222 turned off, possible causing trouble later; (2) The flag was not being
3223 set at all for the header_verify test, causing "group"-style headers to
3224 be rejected. I have now set it in this case, and also caused header_
3225 verify to ignore an empty address taken from a group. While doing this, I
3226 came across some other cases where the code for allowing group syntax
3227 while scanning a header line wasn't quite right (mostly, not resetting
3228 the flag correctly in the right place). These bugs could have caused
3229 trouble for malformed header lines. I hope it is now all correct.
3231 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3232 with the "reply" argument non-NULL. The code, however (which originally
3233 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3234 but it didn't always do it. This confused somebody who was copying the
3235 code for some other use. I have removed all the tests.
3237 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3238 feature that was used to support insecure browsers during the U.S. crypto
3239 embargo. It requires special client support, and Exim is probably the
3240 only MTA that supported it -- and would never use it because real RSA is
3241 always available. This code has been removed, because it had the bad
3242 effect of slowing Exim down by computing (never used) parameters for the
3243 RSA_EXPORT functionality.
3245 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3246 authenticator to fail if there's a tab character in the incoming data
3247 (there should never be unless someone is messing about, as it's supposed
3248 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3249 if the connection is using TLS or if the remote IP is the same as the
3250 local IP, and the "valid-client-cert option" if a client certificate has
3253 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3254 authenticators. This can be used for authorization after authentication
3255 succeeds. (In the case of plaintext, it servers for both authentication
3258 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3259 if any retry times were supplied.
3261 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3262 connection, where there is no client IP address to check. In this
3263 situation, the verify now always succeeds.
3265 PH/25 Applied John Jetmore's -Mset patch.
3267 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3269 PH/27 In a string expansion for a processed (not raw) header when multiple
3270 headers of the same name were present, leading whitespace was being
3271 removed from all of them, but trailing whitespace was being removed only
3272 from the last one. Now trailing whitespace is removed from each header
3273 before concatenation. Completely empty headers in a concatenation (as
3274 before) are ignored.
3276 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3277 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3279 PH/29 [Removed. This was a change that I later backed out, and forgot to
3280 correct the ChangeLog entry (that I had efficiently created) before
3281 committing the later change.]
3283 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3284 address errors (4xx response to RCPT) over the same connection as other
3285 messages routed to the same hosts. Such deliveries are always "forced",
3286 so retry times are not inspected. This resulted in far too many retries
3287 for the affected addresses. The effect occurred only when there were more
3288 hosts than the hosts_max_try setting in the smtp transport when it had
3289 the 4xx errors. Those hosts that it had tried were not added to the list
3290 of hosts for which the message was waiting, so if all were tried, there
3291 was no problem. Two fixes have been applied:
3293 (i) If there are any address or message errors in an SMTP delivery, none
3294 of the hosts (tried or untried) are now added to the list of hosts
3295 for which the message is waiting, so the message should not be a
3296 candidate for sending over the same connection that was used for a
3297 successful delivery of some other message. This seems entirely
3298 reasonable: after all the message is NOT "waiting for some host".
3299 This is so "obvious" that I'm not sure why it wasn't done
3300 previously. Hope I haven't missed anything, but it can't do any
3301 harm, as the worst effect is to miss an optimization.
3303 (ii) If, despite (i), such a delivery is accidentally attempted, the
3304 routing retry time is respected, so at least it doesn't keep
3305 hammering the server.
3307 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3308 in ${readsocket because some servers need this prod.
3310 PH/32 Added some extra debug output when updating a wait-xxx database.
3312 PH/33 The hint "could be header name not terminated by colon", which has been
3313 given for certain expansion errors for a long time, was not being given
3314 for the ${if def:h_colon_omitted{... case.
3316 PH/34 The spec says: "With one important exception, whenever a domain list is
3317 being scanned, $domain contains the subject domain." There was at least
3318 one case where this was not true.
3320 PH/35 The error "getsockname() failed: connection reset by peer" was being
3321 written to the panic log as well as the main log, but it isn't really
3322 panic-worthy as it just means the connection died rather early on. I have
3323 removed the panic log writing for the ECONNRESET error when getsockname()
3326 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3327 runs only) independently of the message's sender address. This meant
3328 that, if the 4xx error was in fact related to the sender, a different
3329 message to the same recipient with a different sender could confuse
3330 things. In particular, this can happen when sending to a greylisting
3331 server, but other circumstances could also provoke similar problems.
3332 I have changed the default so that the retry time for these errors is now
3333 based a combination of the sender and recipient addresses. This change
3334 can be overridden by setting address_retry_include_sender=false in the
3337 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3338 remote server are returned as part of bounce messages. This was not
3339 happening for LMTP over a pipe (the lmtp transport), but now it is the
3340 same for both kinds of LMTP.
3342 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3343 in header lines that were in fact CNAMEs. This is no longer the case.
3345 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3346 and queue runs started by the daemon processed all messages. This has
3347 been fixed so that -R and -S can now usefully be given with -q<time>.
3349 PH/40 Import PCRE release 6.7 (fixes some bugs).
3351 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3353 PH/42 Give an error if -q is specified more than once.
3355 PH/43 Renamed the variables $interface_address and $interface_port as
3356 $received_ip_address and $received_port, to make it clear that these
3357 values apply to message reception, and not to the outgoing interface when
3358 a message is delivered. (The old names remain recognized, of course.)
3360 PH/44 There was no timeout on the connect() call when using a Unix domain
3361 socket in the ${readsocket expansion. There now is.
3363 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3364 be meaningful with "accept".
3366 SC/01 Eximstats V1.43
3367 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3369 SC/02 Eximstats V1.44
3370 Use a glob alias rather than an array ref in the generated
3371 parser. This improves both readability and performance.
3373 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3374 Collect SpamAssassin and rejection statistics.
3375 Don't display local sender or destination tables unless
3376 there is data to show.
3377 Added average volumes into the top table text output.
3379 SC/04 Eximstats V1.46
3380 Collect data on the number of addresses (recipients)
3381 as well as the number of messages.
3383 SC/05 Eximstats V1.47
3384 Added 'Message too big' to the list of mail rejection
3385 reasons (thanks to Marco Gaiarin).
3387 SC/06 Eximstats V1.48
3388 Mainlog lines which have GMT offsets and are too short to
3389 have a flag are now skipped.
3391 SC/07 Eximstats V1.49 (Alain Williams)
3392 Added the -emptyok flag.
3394 SC/08 Eximstats V1.50
3395 Fixes for obtaining the IP address from reject messages.
3397 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3398 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3399 whitespace changes from 4.64-PH/27
3401 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3404 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3405 are found, allow negative numbers in numeric criteria)
3407 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3409 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3410 to match changes made in 4.64-PH/43
3412 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3414 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3415 "domain" is actually an IP address, and force a failure. This locks out
3416 those revolvers/nameservers that support "A-for-A" lookups, in
3417 contravention of the specifications.
3419 PH/48 When a host name was looked up from an IP address, and the subsequent
3420 forward lookup of the name timed out, the host name was left in
3421 $sender_host_name, contrary to the specification.
3423 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3424 restricted to single-key lookups, Exim was not diagnosing an error if
3425 * or *@ was used with a query-style lookup.
3427 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3429 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3430 long ago, but noone interested enough thought of it. Let's just say that
3431 the "1.1" means that there are some new functions that weren't there at
3432 some point in the past.
3434 PH/51 Error processing for expansion failure of helo_data from an smtp
3435 transport during callout processing was broken.
3437 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3438 tested/used via the -bh/-bhc/-bs options.
3440 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3441 bug, fixed in subsequent PCRE releases).
3443 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3444 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3446 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3452 SC/01 Use a glob alias rather than an array ref in eximstats generated
3453 parser. This improves both readability and performance.
3455 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3456 Don't display local sender or destination tables in eximstats unless
3457 there is data to show.
3458 Added average volumes into the eximstats top table text output.
3460 SC/03 Collect data on the number of addresses (recipients) as well
3461 as the number of messages in eximstats.
3463 TF/01 Correct an error in the documentation for the redirect router. Exim
3464 does (usually) call initgroups() when daemonizing.
3466 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3467 with consistent privilege compared to when running as a daemon.
3469 TF/03 Note in the spec that $authenticated_id is not set for local
3470 submissions from trusted users.
3472 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3473 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3475 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3476 by adding some example configuration directives to the default
3477 configuration file. A little bit of work is required to uncomment the
3478 directives and define how usernames and passwords are checked, but
3479 there is now a framework to start from.
3481 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3482 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3483 without this. I don't know how relevant this is to other LDAP libraries.
3485 PH/02 Add the verb name to the "unknown ACL verb" error.
3487 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3489 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3491 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3492 directory not expanded when it should be if an expanded home directory
3493 was set for the address (which is overridden by the transport).
3495 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3498 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3499 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3500 because it is too late at that time, and has no effect.
3502 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3503 security issue with \' (bugzilla #107). I could not use the
3504 PQescapeStringConn() function, because it needs a PGconn value as one of
3507 PH/08 When testing addresses using -bt, indicate those final addresses that
3508 are duplicates that would not cause an additional delivery. At least one
3509 person was confused, thinking that -bt output corresponded to deliveries.
3510 (Suppressing duplicates isn't a good idea as you lose the information
3511 about possibly different redirections that led to the duplicates.)
3513 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3514 systems where poll() doesn't work, in particular OS X.
3516 PH/10 Added more information to debugging output for retry time not reached.
3518 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3519 operations in malware.c.
3521 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3524 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3525 syslog_duplication set false, log lines that would normally be written
3526 both the the main log and to the reject log were not written to syslog at
3529 PH/14 In the default configuration, change the use of "message" in ACL warn
3530 statements to "add_header".
3532 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3533 not followed by a command (e.g. "seen endif").
3535 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3536 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3539 PH/17 Added extra conditions to the default value of delay_warning_condition
3543 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3544 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3545 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3548 The Auto-Submitted: and various List- headers are standardised, whereas I
3549 don't think Precedence: ever was.
3551 PH/18 Refactored debugging code in route_finduser() to show more information,
3552 in particular, the error code if getpwnam() issues one.
3554 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3555 This is apparently needed in addition to the PH/07 change above to avoid
3556 any possible encoding problems.
3558 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3559 but not after initializing Perl.
3561 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3562 output them only if debugging. By default they are written stderr,
3563 apparently, which is not desirable.
3565 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3568 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3571 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3573 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3574 authenticated or an ident call has been made. Suppress the default
3575 values for $authenticated_id and $authenticated_sender (but permit -oMai
3576 and -oMas) when testing with -bh.
3578 PH/24 Re-jigged the order of the tests in the default configuration so that the
3579 tests for valid domains and recipients precede the DNS black list and CSA
3580 tests, on the grounds that those ones are more expensive.
3582 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3583 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3584 This bug exists in every version of Exim that I still have, right back to
3587 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3588 However, an attempt to turn on case-sensitivity in a regex key by
3589 including (?-i) didn't work because the subject string was already
3590 lowercased, and the effects were non-intuitive. It turns out that a
3591 one-line patch can be used to allow (?-i) to work as expected.
3597 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3598 other effects) broke the use of negated acl sub-conditions.
3600 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3603 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3604 "Deny" causes Exim to reject the incoming connection with a 554 error.
3605 Unfortunately, if there is a major crisis, such as a disk failure,
3606 tcp-wrappers gives "deny", whereas what one would like would be some
3607 kind of temporary error. A kludge has been added to help with this.
3608 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3609 554 error is used if errno is still zero or contains ENOENT (which occurs
3610 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3613 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3615 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3616 errors. Otherwise a message that provokes a temporary error (when other
3617 messages do not) can cause a whole host to time out.
3619 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3620 addresses were routed directly to files or pipes from a redirect router.
3621 File deliveries just didn't batch; pipe deliveries might have suffered
3624 PH/06 A failure to get a lock for a hints database would erroneously always say
3625 "Failed to get write lock", even when it was really a read lock.
3627 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3628 of 0600. This has been changed to use the value of the lockfile_mode
3629 option (which defaults to 0600).
3631 PH/08 Applied small patch from the Sieve maintainer.
3633 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3634 folder from quota calculations, a direct delivery into this folder messed
3635 up the contents of the maildirsize file. This was because the regex was
3636 used only to exclude .Trash (or whatever) when the size of the mailbox
3637 was calculated. There was no check that a delivery was happening into an
3638 excluded directory. This bug has been fixed by ignoring all quota
3639 processing for deliveries into excluded directories.
3641 PH/10 Added the maildirfolder_create_regex option to appendfile.
3647 PH/01 The code for finding all the local interface addresses on a FreeBSD
3648 system running IPv6 was broken. This may well have applied to all BSD
3649 systems, as well as to others that have similar system calls. The broken
3650 code found IPv4 interfaces correctly, but gave incorrect values for the
3651 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3652 that it would not match correctly against @[] and not recognize the IPv6
3655 PH/02 The ipliteral router was not recognizing addresses of the form user@
3656 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3658 PH/03 Added disable_ipv6.
3660 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3661 decoded form, because it is most often used to construct To: headers
3662 lines in autoreplies, and the decoded form may well be syntactically
3663 invalid. However, $reply_address has leading white space removed, and all
3664 newlines turned into spaces so that the autoreply transport does not
3667 PH/05 If group was specified without a user on a router, and no group or user
3668 was specified on a transport, the group from the router was ignored.
3670 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3671 for visible compile-time settings that can be used to change these
3672 numbers, for those that want even more. Backwards compatibility with old
3673 spool files has been maintained. However, going back to a previous Exim
3674 release will lost any variables that are in spool files.
3676 PH/07 Two small changes when running in the test harness: increase delay when
3677 passing a TCP/IP connection to a new process, in case the original
3678 process has to generate a bounce, and remove special handling of
3679 127.0.0.2 (sic), which is no longer necessary.
3681 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3682 be the same on different OS.
3684 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3687 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3688 whether --show-vars was specified or not
3690 JJ/02 exipick: Added support for new ACL variable spool format introduced
3693 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3694 syntactically invalid From: or Reply-to: line, and a filter used this to
3695 generate an autoreply, and therefore failed to obtain an address for the
3696 autoreply, Exim could try to deliver to a non-existent relative file
3697 name, causing unrelated and misleading errors. What now happens is that
3698 it logs this as a hard delivery error, but does not attempt to create a
3701 PH/11 The exinext utility has a -C option for testing purposes, but although
3702 the given file was scanned by exinext itself; it wasn't being passed on
3703 when Exim was called.
3705 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3706 an end-of-file indication when reading a command response.
3708 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3709 compiled. In many other places in Exim, IPv6 addresses are always
3710 recognized, so I have changed this. It also means that IPv4 domain
3711 literals of the form [IPV4:n.n.n.n] are now always recognized.
3713 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3714 used if the router is not running as root, for example, when verifying at
3715 ACL time, or when using -bh. The debugging output from this situation was
3716 non-existent - all you got was a failure to exec. I have made two
3719 (a) Failures to set uid/gid, the current directory, or a process leader
3720 in a subprocess such as that created by queryprogram now generate
3721 suitable debugging output when -d is set.
3723 (b) The queryprogram router detects when it is not running as root,
3724 outputs suitable debugging information if -d is set, and then runs
3725 the subprocess without attempting to change uid/gid.
3727 PH/15 Minor change to Makefile for building test_host (undocumented testing
3730 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3731 additional section of a DNS packet that returns MX or SRV records.
3732 Instead, it always explicitly searches for A/AAAA records. This avoids
3733 major problems that occur when a DNS server includes only records of one
3734 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3735 fixed another bug: if SRV records were looked up and the corresponding
3736 address records were *not* found in the additional section, the port
3737 values from the SRV records were lost.
3739 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3740 using the correct key (the original address) when searching the retry
3741 rules in order to find which one to use for generating the retry hint.
3743 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3744 adding the default one. Similarly, if it contains a Reply-To: header, the
3745 errors_reply_to option, if set, is not used.
3747 PH/19 When calculating a retry time, Exim used to measure the "time since
3748 failure" by looking at the "first failed" field in the retry record. Now
3749 it does not use this if it is later than than the arrival time of the
3750 message. Instead it uses the arrival time. This makes for better
3751 behaviour in cases where some deliveries succeed, thus re-setting the
3752 "first failed" field. An example is a quota failure for a huge message
3753 when small messages continue to be delivered. Without this change, the
3754 "time since failure" will always be short, possible causing more frequent
3755 delivery attempts for the huge message than are intended.
3756 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3758 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3759 $1, $2, $3) because the numerical variables can be reset during some
3760 expansion items (e.g. "match"), thereby losing the authentication data.
3762 PH/21 Make -bV show the size of off_t variables so that the test suite can
3763 decide whether to run tests for quotas > 2G.
3765 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3766 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3767 filecount value is greater than 2G or if a quota value is greater than 2G
3768 on a system where the size of off_t is not greater than 4, a panic error
3771 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3772 never match. The debug and -bh output now contains an explicit error
3773 message indicating a malformed IPv4 address or mask.
3775 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3776 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3777 PH/23 above applies.
3779 PH/25 Do not write to syslog when running in the test harness. The only
3780 occasion when this arises is a failure to open the main or panic logs
3781 (for which there is an explicit test).
3783 PH/26 Added the /no_tell option to "control=freeze".
3785 PH/27 If a host name lookup failed very early in a connection, for example, if
3786 the IP address matched host_lookup and the reverse lookup yielded a name
3787 that did not have a forward lookup, an error message of the form "no IP
3788 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3789 could be logged. Now it outputs the IP address instead of "NULL".
3791 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3792 allows the sender and the authenticated sender to be set when
3793 submitting a message from within Exim. Since child_open_exim() is
3794 documented for local_scan(), the new function should be too.
3796 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3797 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3798 results in an empty string is now treated as unset.
3800 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3802 PH/31 Added sender_verify_fail logging option.
3804 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3805 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3806 tidied the source and removed it altogether.
3808 PH/33 When a queue run was abandoned because the load average was too high, a
3809 log line was always written; now it is written only if the queue_run log
3810 selector is set. In addition, the log line for abandonment now contains
3811 information about the queue run such as the pid. This is always present
3812 in "start" and "stop" lines but was omitted from the "abandon" line.
3814 PH/34 Omit spaces between a header name and the colon in the error message that
3815 is given when verify = headers_syntax fails (if there are lots of them,
3816 the message gets confusing).
3818 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3819 names, as there are now some PTR records that contain slashes. This check
3820 is only to protect against broken name servers that fall over on strange
3821 characters, so the fact that it applies to all lookups doesn't matter.
3823 PH/36 Now that the new test suite is complete, we can remove some of the
3824 special code in Exim that was needed for the old test suite. For example,
3825 sorting DNS records because real resolvers return them in an arbitrary
3826 order. The new test suite's fake resolver always returns records in the
3829 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3830 bounces) except when queue_only is set, to avoid logging races between
3831 the different processes.
3833 PH/38 Panic-die if .include specifies a non-absolute path.
3835 PH/39 A tweak to the "H" retry rule from its user.
3837 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3838 a label. They prevented compilation on older perls.
3840 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3841 a warning to be raised on newish perls.
3843 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3844 on queue. Changes to match documented behaviour of showing count of
3845 messages matching specified criteria.
3847 PH/40 Changed the default ident timeout from 30s to 5s.
3849 PH/41 Added support for the use of login_cap features, on those BSD systems
3850 that have them, for controlling the resources used by pipe deliveries.
3852 PH/42 The content-scanning code uses fopen() to create files in which to put
3853 message data. Previously it was not paying any attention to the mode of
3854 the files. Exim runs with umask(0) because the rest of the code creates
3855 files with open(), and sets the required mode explicitly. Thus, these
3856 files were ending up world-writeable. This was not a big issue, because,
3857 being within the spool directory, they were not world-accessible. I have
3858 created a function called modefopen, which takes an additional mode
3859 argument. It sets umask(777), creates the file, chmods it to the required
3860 mode, then resets the umask. All the relevant calls to fopen() in the
3861 content scanning code have been changed to use this function.
3863 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3864 to 24 hours. This avoids potential overflow problems when processing G
3865 and H retry rules. I suspect nobody ever tinkers with this value.
3867 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3869 PH/45 When the plaintext authenticator is running as a client, the server's
3870 challenges are checked to ensure they are valid base64 strings. By
3871 default, the authentication attempt is cancelled if an invalid string is
3872 received. Setting client_ignore_invalid_base64 true ignores these errors.
3873 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3874 they are received. Thus, the responses can be made to depend on the
3875 challenges. If an invalid string is ignored, an empty string is placed in
3878 PH/46 Messages that are created by the autoreply transport now contains a
3879 References: header, in accordance with RFCs 2822 and 3834.
3881 PH/47 Added authenticated_sender_force to the smtp transport.
3883 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3885 PH/49 Installed latest patch from the Sieve maintainer.
3887 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3888 was also set, the appendfile transport was unnecessarily scanning a
3889 directory of message files (e.g. for maildir delivery) to find the count
3890 of files (along with the size), even though it did not need this
3891 information. It now does the scan only if it needs to find either the
3892 size of the count of files.
3894 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3896 PH/52 Two bugs concerned with error handling when the smtp transport is
3899 (i) Exim was not creating retry information for temporary errors given
3900 for individual recipients after the DATA command when the smtp transport
3901 was used in LMTP mode. This meant that they could be retried too
3902 frequently, and not timed out correctly.
3904 (ii) Exim was setting the flag that allows error details to be returned
3905 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3906 recipients that were returned after the DATA command.
3908 PH/53 This is related to PH/52, but is more general: for any failing address,
3909 when detailed error information was permitted to be returned to the
3910 sender, but the error was temporary, then after the final timeout, only
3911 "retry timeout exceeded" was returned. Now it returns the full error as
3912 well as "retry timeout exceeded".
3914 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3915 do this, and (what is worse) MTAs that accept it.
3917 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3918 will now be deprecated.
3920 PH/56 New os.c-cygwin from the Cygwin maintainer.
3922 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3923 formats (previously only available in exim formats via -bpr, -bpru,
3924 and -bpra. Now also available in native and exiqgrep formats)
3926 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3927 with very large, slow to parse queues
3929 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3931 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3933 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3934 responses to authentication challenges, though it was showing the
3935 challenges; (ii) I've removed the CR characters from the debug output for
3938 PH/58 Allow for the insertion of a newline as well as a space when a string
3939 is turned into more than one encoded-word during RFC 2047 encoding. The
3940 Sieve code now uses this.
3942 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3943 data_4xx, lost_connection, tls_required.
3945 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3946 message was being sent as an SMTP response.
3948 PH/61 Add -l and -k options to exicyclog.
3950 PH/62 When verifying, if an address was redirected to one new address, so that
3951 verification continued, and the new address failed or deferred after
3952 having set something in $address_data, the value of $address_data was not
3953 passed back to the ACL. This was different to the case when no
3954 redirection occurred. The value is now passed back in both cases.
3956 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3957 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3958 use login_cap.h, so on its own it isn't the distinguishing feature. The
3959 new name refers directly to the setclassresources() function.
3961 PH/65 Added configuration files for NetBSD3.
3963 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3965 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3966 is preferred over IPv4.
3968 PH/68 The bounce_return_message and bounce_return_body options were not being
3969 honoured for bounces generated during the reception of non-SMTP messages.
3970 In particular, this applied to messages rejected by the ACL. This bug has
3971 been fixed. However, if bounce_return_message is true and bounce_return_
3972 body is false, the headers that are returned for a non-SMTP message
3973 include only those that have been read before the error was detected.
3974 (In the case of an ACL rejection, they have all been read.)
3976 PH/69 The HTML version of the specification is now built in a directory called
3977 spec_html instead of spec.html, because the latter looks like a path with
3978 a MIME-type, and this confuses some software.
3980 PH/70 Catch two compiler warnings in sieve.c.
3982 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3983 function verify_get_ident() calls ip_connect() to connect a socket, but
3984 if the "connect()" function timed out, ip_connect() used to close the
3985 socket. However, verify_get_ident() also closes the socket later, and in
3986 between Exim writes to the log, which may get opened at this point. When
3987 the socket was closed in ip_connect(), the log could get the same file
3988 descriptor number as the socket. This naturally causes chaos. The fix is
3989 not to close the socket in ip_connect(); the socket should be closed by
3990 the function that creates it. There was only one place in the code where
3991 this was missing, in the iplookup router, which I don't think anybody now
3992 uses, but I've fixed it anyway.
3994 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3995 well as to direct DNS lookups. Otherwise the handling of names in host
3996 lists is inconsistent and therefore confusing.
4002 PH/01 Two changes to the default runtime configuration:
4004 (1) Move the checks for relay_from_hosts and authenticated clients from
4005 after to before the (commented out) DNS black list checks.
4007 (2) Add control=submission to the relay_from_hosts and authenticated
4008 clients checks, on the grounds that messages accepted by these
4009 statements are most likely to be submissions.
4011 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4013 (1) Generate an error if the third argument for the ${prvs expansion is
4016 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4019 (3) Reset the variables that are obtained from the first argument of
4020 ${prvscheck and used in the second argument before leaving the code,
4021 because their memory is reclaimed, so using them afterwards may do
4024 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4025 one (it's much easier than Tom thought :-).
4027 (5) Because of (4), we can now allow for the use of $prvscheck_result
4028 inside the third argument.
4030 PH/03 For some reason, the default setting of PATH when running a command from
4031 a pipe transport was just "/usr/bin". I have changed it to
4034 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4035 anything to be listed in the output from -bV.
4037 PH/05 When a filter generated an autoreply, the entire To: header line was
4038 quoted in the delivery log line, like this:
4040 => >A.N.Other <ano@some.domain> <original@ddress> ...
4042 This has been changed so that it extracts the operative address. There
4043 may be more than one such address. If so, they are comma-separated, like
4046 => >ano@some.domain,ona@other.domain <original@ddress> ...
4048 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4049 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4050 not being looked up in the rDNS to get a host name, Exim was showing the
4051 IP address twice in Received: lines, even though the IP addresses were
4052 identical. For example:
4054 Received: from [1.2.3.4] (helo=[1.2.3.4])
4056 However, if the real host name was known, it was omitting the HELO data
4057 if it matched the actual IP address. This has been tidied up so that it
4058 doesn't show the same IP address twice.
4060 PH/07 When both +timestamp and +memory debugging was on, the value given by
4061 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4062 called by the debug printing, thereby overwriting the timestamp buffer.
4063 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4065 PH/08 When the original message was included in an autoreply transport, it
4066 always said "this is a copy of the message, including all the headers",
4067 even if body_only or headers_only was set. It now gives an appropriate
4070 PH/09 Applied a patch from the Sieve maintainer which:
4072 o fixes some comments
4073 o adds the (disabled) notify extension core
4074 o adds some debug output for the result of if/elsif tests
4075 o points to the current vacation draft in the documentation
4076 and documents the missing references header update
4080 o fixes a bug in processing the envelope test (when testing
4081 multiple envelope elements, the last element determined the
4084 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4085 Electronic Mail") by including:
4087 Auto-submitted: auto-generated
4089 in the messages that it generates (bounce messages and others, such as
4090 warnings). In the case of bounce messages for non-SMTP messages, there was
4091 also a typo: it was using "Auto_submitted" (underscore instead of
4092 hyphen). Since every message generated by Exim is necessarily in response
4093 to another message, thes have all been changed to:
4095 Auto-Submitted: auto-replied
4097 in accordance with these statements in the RFC:
4099 The auto-replied keyword:
4101 - SHOULD be used on messages sent in direct response to another
4102 message by an automatic process,
4104 - MUST NOT be used on manually-generated messages,
4106 - MAY be used on Delivery Status Notifications (DSNs) and Message
4107 Disposition Notifications (MDNs),
4109 - MUST NOT be used on messages generated by automatic or periodic
4110 processes, except for messages which are automatic responses to
4113 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4114 to the default Received: header definition.
4116 PH/12 Added log selector acl_warn_skipped (default on).
4118 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4119 variables because (a) they are in the wrong storage pool and (b) even if
4120 they were copied, it wouldn't work properly because of the caching.
4122 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4123 checking when decoding. Apparently there are clients that generate
4124 overlong encoded strings. Why am I not surprised?
4126 PH/15 If the first argument of "${if match_address" was not empty, but did not
4127 contain an "@" character, Exim crashed. Now it writes a panic log message
4128 and treats the condition as false.
4130 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4132 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4133 extension "envelope-auth". The code is finished and in agreement with
4134 other implementations, but there is no documentation so far and in fact,
4135 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4136 not changing the active code.
4138 Print executed "if" and "elsif" statements when debugging is used. This
4139 helps a great deal to understand what a filter does.
4141 Document more things not specified clearly in RFC3028. I had all this
4142 sorted out, when out of a sudden new issues came to my mind. Oops."
4144 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4147 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4148 canonical form (as documented). However, after a host name lookup from
4149 the IP address, check_host() was doing a simple string comparison with
4150 addresses acquired from the DNS when checking that the found name did
4151 have the original IP as one of its addresses. Since any found IPv6
4152 addresses are likely to be in abbreviated form, the comparison could
4153 fail. Luckily, there already exists a function for doing the comparison
4154 by converting both addresses to binary, so now that is used instead of
4155 the text comparison.
4157 PH/20 There was another similar case to PH/19, when a complete host name was
4158 given in a host list; looking up its IP address could give an abbreviated
4159 form, whereas the current host's name might or might not be abbreviated.
4160 The same fix has been applied.
4166 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4167 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4170 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4171 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4173 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4175 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4176 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4177 Also, for "Auto-Submitted", treat anything other than "no" as
4178 non-personal, in accordance with RFC 3834. (Previously it treated
4179 anything starting "auto-" as non-personal.)
4181 TF/01 The control=submission/name=... option had a problem with syntax
4182 errors if the name included a slash character. The /name= option
4183 now slurps the rest of the string, so it can include any characters
4184 but it must come last in the list of options (after /sender_retain
4187 PH/05 Some modifications to the interface to the fake nameserver for the new
4195 TK/01 Added the "success_on_redirect" address verification option. See
4196 NewStuff for rationale and an example.
4198 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4200 PH/02 Patch to exigrep to allow it to work on syslog lines.
4202 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4203 fread() to skip over the body file's header line, because in Cygwin the
4204 header line is locked and is inaccessible.
4206 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4207 co-exist for some time) to make it clear that it is the Exim ID that is
4208 referenced, not the Message-ID: header line.
4210 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4211 string_format() function, because snprintf() does not exist on all
4214 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4215 ${stat: expansion item.
4217 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4218 protocol synchronization error", to keep the pedants happy.
4220 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4221 well as for IRIX systems, when gcc is being used. See the host.c source
4224 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4226 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4229 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4230 transport and to the smtp transport in LMTP mode.
4232 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4234 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4235 run a filter in a subprocess. This could lead to confusion in subsequent
4236 lookups in the parent process. There should also be a search_tidyup() at
4237 the end of the subprocess.
4239 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4240 only if the host matched helo_try_verify_hosts, which caused the
4241 verification to occur when the EHLO/HELO command was issued. The ACL just
4242 tested the remembered result. Now, if a previous verification attempt has
4243 not happened, "verify = helo" does it there and then.
4245 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4247 TK/03 Fix log output including CR from clamd.
4249 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4250 exist provoked a memory error which could cause a segfault.
4252 PH/15 Installed PCRE 6.2
4254 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4256 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4257 of the problem. Specifically, suggested +O2 rather than +O1 for the
4260 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4262 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4263 record, the port was not correctly logged when the outgoing_port log
4264 selector was set (it logged the transort's default port).
4266 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4267 fallback_hosts, and "hosts" in the smtp transport.
4269 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4270 host errors such as "Connection refused".
4272 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4273 authentication with radiusclient 0.4.9:
4275 - Error returned from rc_read_config was caught wrongly
4276 - Username/password not passed on to radius server due to wrong length.
4278 The presumption is that some radiusclient API changes for 4.51/PH/17
4279 were not taken care of correctly. The code is still untested by me (my
4280 Linux distribution still has 0.3.2 of radiusclient), but it was
4281 contributed by a Radius user.
4283 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4284 expanding the "port" option of the smtp transport.
4286 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4287 while reading a MIME header. Thanks to Tom Hughes for a patch.
4289 PH/24 Include config.h inside local_scan.h so that configuration settings are
4292 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4293 that in an ACL for RCPT (for example), you can examine exactly what was
4296 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4297 commands, but it was not correctly comparing the address with the actual
4298 client host address. Thus, it would show the EHLO address in Received:
4299 header lines when this was not necessary.
4301 PH/27 Added the % operator to ${eval:}.
4303 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4304 it should be ignoring failures (because with -C, for example, it has lost
4305 privilege). It wasn't ignoring creation failures other than "already
4308 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4311 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4312 because an input line was too long, either on its own, or by virtue of
4313 too many continuations, the temporary file was not being removed, and the
4314 return code was incorrect.
4316 PH/31 Missing "BOOL" in function definition in filtertest.c.
4318 PH/32 Applied Sieve patches from the maintainer.
4320 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4322 PH/33 Added "verify = not_blind".
4324 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4325 Local/Makefile (with some defaults set). These are used in built scripts
4326 such as exicyclog, but they have never been used in the exim_install
4327 script (though there are many overriding facilities there). I have
4328 arranged that the exim_install script now takes note of these two
4331 PH/35 Installed configuration files for Dragonfly.
4333 PH/36 When a locally submitted message by a trusted user did not contain a
4334 From: header, and the sender address was obtained from -f or from an SMTP
4335 MAIL command, and the trusted user did not use -F to supply a sender
4336 name, $originator_name was incorrectly used when constructing a From:
4337 header. Furthermore, $originator_name was used for submission mode
4338 messages from external hosts without From: headers in a similar way,
4339 which is clearly wrong.
4341 PH/37 Added control=suppress_local_fixups.
4343 PH/38 When log_selector = +received_sender was set, and the addition of the
4344 sender made the log line's construction buffer exactly full, or one byte
4345 less than full, an overflow happened when the terminating "\n" was
4348 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4349 when the result of a list match is failure because a DNS lookup failed.
4351 PH/40 RM_COMMAND is now used in the building process.
4353 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4354 the "build-* directories that it finds.
4356 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4357 address in a domain literal was a prefix of an interface address.
4359 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4360 when verifying a sender address, unless rewrite_headers is false.
4362 PH/44 Wrote a long comment about why errors_to addresses are verified as
4363 recipients, not senders.
4365 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4366 the ratelimit ACL was added.
4368 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4370 PH/46 Added extra information about PostgreSQL errors to the error string.
4372 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4373 avoiding the need to install special zones in a real server. This is
4374 backwards compatible; if it can't find the fake resolver, it drops back.
4375 Thus, both old and new test suites can be run.
4377 TF/02 Added util/ratelimit.pl
4379 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4380 clock is set back in time.
4382 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4383 Candler <B.Candler@pobox.com>.
4385 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4386 OK for addresses that are the result of redirections.
4388 PH/48 A number of further additions for the benefit of the new test suite,
4389 including a fake gethostbyname() that interfaces to the fake DNS resolver
4392 TF/06 The fix for widen_domains has also been applied to qualify_single and
4393 search_parents which are the other dnslookup options that can cause
4396 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4399 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4401 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4402 service" if it encountered an MX record with an empty target hostname.
4403 The message is now "an MX or SRV record indicated no SMTP service".
4405 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4406 if the DNS of the sending site is misconfigured. This is quite a
4407 common situation. This change restores the behaviour of treating a
4408 helo verification defer as a failure.
4410 PH/51 If self=fail was set on a router, the bounce message did not include the
4411 actual error message.
4417 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4419 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4420 command itself ended in error, the underlying message about the transport
4421 filter timeout was being overwritten with the pipe command error. Now the
4422 underlying error message should be appended to the second error message.
4424 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4426 PH/02 Reduce the amount of output that "make" produces by default. Full output
4427 can still be requested.
4429 PH/03 The warning log line about a condition test deferring for a "warn" verb
4430 was being output only once per connection, rather than after each
4431 occurrence (because it was using the same function as for successful
4432 "warn" verbs). This seems wrong, so I have changed it.
4434 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4435 it should not have, which might have caused a crash in the right
4436 circumstances, but probably never did.
4438 PH/04 Installed a modified version of Tony Finch's patch to make submission
4439 mode fix the return path as well as the Sender: header line, and to
4440 add a /name= option so that you can make the user's friendly name appear
4443 TF/03 Added the control = fakedefer ACL modifier.
4445 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4446 Mark Lowes for thorough testing.
4448 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4450 TK/03 Merged latest SRS patch from Miles Wilton.
4452 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4453 with the definition in sysexits.h (which is #included earlier).
4454 Fortunately, Exim does not actually use EX_OK. The code used to try to
4455 preserve the sysexits.h value, by assuming that macro definitions were
4456 scanned for macro replacements. I have been disabused of this notion,
4457 so now the code just undefines EX_OK before #including unistd.h.
4459 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4460 in the smtp transport. When a block could not be written in a single
4461 write() function, the timeout was being re-applied to each part-write.
4462 This seems wrong - if the receiver was accepting one byte at a time it
4463 would take for ever. The timeout is now adjusted when this happens. It
4464 doesn't have to be particularly precise.
4466 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4467 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4469 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4470 without a domain if the check to <postmaster@domain> fails.
4472 SC/01 Eximstats: added -xls and the ability to specify output files
4473 (patch written by Frank Heydlauf).
4475 SC/02 Eximstats: use FileHandles for outputting results.
4477 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4479 SC/04 Eximstats: fixed display of large numbers with -nvr option
4481 SC/05 Eximstats: fixed merging of reports with empty tables.
4483 SC/06 Eximstats: added the -include_original_destination flag
4485 SC/07 Eximstats: removed tabs and trailing whitespace.
4487 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4489 TK/06 MBOX spool code: Add real "From " MBOX separator line
4490 so the .eml file is really in mbox format (even though
4491 most programs do not really care). Patch from Alex Miller.
4493 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4494 The latter is generated from $received_to and is only set if the
4495 message has one envelope recipient. SA can use these headers,
4496 obviously out-of-the-box. Patch from Alex Miller.
4498 PH/08 The ${def test on a variable was returning false if the variable's
4499 value was "0", contrary to what the specification has always said!
4500 The result should be true unless the variable is empty.
4502 PH/09 The syntax error of a character other than { following "${if
4503 def:variable_name" (after optional whitespace) was not being diagnosed.
4504 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4505 accidental colon was present, for example, could give incorrect results.
4507 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4508 result is used (not including appendfile, where other changes are about
4511 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4512 This involved changing a lot of size variables from int to off_t. It
4513 should work with maildirs and everything.
4515 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4516 spamd dying while we are connected to it.
4518 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4521 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4522 function for BDB 4.3.
4524 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4526 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4527 the sources for PCRE have been re-arranged and more files are now
4530 PH/15 The code I had for printing potentially long long variables in PH/11
4531 above was not the best (it lost precision). The length of off_t variables
4532 is now inspected at build time, and an appropriate printing format (%ld
4533 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4534 to be "long long int" or "long int". This is needed for the internal
4535 formatting function string_vformat().
4537 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4538 the configuration file to be ":syslog", then the script "guesses" where
4539 the logs files are, rather than using the compiled in default. In our
4540 case the guess is not the same as the compiled default, so the script
4541 suddenly stopped working when I started to use syslog. The patch checks
4542 to see if log_file_path is "". If so, it attempts to read it from exim
4543 with no configuration file to get the compiled in version, before it
4544 falls back to the previous guessing code."
4546 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4547 implementing BATV in an Exim configuration. See NewStuff for the gory
4550 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4551 Makefile that are specific to HP-UX.
4553 PH/18 If the "use_postmaster" option was set for a recipient callout together
4554 with the "random" option, the postmaster address was used as the MAIL
4555 FROM address for the random test, but not for the subsequent recipient
4556 test. It is now used for both.
4558 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4559 patch removes a few documentation additions to RFC 3028, because the
4560 latest draft now contains them. It adds the new en;ascii-case comparator
4561 and a new error check for 8bit text in MIME parts. Comparator and
4562 require names are now matched exactly. I enabled the subaddress
4563 extension, but it is not well tested yet (read: it works for me)."
4565 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4566 rework some of the code of TK/09 above to avoid the hardwired use of
4567 "%lld" and "long long". Replaced the call to snprintf() with a call to
4570 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4571 records point to non-existent hosts", "retry timeout exceeded", and
4572 "retry time not reached for any host after a long failure period".
4574 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4575 experimental DomainKeys support:
4577 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4578 (2) On an error such as an illegally used "control", the wrong name for
4579 the control was given.
4581 These problems did NOT occur unless DomainKeys support was compiled.
4583 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4585 PH/24 Added ${if match_ip condition.
4587 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4588 databases so that it will be absolutely obvious if a crash occurs in the
4589 DB library. This is a regular occurrence (often caused by mis-matched
4592 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4593 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4594 picked up on a user's system that detects such things. There doesn't seem
4595 to be a gcc warning option for this - only an attribute that has to be
4596 put on the function's prototype. It seems that in Fedora Core 4 they have
4597 set this on a number of new functions. No doubt there will be more in due
4600 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4601 specify a transport. However, if an address that was verified by such a
4602 router was the subject of a callout, Exim crashed because it tried to
4603 read the rcpt_include_affixes from the non-existent transport. Now it
4604 just assumes that the setting of that option is false. This bug was
4605 introduced by 4.51/PH/31.
4607 PH/28 Changed -d+all to exclude +memory, because that information is very
4608 rarely of interest, but it makes the output a lot bigger. People tend to
4609 do -d+all out of habit.
4611 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4612 code in os-type was giving problems when libc.so lives in lib64, like on
4615 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4616 aren't the modern standard, and it seems that some systems' include files
4617 don't always have them. Exim was already checking for some of the newer
4618 ones like T_AAAA, and defining it itself. I've added checks for all the
4619 record types that Exim uses.
4621 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4622 not automatically generating a new one, as it is supposed to. This
4623 prevented TLS from working. If the file did exist, but contained invalid
4624 data, a new version was generated, as expected. It was only the case of a
4625 non-existent file that was broken.
4627 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4628 with a change in libdomainkeys > 0.64.
4630 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4631 from DNS. If the selector record carries the flag, it now has
4632 precedence over the domain-wide flag.
4634 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4636 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4637 the use of a port name that isn't defined in /etc/services) occurred, the
4638 message was deferred as in a normal delivery, and thus remained on the
4639 spool, instead of being failed because of the mua_wrapper setting. This
4640 is now fixed, and I tidied up some of the mua_wrapper messages at the
4643 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4644 the messages in a hash of arrays rather than using individual hashes.
4645 This is a bit cleaner and results in dramatic memory savings, albeit
4646 at a slight CPU cost.
4648 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4649 as requested by Marc Sherman.
4651 SC/10 Eximstats: added histograms for user specified patterns as requested
4654 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4656 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4657 fopen() in the content-scanning modules that did not already have it.
4663 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4664 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4666 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4668 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4670 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4671 to test Sieve filters that use "vacation".
4673 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4674 that changes the way the GnuTLS parameters are stored in the cache file.
4675 The new format can be generated externally. For backward compatibility,
4676 if the data in the cache doesn't make sense, Exim assumes it has read an
4677 old-format file, and it generates new data and writes a new file. This
4678 means that you can't go back to an older release without removing the
4681 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4682 work if there are any delivery delays because "one_time" forces the
4683 parent to be marked "delivered", so its unseen clone is never tried
4684 again. For this reason, Exim now forbids the simultaneous setting of
4687 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4688 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4689 entry if you want to know the details. The fix, however, neglected to
4690 consider the case when local delivery batching is involved. The test for
4691 "previously delivered" was not happening when checking to see if an
4692 address could be batched with a previous (undelivered) one; under
4693 certain circumstances this could lead to multiple deliveries to the same
4696 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4697 in its include files, and this causes problems building Exim.
4699 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4700 header_syntax) but Exim was just ignoring anything given after a slash.
4701 In particular, this caused confusion with an attempt to use "verify =
4702 reverse_host_lookup/defer_ok". An error is now given when options are
4703 supplied for verify items that do not have them. (Maybe reverse_host_
4704 lookup should have a defer_ok option, but that's a different point.)
4706 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4707 defined by RFC 821) to 2048, because there were problems with some AUTH
4708 commands, and RFC 1869 says the size should be increased for extended
4709 SMTP commands that take arguments.
4711 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4714 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4715 "unknown" error; now it says that the functionality isn't in the binary.
4717 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4718 an address' error message when a string expansion fails (syntax or
4719 whatever). Otherwise the password may appear in the log. Following change
4720 PH/42 below, there is no longer a chance of it appearing in a bounce
4723 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4725 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4726 of its addresses was ever tried. (Bugzilla bug #2.)
4728 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4729 the result incorrectly in the debug output. (It correctly added a newline
4730 to what was transported.)
4732 TF/01 Added $received_time.
4734 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4735 commented out examples of how to interface to a virus scanner and to
4736 SpamAssassin. Also added commented examples of av_scanner and
4737 spamd_address settings.
4739 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4740 and controls are allowed in which ACLs. There were a couple of minor
4741 errors. Some of the entries in the conditions table (which is a table of
4742 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4743 negation of where the condition IS allowed.
4745 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4747 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4748 header file does not have a version number, so I've had to invent a new
4749 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4750 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4751 radiusclient), but it was contributed by a Radius user.
4753 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4754 files or directories, for OpenSSL.
4756 PH/19 When an Exim process that is running as root has to create an Exim log
4757 file, it does so in a subprocess that runs as exim:exim so as to get the
4758 ownership right at creation (otherwise, other Exim processes might see
4759 the file with the wrong ownership). There was no test for failure of this
4760 fork() call, which would lead to the process getting stuck as it waited
4761 for a non-existent subprocess. Forks do occasionally fail when resources
4762 run out. I reviewed all the other calls to fork(); they all seem to check
4765 PH/20 When checking for unexpected SMTP input at connect time (before writing
4766 the banner), Exim was not dealing correctly with a non-positive return
4767 from the read() function. If the client had disconnected by this time,
4768 the result was a log entry for a synchronization error with an empty
4769 string after "input=" when read() returned zero. If read() returned -1
4770 (an event I could not check), uninitialized data bytes were printed.
4771 There were reports of junk text (parts of files, etc) appearing after
4774 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4776 PH/22 Added support for macro redefinition, and (re)definition in between
4777 driver and ACL definitions.
4779 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4780 forgetting to use the resulting value; it was using the unexpanded value.
4782 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4783 hadn't been configured. The fix is from Juergen Kreileder, who
4784 understands it better than I do:
4786 "Here's what I see happening with three configured cyrus_sasl
4787 authenticators configured (plain, login, cram-md5):
4789 On startup auth_cyrus_sasl_init() gets called for each of these.
4790 This means three calls to sasl_listmech() without a specified mech_list.
4791 => SASL tests which mechs of all available mechs actually work
4792 => three warnings about OTP not working
4793 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4795 With the patch, sasl_listmech() also gets called three times. But now
4796 SASL's mech_list option is set to the server_mech specified in the the
4797 authenticator. Or in other words, the answer from sasl_listmech()
4798 gets limited to just the mech you're testing for (which is different
4800 => the return list contains just 'plain' or 'login', 'cram-md5' or
4801 nothing depending on the value of ob->server_mech.
4803 I've just tested the patch: Authentication still works fine,
4804 unavailable mechs specified in the exim configuration are still
4805 caught, and the auth.log warnings about OTP are gone."
4807 PH/25 When debugging is enabled, the contents of the command line are added
4808 to the debugging output, even when log_selector=+arguments is not
4811 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4812 answer is "GNU", and only if the return is "GNU/something" is the answer
4815 PH/27 $acl_verify_message is now set immediately after the failure of a
4816 verification in an ACL, and so is available in subsequent modifiers. In
4817 particular, the message can be preserved by coding like this:
4819 warn !verify = sender
4820 set acl_m0 = $acl_verify_message
4822 Previously, $acl_verify_message was set only while expanding "message"
4823 and "log_message" when a very denied access.
4825 PH/28 Modified OS/os.c-Linux with
4827 -#ifndef OS_LOAD_AVERAGE
4828 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4830 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4831 nomenclature these days.)
4833 PH/29 Installed patch from the Sieve maintainer that adds the options
4834 sieve_useraddress and sieve_subaddress to the redirect router.
4836 PH/30 In these circumstances:
4837 . Two addresses routed to the same list of hosts;
4838 . First host does not offer TLS;
4839 . First host accepts first address;
4840 . First host gives temporary error to second address;
4841 . Second host offers TLS and a TLS session is established;
4842 . Second host accepts second address.
4843 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4844 and peerdn, if requested) that were in fact used only for the second
4847 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4848 attention to any local part prefix or suffix that was matched by the
4849 router that accepted the address. It now behaves in the same way as it
4850 does for delivery: the affixes are removed from the local part unless
4851 rcpt_include_affixes is set on the transport.
4853 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4854 timeout during the DATA phase of an incoming message.
4856 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4857 applied a patch sanctioned by the Sieve maintainer.
4859 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4860 the uid or gid is negative. A case of a negative gid caused this to be
4861 noticed. The fix allows for either to be negative.
4863 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4864 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4867 PH/36 The change PH/12 above was broken. Fixed it.
4869 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4870 the grounds that routing the same address twice would always produce the
4871 same answer. This might have been true once, but it is certainly no
4872 longer true now. Routing a child address may depend on the previous
4873 routing that produced that child. Some complicated redirection strategies
4874 went wrong when messages had multiple recipients, and made Exim's
4875 behaviour dependent on the order in which the addresses were given.
4877 I have moved the duplicate checking until after the routing is complete.
4878 Exim scans the addresses that are assigned to local and remote
4879 transports, and removes any duplicates. This means that more work will be
4880 done, as duplicates will always all be routed, but duplicates are
4881 presumably rare, so I don't expect this is of any significance.
4883 For deliveries to pipes, files, and autoreplies, the duplicate checking
4884 still happens during the routing process, since they are not going to be
4887 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4888 It corrects a timeout issue with spamd. This is Ian's comment: "The
4889 background is that sometimes spamd either never reads data from a
4890 connection it has accepted, or it never writes response data. The exiscan
4891 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4892 blindly assumes that writes won't block so it may never time out."
4894 PH/39 Allow G after quota size as well as K and M.
4896 PH/40 The value set for $authenticated_id in an authenticator may not contain
4897 binary zeroes or newlines because the value is written to log lines and
4898 to spool files. There was no check on this. Now the value is run through
4899 the string_printing() function so that such characters are converted to
4900 printable escape sequences.
4902 PH/41 $message_linecount is a new variable that contains the total number of
4903 lines in the message. Compare $body_linecount, which is the count for the
4906 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4907 bounce and delay warning messages, except in certain special cases, which
4910 (a) An SMTP error message from a remote host;
4911 (b) A message specified in a :fail: redirection;
4912 (c) A message specified in a "fail" command in a system filter;
4913 (d) A message specified in a FAIL return from the queryprogram router;
4914 (e) A message specified by the cannot_route_message router option.
4916 In these cases only, Exim does include the error details in bounce and
4917 warning messages. There are also a few cases where bland messages such
4918 as "unrouteable address" or "local delivery error" are given.
4920 PH/43 $value is now also set for the "else" part of a ${run expansion.
4922 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4923 being worked on, but at least Exim now implements the latest version to
4926 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4927 process to complete was treated as a delivery failure, a timeout while
4928 writing the message to the pipe was logged, but erroneously treated as a
4929 successful delivery. Such timeouts include transport filter timeouts. For
4930 consistency with the overall process timeout, these timeouts are now
4931 treated as errors, giving rise to delivery failures by default. However,
4932 there is now a new Boolean option for the pipe transport called
4933 timeout_defer, which, if set TRUE, converts the failures into defers for
4934 both kinds of timeout. A transport filter timeout is now identified in
4937 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4938 systems where "make" and "gmake" are different, calling "gmake" at top
4939 level broke things. I've arranged for the value of $(MAKE) to be passed
4940 from the Makefile to this script so that it can call the same version of
4944 A note about Exim versions 4.44 and 4.50
4945 ----------------------------------------
4947 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4948 changes of various kinds. As a consequence, a big documentation update was
4949 needed. This delayed the release for rather longer than seemed good, especially
4950 in the light of a couple of (minor) security issues. Therefore, the changes
4951 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4952 release. So 4.44 and 4.50 are in effect two different branches that both start
4955 I have left the 4.50 change log unchanged; it contains all the changes since
4956 4.43. The change log for 4.44 is below; many of its items are identical to
4957 those for 4.50. This seems to be the most sensible way to preserve the
4958 historical information.
4964 1. Minor wording change to the doc/README.SIEVE file.
4966 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4967 computation of the current number of files was incorrect.
4969 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4970 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4973 4. Give more explanation in the error message when the command for a transport
4974 filter fails to execute.
4976 5. There are several places where Exim runs a non-Exim command in a
4977 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4978 was being done only for the command run by the queryprogram router. It is
4979 now done for all such subprocesses. The other cases are: ${run, transport
4980 filters, and the commands run by the lmtp and pipe transports.
4982 6. Added CONFIGURE_GROUP build-time option.
4984 7. Some older OS have a limit of 256 on the maximum number of file
4985 descriptors. Exim was using setrlimit() to set 1000 as a large value
4986 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4987 systems. I've change it so that if it can't get 1000, it tries for 256.
4989 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4990 was an oversight, and furthermore, ever since the addition of extra
4991 controls (e.g. 4.43/32), the checks on when to allow different forms of
4992 "control" were broken. There should now be diagnostics for all cases when a
4993 control that does not make sense is encountered.
4995 9. Added the /retain_sender option to "control=submission".
4997 10. $recipients is now available in the predata ACL (oversight).
4999 11. Tidy the search cache before the fork to do a delivery from a message
5000 received from the command line. Otherwise the child will trigger a lookup
5001 failure and thereby defer the delivery if it tries to use (for example) a
5002 cached ldap connection that the parent has called unbind on.
5004 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5005 of $address_data from the recipient verification was clobbered by the
5006 sender verification.
5008 13. The value of address_data from a sender verification is now available in
5009 $sender_address_data in subsequent conditions in the ACL statement.
5011 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5013 15. Added a new option "connect=<time>" to callout options, to set a different
5016 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5017 was its contents. (It was OK if the option was not defined at all.)
5019 17. A "Completed" log line is now written for messages that are removed from
5020 the spool by the -Mrm option.
5022 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5023 information about exactly what failed.
5025 19. Added -dd to debug only the daemon process.
5027 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5028 handles timeouts, both on the server side and network timeouts. Renamed the
5029 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5031 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5032 It is now set to "smtps".
5034 22. $host_address is now set to the target address during the checking of
5035 ignore_target_hosts.
5037 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5038 being passed; this would have caused $sender_host_name to have been used if
5039 matching the list had actually called for a host name (not very likely,
5040 since this list is usually IP addresses). A host name is now passed as
5043 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5044 code that specifies a non-restarting handler (typically sigaction() in
5045 modern systems) in an attempt to fix a rare and obscure crash bug.
5047 25. Narrowed the window for a race in the daemon that could cause it to ignore
5048 SIGCHLD signals. This is not a major problem, because they are used only to
5049 wake it up if nothing else does.
5051 26. A malformed maildirsize file could cause Exim to calculate negative values
5052 for the mailbox size or file count. Odd effects could occur as a result.
5053 The maildirsize information is now recalculated if the size or filecount
5056 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5057 support for a long time. Removed HAVE_SYS_VFS_H.
5059 28. Installed the latest version of exipick from John Jetmore.
5061 29. In an address list, if the pattern was not a regular expression, an empty
5062 subject address (from a bounce message) matched only if the pattern was an
5063 empty string. Non-empty patterns were not even tested. This was the wrong
5064 because it is perfectly reasonable to use an empty address as part of a
5065 database query. An empty address is now tested by patterns that are
5066 lookups. However, all the other forms of pattern expect the subject to
5067 contain a local part and a domain, and therefore, for them, an empty
5068 address still always fails if the pattern is not itself empty.
5070 30. Exim went into a mad DNS loop when attempting to do a callout where the
5071 host was specified on an smtp transport, and looking it up yielded more
5072 than one IP address.
5074 31. Re-factored the code for checking spool and log partition space into a
5075 function that finds that data and another that does the check. The former
5076 is then used to implement four new variables: $spool_space, $log_space,
5077 $spool_inodes, and $log_inodes.
5079 32. The RFC2047 encoding function was originally intended for short strings
5080 such as real names; it was not keeping to the 75-character limit for
5081 encoded words that the RFC imposes. It now respects the limit, and
5082 generates multiple encoded words if necessary. To be on the safe side, I
5083 have increased the buffer size for the ${rfc2047: expansion operator from
5086 33. It is now permitted to omit both strings after an "if" condition; if the
5087 condition is true, the result is "true". As before, when the second string
5088 is omitted, a false condition yields an empty string. This makes it less
5089 cumbersome to write custom ACL and router conditions.
5091 34. Failure to deliver a bounce message always caused it to be frozen, even if
5092 there was an errors_to setting on the router. The errors_to setting is now
5095 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5096 canonical form (fully expanded) before being placed in
5097 $sender_host_address.
5099 36. The table in the code that translates DNS record types into text (T_A to
5100 "A" for instance) was missing entries for NS and CNAME. It is just possible
5101 that this could have caused confusion if both these types were looked up
5102 for the same domain, because the text type is used as part of Exim's
5103 per-process caching. But the chance of anyone hitting this buglet seems
5106 37. The dnsdb lookup has been extended in a number of ways.
5108 (1) There is a new type, "zns", which walks up the domain tree until it
5109 finds some nameserver records. It should be used with care.
5111 (2) There is a new type, "mxh", which is like "mx" except that it returns
5112 just the host names, not the priorities.
5114 (3) It is now possible to give a list of domains (or IP addresses) to be
5115 looked up. The behaviour when one of the lookups defers can be
5116 controlled by a keyword.
5118 (4) It is now possible to specify the separator character for use when
5119 multiple records are returned.
5121 38. The dnslists ACL condition has been extended: it is now possible to supply
5122 a list of IP addresses and/or domains to be looked up in a particular DNS
5125 39. Added log_selector=+queue_time_overall.
5127 40. When running the queue in the test harness, wait just a tad after forking a
5128 delivery process, to get repeatability of debugging output.
5130 41. Include certificate and key file names in error message when GnuTLS fails
5131 to set them up, because the GnuTLS error message doesn't include the name
5132 of the failing file when there is a problem reading it.
5134 42. Allow both -bf and -bF in the same test run.
5136 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5138 44. The "Exiscan patch" is now merged into the mainline Exim source.
5140 45. Sometimes the final signoff response after QUIT could fail to get
5141 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5142 < 0 before doing a fflush(). This bug looks as though it goes back to the
5143 introduction of TLS in release 3.20, but "sometimes" must have been rare
5144 because the tests only now provoked it.
5146 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5147 (this can affect the format of dates).
5149 47. exim_tidydb, when checking for the continued existence of a message for
5150 which it has found a message-specific retry record, was not finding
5151 messages that were in split spool directories. Consequently, it was
5152 deleting retry records that should have stayed in existence.
5154 48. Steve fixed some bugs in eximstats.
5156 49. The SPA authentication driver was not abandoning authentication and moving
5157 on to the next authenticator when an expansion was forced to fail,
5158 contradicting the general specification for all authenticators. Instead it
5159 was generating a temporary error. It now behaves as specified.
5161 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5162 (the order specifies the preference for clients). The order is now AES256,
5163 AES128, 3DES, ARCFOUR128.
5165 51. Small patch to Sieve code - explicitly set From: when generating an
5168 52. Exim crashed if a remote delivery caused a very long error message to be
5169 recorded - for instance if somebody sent an entire SpamAssassin report back
5170 as a large number of 550 error lines. This bug was coincidentally fixed by
5171 increasing the size of one of Exim's internal buffers (big_buffer) that
5172 happened as part of the Exiscan merge. However, to be on the safe side, I
5173 have made the code more robust (and fixed the comments that describe what
5176 53. Now that there can be additional text after "Completed" in log lines (if
5177 the queue_time_overall log selector is set), a one-byte patch to exigrep
5178 was needed to allow it to recognize "Completed" as not the last thing in
5181 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5182 patch that reportedly fixes this has been added. I am not expert enough to
5183 create a test for it. This is what the patch creator wrote:
5185 "I found a little strange behaviour of ldap code when working with
5186 Windows 2003 AD Domain, where users was placed in more than one
5187 Organization Units. When I tried to give exim partial DN, the exit code
5188 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5189 But simultaneously result of request was absolutely normal ldap result,
5190 so I produce this patch..."
5192 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5193 REFERENCE, so I have modified the code to exclude the patch when that macro
5196 55. Some experimental protocols are using DNS PTR records for new purposes. The
5197 keys for these records are domain names, not reversed IP addresses. The
5198 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5199 leaves it alone. Component reversal etc. now happens only for IP addresses.
5202 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5204 57. Double the size of the debug message buffer (to 2048) so that more of very
5205 long debug lines gets shown.
5207 58. The exicyclog utility now does better if the number of log files to keep
5208 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5210 59. Two changes related to the smtp_active_hostname option:
5212 (1) $smtp_active_hostname is now available as a variable.
5213 (2) The default for smtp_banner uses $smtp_active_hostname instead
5214 of $primary_hostname.
5216 60. The host_aton() function is supposed to be passed a string that is known
5217 to be a valid IP address. However, in the case of IPv6 addresses, it was
5218 not checking this. This is a hostage to fortune. Exim now panics and dies
5219 if the condition is not met. A case was found where this could be provoked
5220 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5221 components; fortuitously, this particular loophole had already been fixed
5222 by change 4.50/55 above.
5224 If there are any other similar loopholes, the new check in host_aton()
5225 itself should stop them being exploited. The report I received stated that
5226 data on the command line could provoke the exploit when Exim was running as
5227 exim, but did not say which command line option was involved. All I could
5228 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5229 running as the user.
5232 61. There was a buffer overflow vulnerability in the SPA authentication code
5233 (which came originally from the Samba project). I have added a test to the
5234 spa_base64_to_bits() function which I hope fixes it.
5237 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5238 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5240 63. The daemon start-up calls getloadavg() while still root for those OS that
5241 need the first call to be done as root, but it missed one case: when
5242 deliver_queue_load_max is set with deliver_drop_privilege. This is
5243 necessary for the benefit of the queue runner, because there is no re-exec
5244 when deliver_drop_privilege is set.
5246 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5247 This has been fixed.
5249 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5250 was in use, was not putting the data itself into the right store pool;
5251 consequently, it could be overwritten for a subsequent message in the same
5252 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5255 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5257 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5258 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5259 the function were treating the return as a boolean value, which happened to
5260 work because 0=false and not-0=true, but is not correct code.
5262 68. The host_aton() function was not handling scoped IPv6 addresses (those
5263 with, for example, "%eth0" on the end) correctly.
5265 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5266 negated items (that is, ~something) in unsigned ints. Some compilers
5267 apparently mutter when there is no cast.
5269 70. If an address verification called from an ACL failed, and did not produce a
5270 user-specific message (i.e. there was only a "system" message), nothing was
5271 put in $acl_verify_message. In this situation, it now puts the system
5274 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5275 session; change 4.31/43 added the unwanted input to the log line - except
5276 that it did not do this in the start of session case. It now does.
5278 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5279 This is wrong and can cause the other end to generate a synchronization
5280 error if it is another Exim or anything else that does the synchronization
5281 check. A QUIT command is no longer sent after a timeout.
5283 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5284 during host lookups.
5286 74. The defer_ok option of callout verification was not working if it was used
5287 when verifying addresses in header lines, that is, for this case:
5289 verify = header_sender/callout=defer_ok
5291 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5292 those file descriptors could be used for SMTP connections. If anything
5293 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5294 could be sent to the SMTP client, causing chaos. The daemon now opens
5295 stdin, stdout, and stderr to /dev/null when it puts itself into the
5298 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5299 log by default. The user can override this with suitable Perl magic.
5301 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5302 the log message when discard triggers, was not working for the DATA ACL or
5303 for the non-SMTP ACL.
5305 78. Error message wording change in sieve.c.
5307 79. If smtp_accept_max_per_host was set, the number of connections could be
5308 restricted to fewer than expected, because the daemon was trying to set up
5309 a new connection before checking whether the processes handling previous
5310 connections had finished. The check for completed processes is now done
5311 earlier. On busy systems, this bug wouldn't be noticed because something
5312 else would have woken the daemon, and it would have reaped the completed
5315 80. If a message was submitted locally by a user whose login name contained one
5316 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5317 It caused a spool format error. I have fixed the spool reading code. A
5318 related problem was that the "from" clause in the Received: line became
5319 illegal because of the space(s). It is now covered by ${quote_local_part.
5321 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5324 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5326 83. Patch from Sieve maintainer to fix unterminated string problem in
5327 "vacation" handling.
5329 84. Some minor changes to the Linux configuration files to help with other
5330 OS variants using glibc.
5332 85. One more patch for Sieve to update vacation handling to latest spec.
5335 ----------------------------------------------------
5336 See the note above about the 4.44 and 4.50 releases.
5337 ----------------------------------------------------
5343 1. Change 4.43/35 introduced a bug that caused file counts to be
5344 incorrectly computed when quota_filecount was set in an appendfile
5347 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5348 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5351 3. Give more explanation in the error message when the command for a transport
5352 filter fails to execute.
5354 4. There are several places where Exim runs a non-Exim command in a
5355 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5356 was being done only for the command run by the queryprogram router. It is
5357 now done for all such subprocesses. The other cases are: ${run, transport
5358 filters, and the commands run by the lmtp and pipe transports.
5360 5. Some older OS have a limit of 256 on the maximum number of file
5361 descriptors. Exim was using setrlimit() to set 1000 as a large value
5362 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5363 systems. I've change it so that if it can't get 1000, it tries for 256.
5365 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5366 was an oversight, and furthermore, ever since the addition of extra
5367 controls (e.g. 4.43/32), the checks on when to allow different forms of
5368 "control" were broken. There should now be diagnostics for all cases when a
5369 control that does not make sense is encountered.
5371 7. $recipients is now available in the predata ACL (oversight).
5373 8. Tidy the search cache before the fork to do a delivery from a message
5374 received from the command line. Otherwise the child will trigger a lookup
5375 failure and thereby defer the delivery if it tries to use (for example) a
5376 cached ldap connection that the parent has called unbind on.
5378 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5379 of $address_data from the recipient verification was clobbered by the
5380 sender verification.
5382 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5383 was its contents. (It was OK if the option was not defined at all.)
5385 11. A "Completed" log line is now written for messages that are removed from
5386 the spool by the -Mrm option.
5388 12. $host_address is now set to the target address during the checking of
5389 ignore_target_hosts.
5391 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5392 being passed; this would have caused $sender_host_name to have been used if
5393 matching the list had actually called for a host name (not very likely,
5394 since this list is usually IP addresses). A host name is now passed as
5397 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5398 code that specifies a non-restarting handler (typically sigaction() in
5399 modern systems) in an attempt to fix a rare and obscure crash bug.
5401 15. Narrowed the window for a race in the daemon that could cause it to ignore
5402 SIGCHLD signals. This is not a major problem, because they are used only to
5403 wake it up if nothing else does.
5405 16. A malformed maildirsize file could cause Exim to calculate negative values
5406 for the mailbox size or file count. Odd effects could occur as a result.
5407 The maildirsize information is now recalculated if the size or filecount
5410 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5411 support for a long time. Removed HAVE_SYS_VFS_H.
5413 18. Updated exipick to current release from John Jetmore.
5415 19. Allow an empty sender to be matched against a lookup in an address list.
5416 Previously the only cases considered were a regular expression, or an
5419 20. Exim went into a mad DNS lookup loop when doing a callout where the
5420 host was specified on the transport, if the DNS lookup yielded more than
5423 21. The RFC2047 encoding function was originally intended for short strings
5424 such as real names; it was not keeping to the 75-character limit for
5425 encoded words that the RFC imposes. It now respects the limit, and
5426 generates multiple encoded words if necessary. To be on the safe side, I
5427 have increased the buffer size for the ${rfc2047: expansion operator from
5430 22. Failure to deliver a bounce message always caused it to be frozen, even if
5431 there was an errors_to setting on the router. The errors_to setting is now
5434 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5435 canonical form (fully expanded) before being placed in
5436 $sender_host_address.
5438 24. Updated eximstats to version 1.33
5440 25. Include certificate and key file names in error message when GnuTLS fails
5441 to set them up, because the GnuTLS error message doesn't include the name
5442 of the failing file when there is a problem reading it.
5444 26. Expand error message when OpenSSL has problems setting up cert/key files.
5447 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5448 (this can affect the format of dates).
5450 28. exim_tidydb, when checking for the continued existence of a message for
5451 which it has found a message-specific retry record, was not finding
5452 messages that were in split spool directories. Consequently, it was
5453 deleting retry records that should have stayed in existence.
5455 29. eximstats updated to version 1.35
5456 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5457 1.35 - bugfix such that pie charts by volume are generated correctly
5459 30. The SPA authentication driver was not abandoning authentication and moving
5460 on to the next authenticator when an expansion was forced to fail,
5461 contradicting the general specification for all authenticators. Instead it
5462 was generating a temporary error. It now behaves as specified.
5464 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5465 (the order specifies the preference for clients). The order is now AES256,
5466 AES128, 3DES, ARCFOUR128.
5468 31. Small patch to Sieve code - explicitly set From: when generating an
5471 32. Exim crashed if a remote delivery caused a very long error message to be
5472 recorded - for instance if somebody sent an entire SpamAssassin report back
5473 as a large number of 550 error lines. This bug was coincidentally fixed by
5474 increasing the size of one of Exim's internal buffers (big_buffer) that
5475 happened as part of the Exiscan merge. However, to be on the safe side, I
5476 have made the code more robust (and fixed the comments that describe what
5479 33. Some experimental protocols are using DNS PTR records for new purposes. The
5480 keys for these records are domain names, not reversed IP addresses. The
5481 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5482 leaves it alone. Component reversal etc. now happens only for IP addresses.
5485 34. The host_aton() function is supposed to be passed a string that is known
5486 to be a valid IP address. However, in the case of IPv6 addresses, it was
5487 not checking this. This is a hostage to fortune. Exim now panics and dies
5488 if the condition is not met. A case was found where this could be provoked
5489 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5490 components; fortuitously, this particular loophole had already been fixed
5491 by change 4.50/55 or 4.44/33 above.
5493 If there are any other similar loopholes, the new check in host_aton()
5494 itself should stop them being exploited. The report I received stated that
5495 data on the command line could provoke the exploit when Exim was running as
5496 exim, but did not say which command line option was involved. All I could
5497 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5498 running as the user.
5501 35. There was a buffer overflow vulnerability in the SPA authentication code
5502 (which came originally from the Samba project). I have added a test to the
5503 spa_base64_to_bits() function which I hope fixes it.
5506 36. The daemon start-up calls getloadavg() while still root for those OS that
5507 need the first call to be done as root, but it missed one case: when
5508 deliver_queue_load_max is set with deliver_drop_privilege. This is
5509 necessary for the benefit of the queue runner, because there is no re-exec
5510 when deliver_drop_privilege is set.
5512 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5513 was in use, was not putting the data itself into the right store pool;
5514 consequently, it could be overwritten for a subsequent message in the same
5515 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5518 38. Sometimes the final signoff response after QUIT could fail to get
5519 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5520 < 0 before doing a fflush(). This bug looks as though it goes back to the
5521 introduction of TLS in release 3.20, but "sometimes" must have been rare
5522 because the tests only now provoked it.
5528 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5529 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5530 Now it returns an integer. A number of calls were still expecting a T/F
5531 return. Fortuitously, in all cases, the tests worked in OK situations,
5532 which is the norm. However, things would have gone wrong on any write
5533 failures on the smtp file descriptor. This function is used when sending
5534 messages over SMTP and also when doing verify callouts.
5536 2. When Exim is called to do synchronous delivery of a locally submitted
5537 message (the -odf or -odi options), it no longer closes stderr before doing
5540 3. Implemented the mua_wrapper option.
5542 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5544 5. Implemented the functions header_remove(), header_testname(),
5545 header_add_at_position(), and receive_remove_recipient(), and exported them
5548 6. If an ACL "warn" statement specified the addition of headers, Exim already
5549 inserted X-ACL-Warn: at the start if there was no header name. However, it
5550 was not making this test for the second and subsequent header lines if
5551 there were newlines in the string. This meant that an invalid header could
5552 be inserted if Exim was badly configured.
5554 7. Allow an ACL "warn" statement to add header lines at the start or after all
5555 the Received: headers, as well as at the end.
5557 8. Added the rcpt_4xx retry error code.
5559 9. Added postmaster_mailfrom=xxx to callout verification option.
5561 10. Added mailfrom=xxxx to the callout verification option, for verify=
5564 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5565 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5567 12. Inserted some casts to stop certain compilers warning when using pointer
5568 differences as field lengths or precisions in printf-type calls (mostly
5569 affecting debugging statements).
5571 13. Added optional readline() support for -be (dynamically loaded).
5573 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5574 same clock tick as a message's arrival, so that its received time was the
5575 same as the "first fail" time on the retry record, and that message
5576 remained on the queue past the ultimate address timeout, every queue runner
5577 would try a delivery (because it was past the ultimate address timeout) but
5578 after another failure, the ultimate address timeout, which should have then
5579 bounced the address, did not kick in. This was a "< instead of <=" error;
5580 in most cases the first failure would have been in the next clock tick
5581 after the received time, and all would be well.
5583 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5584 being recognized when the domain list was tested by the match_domain
5585 condition in an expansion string.
5587 16. Added the ${str2b64: operator.
5589 17. Exim was always calling setrlimit() to set a large limit for the number of
5590 processes, without checking whether the existing limit was already
5591 adequate. (It did check for the limit on file descriptors.) Furthermore,
5592 errors from getrlimit() and setrlimit() were being ignored. Now they are
5593 logged to the main and panic logs, but Exim does carry on, to try to do its
5594 job under whatever limits there are.
5596 18. Imported PCRE 5.0.
5598 19. Trivial typo in log message " temporarily refused connection" (the leading
5601 20. If the log selector return_path_on_delivery was set and an address was
5602 redirected to /dev/null, the delivery process crashed because it assumed
5603 that a return path would always be set for a "successful" delivery. In this
5604 case, the whole delivery is bypassed as an optimization, and therefore no
5607 21. Internal re-arrangement: the function for sending a challenge and reading
5608 a response while authentication was assuming a zero-terminated challenge
5609 string. It's now changed to take a pointer and a length, to allow for
5610 binary data in such strings.
5612 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5614 23. Exim was not respecting finduser_retries when seeking the login of the
5615 uid under which it was called; it was always trying 10 times. (The default
5616 setting of finduser_retries is zero.) Also, it was sleeping after the final
5617 failure, which is pointless.
5619 24. Implemented tls_on_connect_ports.
5621 25. Implemented acl_smtp_predata.
5623 26. If the domain in control=submission is set empty, Exim assumes that the
5624 authenticated id is a complete email address when it generates From: or
5625 Sender: header lines.
5627 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5628 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5629 chown and chgrp in /bin and hostname in /usr/bin.
5631 28. Exim was keeping the "process log" file open after each use, just as it
5632 does for the main log. This opens the possibility of it remaining open for
5633 long periods when the USR1 signal hits a daemon. Occasional processlog
5634 errors were reported, that could have been caused by this. Anyway, it seems
5635 much more sensible not to leave this file open at all, so that is what now
5638 29. The long-running daemon process does not normally write to the log once it
5639 has entered its main loop, and it closes the log before doing so. This is
5640 so that log files can straightforwardly be renamed and moved. However,
5641 there are a couple of unusual error situations where the daemon does write
5642 log entries, and I had neglected to close the log afterwards.
5644 30. The text of an SMTP error response that was received during a remote
5645 delivery was being truncated at 512 bytes. This is too short for some of
5646 the long messages that one sometimes sees. I've increased the limit to
5649 31. It is now possible to make retry rules that apply only when a message has a
5650 specific sender, in particular, an empty sender.
5652 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5653 it possible to be selective about when SMTP synchronization is enforced.
5655 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5657 32. Implemented hosts_connection_nolog.
5659 33. Added an ACL for QUIT.
5661 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5664 35. Added mailbox_size and mailbox_filecount to appendfile.
5666 36. Added control = no_multiline_responses to ACLs.
5668 37. There was a bug in the logic of the code that waits for the clock to tick
5669 in the case where the clock went backwards by a substantial amount such
5670 that the microsecond fraction of "now" was more than the microsecond
5671 fraction of "then" (but the whole seconds number was less).
5673 38. Added support for the libradius Radius client library this is found on
5674 FreeBSD (previously only the radiusclient library was supported).
5680 1. When certain lookups returned multiple values in the form name=value, the
5681 quoting of the values was not always being done properly. Specifically:
5682 (a) If the value started with a double quote, but contained no whitespace,
5684 (b) If the value contained whitespace other than a space character (i.e.
5685 tabs or newlines or carriage returns) it was not quoted.
5686 This fix has been applied to the mysql and pgsql lookups by writing a
5687 separate quoting function and calling it from the lookup code. The fix
5688 should probably also be applied to nisplus, ibase and oracle lookups, but
5689 since I cannot test any of those, I have not disturbed their existing code.
5691 2. A hit in the callout cache for a specific address caused a log line with no
5692 reason for rejecting RCPT. Now it says "Previous (cached) callout
5693 verification failure".
5695 3. There was an off-by-one bug in the queryprogram router. An over-long
5696 return line was truncated at 256 instead of 255 characters, thereby
5697 overflowing its buffer with the terminating zero. As well as fixing this, I
5698 have increased the buffer size to 1024 (and made a note to document this).
5700 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5701 when Exim is waiting for an SMTP response from a remote server, Exim
5702 restarts its select() call on the socket, thereby resetting its timeout.
5703 This is not a problem when such interrupts are rare. Somebody set up a cron
5704 job to run exiwhat every 2 minutes, which is less than the normal select()
5705 timeout (5 or 10 minutes). This meant that the select() timeout never
5706 kicked in because it was always reset. I have fixed this by comparing the
5707 time when an interrupt arrives with the time at the start of the first call
5708 to select(). If more time than the timeout has elapsed, the interrupt is
5709 treated as a timeout.
5711 5. Some internal re-factoring in preparation for the addition of Sieve
5712 extensions (by MH). In particular, the "personal" test is moved to a
5713 separate function, and given an option for scanning Cc: and Bcc: (which is
5714 not set for Exim filters).
5716 6. When Exim created an email address using the login of the caller as the
5717 local part (e.g. when creating a From: or Sender: header line), it was not
5718 quoting the local part when it contained special characters such as @.
5720 7. Installed new OpenBSD configuration files.
5722 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5723 try to make them clearer.
5725 9. Callout options, other than the timeout value, were being ignored when
5726 verifying sender addresses in header lines. For example, when using
5728 verify = header_sender/callout=no_cache
5730 the cache was (incorrectly) being used.
5732 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5733 only the Cygwin environment.
5735 11. When return_path_on_delivery was set as a log selector, if different remote
5736 addresses in the same message used different return paths and parallel
5737 remote delivery occurred, the wrong values would sometimes be logged.
5738 (Whenever a remote delivery process finished, the return path value from
5739 the most recently started remote delivery process was logged.)
5741 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5742 lines when AUTH and/or TLS are in use. This is the "received protocol"
5743 field. Exim used to use "asmtp" for authenticated SMTP, without any
5744 indication (in the protocol name) for TLS use. Now it follows the RFC and
5745 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5746 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5747 names appear in log lines as well as in Received: header lines.
5749 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5750 extensions, and comparison tests, and to fix some bugs.
5752 14. Changes to the "personal" filter test:
5754 (1) The test was buggy in that it was just doing the equivalent of
5755 "contains" tests on header lines. For example, if a user's address was
5756 anne@some.where, the "personal" test would incorrectly be true for
5758 To: susanne@some.where
5760 This test is now done by extracting each address from the header in turn,
5761 and checking the entire address. Other tests that are part of "personal"
5762 are now done using regular expressions (for example, to check local parts
5763 of addresses in From: header lines).
5765 (2) The list of non-personal local parts in From: addresses has been
5766 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5767 taken from the Sieve specification recommendations.
5769 (3) If the message contains any header line starting with "List-" it is
5770 treated as non-personal.
5772 (4) The test for "circular" in the Subject: header line has been removed
5773 because it now seems ill-conceived.
5775 15. Minor typos in src/EDITME comments corrected.
5777 16. Installed latest exipick from John Jetmore.
5779 17. If headers_add on a router specified a text string that was too long for
5780 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5781 of string_sprintf() is now avoided.
5783 18. $message_body_size was not set (it was always zero) when running the DATA
5784 ACL and the local_scan() function.
5786 19. For the "mail" command in an Exim filter, no default was being set for
5787 the once_repeat time, causing a random time value to be used if "once" was
5788 specified. (If the value happened to be <= 0, no repeat happened.) The
5789 default is now 0s, meaning "never repeat". The "vacation" command was OK
5790 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5791 (I found it when inspecting the code).
5793 20. There is now an overall timeout for performing a callout verification. It
5794 defaults to 4 times the callout timeout, which applies to individual SMTP
5795 commands during the callout. The overall timeout applies when there is more
5796 than one host that can be tried. The timeout is checked before trying the
5797 next host. This prevents very long delays if there are a large number of
5798 hosts and all are timing out (e.g. when the network connections are timing
5799 out). The value of the overall timeout can be changed by specifying an
5800 additional sub-option for "callout", called "maxwait". For example:
5802 verify = sender/callout=5s,maxwait=20s
5804 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5805 to the end before writing, but this should make it even safer).
5807 22. Exim was forgetting that it had advertised PIPELINING for the second and
5808 subsequent messages on an SMTP connection. It was also not resetting its
5809 memory on STARTTLS and an internal HELO.
5811 23. When Exim logs an SMTP synchronization error within a session, it now
5812 records whether PIPELINING has been advertised or not.
5814 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5815 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5816 rather than long int.
5818 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5820 26. Added the never_mail option to autoreply.
5826 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5827 crash if the getsockname() call failed; this can happen if a connection is
5828 closed very soon after it is established. The problem was simply in the
5829 order in which certain operations were done, causing Exim to try to write
5830 to the SMTP stream before it had set up the file descriptor. The bug has
5831 been fixed by making things happen in the correct order.
5837 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5838 before the connection was closed, thus losing the rejection response.
5840 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5841 some early Solaris releases, but causes trouble in current releases where
5842 socklen_t is defined.
5844 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5847 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5850 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5851 that releasing the top of it at the end releases what was used for sub-
5852 expansions (unless the block got too big). However, discard this block if
5853 the first thing is a variable or header, so that we can use its block when
5854 it is dynamic (useful for very large $message_headers, for example).
5856 6. Lookups now cache *every* query, not just the most recent. A new, separate
5857 store pool is used for this. It can be recovered when all lookup caches are
5858 flushed. Lookups now release memory at the end of their result strings.
5859 This has involved some general refactoring of the lookup sources.
5861 7. Some code has been added to the store_xxx() functions to reduce the amount
5862 of flapping under certain conditions.
5864 8. log_incoming_interface used to affect only the <= reception log lines. Now
5865 it causes the local interface and port to be added to several more SMTP log
5866 lines, for example "SMTP connection from", and rejection lines.
5868 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5870 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5872 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5873 could be overwritten at the end of the current message (or the start of a
5874 new message if it was set in a HELO ACL). The value is now preserved for
5875 the duration of the SMTP connection.
5877 12. If a transport had a headers_rewrite setting, and a matching header line
5878 contained an unqualified address, that address was qualified, even if it
5879 did not match any rewriting rules. The underlying bug was that the values
5880 of the flags that permit the existence of unqualified sender and recipient
5881 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5882 non-local messages, and by -bnq for local messages) were not being
5883 preserved with the message after it was received.
5885 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5886 "next input=" as part of the text comprising the host identity instead of
5887 the correct text. The code was using the same buffer for two different
5888 strings. However, depending on which order the printing function evaluated
5889 its arguments, the bug did not always show up. Under Linux, for example, my
5890 test suite worked just fine.
5892 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5893 doesn't work with some older versions of Perl. It has been changed to "my",
5894 which in any case is probably the better facility to use.
5896 15. A really picky compiler found some instances of statements for creating
5897 error messages that either had too many or two few arguments for the format
5900 16. The size of the buffer for calls to the DNS resolver has been increased
5901 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5902 for addresses that have a lot of PTR records. This alleviates a problem; it
5903 does not fully solve it.
5905 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5906 buffer now truncates the list and logs the incident, which is the same
5907 action as happens when Exim is looking up a host name and its aliases.
5908 Previously in this situation something unpredictable would happen;
5909 sometimes it was "internal error: store_reset failed".
5911 18. If a server dropped the connection unexpectedly when an Exim client was
5912 using GnuTLS and trying to read a response, the client delivery process
5913 crashed while trying to generate an error log message.
5915 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5916 string, for example:
5918 warn message = H1: something\nH2: something
5920 the text was added as a single header line from Exim's point of view
5921 though it ended up OK in the delivered message. However, searching for the
5922 second and subsequent header lines using $h_h2: did not work. This has been
5923 fixed. Similarly, if a system filter added multiple headers in this way,
5924 the routers could not see them.
5926 20. Expanded the error message when iplsearch is called with an invalid key to
5927 suggest using net-iplsearch in a host list.
5929 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5930 ACLs are no longer actually imposed (and a message to that effect is
5933 22. If a "gecos" field in a passwd entry contained escaped characters, in
5934 particular, if it contained a \" sequence, Exim got it wrong when building
5935 a From: or a Sender: header from that name. A second bug also caused
5936 incorrect handling when an unquoted " was present following a character
5937 that needed quoting.
5939 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5940 was not being matched caselessly.
5942 24. Arranged for all hyphens in the exim.8 source to be escaped with
5945 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5946 properly. Recipient callout cache records were still being keyed to include
5947 the sender, even when use_sender was set false. This led to far more
5948 callouts that were necessary. The sender is no longer included in the key
5949 when use_sender is false.
5951 26. Added "control = submission" modifier to ACLs.
5953 27. Added the ${base62d: operator to decode base 62 numbers.
5955 28. dnsdb lookups can now access SRV records.
5957 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5958 the configuration file.
5960 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5961 (-v) mode. This makes the output for a verbose queue run more intelligible.
5963 31. Added a use_postmaster feature to recipient callouts.
5965 32. Added the $body_zerocount variable, containing the number of binary zero
5966 bytes in the message body.
5968 33. The time of last modification of the "new" subdirectory is now used as the
5969 "mailbox time last read" when there is a quota error for a maildir
5972 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5974 35. Added +ignore_unknown as a special item in host lists.
5976 36. Code for decoding IPv6 addresses in host lists is now included, even if
5977 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5978 address was recognized as an IP address, but was then not correctly decoded
5979 into binary, causing unexpected and incorrect effects when compared with
5986 1. Very minor rewording of debugging text in manualroute to say "list of
5987 hosts" instead of "hostlist".
5989 2. If verify=header_syntax was set, and a header line with an unqualified
5990 address (no domain) and a large number of spaces between the end of the
5991 name and the colon was received, the reception process suffered a buffer
5992 overflow, and (when I tested it) crashed. This was caused by some obsolete
5993 code that should have been removed. The fix is to remove it!
5995 3. When running in the test harness, delay a bit after writing a bounce
5996 message to get a bit more predictability in the log output.
5998 4. Added a call to search_tidyup() just before forking a reception process. In
5999 theory, someone could use a lookup in the expansion of smtp_accept_max_
6000 per_host which, without the tidyup, could leave open a database connection.
6002 5. Added the variables $recipient_data and $sender_data which get set from a
6003 lookup success in an ACL "recipients" or "senders" condition, or a router
6004 "senders" option, similar to $domain_data and $local_part_data.
6006 6. Moved the writing of debug_print from before to after the "senders" test
6009 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6010 problems for message scanning, either using a data ACL, or using
6011 local_scan() because the Received: header was not generated till after they
6012 were called (in order to set the time as the time of reception completion).
6013 I have revised the way this works. The header is now generated after the
6014 body is received, but before the ACL or local_scan() are called. After they
6015 are run, the timestamp in the header is updated.
6021 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6022 before starting a queue runner without re-exec. This happened only when
6023 deliver_drop_privilege was set or when the Exim user was set to root. The
6024 effect of the bug was that timeouts during subsequent deliveries caused
6025 crashes instead of being properly handled. The handler is now left at its
6026 default (and expected) setting.
6028 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6029 message, again when deliver_drop_privilege is set or Exim is run as root.
6030 The bug described in (1) was not present in this case, but the tidying up
6031 of the other signals was missing. I have made the two cases consistent.
6033 3. The ignore_target_hosts setting on a manualroute router was being ignored
6034 for hosts that were looked up using the /MX notation.
6036 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6039 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6040 operated on the sender address. After changing the $sender_address to <>
6041 for the sender address verify, Exim was re-instated it as the original
6042 (before rewriting) address, but remembering that it had rewritten it, so it
6043 wasn't rewriting it again. This bug also had the effect of breaking the
6044 sender address verification caching when the sender address was rewritten.
6046 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6047 This has been changed so that if the ip literal address matches
6048 ignore_target_hosts, the router declines.
6050 7. Added expansion conditions match_domain, match_address, and match_local_
6051 part (NOT match_host).
6053 8. The placeholder for the Received: header didn't have a length field set.
6055 9. Added code to Exim itself and to exim_lock to test for a specific race
6056 condition that could lead to file corruption when using MBX delivery. The
6057 issue is with the lockfile that is created in /tmp. If this file is removed
6058 after a process has opened it but before that process has acquired a lock,
6059 there is the potential for a second process to recreate the file and also
6060 acquire a lock. This could lead to two Exim processes writing to the file
6061 at the same time. The added code performs the same test as UW imapd; it
6062 checks after acquiring the lock that its file descriptor still refers to
6063 the same named file.
6065 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6066 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6069 11. Added the smtp_active_hostname option. If used, this will typically be made
6070 to depend on the incoming interface address. Because $interface_address is
6071 not set up until the daemon has forked a reception process, error responses
6072 that can happen earlier (such as "too many connections") no longer contain
6075 12. If an expansion in a condition on a "warn" statement fails because a lookup
6076 defers, the "warn" statement is abandoned, and the next ACL statement is
6077 processed. Previously this caused the whole ACL to be aborted.
6079 13. Added the iplsearch lookup type.
6081 14. Added ident_timeout as a log selector.
6083 15. Added tls_certificate_verified as a log selector.
6085 16. Added a global option tls_require_ciphers (compare the smtp transport
6086 option of the same name). This controls incoming TLS connections.
6088 17. I finally figured out how to make tls_require_ciphers do a similar thing
6089 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6090 before starting the TLS session.
6092 18. Tabs are now shown as \t in -bP output.
6094 19. If the log selector return_path_on_delivery was set, Exim crashed when
6095 bouncing a message because it had too many Received: header lines.
6097 20. If two routers both had headers_remove settings, and the first one included
6098 a superfluous trailing colon, the final name in the first list and the
6099 first name in the second list were incorrectly joined into one item (with a
6100 colon in the middle).
6106 1. Added -C and -D options to the exinext utility, mainly to make it easier
6107 to include in the automated testing, but these could be helpful when
6108 multiple configurations are in use.
6110 2. The exinext utility was not formatting the output nicely when there was
6111 an alternate port involved in the retry record key, nor when there was a
6112 message id as well (for retries that were specific to a specific message
6113 and a specific host). It was also confused by IPv6 addresses, because of
6114 the additional colons they contain. I have fixed the IPv4 problem, and
6115 patched it up to do a reasonable job for IPv6.
6117 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6118 delivery, the log line now contains "pipelined" if PIPELINING was used.
6120 4. An SMTP transport process used to panic and die if the bind() call to set
6121 an explicit outgoing interface failed. This has been changed; it is now
6122 treated in the same way as a connect() failure.
6124 5. A reference to $sender_host_name in the part of a conditional expansion
6125 that was being skipped was still causing a DNS lookup. This no longer
6128 6. The def: expansion condition was not recognizing references to header lines
6129 that used bh_ and bheader_.
6131 7. Added the _cache feature to named lists.
6133 8. The code for checking quota_filecount in the appendfile transport was
6134 allowing one more file than it should have been.
6136 9. For compatibility with Sendmail, the command line option
6144 and sets the incoming protocol and host name (for trusted callers). The
6145 host name and its colon can be omitted when only the protocol is to be set.
6146 Note the Exim already has two private options, -pd and -ps, that refer to
6147 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6148 "s", but I don't think that's a major issue.
6150 10. A number of refactoring changes to the code, none of which should affect
6153 (a) The number of logging options was getting close to filling up the
6154 32-bit word that was used as a bit map. I have split them into two classes:
6155 those that are passed in the argument to log_write(), and those that are
6156 only ever tested independently outside of that function. These are now in
6157 separate 32-bit words, so there is plenty of room for expansion again.
6158 There is no change in the user interface or the logging behaviour.
6160 (b) When building, for example, log lines, the code previously used a
6161 macro that called string_cat() twice, in order to add two strings. This is
6162 not really sufficiently general. Furthermore, there was one instance where
6163 it was actually wrong because one of the argument was used twice, and in
6164 one call a function was used. (As it happened, calling the function twice
6165 did not affect the overall behaviour.) The macro has been replaced by a
6166 function that can join an arbitrary number of extra strings onto a growing
6169 (c) The code for expansion conditions now uses a table and a binary chop
6170 instead of a serial search (which was left over from when there were very
6171 few conditions). Also, it now recognizes conditions like "pam" even when
6172 the relevant support is not compiled in: a suitably worded error message is
6173 given if an attempt is made to use such a condition.
6175 11. Added ${time_interval:xxxxx}.
6177 12. A bug was causing one of the ddress fields not to be passed back correctly
6178 from remote delivery subprocesses. The field in question was not being
6179 subsequently used, so this caused to problems in practice.
6181 13. Added new log selectors queue_time and deliver_time.
6183 14. Might have fixed a bug in maildirsizefile handling that threw up
6184 "unexpected character" debug warnings, and recalculated the data
6185 unnecessarily. In any case, I expanded the warning message to give more
6188 15. Added the message "Restricted characters in address" to the statements in
6189 the default ACL that block characters like @ and % in local parts.
6191 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6192 Three changes have been made:
6194 (a) There was a serious bug; a negative response to MAIL caused the whole
6195 recipient domain to be cached as invalid, thereby blocking all messages
6196 to all local parts at the same domain, from all senders. This bug has
6197 been fixed. The domain is no longer cached after a negative response to
6198 MAIL if the sender used is not empty.
6200 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6203 (c) A new callout option, "use_sender" has been added for people who want
6204 the modified behaviour.
6210 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6213 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6214 indeed breaks things for older releases.
6216 3. Added additional logging to the case where there is a problem reading data
6217 from a filter that is running in a subprocess using a pipe, in order to
6218 try to track down a specific problem.
6220 4. Testing facility fudge: when running in the test harness and attempting
6221 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6222 getting "No route to host". Convert this to a timeout.
6224 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6227 6. Some OS don't have socklen_t but use size_t instead. This affects the
6228 fifth argument of getsockopt() amongst other things. This is now
6229 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6230 can be set for individual OS. I have set it for SunOS5, OSF1, and
6231 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6232 some earlier ones do not.
6234 7. Change 4.30/15 was not doing the test caselessly.
6236 8. The standard form for an IPv6 address literal was being rejected by address
6237 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6238 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6239 this, as well as the form without the "IPv6" on the front (but only when
6240 address literals are enabled, of course).
6242 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6244 10. Exim crashed if a message with an empty sender address specified by -f
6245 encountered a router with an errors_to setting. This could be provoked only
6246 by a command such as
6250 where an empty string was supplied; "<>" did not hit this bug.
6252 11. Installed PCRE release 4.5.
6254 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6255 remained set. It is now erased.
6257 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6258 times from message ids (which are base 36 rather than the normal 62).
6260 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6261 were being counted as actual protocol errors, and logged if the log
6262 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6263 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6264 and DATA following a set of rejected RCPTs do not count as protocol errors.
6265 In other words, Exim assumes they were pipelined, though this may not
6266 actually be the case. Of course, in all cases the client gets an
6267 appropriate error code.
6269 15. If a lookup fails in an ACL condition, a message about the failure may
6270 be available; it is used if testing the ACL cannot continue, because most
6271 such messages specify what the cause of the deferral is. However, some
6272 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6273 that caused an old message to be retained and used if a later statement
6274 caused a defer, replacing the real cause of the deferral.
6276 16. If an IP address had so many PTR records that the DNS lookup buffer
6277 was not large enough to hold them, Exim could crash while trying to process
6278 the truncated data. It now detects and logs this case.
6280 17. Further to 4.21/58, another change has been made: if (and only if) the
6281 first line of a message (the first header line) ends with CRLF, a bare LF
6282 in a subsequent header line has a space inserted after it, so as not to
6283 terminate the header.
6285 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6286 unnecessarily, used atoi() instead of strtol(), and didn't check the
6287 termination when getting file sizes from file names by regex.
6289 19. Completely re-implemented the support for maildirsize files, in the light
6290 of a number of problems with the previous contributed implementation
6291 (4.30/29). In particular:
6293 . If the quota is zero, the maildirsize file is maintained, but no quota is
6296 . If the maildir directory does not exist, it is created before any attempt
6297 to write a maildirsize file.
6299 . The quota value in the file is just a cache; if the quota is changed in
6300 the transport, the new value overrides.
6302 . A regular expression is available for excluding directories from the
6305 20. The autoreply transport checks the characters in options that define the
6306 message's headers; it allows continued headers, but it was checking with
6307 isspace() after an embedded newline instead of explicitly looking for a
6310 21. If all the "regular" hosts to which an address was routed had passed their
6311 expiry times, and had not reached their retry times, the address was
6312 bounced, even if fallback hosts were defined. Now Exim should go on to try
6315 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6316 equivalent code in the SMTP transport. Some hosts send humungous responses
6317 to HELO/EHLO, more than 1024 it seems.
6319 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6320 gives compiler warnings in some environments. I've now done it "properly",
6323 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6324 (because of problems with the built-in one) was declared to return uschar *
6325 instead of char *, causing compiler failure.
6327 25. Fixed a file descriptor leak when processing alias/forward files.
6329 26. Fixed a minor format string issue in dbfn.c.
6331 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6333 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6334 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6337 29. When a sender address is verified, it is cached, to save repeating the test
6338 when there is more than one recipient in a message. However, when the
6339 verification involves a callout, it is possible for different callout
6340 options to be set for different recipients. It is too complicated to keep
6341 track of this in the cache, so now Exim always runs a verification when a
6342 callout is required, relying on the callout cache for the optimization.
6343 The overhead is duplication of the address routing, but this should not be
6346 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6347 to be verified with callout=postmaster, and the main callout worked but the
6348 postmaster check failed, the verification correctly failed. However, if a
6349 subsequent RCPT command asked for sender verification *without* the
6350 postmaster check, incorrect caching caused this verification also to fail,
6353 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6354 it was not caching the DNS options (qualify_single, search_parents) that
6355 were used when the lookup failed. A subsequent lookup with different
6356 options therefore always gave the same answer, though there were cases
6357 where it should not have. (Example: a "domains = !$mx_any" option on a
6358 dnslookup router: the "domains" option is always processed without any
6359 widening, but the router might have qualify_single set.) Now Exim uses the
6360 cached value only when the same options are set.
6362 32. Added John Jetmore's "exipick" utility to the distribution.
6364 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6365 than a timeout (e.g. a certificate is required, and is not provided), an
6366 Exim server now closes the connection immediately. Previously it waited for
6367 the client to close - but if the client is SSL, it seems that they each
6368 wait for each other, leading to a delay before one of them times out.
6370 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6371 maintained 0.8.x compatibility because I don't think many are using it, and
6372 it is clearly obsolete.
6374 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6377 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6378 client certificate was expired. A simple patch fixes this, though I don't
6379 understand the full logic of why the verify callback is called multiple
6382 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6383 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6384 which causes problems with some clients (such as the Certicom SSL Plus
6385 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6386 disables the coutermeasure allowing Eudora to connect."
6388 38. Exim was not checking that a write() to a log file succeeded. This could
6389 lead to Bad Things if a log got too big, in particular if it hit a file
6390 size limit. Exim now panics and dies if it cannot write to a log file, just
6391 as it does if it cannot open a log file.
6393 39. Modified OS/Makefile-Linux so that it now contains
6395 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6397 The two -D definitions ensure that Exim is compiled with large file
6398 support, which makes it possible to handle log files that are bigger than
6401 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6402 instance) a domain was checked against a named list that involved a lookup,
6403 causing $domain_data to be set, then another domain was checked against the
6404 same list, then the first domain was re-checked, the value of $domain_data
6405 after the final check could be wrong. In particular, if the second check
6406 failed, it could be set empty. This bug probably also applied to
6409 41. The strip_trailing_dot option was not being applied to the address given
6410 with the -f command-line option.
6412 42. The code for reading a message's header from the spool was incrementing
6413 $received_count, but never initializing it. This meant that the value was
6414 incorrect (doubled) while delivering a message in the same process in which
6415 it was received. In the most common configuration of Exim, this never
6416 happens - a fresh exec is done - but it can happen when
6417 deliver_drop_privilege is set.
6419 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6420 it now includes up to 150 characters of the unexpected data in the log
6423 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6424 and building data strings. The size of both of these buffers was 10 000
6425 bytes - far larger than anybody would *ever* want, thought I. Needless to
6426 say, somebody hit the limit. I have increased the maximum line length to
6427 20 000 and the maximum data length of concatenated lines to 100 000. I have
6428 also fixed two bugs, because there was no checking on these buffers. Tsk,
6429 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6430 buffer is too small.
6432 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6433 lsearch lookups. Now it does.
6435 46. When parsing a route_list item in a manualroute router, a fixed-length
6436 buffer was used for the list of hosts. I made this 1024 bytes long,
6437 thinking that nobody would ever have a list of hosts that long. Wrong.
6438 Somebody had a whole pile of complicated expansion conditions, and the
6439 string was silently truncated, leading to an expansion error. It turns out
6440 that it is easier to change to an unlimited length (owing to other changes
6441 that have happened since this code was originally written) than to build
6442 structure for giving a limitation error. The length of the item that
6443 expands into the list of hosts is now unlimited.
6445 47. The lsearch lookup could not handle data where the length of text line was
6446 more than 4095 characters. Such lines were truncated, leading to shortened
6447 data being returned. It should now handle lines of any length.
6449 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6450 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6453 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6454 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6455 no longer happens in comment lines. A list of replacements is now placed
6456 at the head of all of the source files, except those whose only change is
6457 to replace PERL_COMMAND in the very first #! line.
6459 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6460 messages on the queue, with a bottom-up merge sort, using code contributed
6461 by Michael Haardt. This should make operations like -bp somewhat faster on
6462 large queues. It won't affect queue runners, except when queue_run_in_order
6465 51. Installed eximstats 1.31 in the distribution.
6467 52. Added support for SRV lookups to the dnslookup router.
6469 53. If an ACL referred to $message_body or $message_body_end, the value was not
6470 reset for any messages that followed in the same SMTP session.
6472 54. The store-handling optimization for building very long strings was not
6473 differentiating between the different store pools. I don't think this
6474 actually made any difference in practice, but I've tidied it.
6476 55. While running the routers to verify a sender address, $sender_address
6477 was still set to the sender address. This is wrong, because when routing to
6478 send a bounce to the sender, it would be empty. Therefore, I have changed
6479 it so that, while verifying a sender address, $sender_address is set to <>.
6480 (There is no change to what happens when verifying a recipient address.)
6482 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6483 target A or AAAA records (if not already returned) without resetting the
6484 qualify_single or search_parents options of the DNS resolver. These are
6485 inappropriate in this case because the targets of MX and SRV records must
6486 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6487 target that, when qualified, matched something in the local domain. These
6488 two options are now turned off when doing these lookups.
6490 57. It seems that at least some releases of Reiserfs (which does not have the
6491 concept of a fixed number of inodes) returns zero and not -1 for the
6492 number of available inodes. This interacted badly with check_spool_inodes,
6493 which assumed that -1 was the "no such thing" setting. What I have done is
6494 to check that the total number of inodes is greater than zero before doing
6495 the test of how many are available.
6497 58. When a "warn" ACL statement has a log_message modifier, the message is
6498 remembered, and not repeated. This is to avoid a lot of repetition when a
6499 message has many recipients that cause the same warning to be written.
6500 However, Exim was preserving the list of already written lines for an
6501 entire SMTP session, which doesn't seem right. The memory is now reset if a
6502 new message is started.
6504 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6505 debugging output unless log_rewrite was also set.
6507 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6508 of (int)(handle) when we know that handle contains (void *)(-1).
6510 61. The Exim daemon panic-logs an error return when it closes the incoming
6511 connection. However "connection reset by peer" seems to be common, and
6512 isn't really an error worthy of noting specially, so that particular error
6515 62. When Exim is trying to find all the local interfaces, it used to panic and
6516 die if the ioctl to get the interface flags failed. However, it seems that
6517 on at least one OS (Solaris 9) it is possible to have an interface that is
6518 included in the list of interfaces, but for which you get a failure error
6519 for this call. This happens when the interface is not "plumbed" into a
6520 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6521 failure of the "get flags" call assumes that the interface is down.
6523 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6524 makes life easier for people who are doing arithmetic on fields extracted
6525 from dates, where you often get leading zeros that should not be
6526 interpreted as octal.
6528 64. Added qualify_domain to the redirect router, to override the global
6531 65. If a pathologically long header line contained very many addresses (the
6532 report of this problem mentioned 10 000) and each of them was rewritten,
6533 Exim could use up a very large amount of memory. (It kept on making new
6534 copies of the header line as it rewrote, and never released the old ones.)
6535 At the expense of a bit more processing, the header rewriting function has
6536 been changed so that it no longer eats memory in this way.
6538 66. The generation of the Received: header has been moved from the time that a
6539 message starts to be received, to the time that it finishes. The timestamp
6540 in the Received: header should now be very close to that of the <= log
6541 line. There are two side-effects of this change:
6543 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6544 logged header lines no longer include the local Received: line, because
6545 it has not yet been created. The same applies to a copy of the message
6546 that is returned to a non-SMTP sender when a message is rejected.
6548 (b) When a filter file is tested using -bf, no additional Received: header
6549 is added to the test message. After some thought, I decided that this
6552 This change does not affect the value of $received_for. It is still set
6553 after address rewriting, but before local_scan() is called.
6555 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6557 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6558 gave an unhelpful panic error message, and a defer error. I have managed to
6559 change this behaviour so that it now rejects any supplied certificate,
6560 which seems right, as the list of acceptable certificates is empty.
6562 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6563 gave an unhelpful defer error. I have not managed to make this reject any
6564 supplied certificates, but the error message it gives is "no certificate
6565 supplied", which is not helpful.
6567 70. exigrep's output now also includes lines that are not associated with any
6568 message, but which match the given pattern. Implemented by a patch from
6569 Martin Sluka, which also tidied up the Perl a bit.
6571 71. Recipient callout verification, like sender verification, was using <> in
6572 the MAIL FROM command. This isn't really the right thing, since the actual
6573 sender may affect whether the remote host accepts the recipient or not. I
6574 have changed it to use the actual sender in the callout; this means that
6575 the cache record is now keyed on a recipient/sender pair, not just the
6576 recipient address. There doesn't seem to be a real danger of callout loops,
6577 since a callout by the remote host to check the sender would use <>.
6578 [SEE ABOVE: changed after hitting problems.]
6580 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6581 temporary errors. However, in the case of such a code being given after
6582 the end of a data transmission (i.e. after ".") Exim was failing to write
6583 a retry record for the message. (Yes, there was some broken host that was
6584 actually sending 8xx at this point.)
6586 73. An unknown lookup type in a host list could cause Exim to panic-die when
6587 the list was checked. (An example that provoked this was putting <; in the
6588 middle of a list instead of at the start.) If this happened during a DATA
6589 ACL check, a -D file could be left lying around. This kind of configuration
6590 error no longer causes Exim to die; instead it causes a defer error. The
6591 incident is still logged to the main and panic logs.
6593 74. Buglet left over from Exim 3 conversion. The message "too many messages
6594 in one connection" was written to the rejectlog but not the mainlog, except
6595 when address rewriting (yes!) was being logged.
6597 75. Added write_rejectlog option.
6599 76. When a system filter was run not as root (that is, when system_filter_user
6600 was set), the values of the $n variables were not being returned to the
6601 main process; thus, they were not subsequently available in the $sn
6604 77. Added +return_path_on_delivery log selector.
6606 78. A connection timeout was being treated differently from recipients deferred
6607 when testing hosts_max_try with a message that was older than the host's
6608 retry timeout. (The host should not be counted, thus allowing all hosts to
6609 be tried at least once before bouncing.) This may have been the cause of an
6610 occasionally reported bug whereby a message would remain on the queue
6611 longer than the retry timeout, but would be bounced if a delivery was
6612 forced. I say "may" because I never totally pinned down the problem;
6613 setting up timeout/retry tests is difficult. See also the next item.
6615 79. The ultimate address timeout was not being applied to errors that involved
6616 a combination of host plus message (for example, a timeout on a MAIL
6617 command). When an address resolved to a number of possible hosts, and they
6618 were not all tried for each delivery (e.g. because of hosts_max_try), a
6619 message could remain on the queue longer than the retry timeout.
6621 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6624 81. Fixed an obscure SMTP outgoing bug which required at least the following
6625 conditions: (a) there was another message waiting for the same server;
6626 (b) the server returned 5xx to all RCPT commands in the first message so
6627 that the message was not completed; (c) the server dropped the connection
6628 or gave a negative response to the RSET that Exim sends to abort the
6629 transaction. The observed case was a dropped connection after DATA that had
6630 been sent in pipelining mode. That is, the server had advertised PIPELINING
6631 but was not implementing it correctly. The effect of the bug was incorrect
6632 behaviour, such as trying another host, and this could lead to a crash.
6638 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6639 and daemon.c were passed as pointers to ints; they should have been
6640 pointers to socklen_t variables (which are typically unsigned ints).
6642 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6645 3. Fixed a really odd bug that affected only the testing scheme; patching a
6646 certain fixed string in the binary changed the value of another string that
6647 happened to be identical to the end of the original first string.
6649 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6650 name", it returns that address as the IP address. On some operating
6651 systems (e.g. Solaris), it also passes back the IP address string as the
6652 "host name". However, on others (e.g. Linux), it passes back an empty
6653 string. Exim wasn't checking for this, and was changing the host name to an
6654 empty string, assuming it had been canonicalized.
6656 5. Although rare, it is permitted to have more than one PTR record for a given
6657 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6658 all the names associated with an address, because they do in Solaris.
6659 However, it seems that they do not in Linux for data that comes from the
6660 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6661 I found this out when I moved to a new Linux workstation and tried to run
6662 the Exim test suite.
6664 To get round this problem I have changed the code so that it now does its
6665 own call to the DNS to look up PTR records when searching for a host name.
6666 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6667 addresses that are only in /etc/hosts are still found.
6669 This behaviour is, however, controlled by an option called host_lookup_
6670 order, which defaults to "bydns:byaddr". If people want to use the other
6671 order, or indeed, just use one or the other means of lookup, they can
6672 specify it in this variable.
6674 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6675 some operating systems, this comes back from gethostbyaddr() as an empty
6676 string, and this is what Exim used to test for. However, it seems that in
6677 other systems, "." is yielded. Exim now tests for this case too.
6679 7. The values of check_spool_space and check_log_space are now held internally
6680 as a number of kilobytes instead of an absolute number of bytes. If a
6681 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6682 kilobyte. This means that much larger values can be stored.
6684 8. Exim monitor: an attempt to get the action menu when not actually pointing
6685 at a message produces an empty menu entitled "No message selected". This
6686 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6687 no entries in it ("Shell widget menu has zero width and/or height"). So I
6688 have added a single, blank menu entry in this case.
6690 9. Added ${quote_local_part.
6692 10. MIME decoding is now applied to the contents of Subject: header lines when
6695 11. Now that a reference to $sender_host_address automatically causes a reverse
6696 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6697 host lookup before query-style lookups in lists that might use this
6698 variable. This has therefore been abolished, and the "net-" prefix is no
6699 longer necessary for query-style lookups.
6701 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6702 have been a typo for LFLAGS, so it has been changed.
6704 13. The install script calls Exim with "-C /dev/null" in order to find the
6705 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6706 to be output. However, since Exim outputs its version number before the
6707 error, it didn't break the script. It just looked ugly. I fixed this by
6708 always allowing "-C /dev/null" if the caller is root.
6710 14. Ignore overlarge ACL variable number when reading spool file - insurance
6711 against a later release with more variables having written the file.
6713 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6714 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6715 this, as well as the form without the "IPv6" on the front.
6717 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6718 OS/Makefile-Darwin file.
6720 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6721 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6723 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6724 variable. However, some OS use ints for inodes. I've added cast to long int
6725 to get rid of the compiler warning.
6727 19. I had forgotten to lock out "/../" in configuration file names when
6728 ALT_CONFIG_PREFIX was set.
6730 20. Routers used for verification do not need to specify transports. However,
6731 if such a router generated a host list, and callout was configured, Exim
6732 crashed, because it could not find a port number from the (non-existent)
6733 transport. It now assumes port 25 in this circumstance.
6735 21. Added the -t option to exigrep.
6737 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6738 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6739 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6740 them is set, LOOKUP_LSEARCH is forced.
6742 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6744 24. Added sender and host information to the "rejected by local_scan()" log
6745 line; previously there was no indication of these.
6747 25. Added .include_if_exists.
6749 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6750 Linux. It turns out that not all file systems support this. Apparently some
6751 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6752 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6753 not supported on the file descriptor, is now ignored when Exim is trying to
6754 sync a directory. This applies only to Linux.
6756 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6758 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6759 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6762 29. Added support for maildirsize files from supplied patch (modified a bit).
6764 30. The use of :fail: followed by an empty string could lead Exim to respond to
6765 sender verification failures with (e.g.):
6767 550 Verification failed for <xxx>
6768 550 Sender verify failed
6770 where the first response line was missing the '-' that indicates it is not
6771 the final line of the response.
6773 31. The loop for finding the name of the user that called Exim had a hardwired
6774 limit of 10; it now uses the value of finduser_retries, which is used for
6775 all other user lookups.
6777 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6780 33. Exim was neglecting to zero errno before one call of strtol() when
6781 expanding a string and expecting an integer value. On some systems this
6782 resulted in spurious "integer overflow" errors. Also, it was casting the
6783 result into an int without checking.
6785 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6786 did not work. The code looks as if it has *never* worked, though it appears
6787 to have been documented since at least release 1.62. I have made it work.
6789 35. The "timeout_DNS" error in retry rules, also documented since at least
6790 1.62, also never worked. As it isn't clear exactly what this means, and
6791 clearly it isn't a major issue, I have abolished the feature by treating it
6792 as "timeout", and writing a warning to the main and panic logs.
6794 36. The display of retry rules for -brt wasn't always showing the error code
6797 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6798 timeout_connect_A, timeout_connect_MX.
6800 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6801 to the empty sender.
6803 39. The daemon was not analysing the content of -oX till after it had closed
6804 stderr and disconnected from the controlling terminal. This meant that any
6805 syntax errors were only noted on the panic log, and the return code from
6806 the command was 0. By re-arranging the code a little, I've made the
6807 decoding happen first, so such errors now appear on stderr, and the return
6808 code is 1. However, the actual setting up of the sockets still happens in
6809 the disconnected process, so errors there are still only recorded on the
6812 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6813 connections (as happens on some IP stacks) was logged at start up time as
6814 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6815 it from "IPv6 and IPv4", which means that two separate sockets are being
6818 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6819 says whether AF_INET or AF_INET6 was passed as an argument.
6821 42. Exiwhat output was messed up when time zones were included in log
6824 43. Exiwhat now gives more information about the daemon's listening ports,
6825 and whether -tls-on-connect was used.
6827 44. The "port" option of the smtp transport is now expanded.
6829 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6830 silently ignored. Now an error message is written to the main and panic
6833 46. There's a new ACL modifier called "logwrite" which writes to a log file
6834 as soon as it is encountered.
6836 47. Added $local_user_uid and $local_user_gid at routing time.
6838 48. Exim crashed when trying to verify a sender address that was being
6841 49. Exim was recognizing only a space character after ".include". It now also
6842 recognizes a tab character.
6844 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6845 extracting the relevant information from the specification. The man page no
6846 longer contains scrambled data for the -d option, and I've added a section
6847 at the front about calling Exim under different names.
6849 51. Added "extra_headers" argument to the "mail" command in filter files.
6851 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6854 53. Installed eximstats 1.29.
6856 54. Added transport_filter_timeout as a generic transport option.
6858 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6859 Cc: header lines. This was required by RFC 822, but it not required by RFC
6862 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6863 incoming messages that did not have them. Now it does so only if the
6864 message originates locally, that is, if there is no associated remote host
6865 address. When Resent- header lines are present, this applies to the Resent-
6866 lines rather than the non-Resent- lines.
6868 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6869 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6871 58. Messages for configuration errors now include the name of the main
6872 configuration file - useful now that there may be more than one file in a
6873 list (.included file names were always shown).
6875 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6876 for those rare installations that do not start the daemon as root or run it
6877 setuid root. I've cut out the call to initgroups() if the daemon is not
6880 60. The Exim user and group can now be bound into the binary as text strings
6881 that are looked up at the start of Exim's processing.
6883 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6885 62. Added $mailstore_basename variable.
6887 63. Installed patch to sieve.c from Michael Haardt.
6889 64. When Exim failed to open the panic log after failing to open the main log,
6890 the original message it was trying to log was written to stderr and debug
6891 output, but if they were not available (the usual case in production), it
6892 was lost. Now it is written to syslog before the two lines that record the
6893 failures to open the logs.
6895 65. Users' Exim filters run in subprocesses under the user's uid. It is
6896 possible for a "deliver" command or an alias in a "personal" command to
6897 provoke an address rewrite. If logging of address rewriting is configured,
6898 this fails because the process is not running as root or exim. There may be
6899 a better way of dealing with this, but for the moment (because 4.30 needs
6900 to be released), I have disabled address rewrite logging when running a
6901 filter in a non-root, non-exim process.
6907 1. The buildconfig auxiliary program wasn't quoting the value set for
6908 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6909 not defined. This bug was masked in 4.22 by the effect that was fixed in
6912 2. Some messages that were rejected after a message id was allocated were
6913 shown as "incomplete" by exigrep. It no longer does this for messages that
6914 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6916 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6917 have allow_domain_literals set, the ID did not get logged in the <= line.
6918 Domain literals are now always recognized in Message-ID: header lines.
6920 4. The first argument for a ${extract expansion item is the key name or field
6921 number. Leading and trailing spaces in this item were not being ignored,
6922 causing some misleading effects.
6924 5. When deliver_drop_privilege was set, single queue runner processes started
6925 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6926 same command in the process it spins off) were not dropping privilege.
6928 6. When the daemon running as "exim" started a queue runner, it always
6929 re-executed Exim in the spun-off process. This is a waste of effort when
6930 deliver_drop_privilege is set. The new process now just calls the
6931 queue-runner function directly.
6937 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6940 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6941 ignored. Though the use of -oP was forcing the writing of a pid file, it
6942 was always written to the default place.
6944 3. If the message "no IP address found for host xxxx" is generated during
6945 incoming verification, it is now followed by identification of the incoming
6946 connection (so you can more easily find what provoked it).
6948 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6950 5. Added some features to "harden" Exim a bit more against certain attacks:
6952 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6953 be put in Local/Makefile. This is like the never_users runtime option,
6954 but it cannot be overridden. The default setting is "root".
6956 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6957 prefix string with which any file named in a -C command line option
6960 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6961 is retained for -C and -D only if the caller of Exim is root. Without
6962 it, the exim user may also use -C and -D and retain privilege.
6964 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6965 command line option is disabled.
6967 6. Macro names set by the -D option must start with an upper case letter, just
6968 like macro names defined in the configuration file.
6970 7. Added "dereference=" facility to LDAP.
6972 8. Two instances of the typo "uknown" in the source files are fixed.
6974 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6975 the Configure-Makefile script screwed up while processing it.
6977 10. Incorporated PCRE 4.4.
6979 11. The SMTP synchronization check was not operating right at the start of an
6980 SMTP session. For example, it could not catch a HELO sent before the client
6981 waited for the greeting. There is now a check for outstanding input at the
6982 point when the greeting is written. Because of the duplex, asynchronous
6983 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6984 way, but not yet received, when the check is performed.
6986 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6987 on TCP/IP sockets, because this apparently causes some broken clients to
6990 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6991 unchanged) from the Cygwin maintainer.
6993 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6994 was supported instead of testing for mbx. Effectively a typo.
6996 15. The spa authenticator server code was not checking that the input it
6997 received was valid base64.
6999 16. The debug output line for the "set" modifier in ACLs was not showing the
7000 name of the variable that was being set.
7002 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7004 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7005 lookup on its own. Something else was needed to trigger the lookup. For
7006 example, a match in host_lookup or the need for a host name in a host list.
7007 Now, if $sender_host_name is referenced and the host name has not yet been
7008 looked up, a lookup is performed. If the lookup fails, the variable remains
7009 empty, and $host_lookup_failed is set to "1".
7011 19. Added "eqi" as a case-independent comparison operator.
7013 20. The saslauthd authentication condition could segfault if neither service
7014 nor realm was specified.
7016 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7017 error message that was logged was misleading, and incoming SMTP
7018 connections were dropped. The message is now more accurate, and temporary
7019 errors are given to SMTP connections.
7021 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7022 (except RSET) with a 421 error, until QUIT is received. However, it was
7023 failing to send a response to QUIT.
7025 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7026 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7027 ACL is now run after the helo_try_verify_hosts code.
7029 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7030 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7031 case-independent, so other case variants are also recognized). Apparently
7032 some systems use these upper case variants.
7034 25. If more than two messages were waiting for the same host, and a transport
7035 filter was specified for the transport, Exim sent two messages over the
7036 same TCP/IP connection, and then failed with "socket operation on non-
7037 socket" when it tried to send the third.
7039 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7041 27. The extern definition of crypt16() in expand.c was not being excluded when
7042 the OS had its own crypt16() function.
7044 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7045 as a preferred synonym for return_size_limit, both as an option and as an
7048 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7050 30. Changed the default configuration ACL to relax the local part checking rule
7051 for addresses that are not in any local domains. For these addresses,
7052 slashes and pipe symbols are allowed within local parts, but the sequence
7053 /../ is explicitly forbidden.
7055 31. SPA server authentication was not clearing the challenge buffer before
7058 32. log_message in a "warn" ACL statement was writing to the reject log as
7059 well as to the main log, which contradicts the documentation and doesn't
7060 seem right (because no rejection is happening). So I have stopped it.
7062 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7063 However, I am unable to do any testing of this.
7065 34. Fixed an infelicity in the appendfile transport. When checking directories
7066 for a mailbox, to see if any needed to be created, it was accidentally
7067 using path names with one or more superfluous leading slashes; tracing
7068 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7070 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7071 contents are added to the log line that is written for every discarded
7072 recipient. (Previously a log_message setting was ignored.)
7074 36. The ${quote: operator now quotes the string if it is empty.
7076 37. The install script runs exim in order to find its version number. If for
7077 some reason other than non-existence or emptiness, which it checks, it
7078 could not run './exim', it was installing it with an empty version number,
7079 i.e. as "exim-". This error state is now caught, and the installation is
7082 38. An argument was missing from the function that creates an error message
7083 when Exim fails to connect to the socket for saslauthd authentication.
7084 This could cause Exim to crash, or give a corrupted message.
7086 39. Added isip, isip4, and isip6 to ${if conditions.
7088 40. The ACL variables $acl_xx are now saved with the message, and can be
7089 accessed later in routers, transports, and filters.
7091 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7092 strings in the file are not string-expanded.
7094 42. If a MAIL command specified a SIZE value that was too large to fit into an
7095 int variable, the check against message_size_limit failed. Such values are
7096 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7097 day this will have to be increased, but I don't think I want to be around
7098 when emails are that large.
7105 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7106 iconv() is not standard in FreeBSD.
7108 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7109 IPv6 enabled. The observed symptom was a segmentation fault on return from
7110 the function os_common_find_running_interfaces() in src/os.c.
7112 3. In the check_special_case() function in daemon.c I had used "errno" as an
7113 argument name, which causes warnings on some systems. This was basically a
7114 typo, since it was named "eno" in the comments!
7116 4. The code that waits for the clock to tick (at a resolution of some fraction
7117 of a second) so as to ensure message-id uniqueness was always waiting for
7118 at least one whole tick, when it could have waited for less. [This is
7119 almost certainly not relevant at current processor speeds, where it is
7120 unlikely to ever wait at all. But we try to future-proof.]
7122 5. The function that sleeps for a time interval that includes fractions of a
7123 second contained a race. It did not block SIGALRM between setting the
7124 timer, and suspending (a couple of lines later). If the interval was short
7125 and the sigsuspend() was delayed until after it had expired, the suspension
7126 never ended. On busy systems this could lead to processes getting stuck for
7129 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7130 process, before it forks any delivery processes. The open lookup caching
7131 mechanism meant that the open file or database connection was passed into
7132 the delivery process. The problem was that delivery processes always tidy
7133 up cached lookup data. This could cause a problem for the next delivery
7134 process started by the queue runner, because the external queue runner
7135 process does not know about the closure. So the next delivery process
7136 still has data in the lookup cache. In the case of a file lookup, there was
7137 no problem because closing a file descriptor in a subprocess doesn't affect
7138 the parent. However, if the lookup was caching a connection to a database,
7139 the connection was closed, and the second delivery process was likely to
7140 see errors such as "PGSQL: query failed: server closed the connection
7141 unexpectedly". The problem has been fixed by closing all cached lookups
7142 in a queue runner before running a delivery process.
7144 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7145 seem to have the "const" qualifier which it has on other OS. I've
7148 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7149 *of the same type* (client or server) with the same public name that an
7150 error should be diagnosed.
7152 9. When Exim looked up a host name for an IP address, but failed to find the
7153 original IP address when looking up the host name (a safety check), it
7154 output the message "<ip address> does not match any IP for NULL", which was
7155 confusing, to say the least. The bug was that the host name should have
7156 appeared instead of "NULL".
7158 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7159 user that is built into the binary, and the -C or -D options is used, root
7160 privilege is dropped before the configuration file is read. In addition,
7161 logging is switched to stderr instead of the normal log files. If the
7162 configuration then re-defines the Exim user, the unprivileged environment
7163 is probably not what is expected, so Exim logs a panic warning message (but
7166 However, if deliver_drop_privilege is set, the unprivileged state may well
7167 be exactly what is intended, so the warning has been cut out in that case,
7168 and Exim is allowed to try to write to its normal log files.
7174 1. smtp_return_error_details was not giving details for temporary sender
7175 or receiver verification errors.
7177 2. Diagnose a configuration error if two authenticators have the same public
7180 3. Exim used not to create the message log file for a message until the first
7181 delivery attempt. This could be confusing when incoming messages were held
7182 for policy or load reasons. The message log file is now created at the time
7183 the message is received, and an initial "Received" line is written to it.
7185 4. The automatically generated man page for command line options had a minor
7186 bug that caused no ill effects; however, a more serious problem was that
7187 the procedure for building the man page automatically didn't always
7188 operate. Consequently, release 4.20 contains an out-of-date version. This
7189 shouldn't happen again.
7191 5. When building Exim with embedded Perl support, the script that builds the
7192 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7193 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7195 6. The freeze_tell option was not being used for messages that were frozen on
7196 arrival, either by an ACL or by local_scan().
7198 7. Added the smtp_incomplete_transaction log selector.
7200 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7201 was accepting AUTH without a new EHLO.
7203 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7204 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7207 10. Logging of TCP/IP connections (when configured) now happens in the main
7208 daemon process instead of the child process, so that the TCP/IP connection
7209 count is more accurate (but it can never be perfect).
7211 11. The use of "drop" in a nested ACL was not being handled correctly in the
7212 outer ACL. Now, if condition failure induced by the nested "drop" causes
7213 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7214 or "require"), the connection is dropped.
7216 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7217 that yield "discard" can now be used with an "accept" or a "discard" verb,
7218 but an error is generated for any others (because I can't see a useful way
7219 to define what should happen).
7221 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7222 are now processed in the same way as lines in the Exim configuration file.
7223 In particular, continuation lines are supported.
7225 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7227 15. Added -ti meaning -t -i.
7229 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7230 domains, and warn by logging if others are found.
7232 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7233 returned from GIFCONF when Exim is trying to find the list of interfaces on
7234 a host. The code in os.c has been modified to copy each ifreq to an aligned
7235 structure in all cases.
7237 Also, in some cases, the returned ifreq's were being copied to a 'struct
7238 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7239 means the last couple of bytes of an IPv6 address could be chopped if the
7240 ifreq contained only a normal sockaddr (14 bytes storage).
7242 18. Named domain lists were not supported in the hosts_treat_as_local option.
7243 An entry such as +xxxx was not recognized, and was treated as a literal
7246 19. Ensure that header lines added by a DATA ACL are included in the reject log
7247 if the ACL subsequently rejects the message.
7249 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7250 MD5 (which is deprecated).
7252 21. When testing a filter file using -bf, Exim was writing a message when it
7253 took the sender from a "From " line in the message, but it was not doing so
7254 when it took $return_path from a Return-Path: header line. It now does.
7256 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7257 with a valid header line field name (a series of printing characters
7258 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7260 23. Changed "disc" in the source to "disk" to conform to the documentation and
7261 the book and for uniformity.
7263 24. Ignore Sendmail's -Ooption=value command line item.
7265 25. When execve() failed while trying to run a command in a pipe transport,
7266 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7267 could be confused with a return value of 69 from the command itself. This
7268 has been changed to 127, the value the shell returns if it is asked to run
7269 a non-existent command. The wording for the related log line suggests a
7270 non-existent command as the problem.
7272 26. If received_header_text expands to an empty string, do not add a Received:
7273 header line to the message. (Well, it adds a token one on the spool, but
7274 marks it "old" so that it doesn't get used or transmitted.)
7276 27. Installed eximstats 1.28 (addition of -nt option).
7278 28. There was no check for failure on the call to getsockname() in the daemon
7279 code. This can fail if there is a shortage of resources on the system, with
7280 ENOMEM, for example. A temporary error is now given on failure.
7282 29. Contrary to the C standard, it seems that in some environments, the
7283 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7284 program. Exim now does this explicitly; it affects the formatting of
7285 timestamps using strftime().
7287 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7288 complaints. I've now initialized all the variables, to avoid this.
7290 32. Header lines added by a system filter were not being "seen" during
7291 transport-time rewrites.
7293 33. The info_callback() function passed to OpenSSL is set up with type void
7294 (*)(SSL *, int, int), as described somewhere. However, when calling the
7295 function (actually a macro) that sets it up, the type void(*)() is
7296 expected. I've put in a cast to prevent warnings from picky compilers.
7298 34. If a DNS black list lookup found a CNAME record, but there were no A
7299 records associated with the domain it pointed at, Exim crashed.
7301 35. If a DNS black list lookup returned more than one A record, Exim ignored
7302 all but the first. It now scans all returned addresses if a particular IP
7303 value is being sought. In this situation, the contents of the
7304 $dnslist_value variable are a list of all the addresses, separated by a
7307 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7308 to accept a host name and all its aliases if the forward lookup for any of
7309 them yielded the IP address of the incoming connection. Now it accepts only
7310 those names whose forward lookup yields the correct IP address. Any other
7311 names are discarded. This closes a loophole whereby a rogue DNS
7312 administrator could create reverse DNS records to break through a
7313 wildcarded host restriction in an ACL.
7315 37. If a user filter or a system filter that ran in a subprocess used any of
7316 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7317 the wrong values were passed to the pipe command ($thisaddress had the
7318 value of $0, $0 had the value of $1, etc). This bug was introduced by
7319 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7321 38. Improved the line breaking for long SMTP error messages from ACLs.
7322 Previously, if there was no break point between 40 and 75 characters, Exim
7323 left the rest of the message alone. Two changes have been made: (a) I've
7324 reduced the minimum length to 35 characters; (b) if it can't find a break
7325 point between 35 and 75 characters, it looks ahead and uses the first one
7326 that it finds. This may give the occasional overlong line, but at least the
7327 remaining text gets split now.
7329 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7330 file descriptors might be low, and that setting 1000 would always raise it.
7331 It turns out that in some environments, the limit is already over 1000 and
7332 that lowering it causes trouble. So now Exim takes care not to decrease it.
7334 40. When delivering a message, the value of $return_path is set to $sender_
7335 address at the start of routing (routers may change the value). By an
7336 oversight, this default was not being set up when an address was tested by
7337 -bt or -bv, which affected the outcome if any router or filter referred to
7340 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7341 log only when "message" or "log_message" are set. However, if one of the
7342 conditions was an address verification, or a call to a nested ACL, the
7343 messages generated by the underlying test were being passed through. This
7344 no longer happens. The underlying message is available in $acl_verify_
7345 message for both "message" and "log_message" expansions, so it can be
7346 passed through if needed.
7348 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7349 new expansion $bh_ to give the encoded byte string without charset
7350 translation. Translation happens only if iconv() is available; HAVE_ICONV
7351 indicates this at build time. HEADERS_CHARSET gives the charset to
7352 translate to; headers_charset can change it in the configuration, and
7353 "headers charset" can change it in an individual filter file.
7355 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7356 that creates RFC 2047 encoded "words" labels them as that charset instead
7357 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7358 expansion operator; (ii) when Exim creates a From: line for a local
7359 message; (iii) when a header line is rewritten to include a "phrase" part.
7361 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7362 buggy, causing it to skip the first lines of messages whose message ID
7363 ended in 'D'. This would not have bitten before Exim release 4.14, because
7364 message IDs were unlikely to end in 'D' before then. The effect was to have
7365 incorrect size information for certain domains.
7367 45. #include "config.h" was missing at the start of the crypt16.c module. This
7368 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7371 46. If there was a timeout during a "random" callout check, Exim treated it as
7372 a failure of the random address, and carried on sending RSET and the real
7373 address. If the delay was just some slowness somewhere, the response to the
7374 original RCPT would be taken as a response to RSET and so on, causing
7375 mayhem of various kinds.
7377 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7378 when I implemented it. It didn't allow for the fact that some option values
7379 may legitimately be negative (e.g. size_addition), and it didn't even do
7380 the right test for positive values.
7382 48. Domain names in DNS records are case-independent. Exim always looks them up
7383 in lower case. Some resolvers return domain names in exactly the case they
7384 appear in the zone file, that is, they may contain uppercase letters. Not
7385 all resolvers do this - some return always lower case. Exim was treating a
7386 change of case by a resolver as a change of domain, similar to a widening
7387 of a domain abbreviation. This triggered its re-routing code and so it was
7388 trying to route what was effectively the same domain again. This normally
7389 caused routing to fail (because the router wouldn't handle the domain
7390 twice). Now Exim checks for this case specially, and just changes the
7391 casing of the domain that it ultimately uses when it transmits the message
7394 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7397 50. If a filter generated a file delivery with a non-absolute name (possible if
7398 no home directory exists for the router), the forbid_file option was not
7401 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7402 the existing equality matching.
7404 52. Exim was using ints instead of ino_t variables in some places where it was
7405 dealing with inode numbers.
7407 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7408 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7409 called TMPDIR, and if it finds it is different, it changes its value.
7411 54. The smtp_printf() function is now made available to local_scan() so
7412 additional output lines can be written before returning. There is also an
7413 smtp_fflush() function to enable the detection of a dropped connection.
7414 The variables smtp_input and smtp_batched_input are exported to
7417 55. Changed the default runtime configuration: the message "Unknown user"
7418 has been removed from the ACL, and instead placed on the localuser router,
7419 using the cannot_route_message feature. This means that any verification
7420 failures that generate their own messages won't get overridden. Similarly,
7421 the "Unrouteable address" message that was in the ACL for unverifiable
7422 relay addresses has also been removed.
7424 56. Added hosts_avoid_esmtp to the smtp transport.
7426 57. The exicyclog script was not checking for the esoteric option
7427 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7428 will work only if exicyclog is run under the appropriate euid.
7430 58. Following a discussion on the list, the rules by which Exim recognises line
7431 endings on incoming messages have been changed. The -dropcr and drop_cr
7432 options are now no-ops, retained only for backwards compatibility. The
7433 following line terminators are recognized: LF CRLF CR. However, special
7434 processing applies to CR:
7436 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7437 nor a local message in the state where . is a terminator.
7439 (ii) If a bare CR is encountered in a header line, an extra space is added
7440 after the line terminator so as not to end the header. The reasoning
7441 behind this is that bare CRs in header lines are most likely either
7442 to be mistakes, or people trying to play silly games.
7444 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7445 was being incorrectly given as 18 bytes larger than it should have been.
7446 This is a VOB (very old bug).
7448 60. This may never have affected anything current, but just in case it has:
7449 When the local host is found other than at the start of a list of hosts,
7450 the local host, those with the same MX, and any that follow, are discarded.
7451 When the list in question was part of a longer list of hosts, the following
7452 hosts (not currently being processed) were also being discarded. This no
7453 longer happens. I'm not sure if this situation could ever has previously
7456 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7459 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7460 header. This is something that is recommended in a new Internet Draft, and
7461 is something that is documented as being done by Sendmail. There are two
7462 possible values. For messages generated by the autoreply transport, Exim
7465 Auto-Submitted: auto-replied
7467 whereas for all other generated messages (e.g. bounces) it adds
7469 Auto-Submitted: auto-generated
7471 63. The "personal" condition in filters now includes a test for the
7472 Auto-Submitted: header. If it contains the string "auto-" the message it
7473 not considered personal.
7475 64. Added rcpt_include_affixes as a generic transport option.
7477 65. Added queue_only_override (default true).
7479 66. Added the syslog_duplication option.
7481 67. If what should have been the first header line of a message consisted of
7482 a space followed by a colon, Exim was mis-interpreting it as a header line.
7483 It isn't of course - it is syntactically invalid and should therefore be
7484 treated as the start of the message body. The misbehaviour could have
7485 caused a number of strange effects, including loss of data in subsequent
7486 header lines, and spool format errors.
7488 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7489 client host had authenticated. This control can now be exercised by an ACL
7490 for more flexibility.
7492 69. By default, callouts do not happen when testing with -bh. There is now a
7493 variant, -bhc, which does actually run the callout code, including
7494 consulting and updating the callout cache.
7496 70. Added support for saslauthd authentication, courtesy of Alexander
7499 71. If statvfs() failed on the spool or log directories while checking their
7500 size for availability, Exim confusingly gave the error "space shortage".
7501 Furthermore, in debugging mode it crashed with a floating point exception.
7502 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7503 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7504 serious problem, Exim now writes to the main and panic logs when this
7505 happens, with details of the failure. It then refuses to accept the
7506 incoming message, giving the message "spool directory problem" or "log
7507 directory problem" with a 421 code for SMTP messages.
7509 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7510 0, 1, and 2 exist, because some OS complain for execs without them (see
7511 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7512 descriptors. However, the code omitted to check that the open succeeded,
7513 causing mysterious errors if for some reason the permissions on /dev/null
7514 got screwed. Now Exim writes a message to the main and panic logs, and
7515 bombs out if it can't open /dev/null.
7517 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7518 interact so that it is all more flexible. It is supposed to remain
7519 backwards compatible. Also added extra_local_interfaces.
7521 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7522 to bomb out with an assertion failure - to the client this appears as a
7523 connection drop. This problem occurs in the part of the code that was taken
7524 from the Samba project. Fortunately, the assertion is in a very simple
7525 function, so I have fixed this by reproducing the function inline in the
7526 one place where it is called, and arranging for authentication to fail
7527 instead of killing the process with assert().
7529 75. The SPA client code was not working when the server requested OEM rather
7530 than Unicode encoding.
7532 76. Added code to make require_files with a specific uid setting more usable in
7533 the case where statting the file as root fails - usually a non-root-mounted
7534 NFS file system. When this happens and the failure is EACCES, Exim now
7535 forks a subprocess and does the per-uid checking as the relevant uid.
7537 77. Added process_log_path.
7539 78. If log_file_path was not explicitly set, a setting of check_log_space or
7540 check_log_inodes was ignored.
7542 79. If a space check for the spool or log partitions fails, the incident is now
7543 logged. Of course, in the latter case the data may get lost...
7545 80. Added the %p formatting code to string_format() so that it can be used to
7546 print addresses in debug_print(). Adjusted all the address printing in the
7547 debugging in store.c to use %p rather than %d.
7549 81. There was a concern that a line of code in smtp_in.c could overflow a
7550 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7551 initially expressed, the concern was not well-founded, because trailing
7552 spaces are removed early. However, if the trailing spaces were followed by
7553 a NULL, they did not get removed, so the overflow was possible. Two fixes
7556 (a) I re-wrote the offending code in a cleaner fashion.
7557 (b) If an incoming SMTP command contains a NULL character, it is rejected
7560 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7561 runs initgroups(), so that if the Exim user is in any additional groups,
7562 they will be used during message reception.
7568 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob