1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Avoid costly startup code when not strictly needed. This reduces time
12 for some exim process initialisations. It does mean that the logging
13 of TLS configuration problems is only done for the daemon startup.
15 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
17 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
18 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
21 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
22 DKIM signing is being done. Previously a transport_filter would always
23 disable CHUNKING, falling back to traditional DATA.
29 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
30 side, to decode a TLS 1.3 packet capture.
32 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
33 Previously the default library behaviour applied, sending two, each in
36 JH/03 Debug output for ACL now gives the config file name and line number for
39 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
41 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
43 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
44 buffer overrun for (non-chunking) other transports.
46 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
47 TLS1.3, means that a server rejecting a client certificate is not visible
48 to the client until the first read of encrypted data (typically the
49 response to EHLO). Add detection for that case and treat it as a failed
50 TLS connection attempt, so that the normal retry-in-clear can work (if
53 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
54 and/or domain. Found and fixed by Jason Betts.
56 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
57 configuration). If a CNAME target was not a wellformed name pattern, a
60 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
61 the OS reports them interleaved with other addresses.
63 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
64 used both for input and for a verify callout, both encrypted, SMTP
65 responses being sent by the server could be lost. This resulted in
66 dropped connections and sometimes bounces generated by a peer sending
69 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
70 string. Previously it was possible to cause undefined behaviour in a
71 library routine (usually a crash). Found by "zerons".
73 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
76 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
77 API was removed, so update to use the newer ones.
79 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
80 any timeout set, is taking a long time. Previously we would hang on to a
81 rotated logfile "forever" if the input was arriving with long gaps
82 (a previous attempt to fix addressed lack, for a long time, of initial
85 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
86 shared (NFS) environment. The length of the tempfile name is now
87 4 + 16 ("hdr.$message_exim_id") which might break on file
88 systems which restrict the file name length to lower values.
91 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
92 shared (NFS) environment.
94 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
95 did for all versions <4.90). Notably -M, -m, --invert, -I may be
98 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
99 on some platforms for bit 31.
101 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
102 to changes apparently associated with TLS1.3 handling some of the APIs
103 previously used were either nonfunctional or inappropriate. Strings
104 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
105 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
106 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
107 This affects log line X= elements, the $tls_{in,out}_cipher variables,
108 and the use of specific cipher names in the encrypted= ACL condition.
110 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
112 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
113 verification result was not updated unless hosts_require_ocsp applied.
115 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
116 queue_list_requires_admin set to false, non-admin users were denied the
119 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
120 directory-of-certs mode. Previously they were advertised despite the
123 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
124 A single TCP connection by a client will now hold a TLS connection open
125 for multiple message deliveries, by default. Previoud the default was to
128 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
129 default. If built with the facility, DANE will be used. The facility
130 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
132 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
133 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
134 must be defined and you must still, unless you define DISABLE_TLS, manage
135 the the include-dir and library-file requirements that go with that
136 choice. Non-TLS builds are still supported.
138 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
141 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
142 default. If the platform supports and has the facility enabled, it will
143 be requested on all coneections.
145 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
146 controlled by the build-time option SUPPORT_PIPE_CONNECT.
148 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
150 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
151 success-DSN messages. Previously the From: header was always the default
152 one for these; the option was ignored.
154 JH/28 Fix the timeout on smtp response to apply to the whole response.
155 Previously it was reset for every read, so a teergrubing peer sending
156 single bytes within the time limit could extend the connection for a
157 long time. Credit to Qualsys Security Advisory Team for the discovery.
159 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
160 delivery address, which leaked information of the results of local
161 forwarding. Change to the original envelope recipient address, per
164 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
165 requested. Previously not bounce was generated and a log entry of
166 error ignored was made.
168 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
170 JH/32 Introduce a general tainting mechanism for values read from the input
171 channel, and values derived from them. Refuse to expand any tainted
172 values, to catch one form of exploit.
174 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
175 was unused and the unexpanded text used for the test. Found and
176 fixed by Ruben Jenster.
178 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
179 an attempt to use a TLS library read routine dereffed a nul pointer,
182 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
183 them in our smtp response.
185 JH/36 Have the general router option retry_use_local_part default to true when
186 any of the restrictive preconditions are set (to anything). Previously it
187 was only for check_local user. The change removes one item of manual
188 configuration which is required for proper retries when a remote router
189 handles a subset of addresses for a domain.
191 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
192 link count into consideration.
194 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
195 caused the extension of big_buffer, the following lines were ignored.
197 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
198 accordance with RFC 2308. Previously there was no expiry, so a longlived
199 receive process (eg. due to ACL delays) versus a short SOA value could
202 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
204 JH/39 Promote DMARC support to mainline.
206 JH/40 Bug 2452: Add a References: header to DSNs.
208 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
209 parameters. The relevant library call is documented as "Deprecated: This
210 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
211 3.6.0, DH parameters are negotiated following RFC7919."
217 JH/01 Remove code calling the customisable local_scan function, unless a new
218 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
220 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
221 non-signal-safe functions being used.
223 JH/03 Bug 2269: When presented with a received message having a stupidly large
224 number of DKIM-Signature headers, disable DKIM verification to avoid
225 a resource-consumption attack. The limit is set at twenty.
227 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
228 report of oldest_pass in ${authres } in consequence, and separate out
229 some descriptions of reasons for verification fail.
231 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
232 files in the spool were present and unlocked. A queue-runner could spot
233 them, resulting in a duplicate delivery. Fix that by doing the unlock
234 after the unlink. Investigation by Tim Stewart. Take the opportunity to
235 add more error-checking on spoolfile handling while that code is being
238 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
239 No known attacks, no CVE, this is defensive hardening.
241 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
242 a queue-runner could start a delivery while other operations were ongoing.
243 Cutthrough delivery was a common victim, resulting in duplicate delivery.
244 Found and investigated by Tim Stewart. Fix by using the open message data
245 file handle rather than opening another, and not locally closing it (which
246 releases a lock) for that case, while creating the temporary .eml format
247 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
249 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
250 $sender_verify_failure/$recipient_verify_failure to "random".
252 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
255 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
256 Previously this would segfault.
258 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
261 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
262 like zero, since the resolver should be doing this for us, But we need one
263 as a CNAME but no MX presence gets the CNAME returned; we need to check
264 that doesn't point to an MX to declare it "no MX returned" rather than
265 "error, loop". A new main option is added so the older capability of
266 following some limited number of chain links is maintained.
268 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
270 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
271 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
272 now-deprecated earlier definitions used only specified the range up to TLS
273 1.2 (in the older-version library docs).
275 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
277 JH/15 Rework TLS client-side context management. Stop using a global, and
278 explicitly pass a context around. This enables future use of TLS for
279 connections to service-daemons (eg. malware scanning) while a client smtp
280 connection is using TLS; with cutthrough connections this is quite likely.
282 JH/16 Fix ARC verification to do AS checks in reverse order.
284 JH/17 Support a "tls" option on the ${readsocket } expansion item.
286 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
287 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
288 Previously the "utf8" would be re-prepended for every additional message.
290 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
291 Previously thery were accepted, resulting in issues when attempting to
292 forward messages to a non-supporting MTA.
294 PP/02 Let -n work with printing macros too, not just options.
296 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
297 one parent address was copied, and bogus data was used at delivery-logging
298 time. Either a crash (after delivery) or bogus log data could result.
299 Discovery and analysis by Tim Stewart.
301 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
302 Previously if the string ended mid-character, we did not insert the
303 promised '?' replacement.
305 PP/04 Documentation: current string operators work on bytes, not codepoints.
307 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
308 should pack well giving a smaller memory footprint so better caching and
309 therefore performance. Group the declarations where this can't be done so
310 that the byte-sized flag variables are not interspersed among pointer
311 variables, giving a better chance of good packing by the compiler.
313 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
314 non-null, to avoid issues with sites running BATV. Previously reports were
315 sent with an empty envelope sender so looked like bounces.
317 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
318 The ignore_error flag wasn't being returned from the filter subprocess so
319 was not set for later routers. Investigation and fix by Matthias Kurz.
321 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
322 and a msg:complete for the whole, when a message is manually removed using
323 -Mrm. Developement by Matthias Kurz, hacked on by JH.
325 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
326 a "Gnu special" function, asprintf() in the DB utility binary builds; I
327 hope that is portable enough.
329 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
330 requiring a known-CA anchor certificate; make it now rely entirely on the
331 TLSA as an anchor. Checking the name on the leaf cert against the name
332 on the A-record for the host is still done for TA (but not for EE mode).
334 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
335 included in delivery lines for non-proxied connections, when compiled with
336 SUPPORT_SOCKS and running with proxy logging enabled.
338 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
339 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
340 move the existing event to fire before the normal logging of message
341 failure so that custom logging is bracketed by normal logging.
343 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
344 msg:fail:internal event. Developement by Matthias Kurz.
346 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
347 far too small for todays use of crypto signatures stored there. Go all
348 the way to the max DNS message size of 64kB, even though this might be
349 overmuch for IOT constrained device use.
351 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
352 copy a string over itself. The library routine is documented as not
353 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
355 JH/32 For main options check_spool_space and check_inode_space, where the
356 platform supports 64b integers, support more than the previous 2^31 kB
357 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
358 the previous G, M, k.
360 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
361 $authenticated_fail_id variable on authentication failure. Previously
364 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
365 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
366 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
367 GNUTLS_SEC_PARAM_MEDIUM.
369 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
370 side. Previously we would continue as if no SNI had been received.
372 JH/36 Harden the handling of string-lists. When a list consisted of a sole
373 "<" character, which should be a list-separator specification, we walked
374 off past the nul-terimation.
376 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
377 causes) even when the retry time is not yet met. Previously they were
378 not, meaning that when (say) an account was over-quota and temp-rejecting,
379 and multiple senders' messages were queued, only one sender would get
380 notified on each configured delay_warning cycle.
382 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
384 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
385 cipher-suites, an error can be left on the stack even for a succeeding
386 accept; this results in impossible error messages when a later operation
389 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
390 return error codes indicating retry. Under TLS1.3 this becomes required.
392 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
393 it only wrote the new authenticators, resulting in a lack of tracking of
394 peer changes of ESMTP extensions until the next cache flush.
396 JH/41 Fix the loop reading a message header line to check for integer overflow,
397 and more-often against header_maxsize. Previously a crafted message could
398 induce a crash of the recive process; now the message is cleanly rejected.
400 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
401 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
407 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
408 When redis_servers is set to a list of > 1 element, and the Redis servers
409 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
410 case of MOVED into a DEFER case instead, thus moving the query onto the
411 next server in the list. For a cluster of N elements, all N servers must
412 be defined in redis_servers.
414 GF/02 Catch and remove uninitialized value warning in exiqsumm
415 Check for existence of @ARGV before looking at $ARGV[0]
417 JH/01 Replace the store_release() internal interface with store_newblock(),
418 which internalises the check required to safely use the old one, plus
419 the allocate and data copy operations duplicated in both (!) of the
420 extant use locations.
422 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
423 modifier. This matches the restriction on the commandline.
425 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
426 Previously only the last row was returned.
428 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
429 we assumed that tags in the header were well-formed, and parsed the
430 element content after inspecting only the first char of the tag.
431 Assumptions at that stage could crash the receive process on malformed
434 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
435 While running the DKIM ACL we operate on the Permanent memory pool so that
436 variables created with "set" persist to the DATA ACL. Also (at any time)
437 DNS lookups that fail create cache records using the Permanent pool. But
438 expansions release any allocations made on the current pool - so a dnsdb
439 lookup expansion done in the DKIM ACL releases the memory used for the
440 DNS negative-cache, and bad things result. Solution is to switch to the
441 Main pool for expansions.
442 While we're in that code, add checks on the DNS cache during store_reset,
443 active in the testsuite.
444 Problem spotted, and debugging aided, by Wolfgang Breyha.
446 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
447 When none of the hosts presented to a transport match an already-open
448 connection, close it and proceed with the list. Previously we would
449 queue the message. Spotted by Lena with Yahoo, probably involving
452 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
453 Previously a spurious "250 OK id=" response was appended to the proper
456 JH/08 The "support for" informational output now, which built with Content
457 Scanning support, has a line for the malware scanner interfaces compiled
458 in. Interface can be individually included or not at build time.
460 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
461 by the template makefile "src/EDITME". The "STREAM" support for an older
462 ClamAV interface method is removed.
464 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
465 rows affected is given instead).
467 JH/11 The runtime Berkeley DB library version is now additionally output by
468 "exim -d -bV". Previously only the compile-time version was shown.
470 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
471 SMTP connection. Previously, when one had more receipients than the
472 first, an abortive onward connection was made. Move to full support for
473 multiple onward connections in sequence, handling cutthrough connection
474 for all multi-message initiating connections.
476 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
477 routers. Previously, a multi-recipient message would fail to match the
478 onward-connection opened for the first recipient, and cause its closure.
480 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
481 a timeout on read on a GnuTLS initiating connection, resulting in the
482 initiating connection being dropped. This mattered most when the callout
483 was marked defer_ok. Fix to keep the two timeout-detection methods
486 JH/15 Relax results from ACL control request to enable cutthrough, in
487 unsupported situations, from error to silently (except under debug)
488 ignoring. This covers use with PRDR, frozen messages, queue-only and
491 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
493 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
494 metadata, resulting in a crash in free().
496 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
497 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
498 Broken also in d185889f4, with init system revamp.
500 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
501 Previously we abruptly closed the connection after reading a malware-
502 found indication; now we go on to read the "scan ok" response line,
505 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
506 ACL. Previously, a crash would result.
508 JH/19 Speed up macro lookups during configuration file read, by skipping non-
509 macro text after a replacement (previously it was only once per line) and
510 by skipping builtin macros when searching for an uppercase lead character.
512 JH/20 DANE support moved from Experimental to mainline. The Makefile control
513 for the build is renamed.
515 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
516 was allocated for every new TLS startup, meaning one per message. Fix
517 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
519 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
520 reported the original. Fix to report (as far as possible) the ACL
521 result replacing the original.
523 JH/23 Fix memory leak during multi-message connections using STARTTLS under
524 OpenSSL. Certificate information is loaded for every new TLS startup,
525 and the resources needed to be freed.
527 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
529 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
532 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
533 DATA response info to the (existing) per-recipient response info for
534 the "C=" log element. It can have useful tracking info from the
535 destination system. Patch from Simon Arlott.
537 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
538 length value. Previously this would segfault.
540 HS/02 Support Avast multiline protoocol, this allows passing flags to
541 newer versions of the scanner.
543 JH/28 Ensure that variables possibly set during message acceptance are marked
544 dead before release of memory in the daemon loop. This stops complaints
545 about them when the debug_store option is enabled. Discovered specifically
546 for sender_rate_period, but applies to a whole set of variables.
547 Do the same for the queue-runner and queue-list loops, for variables set
548 from spool message files. Do the same for the SMTP per-message loop, for
549 certain variables indirectly set in ACL operations.
551 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
552 as a multi-recipient message from a mailinglist manager). The coding had
553 an arbitrary cutoff number of characters while checking for more input;
554 enforced by writing a NUL into the buffer. This corrupted long / fast
555 input. The problem was exposed more widely when more pipelineing of SMTP
556 responses was introduced, and one Exim system was feeding another.
557 The symptom is log complaints of SMTP syntax error (NUL chars) on the
558 receiving system, and refused recipients seen by the sending system
559 (propating to people being dropped from mailing lists).
560 Discovered and pinpointed by David Carter.
562 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
563 replaced by the ${authresults } expansion.
565 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
567 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
568 allows proper process termination in container environments.
570 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
571 Previously the "final dot" had a newline after it; ensure it is CR,LF.
573 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
574 and "err_perm", deprecated since 4.83 when the RFC-defined words
575 "temperror" and "permerror" were introduced.
577 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
578 transport-filters or DKIM-signing. The restriction was lost in the
579 consolidation of verify-callout and delivery SMTP handling.
580 Extend the restriction to also cover ARC-signing.
582 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
583 in defer=pass mode supply a 450 to the initiator. Previously the message
586 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
587 tls_require_ciphers is used as before.
589 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
590 "pass_unscanned". Only tmpfails from the scanner are written to
591 the paniclog, as they may require admin intervention (permission
592 denied, license issues). Other scanner errors (like decompression
593 bombs) do not cause a paniclog entry.
595 JH/36 Fix reinitialisation of DKIM logging variable between messages.
596 Previously it was possible to log spurious information in receive log
599 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
600 triggered odd behaviour from Outlook Express clients.
602 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
605 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
606 since the IETF WG has not yet settled on that versus the original
607 "bare" representation.
609 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
610 Previously the millisecond value corrupted the output.
611 Fix also for syslog_pid=no and log_selector +pid, for which the pid
612 corrupted the output.
618 JH/01 Rework error string handling in TLS interface so that the caller in
619 more cases is responsible for logging. This permits library-sourced
620 string to be attached to addresses during delivery, and collapses
621 pairs of long lines into single ones.
623 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
624 during configuration. Wildcards are allowed and expanded.
626 JH/02 Rework error string handling in DKIM to pass more info back to callers.
627 This permits better logging.
629 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
630 do not close it down and have the child transport start it up again on
631 the passed-on TCP connection. Instead, proxy the child (and any
632 subsequent ones) for TLS via a unix-domain socket channel. Logging is
633 affected: the continued delivery log lines do not have any DNSSEC, TLS
634 Certificate or OCSP information. TLS cipher information is still logged.
636 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
637 identical IP addresses on different listening ports. Will also affect
640 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
641 add noisy ifdef guards to special-case this sillyness.
642 Patch from Bernd Kuhls.
644 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
645 than 255 are no longer allowed.
647 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
648 Disable the session-cache too, which might reduce our load. Since we
649 currrectly use a new context for every connection, both as server and
650 client, there is no benefit for these.
651 GnuTLS appears to not support tickets server-side by default (we don't
652 call gnutls_session_ticket_enable_server()) but client side is enabled
653 by default on recent versions (3.1.3 +) unless the PFS priority string
656 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
657 <https://reproducible-builds.org/specs/source-date-epoch/>.
659 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
660 the check for any unsuccessful recipients did not notice the limit, and
661 erroneously found still-pending ones.
663 JH/08 Pipeline CHUNKING command and data together, on kernels that support
664 MSG_MORE. Only in-clear (not on TLS connections).
666 JH/09 Avoid using a temporary file during transport using dkim. Unless a
667 transport-filter is involved we can buffer the headers in memory for
668 creating the signature, and read the spool data file once for the
669 signature and again for transmission.
671 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
672 4.77 as the kernel support then wasn't solid, having issues in 64bit
673 mode. Now, it's been long enough. Add support for FreeBSD also.
675 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
676 case where the routing stage had gathered several addresses to send to
677 a host before calling the transport for the first, we previously failed
678 to close down TLS in the old transport process before passing the TCP
679 connection to the new process. The new one sent a STARTTLS command
680 which naturally failed, giving a failed delivery and bloating the retry
681 database. Investigation and fix prototype from Wolfgang Breyha.
683 JH/12 Fix check on SMTP command input synchronisation. Previously there were
684 false-negatives in the check that the sender had not preempted a response
685 or prompt from Exim (running as a server), due to that code's lack of
686 awareness of the SMTP input buffering.
688 PP/04 Add commandline_checks_require_admin option.
689 Exim drops privileges sanely, various checks such as -be aren't a
690 security problem, as long as you trust local users with access to their
691 own account. When invoked by services which pass untrusted data to
692 Exim, this might be an issue. Set this option in main configuration
693 AND make fixes to the calling application, such as using `--` to stop
696 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
697 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
698 responses to those, into a single TLS record each way (this usually means
699 a single packet). As a side issue, smtp_enforce_sync now works on TLS
702 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
703 affects you only if you're dancing at the edge of the param size limits.
704 If you are, and this message makes sense to you, then: raise the
705 configured limit or use OpenSSL 1.1. Nothing we can do for older
708 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
709 cmdline element to get the documented default one. Previously it was
712 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
713 are unsuable for non-message-specific reasons - eg. connection timeout,
714 banner-time rejection.
716 JH/16 Fix logging of delivery remote port, when specified by router, under
719 PP/06 Repair manualroute's ability to take options in any order, even if one
720 is the name of a transport.
723 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
725 JH/17 Change the list-building routines interface to use the expanding-string
726 triplet model, for better allocation and copying behaviour.
728 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
729 Previously it was constructed the first time a possibly-matching string
730 was met in the configuration file input during startup; now it is done
733 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
734 compatible one, to avoid the (poorly documented) possibility of a config
735 file in the working directory redirecting the DB files, possibly correpting
736 some existing file. CVE-2017-10140 assigned for BDB.
738 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
739 cache-hot. Previously, although the result was properly cached, the
740 initial verify call returned a defer.
742 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
743 the main verify for receipient in uncached-mode.
745 JH/22 Retire historical build files to an "unsupported" subdir. These are
746 defined as "ones for which we have no current evidence of testing".
748 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
749 if present. Previously it was ignored.
751 JH/24 Start using specified-initialisers in C structure init coding. This is
752 a C99 feature (it's 2017, so now considered safe).
754 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
755 if was a fixed-sized field and bitmask ops via macros; it is now more
758 PP/07 GitHub PR 56: Apply MariaDB build fix.
759 Patch provided by Jaroslav Škarvada.
761 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
762 during Coverity cleanups [4.87 JH/47]
763 Diagnosis and fix provided by Michael Fischer v. Mollard.
765 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
766 the right size to place the terminating semicolon on its own folded
767 line, the header hash was calculated to an incorrect value thanks to
768 the (relaxed) space the fold became.
770 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
771 and confused the parent.
773 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
774 which could crash as a result. This could lead to undeliverable messages.
776 JH/28 Logging: "next input sent too soon" now shows where input was truncated
779 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
780 matters on fast-turnover and PID-randomising systems, which were getting
781 out-of-order delivery.
783 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
784 a possibly-overlapping copy. The symptom was that "Remote host closed
785 connection in response to HELO" was logged instead of the actual 4xx
788 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
789 Previously only that bufferd was discarded, resulting in SYMTP command
792 JH/32 DKIM: when a message has multiple signatures matching an identity given
793 in dkim_verify_signers, run the dkim acl once for each. Previously only
794 one run was done. Bug 2189.
796 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
797 "panic the current process" to "deliberately defer". The panic log is
798 still written with the problem list name; the mail and reject logs now
799 get a temp-reject line for the message that was being handled, saying
800 something like "domains check lookup or other defer". The SMTP 451
801 message is still "Temporary local problem".
803 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
804 A crafted sequence of BDAT commands could result in in-use memory beeing
805 freed. CVE-2017-16943.
807 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
808 from SMTP input. Previously it was always done; now only done for DATA
809 and not BDAT commands. CVE-2017-16944.
811 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
812 to the message (such as an overlong header line). Previously this was
813 not done and we did not exit BDAT mode. Followon from the previous item
814 though a different problem.
820 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
821 than -2003 did; needs libidn2 in addition to libidn.
823 JH/02 The path option on a pipe transport is now expanded before use.
825 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
826 Patch provided by "Björn", documentation fix added too.
828 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
829 missing a wire-to-host endian conversion.
831 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
832 close after a BDAT command line could be taken as a following command,
833 giving a synch failure. Fix by only checking for synch immediately
834 before acknowledging the chunk.
836 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
837 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
838 macro. Patches provided by Josh Soref.
840 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
841 Previously we did not; the RFC seems ambiguous and VRFY is not listed
842 by IANA as a service extension. However, John Klensin suggests that we
845 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
846 the dkim code may be unix-mode line endings rather than smtp wire-format
847 CRLF, so prepend a CR to any bare LF.
849 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
850 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
852 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
853 the first were themselves being wrongly included in the feed into dkim
854 processing; with most chunk sizes in use this resulted in an incorrect
855 body hash calculated value.
857 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
858 DKIM signature block, for verification. Although advised against by
859 standards it is specifically not ruled illegal.
861 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
863 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
864 missing a body hash (the bh= tag).
866 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
867 It seems that HAProxy sends the Proxy Protocol information in clear and
868 only then does a TLS startup, so do the same.
870 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
871 TCP connections (such as for Spamd) unless the daemon successfully set
872 Fast Open mode on its listening sockets. This fixes breakage seen on
873 too-old kernels or those not configured for Fast Open, at the cost of
874 requiring both directions being enabled for TFO, and TFO never being used
875 by non-daemon-related Exim processes.
877 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
878 endings, at least on the first header line. Try to canonify any that get
879 past that check, despite the cost.
881 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
882 now limited to an arbitrary five deep, while parsing addresses with the
883 strip_excess_angle_brackets option enabled.
885 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
886 instead leave the unprompted TLS handshake in socket buffer for the
887 TLS library to consume.
889 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
891 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
893 JH/16 Drop variables when they go out of scope. Memory management drops a whole
894 region in one operation, for speed, and this leaves assigned pointers
895 dangling. Add checks run only under the testsuite which checks all
896 variables at a store-reset and panics on a dangling pointer; add code
897 explicitly nulling out all the variables discovered. Fixes one known
898 bug: a transport crash, where a dangling pointer for $sending_ip_address
899 originally assigned in a verify callout, is re-used.
901 PP/06 Drop '.' from @INC in various Perl scripts.
903 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
905 PP/08 Reduce a number of compilation warnings under clang; building with
906 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
907 should be warning-free.
909 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
911 HS/01 Fix portability problems introduced by PP/08 for platforms where
912 realloc(NULL) is not equivalent to malloc() [SunOS et al].
914 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
915 chunk. This allows us to accept broken chunked messages. We need a more
916 general solution here.
918 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
919 already-broken messages in the queue.
921 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
923 JH/19 Fix reference counting bug in routing-generated-address tracking.
929 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
930 supports it and a size is available (ie. the sending peer gave us one).
932 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
933 years of being deprecated). The replacements are the ACLs
934 acl_smtp_mime and acl_not_smtp_mime.
936 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
937 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
938 or even an in-clear connection were permitted. Now, if the host lookup
939 was dnssec and dane was requested then the host is only used if the
940 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
941 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
942 if one fails this test.
943 This means that a poorly-configured remote DNS will make it incommunicado;
944 but it protects against a DNS-interception attack on it.
946 JH/04 Bug 1810: make continued-use of an open smtp transport connection
947 non-noisy when a race steals the message being considered.
949 JH/05 If main configuration option tls_certificate is unset, generate a
950 self-signed certificate for inbound TLS connections.
952 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
953 in rewrites and routers.
955 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
956 and logged a warning sing 4.83; now they are a configuration file error.
958 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
959 (lacking @domain). Apply the same qualification processing as RCPT.
961 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
963 JH/10 Support ${sha256:} applied to a string (as well as the previous
966 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
967 a cutthrough deliver is pending, as we always want to make a connection.
968 This also avoids re-routing the message when later placing the cutthrough
969 connection after a verify cache hit.
970 Do not update it with the verify result either.
972 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
973 when routing results in more than one destination address.
975 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
976 signing (which inhibits the cutthrough capability). Previously only
977 the presence of an option was tested; now an expansion evaluating as
978 empty is permissible (obviously it should depend only on data available
979 when the cutthrough connection is made).
981 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
982 the relevant preceding SMTP command did not note the pipelining mode.
984 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
985 Previously they were not counted.
987 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
988 as one having no matching records. Previously we deferred the message
989 that needed the lookup.
991 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
992 distinguished as "(=".
994 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
995 for missing MX records. Previously it only worked for missing A records.
997 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
999 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1000 after the data-go-ahead and data-ack. Patch from Jason Betts.
1002 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1003 even for a "none" policy. Patch from Tony Meyer.
1005 JH/22 Fix continued use of a connection for further deliveries. If a port was
1006 specified by a router, it must also match for the delivery to be
1009 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1010 When one of the recipients of a message was unsuitable for the connection
1011 (has no matching addresses), we lost track of needing to mark it
1012 deferred. As a result mail would be lost.
1014 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1016 JH/25 Decoding ACL controls is now done using a binary search; the source code
1017 takes up less space and should be simpler to maintain. Merge the ACL
1018 condition decode tables also, with similar effect.
1020 JH/26 Fix problem with one_time used on a redirect router which returned the
1021 parent address unchanged. A retry would see the parent address marked as
1022 delivered, so not attempt the (identical) child. As a result mail would
1025 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1026 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1027 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1030 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1031 Main config options check_{log,spool}_{inodes,space} are now
1032 100 inodes, 10MB unless set otherwise in the configuration.
1034 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1035 Previously it only applied to the main-section connection policy
1038 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1040 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1041 by me. Added RFC7919 DH primes as an alternative.
1043 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1044 are not in the system include path.
1046 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1047 GnuTLS, when a session startup failed (eg because the client disconnected)
1048 Exim did stdio operations after fclose. This was exposed by a recent
1049 change which nulled out the file handle after the fclose.
1051 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1052 signed directly by the cert-signing cert, rather than an intermediate
1053 OCSP-signing cert. This is the model used by LetsEncrypt.
1055 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1057 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1058 an incoming connection.
1060 HS/02 Bug 1802: Do not half-close the connection after sending a request
1063 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1064 fallback to "prime256v1".
1066 JH/34 SECURITY: Use proper copy of DATA command in error message.
1067 Could leak key material. Remotely exploitable. CVE-2016-9963.
1073 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1074 and 3.4.4 - once the server is enabled to respond to an OCSP request
1075 it does even when not requested, resulting in a stapling non-aware
1076 client dropping the TLS connection.
1078 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1079 support variable-length bit vectors. No functional change.
1081 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1082 The I= interface field on outgoing lines is now after the H= remote
1083 host field, same as incoming lines. There is a separate
1084 outgoing_interface log selector which allows you to disable the
1087 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1088 If not running log_selector +smtp_connection the mainlog would be held
1089 open indefinitely after a "too many connections" event, including to a
1090 deleted file after a log rotate. Leave the per net connection logging
1091 leaving it open for efficiency as that will be quickly detected by the
1092 check on the next write.
1094 HS/01 Bug 1671: Fix post transport crash.
1095 Processing the wait-<transport> messages could crash the delivery
1096 process if the message IDs didn't exist for some reason. When
1097 using 'split_spool_directory=yes' the construction of the spool
1098 file name failed already, exposing the same netto behaviour.
1100 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1101 mime_regex ACL conditions.
1103 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1104 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1105 to HELO, local diagnostic string.
1107 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1108 log line to debug. Even when configured with a tls authenticator many
1109 client connections are expected to not authenticate in this way, so
1110 an authenticate fail is not an error.
1112 HS/02 Add the Exim version string to the process info. This way exiwhat
1113 gives some more detail about the running daemon.
1115 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1116 matter for fast-change records such as DNSBLs.
1118 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1119 retry record, even if constant. There may be multiple transports with
1120 different interface settings and the retry behaviour needs to be kept
1123 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1125 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1127 JH/10 Bug 840: fix log_defer_output option of pipe transport
1129 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1130 hosts_randomize. This matters a lot when combined with mua_wrapper.
1132 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1133 ${quote_pgsql:<string>} operator.
1135 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1137 JH/14 Change header file name for freeradius-client. Relevant if compiling
1138 with Radius support; from the Gentoo tree and checked under Fedora.
1140 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1142 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1143 option transport_filter, meaning no filtering.
1145 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1147 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1148 defaults to "*" (all hosts). The variable is now available when not built
1149 with TLS, default unset, mainly to enable keeping the testsuite sane.
1150 If a server certificate is not supplied (via tls_certificate) an error is
1151 logged, and clients will find TLS connections fail on startup. Presumably
1152 they will retry in-clear.
1153 Packagers of Exim are strongly encouraged to create a server certificate
1154 at installation time.
1156 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1157 with the $config_file variable.
1159 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1160 in transport context, after the attempt, and per-recipient. The latter type
1161 is per host attempted. The event data is the error message, and the errno
1162 information encodes the lookup type (A vs. MX) used for the (first) host,
1163 and the trailing two digits of the smtp 4xx response.
1165 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1166 to write to mainlog (or rejectlog, paniclog) in the window between file
1167 creation and permissions/ownership being changed. Particularly affects
1168 installations where exicyclog is run as root, rather than exim user;
1169 result is that the running daemon panics and dies.
1171 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1173 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1174 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1175 "pri" and "weight". Note that the previous implicit priority given by the
1176 list order is no longer honoured.
1178 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1179 for DKIM processing.
1181 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1182 by defining SUPPORT_SOCKS.
1184 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1185 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1186 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1187 variables are renamed to proxy_{local,external}_{address,port}.
1189 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1190 for a build by defining SUPPORT_I18N
1192 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1193 of the query string, and make ${quote_redis:} do that quoting.
1195 JH/29 Move Events support from Experimental to mainline, enabled by default
1196 and removable for a build by defining DISABLE_EVENT.
1198 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1200 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1201 cached by the daemon.
1203 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1204 by defining LOOKUP_REDIS. The libhiredis library is required.
1206 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1207 keys are given for lookup.
1209 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1210 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1211 only supported when built with TLS support. The PolarSSL SHA routines
1212 are still used when the TLS library is too old for convenient support.
1214 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1215 openssl_options), for security. OpenSSL forces this from version 1.1.0
1216 server-side so match that on older versions.
1218 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1219 allocation for $value could be released as the expansion processing
1220 concluded, but leaving the global pointer active for it.
1222 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1223 and to use the domains and local_parts ACL conditions.
1225 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1226 incorrectly not doubled on cutthrough transmission, hence seen as a
1227 body-termination at the receiving system - resulting in truncated mails.
1228 Commonly the sender saw a TCP-level error, and retransmitted the message
1229 via the normal store-and-forward channel. This could result in duplicates
1230 received - but deduplicating mailstores were liable to retain only the
1231 initial truncated version.
1233 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1235 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1237 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1238 we're in there, support oversigning also; bug 1309.
1240 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1242 HS/04 Add support for keep_environment and add_environment options.
1244 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1245 either intentional arithmetic overflow during PRNG, or testing config-
1248 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1249 delivery resulted in actual delivery. Cancel cutthrough before DATA
1252 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1253 rejects a recipient: pass the reject to the originator.
1255 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1256 Many were false-positives and ignorable, but it's worth fixing the
1259 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1260 for the new environment-manipulation done at startup. Move the routines
1261 from being local to tls.c to being global via the os.c file.
1263 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1264 an extract embedded as result-arg for a map, the first arg for extract
1265 is unavailable so we cannot tell if this is a numbered or keyed
1266 extraction. Accept either.
1272 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1275 JH/02 The smtp transport option "multi_domain" is now expanded.
1277 JH/03 The smtp transport now requests PRDR by default, if the server offers
1280 JH/04 Certificate name checking on server certificates, when exim is a client,
1281 is now done by default. The transport option tls_verify_cert_hostnames
1282 can be used to disable this per-host. The build option
1283 EXPERIMENTAL_CERTNAMES is withdrawn.
1285 JH/05 The value of the tls_verify_certificates smtp transport and main options
1286 default to the word "system" to access the system default CA bundle.
1287 For GnuTLS, only version 3.0.20 or later.
1289 JH/06 Verification of the server certificate for a TLS connection is now tried
1290 (but not required) by default. The verification status is now logged by
1291 default, for both outbound TLS and client-certificate supplying inbound
1294 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1297 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1298 Status Notification (bounce) messages are now MIME format per RFC 3464.
1299 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1300 under the control of the dsn_advertise_hosts option, and routers may
1301 have a dsn_lasthop option.
1303 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1304 default, modifiable by a malware= option. The list separator for
1305 the options can now be changed in the usual way. Bug 68.
1307 JH/10 The smtp_receive_timeout main option is now expanded before use.
1309 JH/11 The incoming_interface log option now also enables logging of the
1310 local interface on delivery outgoing connections.
1312 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1313 if the interface and destination host and port all match.
1315 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1318 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1319 Patch from Andrew Lewis.
1321 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1322 now supports optional time-restrictions, weighting, and priority
1323 modifiers per server. Patch originally by <rommer@active.by>.
1325 JH/16 The spamd_address main option now supports a mixed list of local
1326 and remote servers. Remote servers can be IPv6 addresses, and
1327 specify a port-range.
1329 JH/17 Bug 68: The spamd_address main option now supports an optional
1330 timeout value per server.
1332 JH/18 Bug 1581: Router and transport options headers_add/remove can
1333 now have the list separator specified.
1335 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1338 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1341 JH/21 Support for the A6 type of dns record is withdrawn.
1343 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1344 rather than the verbs used.
1346 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1347 from 255 to 1024 chars.
1349 JH/24 Verification callouts now attempt to use TLS by default.
1351 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1352 are generic router options now. The defaults didn't change.
1354 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1355 Original patch from Alexander Shikoff, worked over by JH.
1357 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1358 files if ZCAT_COMMAND is not executable.
1360 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1362 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1364 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1365 Normally benign, it bites when the pair was led to by a CNAME;
1366 modern usage is to not canonicalize the domain to a CNAME target
1367 (and we were inconsistent anyway for A-only vs AAAA+A).
1369 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1371 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1372 when evaluating $sender_host_dnssec.
1374 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1375 $sender_helo_dnssec variable.
1377 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1379 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1381 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1383 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1384 documented as working, but never had. Support all but $spam_report.
1386 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1387 added for tls authenticator.
1389 HS/03 Add perl_taintmode main config option
1395 TL/01 When running the test suite, the README says that variables such as
1396 no_msglog_check are global and can be placed anywhere in a specific
1397 test's script, however it was observed that placement needed to be near
1398 the beginning for it to behave that way. Changed the runtest perl
1399 script to read through the entire script once to detect and set these
1400 variables, reset to the beginning of the script, and then run through
1401 the script parsing/test process like normal.
1403 TL/02 The BSD's have an arc4random API. One of the functions to induce
1404 adding randomness was arc4random_stir(), but it has been removed in
1405 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1406 function when detected.
1408 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1409 cause callback expansion.
1411 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1412 syntax errors in an expansion can be treated as a string instead of
1413 logging or causing an error, due to the internal use of bool_lax
1414 instead of bool when processing it.
1416 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1417 server certificates when making smtp deliveries.
1419 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1421 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1423 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1425 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1426 Merged patch from Sebastian Wiedenroth.
1428 JH/05 Fix results-pipe from transport process. Several recipients, combined
1429 with certificate use, exposed issues where response data items split
1430 over buffer boundaries were not parsed properly. This eventually
1431 resulted in duplicates being sent. This issue only became common enough
1432 to notice due to the introduction of connection certificate information,
1433 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1435 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1436 size buffer was used, resulting in syntax errors when an expansion
1439 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1440 version 3.3.6 or later.
1442 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1443 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1444 both become "event_action", the variables become $event_name, $event_data
1445 and $event_defer_errno. There is a new variable $verify_mode, usable in
1446 routers, transports and related events. The tls:cert event is now also
1447 raised for inbound connections, if the main configuration event_action
1450 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1451 early OCSP support, but no stapling (appears to be less than 1.0.0).
1453 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1454 server certificate names available under the smtp transport option
1455 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1458 JH/10 Time-related extraction expansions from certificates now use the main
1459 option "timezone" setting for output formatting, and are consistent
1460 between OpenSSL and GnuTLS compilations. Bug 1541.
1462 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1463 encoded parameter in the incoming message. Bug 1558.
1465 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1466 include certificate info, eximon was claiming there were spoolfile
1469 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1471 JH/14 Log delivery-related information more consistently, using the sequence
1472 "H=<name> [<ip>]" wherever possible.
1474 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1475 are problematic for Debian distribution, omit them from the release
1478 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1480 JH/16 Fix string representation of time values on 64bit time_t architectures.
1483 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1484 output list separator was used.
1489 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1490 checkers that were complaining about end of non-void function with no
1493 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1494 This was a regression introduced in 4.83 by another bugfix.
1496 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1498 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1499 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1505 TF/01 Correctly close the server side of TLS when forking for delivery.
1507 When a message was received over SMTP with TLS, Exim failed to clear up
1508 the incoming connection properly after forking off the child process to
1509 deliver the message. In some situations the subsequent outgoing
1510 delivery connection happened to have the same fd number as the incoming
1511 connection previously had. Exim would try to use TLS and fail, logging
1512 a "Bad file descriptor" error.
1514 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1515 utilities have not been installed.
1517 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1518 temporary space as the ACL may create new global variables.
1520 TL/01 LDAP support uses per connection or global context settings, depending
1521 upon the detected version of the libraries at build time.
1523 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1524 to extract and use the src ip:port in logging and expansions as if it
1525 were a direct connection from the outside internet. PPv2 support was
1526 updated based on HAProxy spec change in May 2014.
1528 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1530 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1531 Properly escape header and check for NULL return.
1533 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1536 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1538 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1539 characters in header names, implemented as a verify condition.
1540 Contributed by Michael Fischer v. Mollard.
1542 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1543 results permerror and temperror. Previous values are deprecated but
1544 still accepted. In a future release, err_perm and err_temp will be
1545 completely removed, which will be a backward incompatibility if the
1546 ACL tests for either of these two old results. Patch contributed by
1547 user bes-internal on the mailing list.
1549 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1551 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1552 selectors, in both main and reject logs.
1554 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1557 JH/07 Add malware type "sock" for talking to simple daemon.
1559 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1561 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1562 routers/transports under cutthrough routing.
1564 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1565 numbers. Touch up "bool" conditional to keep the same definition.
1567 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1569 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1571 JH/12 Expand items in router/transport headers_add or headers_remove lists
1572 individually rather than the list as a whole. Bug 1452.
1574 Required for reasonable handling of multiple headers_ options when
1575 they may be empty; requires that headers_remove items with embedded
1576 colons must have them doubled (or the list-separator changed).
1578 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1579 view the policy declared in the DMARC record. Currently, $dmarc_status
1580 is a combined value of both the record presence and the result of the
1583 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1585 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1586 dnslookup router and the smtp transport (applying to the forward
1589 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1590 of ldap servers used for a specific lookup. Patch provided by Heiko
1593 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1594 New variable $lookup_dnssec_authenticated for observability.
1596 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1597 Patch submitted by Lars Timman.
1599 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1601 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1602 Requires trusted mode and valid format message id, aborts otherwise.
1603 Patch contributed by Heiko Schlichting.
1605 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1606 certextract with support for various fields. Bug 1358.
1608 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1609 is requested by default, modifiable by smtp transport option
1612 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1613 operate on certificate variables to give certificate fingerprints
1614 Also new ${sha256:cert_variable}.
1616 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1618 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1619 Christian Aistleitner.
1621 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1623 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1624 file. Patch from Wolfgang Breyha.
1626 JH/25 Expand the coverage of the delivery $host and $host_address to
1627 client authenticators run in verify callout. Bug 1476.
1629 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1630 align with daemon_smtp_ports. Bug 72.
1632 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1633 support and error reporting did not work properly.
1635 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1636 and is readable. Patch from Andrew Colin Kissa.
1638 TL/14 Enhance documentation of ${run expansion and how it parses the
1639 commandline after expansion, particularly in the case when an
1640 unquoted variable expansion results in an empty value.
1642 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1644 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1645 by renaming away from it. Addresses GH issue 15, reported by
1648 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1650 TL/15 SECURITY: prevent double expansion in math comparison functions
1651 (can expand unsanitized data). Not remotely exploitable.
1658 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1660 PP/02 Make -n do something, by making it not do something.
1661 When combined with -bP, the name of an option is not output.
1663 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1666 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1667 $sender_host_name and config options to manage this, and basic check
1670 PP/05 DSCP support for outbound connections and control modifier for inbound.
1672 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1673 (Only plugin which currently uses this is kerberos4, which nobody should
1674 be using, but we should make it available and other future plugins might
1675 conceivably use it, even though it would break NAT; stuff *should* be
1676 using channel bindings instead).
1678 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1679 name; added for Sendmail compatibility; requires admin caller.
1680 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1681 just ignore it); requires trusted caller.
1682 Also parse but ignore: -Ac -Am -X<logfile>
1685 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1687 TL/02 Add +smtp_confirmation as a default logging option.
1689 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1690 Patch by Magnus Holmgren from 2007-02-20.
1692 TL/04 Bugzilla 1281 - Spec typo.
1693 Bugzilla 1283 - Spec typo.
1694 Bugzilla 1290 - Spec grammar fixes.
1696 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1698 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1700 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1701 mailing list by by Dmitry Isaikin.
1703 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1705 JH/02 Support "G" suffix to numbers in ${if comparisons.
1707 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1709 NM/01 Bugzilla 1197 - Spec typo
1710 Bugzilla 1196 - Spec examples corrections
1712 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1714 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1715 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1718 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1719 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1721 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1722 "acl {{name}{arg}...}", and optional args on acl condition
1725 JH/05 Permit multiple router/transport headers_add/remove lines.
1727 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1729 JH/07 Avoid using a waiting database for a single-message-only transport.
1730 Performance patch from Paul Fisher. Bugzilla 1262.
1732 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1735 JH/09 Add $headers_added variable, with content from use of ACL modifier
1736 add_header (but not yet added to the message). Bugzilla 199.
1738 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1739 Pulled from Bugzilla 817 by Wolfgang Breyha.
1741 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1743 (nb: this is the same fix as in Exim 4.80.1)
1745 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1748 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1751 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1753 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1754 advertise SMTP AUTH mechanism to us, instead of a generic
1755 protocol violation error. Also, make Exim more robust to bad
1756 data from the Dovecot auth socket.
1758 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1760 When a queue runner is handling a message, Exim first routes the
1761 recipient addresses, during which it prunes them based on the retry
1762 hints database. After that it attempts to deliver the message to
1763 any remaining recipients. It then updates the hints database using
1766 So if a recipient address works intermittently, it can get repeatedly
1767 deferred at routing time. The retry hints record remains fresh so the
1768 address never reaches the final cutoff time.
1770 This is a fairly common occurrence when a user is bumping up against
1771 their storage quota. Exim had some logic in its local delivery code
1772 to deal with this. However it did not apply to per-recipient defers
1773 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1775 This change adds a proper retry rule check during routing so that the
1776 final cutoff time is checked against the message's age. We only do
1777 this check if there is an address retry record and there is not a
1778 domain retry record; this implies that previous attempts to handle
1779 the address had the retry_use_local_parts option turned on. We use
1780 this as an approximation for the destination being like a local
1781 delivery, as in LMTP.
1783 I suspect this new check makes the old local delivery cutoff check
1784 redundant, but I have not verified this so I left the code in place.
1786 TF/02 Correct gecos expansion when From: is a prefix of the username.
1788 Test 0254 submits a message to Exim with the header
1792 When I ran the test suite under the user fanf2, Exim expanded
1793 the header to contain my full name, whereas it should have added
1794 a Resent-Sender: header. It erroneously treats any prefix of the
1795 username as equal to the username.
1797 This change corrects that bug.
1799 GF/01 DCC debug and logging tidyup
1800 Error conditions log to paniclog rather than rejectlog.
1801 Debug lines prefixed by "DCC: " to remove any ambiguity.
1803 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1805 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1806 Bug spotted by Jeremy Harris; was flawed since initial commit.
1807 Would have resulted in OCSP responses post-SNI triggering an Exim
1808 NULL dereference and crash.
1810 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1812 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1813 Bug detection, analysis and fix by Samuel Thibault.
1814 Bugzilla 1331, Debian bug #698092.
1816 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1818 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1819 Server implementation by Todd Lyons, client by JH.
1820 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1821 config variable "prdr_enable" controls whether the server
1822 advertises the facility. If the client requests PRDR a new
1823 acl_data_smtp_prdr ACL is called once for each recipient, after
1824 the body content is received and before the acl_smtp_data ACL.
1825 The client is controlled by both of: a hosts_try_prdr option
1826 on the smtp transport, and the server advertisement.
1827 Default client logging of deliveries and rejections involving
1828 PRDR are flagged with the string "PRDR".
1830 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1831 fclose(). Diagnosis by Todd Lyons.
1833 PP/17 Update configure.default to handle IPv6 localhost better.
1834 Patch by Alain Williams (plus minor tweaks).
1837 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1838 This is now consistent with GnuTLS, and is now documented: the
1839 previous undocumented portable approach to treating the option as
1840 unset was to force an expansion failure. That still works, and
1841 an empty string is now equivalent.
1843 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1844 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1845 not performing validation itself.
1847 PP/20 Added force_command boolean option to pipe transport.
1848 Patch from Nick Koston, of cPanel Inc.
1850 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1853 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1855 PP/21 Fix eximon continuous updating with timestamped log-files.
1856 Broken in a format-string cleanup in 4.80, missed when I repaired the
1857 other false fix of the same issue.
1858 Report and fix from Heiko Schlichting.
1861 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1862 Report from Prashanth Katuri.
1864 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1865 It's SecureTransport, so affects any MacOS clients which use the
1866 system-integrated TLS libraries, including email clients.
1868 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1869 using a MIME ACL for non-SMTP local injection.
1870 Report and assistance in diagnosis by Warren Baker.
1872 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1874 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1876 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1877 last id that failed so it may be referenced in subsequent ACL's.
1879 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1882 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1883 ldap library initialization, allowing self-signed CA's to be
1884 used. Also properly sets require_cert option later in code by
1885 using NULL (global ldap config) instead of ldap handle (per
1886 session). Bug diagnosis and testing by alxgomz.
1888 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1889 the src/util/ subdirectory.
1891 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1892 renamed to Transport Post Delivery Action by Jeremy Harris, as
1895 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1896 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1897 redis_servers = needs to be configured which will be used by the redis
1898 lookup. Patch from Warren Baker, of The Packet Hub.
1900 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1902 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1903 hostname or reverse DNS when processing a host list. Used suggestions
1904 from multiple comments on this bug.
1906 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1908 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1909 Missed a few lines, added it to make the runtest require no keyboard
1912 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1913 contains upper case chars. Make router use caseful_local_part.
1915 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1916 support when GnuTLS has been built with p11-kit.
1922 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1924 This, or similar/improved, will also be change PP/11 of 4.82.
1930 PP/01 Handle short writes when writing local log-files.
1931 In practice, only affects FreeBSD (8 onwards).
1932 Bugzilla 1053, with thanks to Dmitry Isaikin.
1934 NM/01 Bugzilla 949 - Documentation tweak
1936 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1939 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1941 PP/02 Implemented gsasl authenticator.
1943 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1945 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1946 `pkg-config foo` for cflags/libs.
1948 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1949 with rest of GSASL and with heimdal_gssapi.
1951 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1952 `pkg-config foo` for cflags/libs for the TLS implementation.
1954 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1955 properties get this fed in as external SSF. A number of robustness
1956 and debugging improvements to the cyrus_sasl authenticator.
1958 PP/08 cyrus_sasl server now expands the server_realm option.
1960 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1961 Patch by Jeremy Harris.
1963 PP/10 Added dbmjz lookup type.
1965 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1967 PP/12 MAIL args handles TAB as well as SP, for better interop with
1968 non-compliant senders.
1969 Analysis and variant patch by Todd Lyons.
1971 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1972 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1973 Patch from Dirk Mueller <dmueller@suse.com>
1975 PP/13 tls_peerdn now print-escaped for spool files.
1976 Observed some $tls_peerdn in wild which contained \n, which resulted
1977 in spool file corruption.
1979 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1980 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1981 or write after TLS renegotiation, which otherwise led to messages
1984 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1985 as a tracking header (ie: a signed header comes before the signature).
1986 Patch from Wolfgang Breyha.
1988 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1989 comma-sep list; embedded commas doubled.
1991 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1993 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1995 Report and patch from Dmitry Banschikov.
1997 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1998 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1999 failures. We appear to now support TLS1.1+ with Exim.
2001 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2002 lets Exim select keys and certificates based upon TLS SNI from client.
2003 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2004 before an outbound SMTP session. New log_selector, +tls_sni.
2006 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2007 NULL dereference. Report and patch from Alun Jones.
2009 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2010 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2011 Not seeing resolver debug output on NetBSD, but suspect this is a
2012 resolver implementation change.
2014 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2015 Left warnings. Added "eximon gdb" invocation mode.
2017 PP/21 Defaulting "accept_8bitmime" to true, not false.
2019 PP/22 Added -bw for inetd wait mode support.
2021 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2022 locate the relevant includes and libraries. Made this the default.
2024 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2025 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2027 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2028 This may cause build issues on older platforms.
2030 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2031 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2032 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2033 Added SNI support via GnuTLS too.
2034 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2036 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2038 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2039 Applied second patch from Janne, implementing suggestion to default
2040 multiple-strings-in-record handling to match SPF spec.
2042 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2044 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2045 read-only, out of scope).
2046 Patch from Wolfgang Breyha, report from Stuart Northfield.
2048 PP/29 Fix three issues highlighted by clang analyser static analysis.
2049 Only crash-plausible issue would require the Cambridge-specific
2050 iplookup router and a misconfiguration.
2051 Report from Marcin Mirosław.
2053 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2055 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2056 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2057 As part of this, removing so much warning spew let me fix some minor
2058 real issues in debug logging.
2060 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2061 assignment on my part. Fixed.
2063 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2064 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2065 Janne Snabb (who went above and beyond: thank you).
2067 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2068 string otherwise requires a connection and a bunch more work and it's
2069 relatively easy to get wrong. Should also expose TLS library linkage
2072 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2073 64-bit ${eval} (JH/03).
2075 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2076 GNU libc to support some of the 64-bit stuff, should not lead to
2077 conflicts. Defined before os.h is pulled in, so if a given platform
2078 needs to override this, it can.
2080 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2081 protection layer was required, which is not implemented.
2082 Bugzilla 1254, patch from Wolfgang Breyha.
2084 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2085 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2086 tls_dhparam take prime identifiers. Also unbreak combination of
2087 OpenSSL+DH_params+TLSSNI.
2089 PP/39 Disable SSLv2 by default in OpenSSL support.
2095 PP/01 Solaris build fix for Oracle's LDAP libraries.
2096 Bugzilla 1109, patch from Stephen Usher.
2098 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2100 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2103 TF/02 Fix a couple more cases where we did not log the error message
2104 when unlink() failed. See also change 4.74-TF/03.
2106 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2107 lock up or crash if it happened to be inside a call to libc when it
2108 got a SIGUSR1 from exiwhat.
2110 The SIGUSR1 handler appends the current process status to the process
2111 log which is later printed by exiwhat. It used to use the general
2112 purpose logging code to do this, but several functions it calls are
2113 not safe for signals.
2115 The new output code in the SIGUSR1 handler is specific to the process
2116 log, and simple enough that it's easy to inspect for signal safety.
2117 Removing some special cases also simplifies the general logging code.
2118 Removing the spurious timestamps from the process log simplifies
2121 TF/04 Improved ratelimit ACL condition.
2123 The /noupdate option has been deprecated in favour of /readonly which
2124 has clearer semantics. The /leaky, /strict, and /readonly update modes
2125 are mutually exclusive. The update mode is no longer included in the
2126 database key; it just determines when the database is updated. (This
2127 means that when you upgrade Exim will forget old rate measurements.)
2129 Exim now checks that the per_* options are used with an update mode that
2130 makes sense for the current ACL. For example, when Exim is processing a
2131 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2132 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2133 must specify per_mail/readonly. If you omit the update mode it defaults to
2134 /leaky where that makes sense (as before) or /readonly where required.
2136 The /noupdate option is now undocumented but still supported for
2137 backwards compatibility. It is equivalent to /readonly except that in
2138 ACLs where /readonly is required you may specify /leaky/noupdate or
2139 /strict/noupdate which are treated the same as /readonly.
2141 A useful new feature is the /count= option. This is a generalization
2142 of the per_byte option, so that you can measure the throughput of other
2143 aggregate values. For example, the per_byte option is now equivalent
2144 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2146 The per_rcpt option has been generalized using the /count= mechanism
2147 (though it's more complicated than the per_byte equivalence). When it is
2148 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2149 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2150 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2151 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2152 Note that using per_rcpt with a non-readonly update mode in more than
2153 one ACL will cause the recipients to be double-counted. (The per_mail
2154 and per_byte options don't have this problem.)
2156 The handling of very low rates has changed slightly. If the computed rate
2157 is less than the event's count (usually one) then this event is the first
2158 after a long gap. In this case the rate is set to the same as this event's
2159 count, so that the first message of a spam run is counted properly.
2161 The major new feature is a mechanism for counting the rate of unique
2162 events. The new per_addr option counts the number of different
2163 recipients that someone has sent messages to in the last time period. It
2164 behaves like per_rcpt if all the recipient addresses are different, but
2165 duplicate recipient addresses do not increase the measured rate. Like
2166 the /count= option this is a general mechanism, so the per_addr option
2167 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2168 example, measure the rate that a client uses different sender addresses
2169 with the options per_mail/unique=$sender_address. There are further
2170 details in the main documentation.
2172 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2174 TF/06 Removed a few PCRE remnants.
2176 TF/07 Automatically extract Exim's version number from tags in the git
2177 repository when doing development or release builds.
2179 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2180 Bugzilla 879. Patch from Paul Fisher.
2182 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2183 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2186 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2188 PP/05 Variable $av_failed, true if the AV scanner deferred.
2189 Bugzilla 1078. Patch from John Horne.
2191 PP/06 Stop make process more reliably on build failure.
2192 Bugzilla 1087. Patch from Heiko Schlittermann.
2194 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2195 Bugzilla 1089. Patch from Heiko Schlittermann.
2197 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2198 Bugzilla 1131. Patch from Holger Weiß.
2200 PP/09 Handle IPv6 addresses with SPF.
2201 Bugzilla 860. Patch from Wolfgang Breyha.
2203 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2205 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2208 PP/11 match_* no longer expand right-hand-side by default.
2209 New compile-time build option, EXPAND_LISTMATCH_RHS.
2210 New expansion conditions, "inlist", "inlisti".
2212 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2214 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2216 PP/14 fix log_write() format string regression from TF/03.
2217 Bugzilla 1152. Patch from Dmitry Isaikin.
2223 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2225 PP/02 Harmonised TLS library version reporting; only show if debugging.
2226 Layout now matches that introduced for other libraries in 4.74 PP/03.
2228 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2230 PP/04 New "dns_use_edns0" global option.
2232 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2235 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2236 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2238 TK/01 Updated PolarSSL code to 0.14.2.
2239 Bugzilla 1097. Patch from Andreas Metzler.
2241 PP/07 Catch divide-by-zero in ${eval:...}.
2242 Fixes bugzilla 1102.
2244 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2247 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2248 format-string attack -- SECURITY: remote arbitrary code execution.
2250 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2251 time unintentionally subject to list matching rules, letting the header
2252 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2253 arbitrary string expansion). This allowed for information disclosure.
2255 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2256 INT_MIN/-1 -- value coerced to INT_MAX.
2262 NM/01 Workaround for PCRE version dependency in version reporting
2265 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2266 This fixes portability to compilers other than gcc, notably
2267 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2269 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2270 makefiles for portability to HP-UX and POSIX correctness.
2272 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2273 Also via indented variable definition in the Makefile.
2274 (Debugging by Oliver Heesakkers).
2276 PP/02 Restore caching of spamd results with expanded spamd_address.
2277 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2279 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2280 Improves build reliability. Fix from: Frank Elsner
2282 NM/02 Fix wide character breakage in the rfc2047 coding
2283 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2285 NM/03 Allow underscore in dnslist lookups
2286 Fixes bug 1026. Patch from Graeme Fowler
2288 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2289 Code patches from Adam Ciarcinski of NetBSD.
2291 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2294 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2295 is logged, to avoid truncation. Patch from John Horne.
2297 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2298 Patch from Jakob Hirsch.
2300 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2301 SQL string expansion failure details.
2302 Patch from Andrey Oktyabrski.
2304 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2305 Patch from Simon Arlott.
2307 PP/09 New lookups functionality failed to compile on old gcc which rejects
2308 extern declarations in function scope.
2309 Patch from Oliver Fleischmann
2311 PP/10 Use sig_atomic_t for flags set from signal handlers.
2312 Check getgroups() return and improve debugging.
2313 Fixed developed for diagnosis in bug 927 (which turned out to be
2316 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2317 Patch from Mark Zealey.
2319 PP/12 Bugzilla 1056: Improved spamd server selection.
2320 Patch from Mark Zealey.
2322 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2323 Based on patch from Heiko Schlittermann.
2325 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2326 Patch from Uwe Doering, sign-off by Michael Haardt.
2328 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2329 variable declaration deep within a block. Bug and patch from
2332 PP/15 lookups-Makefile IRIX compatibility coercion.
2334 PP/16 Make DISABLE_DKIM build knob functional.
2336 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2337 Patch by Simon Arlott
2339 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2340 variable argument macros. Our copy now differs from upstream.
2346 TF/01 Failure to get a lock on a hints database can have serious
2347 consequences so log it to the panic log.
2349 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2350 controlled using the smtp_confirmation log selector.
2352 TF/03 Include the error message when we fail to unlink a spool file.
2354 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2355 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2356 for maintaining out-of-tree patches for some time.
2358 PP/01 Bugzilla 139: Documentation and portability issues.
2359 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2360 Handle per-OS dynamic-module compilation flags.
2362 PP/02 Let /dev/null have normal permissions.
2363 The 4.73 fixes were a little too stringent and complained about the
2364 permissions on /dev/null. Exempt it from some checks.
2365 Reported by Andreas M. Kirchwitz.
2367 PP/03 Report version information for many libraries, including
2368 Exim version information for dynamically loaded libraries. Created
2369 version.h, now support a version extension string for distributors
2370 who patch heavily. Dynamic module ABI change.
2372 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2373 privilege escalation vulnerability whereby the Exim run-time user
2374 can cause root to append content of the attacker's choosing to
2377 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2380 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2381 If dropping privileges for untrusted macros, we disabled normal logging
2382 on the basis that it would fail; for the Exim run-time user, this is not
2383 the case, and it resulted in successful deliveries going unlogged.
2384 Fixed. Reported by Andreas Metzler.
2390 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2391 only prepend for the Resent-* case. Fixes regression introduced in
2392 Exim 4.70 by NM/22 for Bugzilla 607.
2394 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2395 increasing numbers of administrators be bitten by this.
2397 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2399 PP/03 Bugzilla 994: added openssl_options main configuration option.
2401 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2403 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2405 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2407 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2408 without a peer certificate, leading to a segfault because of an
2409 assumption that peers always have certificates. Be a little more
2410 paranoid. Problem reported by Martin Tscholak.
2412 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2413 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2414 NB: ClamAV planning to remove STREAM in "middle of 2010".
2415 CL also introduces -bmalware, various -d+acl logging additions and
2416 more caution in buffer sizes.
2418 PP/09 Implemented reverse_ip expansion operator.
2420 PP/10 Bugzilla 937: provide a "debug" ACL control.
2422 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2424 PP/12 Bugzilla 973: Implement --version.
2426 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2428 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2430 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2432 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2433 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2434 ignore trailing whitespace.
2436 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2438 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2441 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2442 Notification from Dr Andrew Aitchison.
2444 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2445 ExtendedDetectionInfo response format.
2446 Notification from John Horne.
2448 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2451 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2452 XSL and documented dependency on system catalogs, with examples of how
2455 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2458 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2459 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2460 configuration file which is writeable by the Exim user or group.
2462 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2463 of configuration files to cover files specified with the -C option if
2464 they are going to be used with root privileges, not just the default
2467 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2468 option (effectively making it always true).
2470 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2471 files to be used while preserving root privileges.
2473 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2474 that rogue child processes cannot use them.
2476 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2477 run-time user, instead of root.
2479 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2480 Exim run-time user without dropping privileges.
2482 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2483 result string, instead of calling string_vformat() twice with the same
2486 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2487 for other users. Others should always drop root privileges if they use
2488 -C on the command line, even for a whitelisted configure file.
2490 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2492 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2498 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2499 $data_path, and $header_path variables; fixed documentation bugs and
2502 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2503 exipick to access non-standard spools, including the "frozen" queue
2506 NM/01 Bugzilla 965: Support mysql stored procedures.
2507 Patch from Alain Williams
2509 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2511 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2512 Patch from Andreas Metzler
2514 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2515 Patch from Kirill Miazine
2517 NM/05 Bugzilla 671: Added umask to procmail example.
2519 JJ/03 installed exipick 20100323.0, fixing doc bug
2521 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2522 directory. Notification and patch from Dan Rosenberg.
2524 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2526 TK/02 Improve log output when DKIM signing operation fails.
2528 MH/01 Treat the transport option dkim_domain as a colon separated
2529 list, not as a single string, and sign the message with each element,
2530 omitting multiple occurences of the same signer.
2532 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2533 Bugzilla 985, 986. Patch by Simon Arlott
2535 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2536 Patch by Simon Arlott
2538 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2539 MBX locking. Notification from Dan Rosenberg.
2545 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2547 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2549 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2551 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2553 NM/04 Bugzilla 915: Flag broken perl installation during build.
2559 TK/01 Added patch by Johannes Berg that expands the main option
2560 "spamd_address" if it starts with a dollar sign.
2562 TK/02 Write list of recipients to X-Envelope-Sender header when building
2563 the mbox-format spool file for content scanning (suggested by Jakob
2566 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2567 (http://www.dcc-servers.net/) support via dccifd. Activated by
2568 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2570 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2571 by Mark Daniel Reidel <mr@df.eu>.
2573 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2574 When building exim an external PCRE library is now needed -
2575 PCRE is a system library on the majority of modern systems.
2576 See entry on PCRE_LIBS in EDITME file.
2578 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2579 conversation. Added nologin parameter to request.
2580 Patch contributed by Kirill Miazine.
2582 TF/01 Do not log submission mode rewrites if they do not change the address.
2584 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2586 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2587 log files in place. Contributed by Roberto Lima.
2589 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2591 TF/03 Bugzilla 615: When checking the local_parts router precondition
2592 after a local_part_suffix or local_part_prefix option, Exim now
2593 does not use the address's named list lookup cache, since this
2594 contains cached lookups for the whole local part.
2596 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2597 Robert Millan. Documentation is in experimental-spec.txt.
2599 TF/04 Bugzilla 668: Fix parallel build (make -j).
2601 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2603 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2604 Patch provided by Jan Srzednicki.
2606 TF/05 Leading white space used to be stripped from $spam_report which
2607 wrecked the formatting. Now it is preserved.
2609 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2610 that they are available at delivery time.
2612 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2614 TF/08 TLS error reporting now respects the incoming_interface and
2615 incoming_port log selectors.
2617 TF/09 Produce a more useful error message if an SMTP transport's hosts
2618 setting expands to an empty string.
2620 NM/06 Bugzilla 744: EXPN did not work under TLS.
2621 Patch provided by Phil Pennock.
2623 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2624 Patch provided by Richard Godbee.
2626 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2627 acl_smtp_notquit, added index entry.
2629 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2630 Patch provided by Eugene Bujak.
2632 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2633 accept(). Patch provided by Maxim Dounin.
2635 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2636 Patch provided by Phil Pennock.
2638 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2640 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2641 Patch provided by Brad "anomie" Jorsch.
2643 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2644 Patch provided by Dean Brooks.
2646 TK/05 Add native DKIM support (does not depend on external libraries).
2648 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2649 Patch provided by Graeme Fowler.
2651 NM/16 Bugzilla 851: Documentation example syntax fix.
2653 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2655 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2658 NM/19 Bugzilla 745: TLS version reporting.
2659 Patch provided by Phil Pennock.
2661 NM/20 Bugzilla 167: bool: condition support.
2662 Patch provided by Phil Pennock.
2664 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2665 clients. Patch provided by Phil Pennock.
2667 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2668 Patch provided by Brad "anomie" Jorsch.
2670 NM/23 Bugzilla 687: Fix misparses in eximstats.
2671 Patch provided by Heiko Schlittermann.
2673 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2674 Patch provided by Heiko Schlittermann.
2676 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2677 plus update to original patch.
2679 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2681 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2682 Patch provided by David Brownlee.
2684 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2686 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2688 NM/30 Bugzilla 888: TLS documentation bugfixes.
2690 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2692 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2693 Unlike the original bugzilla I have changed all shell scripts in src tree.
2695 NM/33 Bugzilla 898: Transport filter timeout fix.
2696 Patch by Todd Rinaldo.
2698 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2699 Patch by Serge Demonchaux.
2701 NM/35 Bugzilla 39: Base64 decode bug fixes.
2702 Patch by Jakob Hirsch.
2704 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2706 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2708 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2710 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2716 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2717 ALT-N's libdkim that I have put here:
2718 http://duncanthrax.net/exim-experimental/
2720 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2721 were called 'true' and it seems that C99 defines that as a reserved
2722 keyword to be used with 'bool' variable types. That means you could
2723 not include C99-style headers which use bools without triggering
2724 build errors in sieve.c.
2726 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2727 as mailq or other aliases. Changed the --help handling significantly
2728 to do whats expected. exim_usage() emits usage/help information.
2730 SC/01 Added the -bylocaldomain option to eximstats.
2732 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2734 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2736 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2742 PH/01 Another patch from the Sieve maintainer.
2744 PH/02 When an IPv6 address is converted to a string for single-key lookup
2745 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2746 dots are used instead of colons so that keys in lsearch files need not
2747 contain colons. This was done some time before quoting was made available
2748 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2749 (notated using the quote facility) so as to distinguish them from IPv4
2750 keys. This meant that lookups for IP addresses in host lists did not work
2751 for iplsearch lookups.
2753 This has been fixed by arranging for IPv6 addresses to be expressed with
2754 colons if the lookup type is iplsearch. This is not incompatible, because
2755 previously such lookups could never work.
2757 The situation is now rather anomalous, since one *can* have colons in
2758 ordinary lsearch keys. However, making the change in all cases is
2759 incompatible and would probably break a number of configurations.
2761 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2764 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2765 conversion specification without a maximum field width, thereby enabling
2766 a rogue spamd server to cause a buffer overflow. While nobody in their
2767 right mind would setup Exim to query an untrusted spamd server, an
2768 attacker that gains access to a server running spamd could potentially
2769 exploit this vulnerability to run arbitrary code as the Exim user.
2771 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2772 $primary_hostname instead of what libspf2 thinks the hosts name is.
2774 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2775 a directory entry by the name of the lookup key. Previously, if a
2776 symlink pointed to a non-existing file or a file in a directory that
2777 Exim lacked permissions to read, a lookup for a key matching that
2778 symlink would fail. Now it is enough that a matching directory entry
2779 exists, symlink or not. (Bugzilla 503.)
2781 PH/03 The body_linecount and body_zerocount variables are now exported in the
2784 PH/04 Added the $dnslist_matched variable.
2786 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2787 This means they are set thereafter only if the connection becomes
2790 PH/06 Added the client_condition to authenticators so that some can be skipped
2791 by clients under certain conditions.
2793 PH/07 The error message for a badly-placed control=no_multiline_responses left
2794 "_responses" off the end of the name.
2796 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2798 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2799 (without spaces) instead of just copying the configuration text.
2801 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2803 PH/11 Added $max_received_linelength.
2805 PH/12 Added +ignore_defer and +include_defer to host lists.
2807 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2808 way in which PCRE > 7.0 is built.
2810 PH/14 Implemented queue_only_load_latch.
2812 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2813 MAIL command. The effect was to mangle the value on 64-bit systems.
2815 PH/16 Another patch from the Sieve maintainer.
2817 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2819 PH/18 If a system quota error occurred while trying to create the file for
2820 a maildir delivery, the message "Mailbox is full" was not appended to the
2821 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2822 only to a quota excession during the actual writing of the file.
2824 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2825 characters?) which causes problems in log lines. The DN values are now
2826 passed through string_printing() before being added to log lines.
2828 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2829 and InterBase are left for another time.)
2831 PH/21 Added message_body_newlines option.
2833 PH/22 Guard against possible overflow in moan_check_errorcopy().
2835 PH/23 POSIX allows open() to be a macro; guard against that.
2837 PH/24 If the recipient of an error message contained an @ in the local part
2838 (suitably quoted, of course), incorrect values were put in $domain and
2839 $local_part during the evaluation of errors_copy.
2845 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2846 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2849 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2850 issue a MAIL command.
2852 PH/02 In an ACL statement such as
2854 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2856 if a client was not listed at all, or was listed with a value other than
2857 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2858 the condition was not true (as it should be), so access was not denied.
2859 The bug was that the ! inversion was incorrectly passed on to the second
2860 item. This has been fixed.
2862 PH/03 Added additional dnslists conditions == and =& which are different from
2863 = and & when the dns lookup returns more than one IP address.
2865 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2866 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2868 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2869 FSYNC, which compiles an option called disable_fsync that allows for
2870 bypassing fsync(). The documentation is heavily laced with warnings.
2872 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2874 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2875 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2876 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2877 including adding "make clean"; (3) Added -fPIC when compiling the test
2878 dynamically loaded module, to get rid of a warning.
2880 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2881 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2882 = 0s. The bug is otherwise harmless.
2884 PH/07 There was a bug in the dovecot authenticator such that the value of
2885 $auth1 could be overwritten, and so not correctly preserved, after a
2886 successful authentication. This usually meant that the value preserved by
2887 the server_setid option was incorrect.
2889 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2891 PH/09 Installed PCRE release 7.0.
2893 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2894 run for batched SMTP input. It is now run at the start of every message
2895 in the batch. While fixing this I discovered that the process information
2896 (output by running exiwhat) was not always getting set for -bs and -bS
2897 input. This is fixed, and it now also says "batched" for BSMTP.
2899 PH/11 Added control=no_pipelining.
2901 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2902 patch, slightly modified), and move the expansion of helo_data till after
2903 the connection is made in the smtp transport (so it can use these
2906 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2908 PH/14 Added log_selector = +pid.
2910 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2912 PH/16 Add ${if forany and ${if forall.
2914 PH/17 Added dsn_from option to vary the From: line in DSNs.
2916 PH/18 Flush SMTP output before performing a callout, unless control =
2917 no_callout_flush is set.
2919 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2920 was true (the default) a successful delivery failed to delete the retry
2921 item, thus causing premature timeout of the address. The bug is now
2924 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2926 PH/21 Long custom messages for fakedefer and fakereject are now split up
2927 into multiline responses in the same way that messages for "deny" and
2928 other ACL rejections are.
2930 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2931 with slight modification.
2933 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2934 draft, changing the syntax and factoring some duplicate code".
2936 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2937 for deliveries of the second and subsequent messages over the same SMTP
2940 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2941 ${reduce, with only minor "tidies".
2943 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2945 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2946 expansion side effects.
2948 PH/27 When a message times out after an over-quota error from an Exim-imposed
2949 quota, the bounce message says "mailbox is full". This message was not
2950 being given when it was a system quota that was exceeded. It now should
2953 MH/03 Made $recipients available in local_scan(). local_scan() already has
2954 better access to the recipient list through recipients_list[], but
2955 $recipients can be useful in postmaster-provided expansion strings.
2957 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2958 in the case of a MAIL command with additional options following the
2959 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2960 were accidentally chopped off.
2962 PH/29 SMTP synchronization checks are implemented when a command is read -
2963 there is a check that no more input is waiting when there shouldn't be
2964 any. However, for some commands, a delay in an ACL can mean that it is
2965 some time before the response is written. In this time, more input might
2966 arrive, invalidly. So now there are extra checks after an ACL has run for
2967 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2968 pipelining has not been advertised.
2970 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2972 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2973 returned in the SMTP response when smtp_return_error_details was false.
2974 This has been fixed.
2976 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2977 instead of the C I/O that was originally supplied, because problems were
2978 reported on Solaris.
2980 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2981 Exim which did not show up earlier: it was assuming that a call to
2982 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2983 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2984 was a macro that became an assignment, so it seemed to work. This has
2985 changed to a proper function call with a void return, hence the compile
2986 error. Exim's code has been fixed.
2988 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2991 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2993 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2995 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2996 because it turns out to be too big or there is a timeout) there is no
2997 "Completed" line in the log. When some messages of this type were
2998 selected by exigrep, they were listed as "not completed". Others were
2999 picked up by some special patterns. I have improved the selection
3000 criteria to be more general.
3002 PH/38 The host_find_failed option in the manualroute router can now be set
3003 to "ignore", to completely ignore a host whose IP address cannot be
3004 found. If all hosts are ignored, the behaviour is controlled by the new
3005 host_all_ignored option.
3007 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3008 homing or because of multiple MX records with /mx) generated more than
3009 one IP address, and the following item turned out to be the local host,
3010 all the secondary addresses of the first item were incorrectly removed
3011 from the list, along with the local host and any following hosts (which
3012 is what is supposed to happen).
3014 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3015 whoever called Exim into the -H file. In the case of the daemon it was
3016 behaving confusingly. When first started, it used values for whoever
3017 started the daemon, but after a SIGHUP it used the Exim user (because it
3018 calls itself on a restart). I have changed the code so that it now always
3021 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3022 message are rejected with the same error (e.g. no authentication or bad
3023 sender address), and a DATA command is nevertheless sent (as can happen
3024 with PIPELINING or a stupid MUA), the error message that was given to the
3025 RCPT commands is included in the rejection of the DATA command. This is
3026 intended to be helpful for MUAs that show only the final error to their
3029 PH/42 Another patch from the Sieve maintainer.
3031 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3032 Eximstats - Fixed some broken HTML links and added missing column headers
3034 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3035 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3037 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3039 PH/43 Yet another patch from the Sieve maintainer.
3041 PH/44 I found a way to check for a TCP/IP connection going away before sending
3042 the response to the final '.' that terminates a message, but only in the
3043 case where the client has not sent further data following the '.'
3044 (unfortunately, this is allowed). However, in many cases there won't be
3045 any further data because there won't be any more messages to send. A call
3046 to select() can be used: if it shows that the input is "ready", there is
3047 either input waiting, or the socket has been closed. An attempt to read
3048 the next input character can distinguish the two cases. Previously, Exim
3049 would have sent an OK response which the client would never have see.
3050 This could lead to message repetition. This fix should cure that, at
3051 least in a lot of common cases.
3053 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3054 advertised in response to EHLO.
3060 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3061 fixed by 4.65/MH/01 (is this a record?) are fixed:
3063 (i) An empty string was always treated as zero by the numeric comparison
3064 operators. This behaviour has been restored.
3066 (ii) It is documented that the numeric comparison operators always treat
3067 their arguments as decimal numbers. This was broken in that numbers
3068 starting with 0 were being interpreted as octal.
3070 While fixing these problems I realized that there was another issue that
3071 hadn't been noticed. Values of message_size_limit (both the global option
3072 and the transport option) were treated as octal if they started with 0.
3073 The documentation was vague. These values are now always treated as
3074 decimal, and I will make that clear in the documentation.
3080 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3081 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3084 MH/01 Don't check that the operands of numeric comparison operators are
3085 integers when their expansion is in "skipping" mode (fixes bug
3086 introduced by 4.64-PH/07).
3088 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3089 child addresses, Exim now panics and dies. Previously, because the count
3090 is held in a short int, deliveries were likely to be lost. As such a
3091 large number of recipients for a single message is ridiculous
3092 (performance will be very, very poor), I have chosen to impose a limit
3093 rather than extend the field.
3099 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3100 leftover -K file (the existence of which was triggered by #402).
3101 While we were at it, introduced process PID as part of the -K
3102 filename. This should rule out race conditions when creating
3105 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3106 processing considerably. Previous code took too long for large mails,
3107 triggering a timeout which in turn triggers #401.
3109 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3110 in the DK code in transports.c. sendfile() is not really portable,
3111 hence the _LINUX specificness.
3113 TF/01 In the add_headers option to the mail command in an Exim filter,
3114 there was a bug that Exim would claim a syntax error in any
3115 header after the first one which had an odd number of characters
3118 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3119 callout verification, Exim cached a "reject" for the entire domain. This
3120 is correct for most verifications, but it is not correct for a recipient
3121 verification with use_sender or use_postmaster set, because in that case
3122 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3123 case of MAIL FROM:<> rejection from other early rejections (e.g.
3124 rejection of HELO). When verifying a recipient using a non-null MAIL
3125 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3126 Whatever the result of the callout, the value of the domain cache is
3127 left unchanged (for any other kind of callout, getting as far as trying
3128 RCPT means that the domain itself is ok).
3130 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3133 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3134 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3135 session, the daemon ignores ECONNECT errors and logs others; it now
3136 ignores EPIPE as well.
3138 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3139 (quoted-printable decoding).
3141 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3142 later the small subsequent patch to fix an introduced bug.
3144 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3146 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3148 PH/08 An error is now given if message_size_limit is specified negative.
3150 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3151 to be given (somewhat) arbitrary names.
3153 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3156 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3157 miscellaneous code fixes
3159 PH/10 Added the log_reject_target ACL modifier to specify where to log
3162 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3163 hostname. This is wrong, because it relates to the incoming message (and
3164 probably the interface on which it is arriving) and not to the outgoing
3165 callout (which could be using a different interface). This has been
3166 changed to use the value of the helo_data option from the smtp transport
3167 instead - this is what is used when a message is actually being sent. If
3168 there is no remote transport (possible with a router that sets up host
3169 addresses), $smtp_active_hostname is used.
3171 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3172 tweaks were necessary in order to get it to work (see also 21 below):
3173 (a) The code assumed that strncpy() returns a negative number on buffer
3174 overflow, which isn't the case. Replaced with Exim's string_format()
3176 (b) There were several signed/unsigned issues. I just did the minimum
3177 hacking in of casts. There is scope for a larger refactoring.
3178 (c) The code used strcasecmp() which is not a standard C function.
3179 Replaced with Exim's strcmpic() function.
3180 (d) The code set only $1; it now sets $auth1 as well.
3181 (e) A simple test gave the error "authentication client didn't specify
3182 service in request". It would seem that Dovecot has changed its
3183 interface. Fortunately there's a specification; I followed it and
3184 changed what the client sends and it appears to be working now.
3186 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3189 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3190 address A is aliased to B and C, where B exists and C does not. Without
3191 -v the output is "A verified" because verification stops after a
3192 successful redirection if more than one address is generated. However,
3193 with -v the child addresses are also verified. Exim was outputting "A
3194 failed to verify" and then showing the successful verification for C,
3195 with its parentage. It now outputs "B failed to verify", showing B's
3196 parentage before showing the successful verification of C.
3198 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3199 look up a TXT record in a specific list after matching in a combined
3202 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3203 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3204 they consult the DNS. I had assumed they would set it the way they
3205 wanted; and indeed my experiments on Linux seem to show that in some
3206 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3207 To be on the safe side, however, I have now made the interface to
3208 host_find_byname() similar to host_find_bydns(), with an argument
3209 containing the DNS resolver options. The host_find_byname() function now
3210 sets these options at its start, just as host_find_bydns() does. The smtp
3211 transport options dns_qualify_single and dns_search_parents are passed to
3212 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3213 of host_find_byname() use the default settings of RES_DEFNAMES
3214 (qualify_single) but not RES_DNSRCH (search_parents).
3216 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3217 spool_read_header() do less string testing, by means of a preliminary
3218 switch on the second character of optional "-foo" lines. (This is
3219 overdue, caused by the large number of possibilities that now exist.
3220 Originally there were few.) While I was there, I also converted the
3221 str(n)cmp tests so they don't re-test the leading "-" and the first
3222 character, in the hope this might squeeze out yet more improvement.
3224 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3225 flag allowing group syntax was set by the header_syntax check but not
3226 turned off, possible causing trouble later; (2) The flag was not being
3227 set at all for the header_verify test, causing "group"-style headers to
3228 be rejected. I have now set it in this case, and also caused header_
3229 verify to ignore an empty address taken from a group. While doing this, I
3230 came across some other cases where the code for allowing group syntax
3231 while scanning a header line wasn't quite right (mostly, not resetting
3232 the flag correctly in the right place). These bugs could have caused
3233 trouble for malformed header lines. I hope it is now all correct.
3235 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3236 with the "reply" argument non-NULL. The code, however (which originally
3237 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3238 but it didn't always do it. This confused somebody who was copying the
3239 code for some other use. I have removed all the tests.
3241 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3242 feature that was used to support insecure browsers during the U.S. crypto
3243 embargo. It requires special client support, and Exim is probably the
3244 only MTA that supported it -- and would never use it because real RSA is
3245 always available. This code has been removed, because it had the bad
3246 effect of slowing Exim down by computing (never used) parameters for the
3247 RSA_EXPORT functionality.
3249 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3250 authenticator to fail if there's a tab character in the incoming data
3251 (there should never be unless someone is messing about, as it's supposed
3252 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3253 if the connection is using TLS or if the remote IP is the same as the
3254 local IP, and the "valid-client-cert option" if a client certificate has
3257 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3258 authenticators. This can be used for authorization after authentication
3259 succeeds. (In the case of plaintext, it servers for both authentication
3262 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3263 if any retry times were supplied.
3265 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3266 connection, where there is no client IP address to check. In this
3267 situation, the verify now always succeeds.
3269 PH/25 Applied John Jetmore's -Mset patch.
3271 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3273 PH/27 In a string expansion for a processed (not raw) header when multiple
3274 headers of the same name were present, leading whitespace was being
3275 removed from all of them, but trailing whitespace was being removed only
3276 from the last one. Now trailing whitespace is removed from each header
3277 before concatenation. Completely empty headers in a concatenation (as
3278 before) are ignored.
3280 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3281 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3283 PH/29 [Removed. This was a change that I later backed out, and forgot to
3284 correct the ChangeLog entry (that I had efficiently created) before
3285 committing the later change.]
3287 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3288 address errors (4xx response to RCPT) over the same connection as other
3289 messages routed to the same hosts. Such deliveries are always "forced",
3290 so retry times are not inspected. This resulted in far too many retries
3291 for the affected addresses. The effect occurred only when there were more
3292 hosts than the hosts_max_try setting in the smtp transport when it had
3293 the 4xx errors. Those hosts that it had tried were not added to the list
3294 of hosts for which the message was waiting, so if all were tried, there
3295 was no problem. Two fixes have been applied:
3297 (i) If there are any address or message errors in an SMTP delivery, none
3298 of the hosts (tried or untried) are now added to the list of hosts
3299 for which the message is waiting, so the message should not be a
3300 candidate for sending over the same connection that was used for a
3301 successful delivery of some other message. This seems entirely
3302 reasonable: after all the message is NOT "waiting for some host".
3303 This is so "obvious" that I'm not sure why it wasn't done
3304 previously. Hope I haven't missed anything, but it can't do any
3305 harm, as the worst effect is to miss an optimization.
3307 (ii) If, despite (i), such a delivery is accidentally attempted, the
3308 routing retry time is respected, so at least it doesn't keep
3309 hammering the server.
3311 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3312 in ${readsocket because some servers need this prod.
3314 PH/32 Added some extra debug output when updating a wait-xxx database.
3316 PH/33 The hint "could be header name not terminated by colon", which has been
3317 given for certain expansion errors for a long time, was not being given
3318 for the ${if def:h_colon_omitted{... case.
3320 PH/34 The spec says: "With one important exception, whenever a domain list is
3321 being scanned, $domain contains the subject domain." There was at least
3322 one case where this was not true.
3324 PH/35 The error "getsockname() failed: connection reset by peer" was being
3325 written to the panic log as well as the main log, but it isn't really
3326 panic-worthy as it just means the connection died rather early on. I have
3327 removed the panic log writing for the ECONNRESET error when getsockname()
3330 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3331 runs only) independently of the message's sender address. This meant
3332 that, if the 4xx error was in fact related to the sender, a different
3333 message to the same recipient with a different sender could confuse
3334 things. In particular, this can happen when sending to a greylisting
3335 server, but other circumstances could also provoke similar problems.
3336 I have changed the default so that the retry time for these errors is now
3337 based a combination of the sender and recipient addresses. This change
3338 can be overridden by setting address_retry_include_sender=false in the
3341 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3342 remote server are returned as part of bounce messages. This was not
3343 happening for LMTP over a pipe (the lmtp transport), but now it is the
3344 same for both kinds of LMTP.
3346 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3347 in header lines that were in fact CNAMEs. This is no longer the case.
3349 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3350 and queue runs started by the daemon processed all messages. This has
3351 been fixed so that -R and -S can now usefully be given with -q<time>.
3353 PH/40 Import PCRE release 6.7 (fixes some bugs).
3355 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3357 PH/42 Give an error if -q is specified more than once.
3359 PH/43 Renamed the variables $interface_address and $interface_port as
3360 $received_ip_address and $received_port, to make it clear that these
3361 values apply to message reception, and not to the outgoing interface when
3362 a message is delivered. (The old names remain recognized, of course.)
3364 PH/44 There was no timeout on the connect() call when using a Unix domain
3365 socket in the ${readsocket expansion. There now is.
3367 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3368 be meaningful with "accept".
3370 SC/01 Eximstats V1.43
3371 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3373 SC/02 Eximstats V1.44
3374 Use a glob alias rather than an array ref in the generated
3375 parser. This improves both readability and performance.
3377 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3378 Collect SpamAssassin and rejection statistics.
3379 Don't display local sender or destination tables unless
3380 there is data to show.
3381 Added average volumes into the top table text output.
3383 SC/04 Eximstats V1.46
3384 Collect data on the number of addresses (recipients)
3385 as well as the number of messages.
3387 SC/05 Eximstats V1.47
3388 Added 'Message too big' to the list of mail rejection
3389 reasons (thanks to Marco Gaiarin).
3391 SC/06 Eximstats V1.48
3392 Mainlog lines which have GMT offsets and are too short to
3393 have a flag are now skipped.
3395 SC/07 Eximstats V1.49 (Alain Williams)
3396 Added the -emptyok flag.
3398 SC/08 Eximstats V1.50
3399 Fixes for obtaining the IP address from reject messages.
3401 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3402 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3403 whitespace changes from 4.64-PH/27
3405 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3408 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3409 are found, allow negative numbers in numeric criteria)
3411 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3413 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3414 to match changes made in 4.64-PH/43
3416 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3418 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3419 "domain" is actually an IP address, and force a failure. This locks out
3420 those revolvers/nameservers that support "A-for-A" lookups, in
3421 contravention of the specifications.
3423 PH/48 When a host name was looked up from an IP address, and the subsequent
3424 forward lookup of the name timed out, the host name was left in
3425 $sender_host_name, contrary to the specification.
3427 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3428 restricted to single-key lookups, Exim was not diagnosing an error if
3429 * or *@ was used with a query-style lookup.
3431 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3433 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3434 long ago, but noone interested enough thought of it. Let's just say that
3435 the "1.1" means that there are some new functions that weren't there at
3436 some point in the past.
3438 PH/51 Error processing for expansion failure of helo_data from an smtp
3439 transport during callout processing was broken.
3441 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3442 tested/used via the -bh/-bhc/-bs options.
3444 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3445 bug, fixed in subsequent PCRE releases).
3447 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3448 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3450 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3456 SC/01 Use a glob alias rather than an array ref in eximstats generated
3457 parser. This improves both readability and performance.
3459 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3460 Don't display local sender or destination tables in eximstats unless
3461 there is data to show.
3462 Added average volumes into the eximstats top table text output.
3464 SC/03 Collect data on the number of addresses (recipients) as well
3465 as the number of messages in eximstats.
3467 TF/01 Correct an error in the documentation for the redirect router. Exim
3468 does (usually) call initgroups() when daemonizing.
3470 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3471 with consistent privilege compared to when running as a daemon.
3473 TF/03 Note in the spec that $authenticated_id is not set for local
3474 submissions from trusted users.
3476 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3477 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3479 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3480 by adding some example configuration directives to the default
3481 configuration file. A little bit of work is required to uncomment the
3482 directives and define how usernames and passwords are checked, but
3483 there is now a framework to start from.
3485 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3486 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3487 without this. I don't know how relevant this is to other LDAP libraries.
3489 PH/02 Add the verb name to the "unknown ACL verb" error.
3491 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3493 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3495 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3496 directory not expanded when it should be if an expanded home directory
3497 was set for the address (which is overridden by the transport).
3499 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3502 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3503 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3504 because it is too late at that time, and has no effect.
3506 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3507 security issue with \' (bugzilla #107). I could not use the
3508 PQescapeStringConn() function, because it needs a PGconn value as one of
3511 PH/08 When testing addresses using -bt, indicate those final addresses that
3512 are duplicates that would not cause an additional delivery. At least one
3513 person was confused, thinking that -bt output corresponded to deliveries.
3514 (Suppressing duplicates isn't a good idea as you lose the information
3515 about possibly different redirections that led to the duplicates.)
3517 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3518 systems where poll() doesn't work, in particular OS X.
3520 PH/10 Added more information to debugging output for retry time not reached.
3522 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3523 operations in malware.c.
3525 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3528 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3529 syslog_duplication set false, log lines that would normally be written
3530 both the the main log and to the reject log were not written to syslog at
3533 PH/14 In the default configuration, change the use of "message" in ACL warn
3534 statements to "add_header".
3536 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3537 not followed by a command (e.g. "seen endif").
3539 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3540 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3543 PH/17 Added extra conditions to the default value of delay_warning_condition
3547 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3548 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3549 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3552 The Auto-Submitted: and various List- headers are standardised, whereas I
3553 don't think Precedence: ever was.
3555 PH/18 Refactored debugging code in route_finduser() to show more information,
3556 in particular, the error code if getpwnam() issues one.
3558 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3559 This is apparently needed in addition to the PH/07 change above to avoid
3560 any possible encoding problems.
3562 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3563 but not after initializing Perl.
3565 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3566 output them only if debugging. By default they are written stderr,
3567 apparently, which is not desirable.
3569 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3572 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3575 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3577 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3578 authenticated or an ident call has been made. Suppress the default
3579 values for $authenticated_id and $authenticated_sender (but permit -oMai
3580 and -oMas) when testing with -bh.
3582 PH/24 Re-jigged the order of the tests in the default configuration so that the
3583 tests for valid domains and recipients precede the DNS black list and CSA
3584 tests, on the grounds that those ones are more expensive.
3586 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3587 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3588 This bug exists in every version of Exim that I still have, right back to
3591 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3592 However, an attempt to turn on case-sensitivity in a regex key by
3593 including (?-i) didn't work because the subject string was already
3594 lowercased, and the effects were non-intuitive. It turns out that a
3595 one-line patch can be used to allow (?-i) to work as expected.
3601 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3602 other effects) broke the use of negated acl sub-conditions.
3604 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3607 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3608 "Deny" causes Exim to reject the incoming connection with a 554 error.
3609 Unfortunately, if there is a major crisis, such as a disk failure,
3610 tcp-wrappers gives "deny", whereas what one would like would be some
3611 kind of temporary error. A kludge has been added to help with this.
3612 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3613 554 error is used if errno is still zero or contains ENOENT (which occurs
3614 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3617 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3619 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3620 errors. Otherwise a message that provokes a temporary error (when other
3621 messages do not) can cause a whole host to time out.
3623 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3624 addresses were routed directly to files or pipes from a redirect router.
3625 File deliveries just didn't batch; pipe deliveries might have suffered
3628 PH/06 A failure to get a lock for a hints database would erroneously always say
3629 "Failed to get write lock", even when it was really a read lock.
3631 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3632 of 0600. This has been changed to use the value of the lockfile_mode
3633 option (which defaults to 0600).
3635 PH/08 Applied small patch from the Sieve maintainer.
3637 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3638 folder from quota calculations, a direct delivery into this folder messed
3639 up the contents of the maildirsize file. This was because the regex was
3640 used only to exclude .Trash (or whatever) when the size of the mailbox
3641 was calculated. There was no check that a delivery was happening into an
3642 excluded directory. This bug has been fixed by ignoring all quota
3643 processing for deliveries into excluded directories.
3645 PH/10 Added the maildirfolder_create_regex option to appendfile.
3651 PH/01 The code for finding all the local interface addresses on a FreeBSD
3652 system running IPv6 was broken. This may well have applied to all BSD
3653 systems, as well as to others that have similar system calls. The broken
3654 code found IPv4 interfaces correctly, but gave incorrect values for the
3655 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3656 that it would not match correctly against @[] and not recognize the IPv6
3659 PH/02 The ipliteral router was not recognizing addresses of the form user@
3660 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3662 PH/03 Added disable_ipv6.
3664 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3665 decoded form, because it is most often used to construct To: headers
3666 lines in autoreplies, and the decoded form may well be syntactically
3667 invalid. However, $reply_address has leading white space removed, and all
3668 newlines turned into spaces so that the autoreply transport does not
3671 PH/05 If group was specified without a user on a router, and no group or user
3672 was specified on a transport, the group from the router was ignored.
3674 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3675 for visible compile-time settings that can be used to change these
3676 numbers, for those that want even more. Backwards compatibility with old
3677 spool files has been maintained. However, going back to a previous Exim
3678 release will lost any variables that are in spool files.
3680 PH/07 Two small changes when running in the test harness: increase delay when
3681 passing a TCP/IP connection to a new process, in case the original
3682 process has to generate a bounce, and remove special handling of
3683 127.0.0.2 (sic), which is no longer necessary.
3685 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3686 be the same on different OS.
3688 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3691 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3692 whether --show-vars was specified or not
3694 JJ/02 exipick: Added support for new ACL variable spool format introduced
3697 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3698 syntactically invalid From: or Reply-to: line, and a filter used this to
3699 generate an autoreply, and therefore failed to obtain an address for the
3700 autoreply, Exim could try to deliver to a non-existent relative file
3701 name, causing unrelated and misleading errors. What now happens is that
3702 it logs this as a hard delivery error, but does not attempt to create a
3705 PH/11 The exinext utility has a -C option for testing purposes, but although
3706 the given file was scanned by exinext itself; it wasn't being passed on
3707 when Exim was called.
3709 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3710 an end-of-file indication when reading a command response.
3712 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3713 compiled. In many other places in Exim, IPv6 addresses are always
3714 recognized, so I have changed this. It also means that IPv4 domain
3715 literals of the form [IPV4:n.n.n.n] are now always recognized.
3717 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3718 used if the router is not running as root, for example, when verifying at
3719 ACL time, or when using -bh. The debugging output from this situation was
3720 non-existent - all you got was a failure to exec. I have made two
3723 (a) Failures to set uid/gid, the current directory, or a process leader
3724 in a subprocess such as that created by queryprogram now generate
3725 suitable debugging output when -d is set.
3727 (b) The queryprogram router detects when it is not running as root,
3728 outputs suitable debugging information if -d is set, and then runs
3729 the subprocess without attempting to change uid/gid.
3731 PH/15 Minor change to Makefile for building test_host (undocumented testing
3734 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3735 additional section of a DNS packet that returns MX or SRV records.
3736 Instead, it always explicitly searches for A/AAAA records. This avoids
3737 major problems that occur when a DNS server includes only records of one
3738 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3739 fixed another bug: if SRV records were looked up and the corresponding
3740 address records were *not* found in the additional section, the port
3741 values from the SRV records were lost.
3743 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3744 using the correct key (the original address) when searching the retry
3745 rules in order to find which one to use for generating the retry hint.
3747 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3748 adding the default one. Similarly, if it contains a Reply-To: header, the
3749 errors_reply_to option, if set, is not used.
3751 PH/19 When calculating a retry time, Exim used to measure the "time since
3752 failure" by looking at the "first failed" field in the retry record. Now
3753 it does not use this if it is later than than the arrival time of the
3754 message. Instead it uses the arrival time. This makes for better
3755 behaviour in cases where some deliveries succeed, thus re-setting the
3756 "first failed" field. An example is a quota failure for a huge message
3757 when small messages continue to be delivered. Without this change, the
3758 "time since failure" will always be short, possible causing more frequent
3759 delivery attempts for the huge message than are intended.
3760 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3762 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3763 $1, $2, $3) because the numerical variables can be reset during some
3764 expansion items (e.g. "match"), thereby losing the authentication data.
3766 PH/21 Make -bV show the size of off_t variables so that the test suite can
3767 decide whether to run tests for quotas > 2G.
3769 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3770 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3771 filecount value is greater than 2G or if a quota value is greater than 2G
3772 on a system where the size of off_t is not greater than 4, a panic error
3775 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3776 never match. The debug and -bh output now contains an explicit error
3777 message indicating a malformed IPv4 address or mask.
3779 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3780 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3781 PH/23 above applies.
3783 PH/25 Do not write to syslog when running in the test harness. The only
3784 occasion when this arises is a failure to open the main or panic logs
3785 (for which there is an explicit test).
3787 PH/26 Added the /no_tell option to "control=freeze".
3789 PH/27 If a host name lookup failed very early in a connection, for example, if
3790 the IP address matched host_lookup and the reverse lookup yielded a name
3791 that did not have a forward lookup, an error message of the form "no IP
3792 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3793 could be logged. Now it outputs the IP address instead of "NULL".
3795 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3796 allows the sender and the authenticated sender to be set when
3797 submitting a message from within Exim. Since child_open_exim() is
3798 documented for local_scan(), the new function should be too.
3800 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3801 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3802 results in an empty string is now treated as unset.
3804 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3806 PH/31 Added sender_verify_fail logging option.
3808 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3809 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3810 tidied the source and removed it altogether.
3812 PH/33 When a queue run was abandoned because the load average was too high, a
3813 log line was always written; now it is written only if the queue_run log
3814 selector is set. In addition, the log line for abandonment now contains
3815 information about the queue run such as the pid. This is always present
3816 in "start" and "stop" lines but was omitted from the "abandon" line.
3818 PH/34 Omit spaces between a header name and the colon in the error message that
3819 is given when verify = headers_syntax fails (if there are lots of them,
3820 the message gets confusing).
3822 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3823 names, as there are now some PTR records that contain slashes. This check
3824 is only to protect against broken name servers that fall over on strange
3825 characters, so the fact that it applies to all lookups doesn't matter.
3827 PH/36 Now that the new test suite is complete, we can remove some of the
3828 special code in Exim that was needed for the old test suite. For example,
3829 sorting DNS records because real resolvers return them in an arbitrary
3830 order. The new test suite's fake resolver always returns records in the
3833 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3834 bounces) except when queue_only is set, to avoid logging races between
3835 the different processes.
3837 PH/38 Panic-die if .include specifies a non-absolute path.
3839 PH/39 A tweak to the "H" retry rule from its user.
3841 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3842 a label. They prevented compilation on older perls.
3844 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3845 a warning to be raised on newish perls.
3847 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3848 on queue. Changes to match documented behaviour of showing count of
3849 messages matching specified criteria.
3851 PH/40 Changed the default ident timeout from 30s to 5s.
3853 PH/41 Added support for the use of login_cap features, on those BSD systems
3854 that have them, for controlling the resources used by pipe deliveries.
3856 PH/42 The content-scanning code uses fopen() to create files in which to put
3857 message data. Previously it was not paying any attention to the mode of
3858 the files. Exim runs with umask(0) because the rest of the code creates
3859 files with open(), and sets the required mode explicitly. Thus, these
3860 files were ending up world-writeable. This was not a big issue, because,
3861 being within the spool directory, they were not world-accessible. I have
3862 created a function called modefopen, which takes an additional mode
3863 argument. It sets umask(777), creates the file, chmods it to the required
3864 mode, then resets the umask. All the relevant calls to fopen() in the
3865 content scanning code have been changed to use this function.
3867 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3868 to 24 hours. This avoids potential overflow problems when processing G
3869 and H retry rules. I suspect nobody ever tinkers with this value.
3871 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3873 PH/45 When the plaintext authenticator is running as a client, the server's
3874 challenges are checked to ensure they are valid base64 strings. By
3875 default, the authentication attempt is cancelled if an invalid string is
3876 received. Setting client_ignore_invalid_base64 true ignores these errors.
3877 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3878 they are received. Thus, the responses can be made to depend on the
3879 challenges. If an invalid string is ignored, an empty string is placed in
3882 PH/46 Messages that are created by the autoreply transport now contains a
3883 References: header, in accordance with RFCs 2822 and 3834.
3885 PH/47 Added authenticated_sender_force to the smtp transport.
3887 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3889 PH/49 Installed latest patch from the Sieve maintainer.
3891 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3892 was also set, the appendfile transport was unnecessarily scanning a
3893 directory of message files (e.g. for maildir delivery) to find the count
3894 of files (along with the size), even though it did not need this
3895 information. It now does the scan only if it needs to find either the
3896 size of the count of files.
3898 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3900 PH/52 Two bugs concerned with error handling when the smtp transport is
3903 (i) Exim was not creating retry information for temporary errors given
3904 for individual recipients after the DATA command when the smtp transport
3905 was used in LMTP mode. This meant that they could be retried too
3906 frequently, and not timed out correctly.
3908 (ii) Exim was setting the flag that allows error details to be returned
3909 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3910 recipients that were returned after the DATA command.
3912 PH/53 This is related to PH/52, but is more general: for any failing address,
3913 when detailed error information was permitted to be returned to the
3914 sender, but the error was temporary, then after the final timeout, only
3915 "retry timeout exceeded" was returned. Now it returns the full error as
3916 well as "retry timeout exceeded".
3918 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3919 do this, and (what is worse) MTAs that accept it.
3921 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3922 will now be deprecated.
3924 PH/56 New os.c-cygwin from the Cygwin maintainer.
3926 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3927 formats (previously only available in exim formats via -bpr, -bpru,
3928 and -bpra. Now also available in native and exiqgrep formats)
3930 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3931 with very large, slow to parse queues
3933 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3935 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3937 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3938 responses to authentication challenges, though it was showing the
3939 challenges; (ii) I've removed the CR characters from the debug output for
3942 PH/58 Allow for the insertion of a newline as well as a space when a string
3943 is turned into more than one encoded-word during RFC 2047 encoding. The
3944 Sieve code now uses this.
3946 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3947 data_4xx, lost_connection, tls_required.
3949 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3950 message was being sent as an SMTP response.
3952 PH/61 Add -l and -k options to exicyclog.
3954 PH/62 When verifying, if an address was redirected to one new address, so that
3955 verification continued, and the new address failed or deferred after
3956 having set something in $address_data, the value of $address_data was not
3957 passed back to the ACL. This was different to the case when no
3958 redirection occurred. The value is now passed back in both cases.
3960 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3961 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3962 use login_cap.h, so on its own it isn't the distinguishing feature. The
3963 new name refers directly to the setclassresources() function.
3965 PH/65 Added configuration files for NetBSD3.
3967 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3969 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3970 is preferred over IPv4.
3972 PH/68 The bounce_return_message and bounce_return_body options were not being
3973 honoured for bounces generated during the reception of non-SMTP messages.
3974 In particular, this applied to messages rejected by the ACL. This bug has
3975 been fixed. However, if bounce_return_message is true and bounce_return_
3976 body is false, the headers that are returned for a non-SMTP message
3977 include only those that have been read before the error was detected.
3978 (In the case of an ACL rejection, they have all been read.)
3980 PH/69 The HTML version of the specification is now built in a directory called
3981 spec_html instead of spec.html, because the latter looks like a path with
3982 a MIME-type, and this confuses some software.
3984 PH/70 Catch two compiler warnings in sieve.c.
3986 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3987 function verify_get_ident() calls ip_connect() to connect a socket, but
3988 if the "connect()" function timed out, ip_connect() used to close the
3989 socket. However, verify_get_ident() also closes the socket later, and in
3990 between Exim writes to the log, which may get opened at this point. When
3991 the socket was closed in ip_connect(), the log could get the same file
3992 descriptor number as the socket. This naturally causes chaos. The fix is
3993 not to close the socket in ip_connect(); the socket should be closed by
3994 the function that creates it. There was only one place in the code where
3995 this was missing, in the iplookup router, which I don't think anybody now
3996 uses, but I've fixed it anyway.
3998 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3999 well as to direct DNS lookups. Otherwise the handling of names in host
4000 lists is inconsistent and therefore confusing.
4006 PH/01 Two changes to the default runtime configuration:
4008 (1) Move the checks for relay_from_hosts and authenticated clients from
4009 after to before the (commented out) DNS black list checks.
4011 (2) Add control=submission to the relay_from_hosts and authenticated
4012 clients checks, on the grounds that messages accepted by these
4013 statements are most likely to be submissions.
4015 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4017 (1) Generate an error if the third argument for the ${prvs expansion is
4020 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4023 (3) Reset the variables that are obtained from the first argument of
4024 ${prvscheck and used in the second argument before leaving the code,
4025 because their memory is reclaimed, so using them afterwards may do
4028 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4029 one (it's much easier than Tom thought :-).
4031 (5) Because of (4), we can now allow for the use of $prvscheck_result
4032 inside the third argument.
4034 PH/03 For some reason, the default setting of PATH when running a command from
4035 a pipe transport was just "/usr/bin". I have changed it to
4038 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4039 anything to be listed in the output from -bV.
4041 PH/05 When a filter generated an autoreply, the entire To: header line was
4042 quoted in the delivery log line, like this:
4044 => >A.N.Other <ano@some.domain> <original@ddress> ...
4046 This has been changed so that it extracts the operative address. There
4047 may be more than one such address. If so, they are comma-separated, like
4050 => >ano@some.domain,ona@other.domain <original@ddress> ...
4052 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4053 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4054 not being looked up in the rDNS to get a host name, Exim was showing the
4055 IP address twice in Received: lines, even though the IP addresses were
4056 identical. For example:
4058 Received: from [1.2.3.4] (helo=[1.2.3.4])
4060 However, if the real host name was known, it was omitting the HELO data
4061 if it matched the actual IP address. This has been tidied up so that it
4062 doesn't show the same IP address twice.
4064 PH/07 When both +timestamp and +memory debugging was on, the value given by
4065 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4066 called by the debug printing, thereby overwriting the timestamp buffer.
4067 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4069 PH/08 When the original message was included in an autoreply transport, it
4070 always said "this is a copy of the message, including all the headers",
4071 even if body_only or headers_only was set. It now gives an appropriate
4074 PH/09 Applied a patch from the Sieve maintainer which:
4076 o fixes some comments
4077 o adds the (disabled) notify extension core
4078 o adds some debug output for the result of if/elsif tests
4079 o points to the current vacation draft in the documentation
4080 and documents the missing references header update
4084 o fixes a bug in processing the envelope test (when testing
4085 multiple envelope elements, the last element determined the
4088 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4089 Electronic Mail") by including:
4091 Auto-submitted: auto-generated
4093 in the messages that it generates (bounce messages and others, such as
4094 warnings). In the case of bounce messages for non-SMTP messages, there was
4095 also a typo: it was using "Auto_submitted" (underscore instead of
4096 hyphen). Since every message generated by Exim is necessarily in response
4097 to another message, thes have all been changed to:
4099 Auto-Submitted: auto-replied
4101 in accordance with these statements in the RFC:
4103 The auto-replied keyword:
4105 - SHOULD be used on messages sent in direct response to another
4106 message by an automatic process,
4108 - MUST NOT be used on manually-generated messages,
4110 - MAY be used on Delivery Status Notifications (DSNs) and Message
4111 Disposition Notifications (MDNs),
4113 - MUST NOT be used on messages generated by automatic or periodic
4114 processes, except for messages which are automatic responses to
4117 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4118 to the default Received: header definition.
4120 PH/12 Added log selector acl_warn_skipped (default on).
4122 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4123 variables because (a) they are in the wrong storage pool and (b) even if
4124 they were copied, it wouldn't work properly because of the caching.
4126 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4127 checking when decoding. Apparently there are clients that generate
4128 overlong encoded strings. Why am I not surprised?
4130 PH/15 If the first argument of "${if match_address" was not empty, but did not
4131 contain an "@" character, Exim crashed. Now it writes a panic log message
4132 and treats the condition as false.
4134 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4136 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4137 extension "envelope-auth". The code is finished and in agreement with
4138 other implementations, but there is no documentation so far and in fact,
4139 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4140 not changing the active code.
4142 Print executed "if" and "elsif" statements when debugging is used. This
4143 helps a great deal to understand what a filter does.
4145 Document more things not specified clearly in RFC3028. I had all this
4146 sorted out, when out of a sudden new issues came to my mind. Oops."
4148 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4151 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4152 canonical form (as documented). However, after a host name lookup from
4153 the IP address, check_host() was doing a simple string comparison with
4154 addresses acquired from the DNS when checking that the found name did
4155 have the original IP as one of its addresses. Since any found IPv6
4156 addresses are likely to be in abbreviated form, the comparison could
4157 fail. Luckily, there already exists a function for doing the comparison
4158 by converting both addresses to binary, so now that is used instead of
4159 the text comparison.
4161 PH/20 There was another similar case to PH/19, when a complete host name was
4162 given in a host list; looking up its IP address could give an abbreviated
4163 form, whereas the current host's name might or might not be abbreviated.
4164 The same fix has been applied.
4170 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4171 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4174 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4175 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4177 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4179 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4180 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4181 Also, for "Auto-Submitted", treat anything other than "no" as
4182 non-personal, in accordance with RFC 3834. (Previously it treated
4183 anything starting "auto-" as non-personal.)
4185 TF/01 The control=submission/name=... option had a problem with syntax
4186 errors if the name included a slash character. The /name= option
4187 now slurps the rest of the string, so it can include any characters
4188 but it must come last in the list of options (after /sender_retain
4191 PH/05 Some modifications to the interface to the fake nameserver for the new
4199 TK/01 Added the "success_on_redirect" address verification option. See
4200 NewStuff for rationale and an example.
4202 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4204 PH/02 Patch to exigrep to allow it to work on syslog lines.
4206 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4207 fread() to skip over the body file's header line, because in Cygwin the
4208 header line is locked and is inaccessible.
4210 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4211 co-exist for some time) to make it clear that it is the Exim ID that is
4212 referenced, not the Message-ID: header line.
4214 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4215 string_format() function, because snprintf() does not exist on all
4218 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4219 ${stat: expansion item.
4221 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4222 protocol synchronization error", to keep the pedants happy.
4224 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4225 well as for IRIX systems, when gcc is being used. See the host.c source
4228 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4230 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4233 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4234 transport and to the smtp transport in LMTP mode.
4236 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4238 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4239 run a filter in a subprocess. This could lead to confusion in subsequent
4240 lookups in the parent process. There should also be a search_tidyup() at
4241 the end of the subprocess.
4243 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4244 only if the host matched helo_try_verify_hosts, which caused the
4245 verification to occur when the EHLO/HELO command was issued. The ACL just
4246 tested the remembered result. Now, if a previous verification attempt has
4247 not happened, "verify = helo" does it there and then.
4249 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4251 TK/03 Fix log output including CR from clamd.
4253 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4254 exist provoked a memory error which could cause a segfault.
4256 PH/15 Installed PCRE 6.2
4258 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4260 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4261 of the problem. Specifically, suggested +O2 rather than +O1 for the
4264 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4266 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4267 record, the port was not correctly logged when the outgoing_port log
4268 selector was set (it logged the transort's default port).
4270 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4271 fallback_hosts, and "hosts" in the smtp transport.
4273 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4274 host errors such as "Connection refused".
4276 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4277 authentication with radiusclient 0.4.9:
4279 - Error returned from rc_read_config was caught wrongly
4280 - Username/password not passed on to radius server due to wrong length.
4282 The presumption is that some radiusclient API changes for 4.51/PH/17
4283 were not taken care of correctly. The code is still untested by me (my
4284 Linux distribution still has 0.3.2 of radiusclient), but it was
4285 contributed by a Radius user.
4287 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4288 expanding the "port" option of the smtp transport.
4290 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4291 while reading a MIME header. Thanks to Tom Hughes for a patch.
4293 PH/24 Include config.h inside local_scan.h so that configuration settings are
4296 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4297 that in an ACL for RCPT (for example), you can examine exactly what was
4300 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4301 commands, but it was not correctly comparing the address with the actual
4302 client host address. Thus, it would show the EHLO address in Received:
4303 header lines when this was not necessary.
4305 PH/27 Added the % operator to ${eval:}.
4307 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4308 it should be ignoring failures (because with -C, for example, it has lost
4309 privilege). It wasn't ignoring creation failures other than "already
4312 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4315 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4316 because an input line was too long, either on its own, or by virtue of
4317 too many continuations, the temporary file was not being removed, and the
4318 return code was incorrect.
4320 PH/31 Missing "BOOL" in function definition in filtertest.c.
4322 PH/32 Applied Sieve patches from the maintainer.
4324 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4326 PH/33 Added "verify = not_blind".
4328 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4329 Local/Makefile (with some defaults set). These are used in built scripts
4330 such as exicyclog, but they have never been used in the exim_install
4331 script (though there are many overriding facilities there). I have
4332 arranged that the exim_install script now takes note of these two
4335 PH/35 Installed configuration files for Dragonfly.
4337 PH/36 When a locally submitted message by a trusted user did not contain a
4338 From: header, and the sender address was obtained from -f or from an SMTP
4339 MAIL command, and the trusted user did not use -F to supply a sender
4340 name, $originator_name was incorrectly used when constructing a From:
4341 header. Furthermore, $originator_name was used for submission mode
4342 messages from external hosts without From: headers in a similar way,
4343 which is clearly wrong.
4345 PH/37 Added control=suppress_local_fixups.
4347 PH/38 When log_selector = +received_sender was set, and the addition of the
4348 sender made the log line's construction buffer exactly full, or one byte
4349 less than full, an overflow happened when the terminating "\n" was
4352 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4353 when the result of a list match is failure because a DNS lookup failed.
4355 PH/40 RM_COMMAND is now used in the building process.
4357 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4358 the "build-* directories that it finds.
4360 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4361 address in a domain literal was a prefix of an interface address.
4363 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4364 when verifying a sender address, unless rewrite_headers is false.
4366 PH/44 Wrote a long comment about why errors_to addresses are verified as
4367 recipients, not senders.
4369 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4370 the ratelimit ACL was added.
4372 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4374 PH/46 Added extra information about PostgreSQL errors to the error string.
4376 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4377 avoiding the need to install special zones in a real server. This is
4378 backwards compatible; if it can't find the fake resolver, it drops back.
4379 Thus, both old and new test suites can be run.
4381 TF/02 Added util/ratelimit.pl
4383 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4384 clock is set back in time.
4386 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4387 Candler <B.Candler@pobox.com>.
4389 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4390 OK for addresses that are the result of redirections.
4392 PH/48 A number of further additions for the benefit of the new test suite,
4393 including a fake gethostbyname() that interfaces to the fake DNS resolver
4396 TF/06 The fix for widen_domains has also been applied to qualify_single and
4397 search_parents which are the other dnslookup options that can cause
4400 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4403 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4405 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4406 service" if it encountered an MX record with an empty target hostname.
4407 The message is now "an MX or SRV record indicated no SMTP service".
4409 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4410 if the DNS of the sending site is misconfigured. This is quite a
4411 common situation. This change restores the behaviour of treating a
4412 helo verification defer as a failure.
4414 PH/51 If self=fail was set on a router, the bounce message did not include the
4415 actual error message.
4421 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4423 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4424 command itself ended in error, the underlying message about the transport
4425 filter timeout was being overwritten with the pipe command error. Now the
4426 underlying error message should be appended to the second error message.
4428 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4430 PH/02 Reduce the amount of output that "make" produces by default. Full output
4431 can still be requested.
4433 PH/03 The warning log line about a condition test deferring for a "warn" verb
4434 was being output only once per connection, rather than after each
4435 occurrence (because it was using the same function as for successful
4436 "warn" verbs). This seems wrong, so I have changed it.
4438 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4439 it should not have, which might have caused a crash in the right
4440 circumstances, but probably never did.
4442 PH/04 Installed a modified version of Tony Finch's patch to make submission
4443 mode fix the return path as well as the Sender: header line, and to
4444 add a /name= option so that you can make the user's friendly name appear
4447 TF/03 Added the control = fakedefer ACL modifier.
4449 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4450 Mark Lowes for thorough testing.
4452 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4454 TK/03 Merged latest SRS patch from Miles Wilton.
4456 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4457 with the definition in sysexits.h (which is #included earlier).
4458 Fortunately, Exim does not actually use EX_OK. The code used to try to
4459 preserve the sysexits.h value, by assuming that macro definitions were
4460 scanned for macro replacements. I have been disabused of this notion,
4461 so now the code just undefines EX_OK before #including unistd.h.
4463 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4464 in the smtp transport. When a block could not be written in a single
4465 write() function, the timeout was being re-applied to each part-write.
4466 This seems wrong - if the receiver was accepting one byte at a time it
4467 would take for ever. The timeout is now adjusted when this happens. It
4468 doesn't have to be particularly precise.
4470 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4471 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4473 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4474 without a domain if the check to <postmaster@domain> fails.
4476 SC/01 Eximstats: added -xls and the ability to specify output files
4477 (patch written by Frank Heydlauf).
4479 SC/02 Eximstats: use FileHandles for outputting results.
4481 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4483 SC/04 Eximstats: fixed display of large numbers with -nvr option
4485 SC/05 Eximstats: fixed merging of reports with empty tables.
4487 SC/06 Eximstats: added the -include_original_destination flag
4489 SC/07 Eximstats: removed tabs and trailing whitespace.
4491 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4493 TK/06 MBOX spool code: Add real "From " MBOX separator line
4494 so the .eml file is really in mbox format (even though
4495 most programs do not really care). Patch from Alex Miller.
4497 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4498 The latter is generated from $received_to and is only set if the
4499 message has one envelope recipient. SA can use these headers,
4500 obviously out-of-the-box. Patch from Alex Miller.
4502 PH/08 The ${def test on a variable was returning false if the variable's
4503 value was "0", contrary to what the specification has always said!
4504 The result should be true unless the variable is empty.
4506 PH/09 The syntax error of a character other than { following "${if
4507 def:variable_name" (after optional whitespace) was not being diagnosed.
4508 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4509 accidental colon was present, for example, could give incorrect results.
4511 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4512 result is used (not including appendfile, where other changes are about
4515 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4516 This involved changing a lot of size variables from int to off_t. It
4517 should work with maildirs and everything.
4519 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4520 spamd dying while we are connected to it.
4522 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4525 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4526 function for BDB 4.3.
4528 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4530 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4531 the sources for PCRE have been re-arranged and more files are now
4534 PH/15 The code I had for printing potentially long long variables in PH/11
4535 above was not the best (it lost precision). The length of off_t variables
4536 is now inspected at build time, and an appropriate printing format (%ld
4537 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4538 to be "long long int" or "long int". This is needed for the internal
4539 formatting function string_vformat().
4541 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4542 the configuration file to be ":syslog", then the script "guesses" where
4543 the logs files are, rather than using the compiled in default. In our
4544 case the guess is not the same as the compiled default, so the script
4545 suddenly stopped working when I started to use syslog. The patch checks
4546 to see if log_file_path is "". If so, it attempts to read it from exim
4547 with no configuration file to get the compiled in version, before it
4548 falls back to the previous guessing code."
4550 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4551 implementing BATV in an Exim configuration. See NewStuff for the gory
4554 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4555 Makefile that are specific to HP-UX.
4557 PH/18 If the "use_postmaster" option was set for a recipient callout together
4558 with the "random" option, the postmaster address was used as the MAIL
4559 FROM address for the random test, but not for the subsequent recipient
4560 test. It is now used for both.
4562 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4563 patch removes a few documentation additions to RFC 3028, because the
4564 latest draft now contains them. It adds the new en;ascii-case comparator
4565 and a new error check for 8bit text in MIME parts. Comparator and
4566 require names are now matched exactly. I enabled the subaddress
4567 extension, but it is not well tested yet (read: it works for me)."
4569 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4570 rework some of the code of TK/09 above to avoid the hardwired use of
4571 "%lld" and "long long". Replaced the call to snprintf() with a call to
4574 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4575 records point to non-existent hosts", "retry timeout exceeded", and
4576 "retry time not reached for any host after a long failure period".
4578 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4579 experimental DomainKeys support:
4581 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4582 (2) On an error such as an illegally used "control", the wrong name for
4583 the control was given.
4585 These problems did NOT occur unless DomainKeys support was compiled.
4587 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4589 PH/24 Added ${if match_ip condition.
4591 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4592 databases so that it will be absolutely obvious if a crash occurs in the
4593 DB library. This is a regular occurrence (often caused by mis-matched
4596 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4597 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4598 picked up on a user's system that detects such things. There doesn't seem
4599 to be a gcc warning option for this - only an attribute that has to be
4600 put on the function's prototype. It seems that in Fedora Core 4 they have
4601 set this on a number of new functions. No doubt there will be more in due
4604 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4605 specify a transport. However, if an address that was verified by such a
4606 router was the subject of a callout, Exim crashed because it tried to
4607 read the rcpt_include_affixes from the non-existent transport. Now it
4608 just assumes that the setting of that option is false. This bug was
4609 introduced by 4.51/PH/31.
4611 PH/28 Changed -d+all to exclude +memory, because that information is very
4612 rarely of interest, but it makes the output a lot bigger. People tend to
4613 do -d+all out of habit.
4615 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4616 code in os-type was giving problems when libc.so lives in lib64, like on
4619 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4620 aren't the modern standard, and it seems that some systems' include files
4621 don't always have them. Exim was already checking for some of the newer
4622 ones like T_AAAA, and defining it itself. I've added checks for all the
4623 record types that Exim uses.
4625 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4626 not automatically generating a new one, as it is supposed to. This
4627 prevented TLS from working. If the file did exist, but contained invalid
4628 data, a new version was generated, as expected. It was only the case of a
4629 non-existent file that was broken.
4631 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4632 with a change in libdomainkeys > 0.64.
4634 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4635 from DNS. If the selector record carries the flag, it now has
4636 precedence over the domain-wide flag.
4638 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4640 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4641 the use of a port name that isn't defined in /etc/services) occurred, the
4642 message was deferred as in a normal delivery, and thus remained on the
4643 spool, instead of being failed because of the mua_wrapper setting. This
4644 is now fixed, and I tidied up some of the mua_wrapper messages at the
4647 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4648 the messages in a hash of arrays rather than using individual hashes.
4649 This is a bit cleaner and results in dramatic memory savings, albeit
4650 at a slight CPU cost.
4652 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4653 as requested by Marc Sherman.
4655 SC/10 Eximstats: added histograms for user specified patterns as requested
4658 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4660 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4661 fopen() in the content-scanning modules that did not already have it.
4667 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4668 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4670 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4672 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4674 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4675 to test Sieve filters that use "vacation".
4677 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4678 that changes the way the GnuTLS parameters are stored in the cache file.
4679 The new format can be generated externally. For backward compatibility,
4680 if the data in the cache doesn't make sense, Exim assumes it has read an
4681 old-format file, and it generates new data and writes a new file. This
4682 means that you can't go back to an older release without removing the
4685 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4686 work if there are any delivery delays because "one_time" forces the
4687 parent to be marked "delivered", so its unseen clone is never tried
4688 again. For this reason, Exim now forbids the simultaneous setting of
4691 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4692 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4693 entry if you want to know the details. The fix, however, neglected to
4694 consider the case when local delivery batching is involved. The test for
4695 "previously delivered" was not happening when checking to see if an
4696 address could be batched with a previous (undelivered) one; under
4697 certain circumstances this could lead to multiple deliveries to the same
4700 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4701 in its include files, and this causes problems building Exim.
4703 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4704 header_syntax) but Exim was just ignoring anything given after a slash.
4705 In particular, this caused confusion with an attempt to use "verify =
4706 reverse_host_lookup/defer_ok". An error is now given when options are
4707 supplied for verify items that do not have them. (Maybe reverse_host_
4708 lookup should have a defer_ok option, but that's a different point.)
4710 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4711 defined by RFC 821) to 2048, because there were problems with some AUTH
4712 commands, and RFC 1869 says the size should be increased for extended
4713 SMTP commands that take arguments.
4715 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4718 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4719 "unknown" error; now it says that the functionality isn't in the binary.
4721 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4722 an address' error message when a string expansion fails (syntax or
4723 whatever). Otherwise the password may appear in the log. Following change
4724 PH/42 below, there is no longer a chance of it appearing in a bounce
4727 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4729 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4730 of its addresses was ever tried. (Bugzilla bug #2.)
4732 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4733 the result incorrectly in the debug output. (It correctly added a newline
4734 to what was transported.)
4736 TF/01 Added $received_time.
4738 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4739 commented out examples of how to interface to a virus scanner and to
4740 SpamAssassin. Also added commented examples of av_scanner and
4741 spamd_address settings.
4743 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4744 and controls are allowed in which ACLs. There were a couple of minor
4745 errors. Some of the entries in the conditions table (which is a table of
4746 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4747 negation of where the condition IS allowed.
4749 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4751 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4752 header file does not have a version number, so I've had to invent a new
4753 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4754 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4755 radiusclient), but it was contributed by a Radius user.
4757 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4758 files or directories, for OpenSSL.
4760 PH/19 When an Exim process that is running as root has to create an Exim log
4761 file, it does so in a subprocess that runs as exim:exim so as to get the
4762 ownership right at creation (otherwise, other Exim processes might see
4763 the file with the wrong ownership). There was no test for failure of this
4764 fork() call, which would lead to the process getting stuck as it waited
4765 for a non-existent subprocess. Forks do occasionally fail when resources
4766 run out. I reviewed all the other calls to fork(); they all seem to check
4769 PH/20 When checking for unexpected SMTP input at connect time (before writing
4770 the banner), Exim was not dealing correctly with a non-positive return
4771 from the read() function. If the client had disconnected by this time,
4772 the result was a log entry for a synchronization error with an empty
4773 string after "input=" when read() returned zero. If read() returned -1
4774 (an event I could not check), uninitialized data bytes were printed.
4775 There were reports of junk text (parts of files, etc) appearing after
4778 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4780 PH/22 Added support for macro redefinition, and (re)definition in between
4781 driver and ACL definitions.
4783 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4784 forgetting to use the resulting value; it was using the unexpanded value.
4786 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4787 hadn't been configured. The fix is from Juergen Kreileder, who
4788 understands it better than I do:
4790 "Here's what I see happening with three configured cyrus_sasl
4791 authenticators configured (plain, login, cram-md5):
4793 On startup auth_cyrus_sasl_init() gets called for each of these.
4794 This means three calls to sasl_listmech() without a specified mech_list.
4795 => SASL tests which mechs of all available mechs actually work
4796 => three warnings about OTP not working
4797 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4799 With the patch, sasl_listmech() also gets called three times. But now
4800 SASL's mech_list option is set to the server_mech specified in the the
4801 authenticator. Or in other words, the answer from sasl_listmech()
4802 gets limited to just the mech you're testing for (which is different
4804 => the return list contains just 'plain' or 'login', 'cram-md5' or
4805 nothing depending on the value of ob->server_mech.
4807 I've just tested the patch: Authentication still works fine,
4808 unavailable mechs specified in the exim configuration are still
4809 caught, and the auth.log warnings about OTP are gone."
4811 PH/25 When debugging is enabled, the contents of the command line are added
4812 to the debugging output, even when log_selector=+arguments is not
4815 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4816 answer is "GNU", and only if the return is "GNU/something" is the answer
4819 PH/27 $acl_verify_message is now set immediately after the failure of a
4820 verification in an ACL, and so is available in subsequent modifiers. In
4821 particular, the message can be preserved by coding like this:
4823 warn !verify = sender
4824 set acl_m0 = $acl_verify_message
4826 Previously, $acl_verify_message was set only while expanding "message"
4827 and "log_message" when a very denied access.
4829 PH/28 Modified OS/os.c-Linux with
4831 -#ifndef OS_LOAD_AVERAGE
4832 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4834 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4835 nomenclature these days.)
4837 PH/29 Installed patch from the Sieve maintainer that adds the options
4838 sieve_useraddress and sieve_subaddress to the redirect router.
4840 PH/30 In these circumstances:
4841 . Two addresses routed to the same list of hosts;
4842 . First host does not offer TLS;
4843 . First host accepts first address;
4844 . First host gives temporary error to second address;
4845 . Second host offers TLS and a TLS session is established;
4846 . Second host accepts second address.
4847 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4848 and peerdn, if requested) that were in fact used only for the second
4851 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4852 attention to any local part prefix or suffix that was matched by the
4853 router that accepted the address. It now behaves in the same way as it
4854 does for delivery: the affixes are removed from the local part unless
4855 rcpt_include_affixes is set on the transport.
4857 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4858 timeout during the DATA phase of an incoming message.
4860 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4861 applied a patch sanctioned by the Sieve maintainer.
4863 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4864 the uid or gid is negative. A case of a negative gid caused this to be
4865 noticed. The fix allows for either to be negative.
4867 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4868 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4871 PH/36 The change PH/12 above was broken. Fixed it.
4873 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4874 the grounds that routing the same address twice would always produce the
4875 same answer. This might have been true once, but it is certainly no
4876 longer true now. Routing a child address may depend on the previous
4877 routing that produced that child. Some complicated redirection strategies
4878 went wrong when messages had multiple recipients, and made Exim's
4879 behaviour dependent on the order in which the addresses were given.
4881 I have moved the duplicate checking until after the routing is complete.
4882 Exim scans the addresses that are assigned to local and remote
4883 transports, and removes any duplicates. This means that more work will be
4884 done, as duplicates will always all be routed, but duplicates are
4885 presumably rare, so I don't expect this is of any significance.
4887 For deliveries to pipes, files, and autoreplies, the duplicate checking
4888 still happens during the routing process, since they are not going to be
4891 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4892 It corrects a timeout issue with spamd. This is Ian's comment: "The
4893 background is that sometimes spamd either never reads data from a
4894 connection it has accepted, or it never writes response data. The exiscan
4895 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4896 blindly assumes that writes won't block so it may never time out."
4898 PH/39 Allow G after quota size as well as K and M.
4900 PH/40 The value set for $authenticated_id in an authenticator may not contain
4901 binary zeroes or newlines because the value is written to log lines and
4902 to spool files. There was no check on this. Now the value is run through
4903 the string_printing() function so that such characters are converted to
4904 printable escape sequences.
4906 PH/41 $message_linecount is a new variable that contains the total number of
4907 lines in the message. Compare $body_linecount, which is the count for the
4910 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4911 bounce and delay warning messages, except in certain special cases, which
4914 (a) An SMTP error message from a remote host;
4915 (b) A message specified in a :fail: redirection;
4916 (c) A message specified in a "fail" command in a system filter;
4917 (d) A message specified in a FAIL return from the queryprogram router;
4918 (e) A message specified by the cannot_route_message router option.
4920 In these cases only, Exim does include the error details in bounce and
4921 warning messages. There are also a few cases where bland messages such
4922 as "unrouteable address" or "local delivery error" are given.
4924 PH/43 $value is now also set for the "else" part of a ${run expansion.
4926 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4927 being worked on, but at least Exim now implements the latest version to
4930 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4931 process to complete was treated as a delivery failure, a timeout while
4932 writing the message to the pipe was logged, but erroneously treated as a
4933 successful delivery. Such timeouts include transport filter timeouts. For
4934 consistency with the overall process timeout, these timeouts are now
4935 treated as errors, giving rise to delivery failures by default. However,
4936 there is now a new Boolean option for the pipe transport called
4937 timeout_defer, which, if set TRUE, converts the failures into defers for
4938 both kinds of timeout. A transport filter timeout is now identified in
4941 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4942 systems where "make" and "gmake" are different, calling "gmake" at top
4943 level broke things. I've arranged for the value of $(MAKE) to be passed
4944 from the Makefile to this script so that it can call the same version of
4948 A note about Exim versions 4.44 and 4.50
4949 ----------------------------------------
4951 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4952 changes of various kinds. As a consequence, a big documentation update was
4953 needed. This delayed the release for rather longer than seemed good, especially
4954 in the light of a couple of (minor) security issues. Therefore, the changes
4955 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4956 release. So 4.44 and 4.50 are in effect two different branches that both start
4959 I have left the 4.50 change log unchanged; it contains all the changes since
4960 4.43. The change log for 4.44 is below; many of its items are identical to
4961 those for 4.50. This seems to be the most sensible way to preserve the
4962 historical information.
4968 1. Minor wording change to the doc/README.SIEVE file.
4970 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4971 computation of the current number of files was incorrect.
4973 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4974 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4977 4. Give more explanation in the error message when the command for a transport
4978 filter fails to execute.
4980 5. There are several places where Exim runs a non-Exim command in a
4981 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4982 was being done only for the command run by the queryprogram router. It is
4983 now done for all such subprocesses. The other cases are: ${run, transport
4984 filters, and the commands run by the lmtp and pipe transports.
4986 6. Added CONFIGURE_GROUP build-time option.
4988 7. Some older OS have a limit of 256 on the maximum number of file
4989 descriptors. Exim was using setrlimit() to set 1000 as a large value
4990 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4991 systems. I've change it so that if it can't get 1000, it tries for 256.
4993 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4994 was an oversight, and furthermore, ever since the addition of extra
4995 controls (e.g. 4.43/32), the checks on when to allow different forms of
4996 "control" were broken. There should now be diagnostics for all cases when a
4997 control that does not make sense is encountered.
4999 9. Added the /retain_sender option to "control=submission".
5001 10. $recipients is now available in the predata ACL (oversight).
5003 11. Tidy the search cache before the fork to do a delivery from a message
5004 received from the command line. Otherwise the child will trigger a lookup
5005 failure and thereby defer the delivery if it tries to use (for example) a
5006 cached ldap connection that the parent has called unbind on.
5008 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5009 of $address_data from the recipient verification was clobbered by the
5010 sender verification.
5012 13. The value of address_data from a sender verification is now available in
5013 $sender_address_data in subsequent conditions in the ACL statement.
5015 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5017 15. Added a new option "connect=<time>" to callout options, to set a different
5020 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5021 was its contents. (It was OK if the option was not defined at all.)
5023 17. A "Completed" log line is now written for messages that are removed from
5024 the spool by the -Mrm option.
5026 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5027 information about exactly what failed.
5029 19. Added -dd to debug only the daemon process.
5031 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5032 handles timeouts, both on the server side and network timeouts. Renamed the
5033 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5035 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5036 It is now set to "smtps".
5038 22. $host_address is now set to the target address during the checking of
5039 ignore_target_hosts.
5041 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5042 being passed; this would have caused $sender_host_name to have been used if
5043 matching the list had actually called for a host name (not very likely,
5044 since this list is usually IP addresses). A host name is now passed as
5047 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5048 code that specifies a non-restarting handler (typically sigaction() in
5049 modern systems) in an attempt to fix a rare and obscure crash bug.
5051 25. Narrowed the window for a race in the daemon that could cause it to ignore
5052 SIGCHLD signals. This is not a major problem, because they are used only to
5053 wake it up if nothing else does.
5055 26. A malformed maildirsize file could cause Exim to calculate negative values
5056 for the mailbox size or file count. Odd effects could occur as a result.
5057 The maildirsize information is now recalculated if the size or filecount
5060 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5061 support for a long time. Removed HAVE_SYS_VFS_H.
5063 28. Installed the latest version of exipick from John Jetmore.
5065 29. In an address list, if the pattern was not a regular expression, an empty
5066 subject address (from a bounce message) matched only if the pattern was an
5067 empty string. Non-empty patterns were not even tested. This was the wrong
5068 because it is perfectly reasonable to use an empty address as part of a
5069 database query. An empty address is now tested by patterns that are
5070 lookups. However, all the other forms of pattern expect the subject to
5071 contain a local part and a domain, and therefore, for them, an empty
5072 address still always fails if the pattern is not itself empty.
5074 30. Exim went into a mad DNS loop when attempting to do a callout where the
5075 host was specified on an smtp transport, and looking it up yielded more
5076 than one IP address.
5078 31. Re-factored the code for checking spool and log partition space into a
5079 function that finds that data and another that does the check. The former
5080 is then used to implement four new variables: $spool_space, $log_space,
5081 $spool_inodes, and $log_inodes.
5083 32. The RFC2047 encoding function was originally intended for short strings
5084 such as real names; it was not keeping to the 75-character limit for
5085 encoded words that the RFC imposes. It now respects the limit, and
5086 generates multiple encoded words if necessary. To be on the safe side, I
5087 have increased the buffer size for the ${rfc2047: expansion operator from
5090 33. It is now permitted to omit both strings after an "if" condition; if the
5091 condition is true, the result is "true". As before, when the second string
5092 is omitted, a false condition yields an empty string. This makes it less
5093 cumbersome to write custom ACL and router conditions.
5095 34. Failure to deliver a bounce message always caused it to be frozen, even if
5096 there was an errors_to setting on the router. The errors_to setting is now
5099 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5100 canonical form (fully expanded) before being placed in
5101 $sender_host_address.
5103 36. The table in the code that translates DNS record types into text (T_A to
5104 "A" for instance) was missing entries for NS and CNAME. It is just possible
5105 that this could have caused confusion if both these types were looked up
5106 for the same domain, because the text type is used as part of Exim's
5107 per-process caching. But the chance of anyone hitting this buglet seems
5110 37. The dnsdb lookup has been extended in a number of ways.
5112 (1) There is a new type, "zns", which walks up the domain tree until it
5113 finds some nameserver records. It should be used with care.
5115 (2) There is a new type, "mxh", which is like "mx" except that it returns
5116 just the host names, not the priorities.
5118 (3) It is now possible to give a list of domains (or IP addresses) to be
5119 looked up. The behaviour when one of the lookups defers can be
5120 controlled by a keyword.
5122 (4) It is now possible to specify the separator character for use when
5123 multiple records are returned.
5125 38. The dnslists ACL condition has been extended: it is now possible to supply
5126 a list of IP addresses and/or domains to be looked up in a particular DNS
5129 39. Added log_selector=+queue_time_overall.
5131 40. When running the queue in the test harness, wait just a tad after forking a
5132 delivery process, to get repeatability of debugging output.
5134 41. Include certificate and key file names in error message when GnuTLS fails
5135 to set them up, because the GnuTLS error message doesn't include the name
5136 of the failing file when there is a problem reading it.
5138 42. Allow both -bf and -bF in the same test run.
5140 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5142 44. The "Exiscan patch" is now merged into the mainline Exim source.
5144 45. Sometimes the final signoff response after QUIT could fail to get
5145 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5146 < 0 before doing a fflush(). This bug looks as though it goes back to the
5147 introduction of TLS in release 3.20, but "sometimes" must have been rare
5148 because the tests only now provoked it.
5150 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5151 (this can affect the format of dates).
5153 47. exim_tidydb, when checking for the continued existence of a message for
5154 which it has found a message-specific retry record, was not finding
5155 messages that were in split spool directories. Consequently, it was
5156 deleting retry records that should have stayed in existence.
5158 48. Steve fixed some bugs in eximstats.
5160 49. The SPA authentication driver was not abandoning authentication and moving
5161 on to the next authenticator when an expansion was forced to fail,
5162 contradicting the general specification for all authenticators. Instead it
5163 was generating a temporary error. It now behaves as specified.
5165 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5166 (the order specifies the preference for clients). The order is now AES256,
5167 AES128, 3DES, ARCFOUR128.
5169 51. Small patch to Sieve code - explicitly set From: when generating an
5172 52. Exim crashed if a remote delivery caused a very long error message to be
5173 recorded - for instance if somebody sent an entire SpamAssassin report back
5174 as a large number of 550 error lines. This bug was coincidentally fixed by
5175 increasing the size of one of Exim's internal buffers (big_buffer) that
5176 happened as part of the Exiscan merge. However, to be on the safe side, I
5177 have made the code more robust (and fixed the comments that describe what
5180 53. Now that there can be additional text after "Completed" in log lines (if
5181 the queue_time_overall log selector is set), a one-byte patch to exigrep
5182 was needed to allow it to recognize "Completed" as not the last thing in
5185 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5186 patch that reportedly fixes this has been added. I am not expert enough to
5187 create a test for it. This is what the patch creator wrote:
5189 "I found a little strange behaviour of ldap code when working with
5190 Windows 2003 AD Domain, where users was placed in more than one
5191 Organization Units. When I tried to give exim partial DN, the exit code
5192 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5193 But simultaneously result of request was absolutely normal ldap result,
5194 so I produce this patch..."
5196 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5197 REFERENCE, so I have modified the code to exclude the patch when that macro
5200 55. Some experimental protocols are using DNS PTR records for new purposes. The
5201 keys for these records are domain names, not reversed IP addresses. The
5202 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5203 leaves it alone. Component reversal etc. now happens only for IP addresses.
5206 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5208 57. Double the size of the debug message buffer (to 2048) so that more of very
5209 long debug lines gets shown.
5211 58. The exicyclog utility now does better if the number of log files to keep
5212 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5214 59. Two changes related to the smtp_active_hostname option:
5216 (1) $smtp_active_hostname is now available as a variable.
5217 (2) The default for smtp_banner uses $smtp_active_hostname instead
5218 of $primary_hostname.
5220 60. The host_aton() function is supposed to be passed a string that is known
5221 to be a valid IP address. However, in the case of IPv6 addresses, it was
5222 not checking this. This is a hostage to fortune. Exim now panics and dies
5223 if the condition is not met. A case was found where this could be provoked
5224 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5225 components; fortuitously, this particular loophole had already been fixed
5226 by change 4.50/55 above.
5228 If there are any other similar loopholes, the new check in host_aton()
5229 itself should stop them being exploited. The report I received stated that
5230 data on the command line could provoke the exploit when Exim was running as
5231 exim, but did not say which command line option was involved. All I could
5232 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5233 running as the user.
5236 61. There was a buffer overflow vulnerability in the SPA authentication code
5237 (which came originally from the Samba project). I have added a test to the
5238 spa_base64_to_bits() function which I hope fixes it.
5241 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5242 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5244 63. The daemon start-up calls getloadavg() while still root for those OS that
5245 need the first call to be done as root, but it missed one case: when
5246 deliver_queue_load_max is set with deliver_drop_privilege. This is
5247 necessary for the benefit of the queue runner, because there is no re-exec
5248 when deliver_drop_privilege is set.
5250 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5251 This has been fixed.
5253 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5254 was in use, was not putting the data itself into the right store pool;
5255 consequently, it could be overwritten for a subsequent message in the same
5256 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5259 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5261 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5262 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5263 the function were treating the return as a boolean value, which happened to
5264 work because 0=false and not-0=true, but is not correct code.
5266 68. The host_aton() function was not handling scoped IPv6 addresses (those
5267 with, for example, "%eth0" on the end) correctly.
5269 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5270 negated items (that is, ~something) in unsigned ints. Some compilers
5271 apparently mutter when there is no cast.
5273 70. If an address verification called from an ACL failed, and did not produce a
5274 user-specific message (i.e. there was only a "system" message), nothing was
5275 put in $acl_verify_message. In this situation, it now puts the system
5278 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5279 session; change 4.31/43 added the unwanted input to the log line - except
5280 that it did not do this in the start of session case. It now does.
5282 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5283 This is wrong and can cause the other end to generate a synchronization
5284 error if it is another Exim or anything else that does the synchronization
5285 check. A QUIT command is no longer sent after a timeout.
5287 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5288 during host lookups.
5290 74. The defer_ok option of callout verification was not working if it was used
5291 when verifying addresses in header lines, that is, for this case:
5293 verify = header_sender/callout=defer_ok
5295 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5296 those file descriptors could be used for SMTP connections. If anything
5297 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5298 could be sent to the SMTP client, causing chaos. The daemon now opens
5299 stdin, stdout, and stderr to /dev/null when it puts itself into the
5302 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5303 log by default. The user can override this with suitable Perl magic.
5305 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5306 the log message when discard triggers, was not working for the DATA ACL or
5307 for the non-SMTP ACL.
5309 78. Error message wording change in sieve.c.
5311 79. If smtp_accept_max_per_host was set, the number of connections could be
5312 restricted to fewer than expected, because the daemon was trying to set up
5313 a new connection before checking whether the processes handling previous
5314 connections had finished. The check for completed processes is now done
5315 earlier. On busy systems, this bug wouldn't be noticed because something
5316 else would have woken the daemon, and it would have reaped the completed
5319 80. If a message was submitted locally by a user whose login name contained one
5320 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5321 It caused a spool format error. I have fixed the spool reading code. A
5322 related problem was that the "from" clause in the Received: line became
5323 illegal because of the space(s). It is now covered by ${quote_local_part.
5325 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5328 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5330 83. Patch from Sieve maintainer to fix unterminated string problem in
5331 "vacation" handling.
5333 84. Some minor changes to the Linux configuration files to help with other
5334 OS variants using glibc.
5336 85. One more patch for Sieve to update vacation handling to latest spec.
5339 ----------------------------------------------------
5340 See the note above about the 4.44 and 4.50 releases.
5341 ----------------------------------------------------
5347 1. Change 4.43/35 introduced a bug that caused file counts to be
5348 incorrectly computed when quota_filecount was set in an appendfile
5351 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5352 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5355 3. Give more explanation in the error message when the command for a transport
5356 filter fails to execute.
5358 4. There are several places where Exim runs a non-Exim command in a
5359 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5360 was being done only for the command run by the queryprogram router. It is
5361 now done for all such subprocesses. The other cases are: ${run, transport
5362 filters, and the commands run by the lmtp and pipe transports.
5364 5. Some older OS have a limit of 256 on the maximum number of file
5365 descriptors. Exim was using setrlimit() to set 1000 as a large value
5366 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5367 systems. I've change it so that if it can't get 1000, it tries for 256.
5369 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5370 was an oversight, and furthermore, ever since the addition of extra
5371 controls (e.g. 4.43/32), the checks on when to allow different forms of
5372 "control" were broken. There should now be diagnostics for all cases when a
5373 control that does not make sense is encountered.
5375 7. $recipients is now available in the predata ACL (oversight).
5377 8. Tidy the search cache before the fork to do a delivery from a message
5378 received from the command line. Otherwise the child will trigger a lookup
5379 failure and thereby defer the delivery if it tries to use (for example) a
5380 cached ldap connection that the parent has called unbind on.
5382 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5383 of $address_data from the recipient verification was clobbered by the
5384 sender verification.
5386 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5387 was its contents. (It was OK if the option was not defined at all.)
5389 11. A "Completed" log line is now written for messages that are removed from
5390 the spool by the -Mrm option.
5392 12. $host_address is now set to the target address during the checking of
5393 ignore_target_hosts.
5395 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5396 being passed; this would have caused $sender_host_name to have been used if
5397 matching the list had actually called for a host name (not very likely,
5398 since this list is usually IP addresses). A host name is now passed as
5401 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5402 code that specifies a non-restarting handler (typically sigaction() in
5403 modern systems) in an attempt to fix a rare and obscure crash bug.
5405 15. Narrowed the window for a race in the daemon that could cause it to ignore
5406 SIGCHLD signals. This is not a major problem, because they are used only to
5407 wake it up if nothing else does.
5409 16. A malformed maildirsize file could cause Exim to calculate negative values
5410 for the mailbox size or file count. Odd effects could occur as a result.
5411 The maildirsize information is now recalculated if the size or filecount
5414 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5415 support for a long time. Removed HAVE_SYS_VFS_H.
5417 18. Updated exipick to current release from John Jetmore.
5419 19. Allow an empty sender to be matched against a lookup in an address list.
5420 Previously the only cases considered were a regular expression, or an
5423 20. Exim went into a mad DNS lookup loop when doing a callout where the
5424 host was specified on the transport, if the DNS lookup yielded more than
5427 21. The RFC2047 encoding function was originally intended for short strings
5428 such as real names; it was not keeping to the 75-character limit for
5429 encoded words that the RFC imposes. It now respects the limit, and
5430 generates multiple encoded words if necessary. To be on the safe side, I
5431 have increased the buffer size for the ${rfc2047: expansion operator from
5434 22. Failure to deliver a bounce message always caused it to be frozen, even if
5435 there was an errors_to setting on the router. The errors_to setting is now
5438 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5439 canonical form (fully expanded) before being placed in
5440 $sender_host_address.
5442 24. Updated eximstats to version 1.33
5444 25. Include certificate and key file names in error message when GnuTLS fails
5445 to set them up, because the GnuTLS error message doesn't include the name
5446 of the failing file when there is a problem reading it.
5448 26. Expand error message when OpenSSL has problems setting up cert/key files.
5451 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5452 (this can affect the format of dates).
5454 28. exim_tidydb, when checking for the continued existence of a message for
5455 which it has found a message-specific retry record, was not finding
5456 messages that were in split spool directories. Consequently, it was
5457 deleting retry records that should have stayed in existence.
5459 29. eximstats updated to version 1.35
5460 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5461 1.35 - bugfix such that pie charts by volume are generated correctly
5463 30. The SPA authentication driver was not abandoning authentication and moving
5464 on to the next authenticator when an expansion was forced to fail,
5465 contradicting the general specification for all authenticators. Instead it
5466 was generating a temporary error. It now behaves as specified.
5468 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5469 (the order specifies the preference for clients). The order is now AES256,
5470 AES128, 3DES, ARCFOUR128.
5472 31. Small patch to Sieve code - explicitly set From: when generating an
5475 32. Exim crashed if a remote delivery caused a very long error message to be
5476 recorded - for instance if somebody sent an entire SpamAssassin report back
5477 as a large number of 550 error lines. This bug was coincidentally fixed by
5478 increasing the size of one of Exim's internal buffers (big_buffer) that
5479 happened as part of the Exiscan merge. However, to be on the safe side, I
5480 have made the code more robust (and fixed the comments that describe what
5483 33. Some experimental protocols are using DNS PTR records for new purposes. The
5484 keys for these records are domain names, not reversed IP addresses. The
5485 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5486 leaves it alone. Component reversal etc. now happens only for IP addresses.
5489 34. The host_aton() function is supposed to be passed a string that is known
5490 to be a valid IP address. However, in the case of IPv6 addresses, it was
5491 not checking this. This is a hostage to fortune. Exim now panics and dies
5492 if the condition is not met. A case was found where this could be provoked
5493 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5494 components; fortuitously, this particular loophole had already been fixed
5495 by change 4.50/55 or 4.44/33 above.
5497 If there are any other similar loopholes, the new check in host_aton()
5498 itself should stop them being exploited. The report I received stated that
5499 data on the command line could provoke the exploit when Exim was running as
5500 exim, but did not say which command line option was involved. All I could
5501 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5502 running as the user.
5505 35. There was a buffer overflow vulnerability in the SPA authentication code
5506 (which came originally from the Samba project). I have added a test to the
5507 spa_base64_to_bits() function which I hope fixes it.
5510 36. The daemon start-up calls getloadavg() while still root for those OS that
5511 need the first call to be done as root, but it missed one case: when
5512 deliver_queue_load_max is set with deliver_drop_privilege. This is
5513 necessary for the benefit of the queue runner, because there is no re-exec
5514 when deliver_drop_privilege is set.
5516 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5517 was in use, was not putting the data itself into the right store pool;
5518 consequently, it could be overwritten for a subsequent message in the same
5519 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5522 38. Sometimes the final signoff response after QUIT could fail to get
5523 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5524 < 0 before doing a fflush(). This bug looks as though it goes back to the
5525 introduction of TLS in release 3.20, but "sometimes" must have been rare
5526 because the tests only now provoked it.
5532 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5533 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5534 Now it returns an integer. A number of calls were still expecting a T/F
5535 return. Fortuitously, in all cases, the tests worked in OK situations,
5536 which is the norm. However, things would have gone wrong on any write
5537 failures on the smtp file descriptor. This function is used when sending
5538 messages over SMTP and also when doing verify callouts.
5540 2. When Exim is called to do synchronous delivery of a locally submitted
5541 message (the -odf or -odi options), it no longer closes stderr before doing
5544 3. Implemented the mua_wrapper option.
5546 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5548 5. Implemented the functions header_remove(), header_testname(),
5549 header_add_at_position(), and receive_remove_recipient(), and exported them
5552 6. If an ACL "warn" statement specified the addition of headers, Exim already
5553 inserted X-ACL-Warn: at the start if there was no header name. However, it
5554 was not making this test for the second and subsequent header lines if
5555 there were newlines in the string. This meant that an invalid header could
5556 be inserted if Exim was badly configured.
5558 7. Allow an ACL "warn" statement to add header lines at the start or after all
5559 the Received: headers, as well as at the end.
5561 8. Added the rcpt_4xx retry error code.
5563 9. Added postmaster_mailfrom=xxx to callout verification option.
5565 10. Added mailfrom=xxxx to the callout verification option, for verify=
5568 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5569 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5571 12. Inserted some casts to stop certain compilers warning when using pointer
5572 differences as field lengths or precisions in printf-type calls (mostly
5573 affecting debugging statements).
5575 13. Added optional readline() support for -be (dynamically loaded).
5577 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5578 same clock tick as a message's arrival, so that its received time was the
5579 same as the "first fail" time on the retry record, and that message
5580 remained on the queue past the ultimate address timeout, every queue runner
5581 would try a delivery (because it was past the ultimate address timeout) but
5582 after another failure, the ultimate address timeout, which should have then
5583 bounced the address, did not kick in. This was a "< instead of <=" error;
5584 in most cases the first failure would have been in the next clock tick
5585 after the received time, and all would be well.
5587 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5588 being recognized when the domain list was tested by the match_domain
5589 condition in an expansion string.
5591 16. Added the ${str2b64: operator.
5593 17. Exim was always calling setrlimit() to set a large limit for the number of
5594 processes, without checking whether the existing limit was already
5595 adequate. (It did check for the limit on file descriptors.) Furthermore,
5596 errors from getrlimit() and setrlimit() were being ignored. Now they are
5597 logged to the main and panic logs, but Exim does carry on, to try to do its
5598 job under whatever limits there are.
5600 18. Imported PCRE 5.0.
5602 19. Trivial typo in log message " temporarily refused connection" (the leading
5605 20. If the log selector return_path_on_delivery was set and an address was
5606 redirected to /dev/null, the delivery process crashed because it assumed
5607 that a return path would always be set for a "successful" delivery. In this
5608 case, the whole delivery is bypassed as an optimization, and therefore no
5611 21. Internal re-arrangement: the function for sending a challenge and reading
5612 a response while authentication was assuming a zero-terminated challenge
5613 string. It's now changed to take a pointer and a length, to allow for
5614 binary data in such strings.
5616 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5618 23. Exim was not respecting finduser_retries when seeking the login of the
5619 uid under which it was called; it was always trying 10 times. (The default
5620 setting of finduser_retries is zero.) Also, it was sleeping after the final
5621 failure, which is pointless.
5623 24. Implemented tls_on_connect_ports.
5625 25. Implemented acl_smtp_predata.
5627 26. If the domain in control=submission is set empty, Exim assumes that the
5628 authenticated id is a complete email address when it generates From: or
5629 Sender: header lines.
5631 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5632 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5633 chown and chgrp in /bin and hostname in /usr/bin.
5635 28. Exim was keeping the "process log" file open after each use, just as it
5636 does for the main log. This opens the possibility of it remaining open for
5637 long periods when the USR1 signal hits a daemon. Occasional processlog
5638 errors were reported, that could have been caused by this. Anyway, it seems
5639 much more sensible not to leave this file open at all, so that is what now
5642 29. The long-running daemon process does not normally write to the log once it
5643 has entered its main loop, and it closes the log before doing so. This is
5644 so that log files can straightforwardly be renamed and moved. However,
5645 there are a couple of unusual error situations where the daemon does write
5646 log entries, and I had neglected to close the log afterwards.
5648 30. The text of an SMTP error response that was received during a remote
5649 delivery was being truncated at 512 bytes. This is too short for some of
5650 the long messages that one sometimes sees. I've increased the limit to
5653 31. It is now possible to make retry rules that apply only when a message has a
5654 specific sender, in particular, an empty sender.
5656 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5657 it possible to be selective about when SMTP synchronization is enforced.
5659 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5661 32. Implemented hosts_connection_nolog.
5663 33. Added an ACL for QUIT.
5665 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5668 35. Added mailbox_size and mailbox_filecount to appendfile.
5670 36. Added control = no_multiline_responses to ACLs.
5672 37. There was a bug in the logic of the code that waits for the clock to tick
5673 in the case where the clock went backwards by a substantial amount such
5674 that the microsecond fraction of "now" was more than the microsecond
5675 fraction of "then" (but the whole seconds number was less).
5677 38. Added support for the libradius Radius client library this is found on
5678 FreeBSD (previously only the radiusclient library was supported).
5684 1. When certain lookups returned multiple values in the form name=value, the
5685 quoting of the values was not always being done properly. Specifically:
5686 (a) If the value started with a double quote, but contained no whitespace,
5688 (b) If the value contained whitespace other than a space character (i.e.
5689 tabs or newlines or carriage returns) it was not quoted.
5690 This fix has been applied to the mysql and pgsql lookups by writing a
5691 separate quoting function and calling it from the lookup code. The fix
5692 should probably also be applied to nisplus, ibase and oracle lookups, but
5693 since I cannot test any of those, I have not disturbed their existing code.
5695 2. A hit in the callout cache for a specific address caused a log line with no
5696 reason for rejecting RCPT. Now it says "Previous (cached) callout
5697 verification failure".
5699 3. There was an off-by-one bug in the queryprogram router. An over-long
5700 return line was truncated at 256 instead of 255 characters, thereby
5701 overflowing its buffer with the terminating zero. As well as fixing this, I
5702 have increased the buffer size to 1024 (and made a note to document this).
5704 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5705 when Exim is waiting for an SMTP response from a remote server, Exim
5706 restarts its select() call on the socket, thereby resetting its timeout.
5707 This is not a problem when such interrupts are rare. Somebody set up a cron
5708 job to run exiwhat every 2 minutes, which is less than the normal select()
5709 timeout (5 or 10 minutes). This meant that the select() timeout never
5710 kicked in because it was always reset. I have fixed this by comparing the
5711 time when an interrupt arrives with the time at the start of the first call
5712 to select(). If more time than the timeout has elapsed, the interrupt is
5713 treated as a timeout.
5715 5. Some internal re-factoring in preparation for the addition of Sieve
5716 extensions (by MH). In particular, the "personal" test is moved to a
5717 separate function, and given an option for scanning Cc: and Bcc: (which is
5718 not set for Exim filters).
5720 6. When Exim created an email address using the login of the caller as the
5721 local part (e.g. when creating a From: or Sender: header line), it was not
5722 quoting the local part when it contained special characters such as @.
5724 7. Installed new OpenBSD configuration files.
5726 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5727 try to make them clearer.
5729 9. Callout options, other than the timeout value, were being ignored when
5730 verifying sender addresses in header lines. For example, when using
5732 verify = header_sender/callout=no_cache
5734 the cache was (incorrectly) being used.
5736 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5737 only the Cygwin environment.
5739 11. When return_path_on_delivery was set as a log selector, if different remote
5740 addresses in the same message used different return paths and parallel
5741 remote delivery occurred, the wrong values would sometimes be logged.
5742 (Whenever a remote delivery process finished, the return path value from
5743 the most recently started remote delivery process was logged.)
5745 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5746 lines when AUTH and/or TLS are in use. This is the "received protocol"
5747 field. Exim used to use "asmtp" for authenticated SMTP, without any
5748 indication (in the protocol name) for TLS use. Now it follows the RFC and
5749 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5750 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5751 names appear in log lines as well as in Received: header lines.
5753 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5754 extensions, and comparison tests, and to fix some bugs.
5756 14. Changes to the "personal" filter test:
5758 (1) The test was buggy in that it was just doing the equivalent of
5759 "contains" tests on header lines. For example, if a user's address was
5760 anne@some.where, the "personal" test would incorrectly be true for
5762 To: susanne@some.where
5764 This test is now done by extracting each address from the header in turn,
5765 and checking the entire address. Other tests that are part of "personal"
5766 are now done using regular expressions (for example, to check local parts
5767 of addresses in From: header lines).
5769 (2) The list of non-personal local parts in From: addresses has been
5770 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5771 taken from the Sieve specification recommendations.
5773 (3) If the message contains any header line starting with "List-" it is
5774 treated as non-personal.
5776 (4) The test for "circular" in the Subject: header line has been removed
5777 because it now seems ill-conceived.
5779 15. Minor typos in src/EDITME comments corrected.
5781 16. Installed latest exipick from John Jetmore.
5783 17. If headers_add on a router specified a text string that was too long for
5784 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5785 of string_sprintf() is now avoided.
5787 18. $message_body_size was not set (it was always zero) when running the DATA
5788 ACL and the local_scan() function.
5790 19. For the "mail" command in an Exim filter, no default was being set for
5791 the once_repeat time, causing a random time value to be used if "once" was
5792 specified. (If the value happened to be <= 0, no repeat happened.) The
5793 default is now 0s, meaning "never repeat". The "vacation" command was OK
5794 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5795 (I found it when inspecting the code).
5797 20. There is now an overall timeout for performing a callout verification. It
5798 defaults to 4 times the callout timeout, which applies to individual SMTP
5799 commands during the callout. The overall timeout applies when there is more
5800 than one host that can be tried. The timeout is checked before trying the
5801 next host. This prevents very long delays if there are a large number of
5802 hosts and all are timing out (e.g. when the network connections are timing
5803 out). The value of the overall timeout can be changed by specifying an
5804 additional sub-option for "callout", called "maxwait". For example:
5806 verify = sender/callout=5s,maxwait=20s
5808 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5809 to the end before writing, but this should make it even safer).
5811 22. Exim was forgetting that it had advertised PIPELINING for the second and
5812 subsequent messages on an SMTP connection. It was also not resetting its
5813 memory on STARTTLS and an internal HELO.
5815 23. When Exim logs an SMTP synchronization error within a session, it now
5816 records whether PIPELINING has been advertised or not.
5818 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5819 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5820 rather than long int.
5822 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5824 26. Added the never_mail option to autoreply.
5830 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5831 crash if the getsockname() call failed; this can happen if a connection is
5832 closed very soon after it is established. The problem was simply in the
5833 order in which certain operations were done, causing Exim to try to write
5834 to the SMTP stream before it had set up the file descriptor. The bug has
5835 been fixed by making things happen in the correct order.
5841 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5842 before the connection was closed, thus losing the rejection response.
5844 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5845 some early Solaris releases, but causes trouble in current releases where
5846 socklen_t is defined.
5848 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5851 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5854 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5855 that releasing the top of it at the end releases what was used for sub-
5856 expansions (unless the block got too big). However, discard this block if
5857 the first thing is a variable or header, so that we can use its block when
5858 it is dynamic (useful for very large $message_headers, for example).
5860 6. Lookups now cache *every* query, not just the most recent. A new, separate
5861 store pool is used for this. It can be recovered when all lookup caches are
5862 flushed. Lookups now release memory at the end of their result strings.
5863 This has involved some general refactoring of the lookup sources.
5865 7. Some code has been added to the store_xxx() functions to reduce the amount
5866 of flapping under certain conditions.
5868 8. log_incoming_interface used to affect only the <= reception log lines. Now
5869 it causes the local interface and port to be added to several more SMTP log
5870 lines, for example "SMTP connection from", and rejection lines.
5872 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5874 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5876 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5877 could be overwritten at the end of the current message (or the start of a
5878 new message if it was set in a HELO ACL). The value is now preserved for
5879 the duration of the SMTP connection.
5881 12. If a transport had a headers_rewrite setting, and a matching header line
5882 contained an unqualified address, that address was qualified, even if it
5883 did not match any rewriting rules. The underlying bug was that the values
5884 of the flags that permit the existence of unqualified sender and recipient
5885 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5886 non-local messages, and by -bnq for local messages) were not being
5887 preserved with the message after it was received.
5889 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5890 "next input=" as part of the text comprising the host identity instead of
5891 the correct text. The code was using the same buffer for two different
5892 strings. However, depending on which order the printing function evaluated
5893 its arguments, the bug did not always show up. Under Linux, for example, my
5894 test suite worked just fine.
5896 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5897 doesn't work with some older versions of Perl. It has been changed to "my",
5898 which in any case is probably the better facility to use.
5900 15. A really picky compiler found some instances of statements for creating
5901 error messages that either had too many or two few arguments for the format
5904 16. The size of the buffer for calls to the DNS resolver has been increased
5905 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5906 for addresses that have a lot of PTR records. This alleviates a problem; it
5907 does not fully solve it.
5909 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5910 buffer now truncates the list and logs the incident, which is the same
5911 action as happens when Exim is looking up a host name and its aliases.
5912 Previously in this situation something unpredictable would happen;
5913 sometimes it was "internal error: store_reset failed".
5915 18. If a server dropped the connection unexpectedly when an Exim client was
5916 using GnuTLS and trying to read a response, the client delivery process
5917 crashed while trying to generate an error log message.
5919 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5920 string, for example:
5922 warn message = H1: something\nH2: something
5924 the text was added as a single header line from Exim's point of view
5925 though it ended up OK in the delivered message. However, searching for the
5926 second and subsequent header lines using $h_h2: did not work. This has been
5927 fixed. Similarly, if a system filter added multiple headers in this way,
5928 the routers could not see them.
5930 20. Expanded the error message when iplsearch is called with an invalid key to
5931 suggest using net-iplsearch in a host list.
5933 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5934 ACLs are no longer actually imposed (and a message to that effect is
5937 22. If a "gecos" field in a passwd entry contained escaped characters, in
5938 particular, if it contained a \" sequence, Exim got it wrong when building
5939 a From: or a Sender: header from that name. A second bug also caused
5940 incorrect handling when an unquoted " was present following a character
5941 that needed quoting.
5943 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5944 was not being matched caselessly.
5946 24. Arranged for all hyphens in the exim.8 source to be escaped with
5949 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5950 properly. Recipient callout cache records were still being keyed to include
5951 the sender, even when use_sender was set false. This led to far more
5952 callouts that were necessary. The sender is no longer included in the key
5953 when use_sender is false.
5955 26. Added "control = submission" modifier to ACLs.
5957 27. Added the ${base62d: operator to decode base 62 numbers.
5959 28. dnsdb lookups can now access SRV records.
5961 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5962 the configuration file.
5964 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5965 (-v) mode. This makes the output for a verbose queue run more intelligible.
5967 31. Added a use_postmaster feature to recipient callouts.
5969 32. Added the $body_zerocount variable, containing the number of binary zero
5970 bytes in the message body.
5972 33. The time of last modification of the "new" subdirectory is now used as the
5973 "mailbox time last read" when there is a quota error for a maildir
5976 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5978 35. Added +ignore_unknown as a special item in host lists.
5980 36. Code for decoding IPv6 addresses in host lists is now included, even if
5981 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5982 address was recognized as an IP address, but was then not correctly decoded
5983 into binary, causing unexpected and incorrect effects when compared with
5990 1. Very minor rewording of debugging text in manualroute to say "list of
5991 hosts" instead of "hostlist".
5993 2. If verify=header_syntax was set, and a header line with an unqualified
5994 address (no domain) and a large number of spaces between the end of the
5995 name and the colon was received, the reception process suffered a buffer
5996 overflow, and (when I tested it) crashed. This was caused by some obsolete
5997 code that should have been removed. The fix is to remove it!
5999 3. When running in the test harness, delay a bit after writing a bounce
6000 message to get a bit more predictability in the log output.
6002 4. Added a call to search_tidyup() just before forking a reception process. In
6003 theory, someone could use a lookup in the expansion of smtp_accept_max_
6004 per_host which, without the tidyup, could leave open a database connection.
6006 5. Added the variables $recipient_data and $sender_data which get set from a
6007 lookup success in an ACL "recipients" or "senders" condition, or a router
6008 "senders" option, similar to $domain_data and $local_part_data.
6010 6. Moved the writing of debug_print from before to after the "senders" test
6013 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6014 problems for message scanning, either using a data ACL, or using
6015 local_scan() because the Received: header was not generated till after they
6016 were called (in order to set the time as the time of reception completion).
6017 I have revised the way this works. The header is now generated after the
6018 body is received, but before the ACL or local_scan() are called. After they
6019 are run, the timestamp in the header is updated.
6025 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6026 before starting a queue runner without re-exec. This happened only when
6027 deliver_drop_privilege was set or when the Exim user was set to root. The
6028 effect of the bug was that timeouts during subsequent deliveries caused
6029 crashes instead of being properly handled. The handler is now left at its
6030 default (and expected) setting.
6032 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6033 message, again when deliver_drop_privilege is set or Exim is run as root.
6034 The bug described in (1) was not present in this case, but the tidying up
6035 of the other signals was missing. I have made the two cases consistent.
6037 3. The ignore_target_hosts setting on a manualroute router was being ignored
6038 for hosts that were looked up using the /MX notation.
6040 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6043 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6044 operated on the sender address. After changing the $sender_address to <>
6045 for the sender address verify, Exim was re-instated it as the original
6046 (before rewriting) address, but remembering that it had rewritten it, so it
6047 wasn't rewriting it again. This bug also had the effect of breaking the
6048 sender address verification caching when the sender address was rewritten.
6050 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6051 This has been changed so that if the ip literal address matches
6052 ignore_target_hosts, the router declines.
6054 7. Added expansion conditions match_domain, match_address, and match_local_
6055 part (NOT match_host).
6057 8. The placeholder for the Received: header didn't have a length field set.
6059 9. Added code to Exim itself and to exim_lock to test for a specific race
6060 condition that could lead to file corruption when using MBX delivery. The
6061 issue is with the lockfile that is created in /tmp. If this file is removed
6062 after a process has opened it but before that process has acquired a lock,
6063 there is the potential for a second process to recreate the file and also
6064 acquire a lock. This could lead to two Exim processes writing to the file
6065 at the same time. The added code performs the same test as UW imapd; it
6066 checks after acquiring the lock that its file descriptor still refers to
6067 the same named file.
6069 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6070 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6073 11. Added the smtp_active_hostname option. If used, this will typically be made
6074 to depend on the incoming interface address. Because $interface_address is
6075 not set up until the daemon has forked a reception process, error responses
6076 that can happen earlier (such as "too many connections") no longer contain
6079 12. If an expansion in a condition on a "warn" statement fails because a lookup
6080 defers, the "warn" statement is abandoned, and the next ACL statement is
6081 processed. Previously this caused the whole ACL to be aborted.
6083 13. Added the iplsearch lookup type.
6085 14. Added ident_timeout as a log selector.
6087 15. Added tls_certificate_verified as a log selector.
6089 16. Added a global option tls_require_ciphers (compare the smtp transport
6090 option of the same name). This controls incoming TLS connections.
6092 17. I finally figured out how to make tls_require_ciphers do a similar thing
6093 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6094 before starting the TLS session.
6096 18. Tabs are now shown as \t in -bP output.
6098 19. If the log selector return_path_on_delivery was set, Exim crashed when
6099 bouncing a message because it had too many Received: header lines.
6101 20. If two routers both had headers_remove settings, and the first one included
6102 a superfluous trailing colon, the final name in the first list and the
6103 first name in the second list were incorrectly joined into one item (with a
6104 colon in the middle).
6110 1. Added -C and -D options to the exinext utility, mainly to make it easier
6111 to include in the automated testing, but these could be helpful when
6112 multiple configurations are in use.
6114 2. The exinext utility was not formatting the output nicely when there was
6115 an alternate port involved in the retry record key, nor when there was a
6116 message id as well (for retries that were specific to a specific message
6117 and a specific host). It was also confused by IPv6 addresses, because of
6118 the additional colons they contain. I have fixed the IPv4 problem, and
6119 patched it up to do a reasonable job for IPv6.
6121 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6122 delivery, the log line now contains "pipelined" if PIPELINING was used.
6124 4. An SMTP transport process used to panic and die if the bind() call to set
6125 an explicit outgoing interface failed. This has been changed; it is now
6126 treated in the same way as a connect() failure.
6128 5. A reference to $sender_host_name in the part of a conditional expansion
6129 that was being skipped was still causing a DNS lookup. This no longer
6132 6. The def: expansion condition was not recognizing references to header lines
6133 that used bh_ and bheader_.
6135 7. Added the _cache feature to named lists.
6137 8. The code for checking quota_filecount in the appendfile transport was
6138 allowing one more file than it should have been.
6140 9. For compatibility with Sendmail, the command line option
6148 and sets the incoming protocol and host name (for trusted callers). The
6149 host name and its colon can be omitted when only the protocol is to be set.
6150 Note the Exim already has two private options, -pd and -ps, that refer to
6151 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6152 "s", but I don't think that's a major issue.
6154 10. A number of refactoring changes to the code, none of which should affect
6157 (a) The number of logging options was getting close to filling up the
6158 32-bit word that was used as a bit map. I have split them into two classes:
6159 those that are passed in the argument to log_write(), and those that are
6160 only ever tested independently outside of that function. These are now in
6161 separate 32-bit words, so there is plenty of room for expansion again.
6162 There is no change in the user interface or the logging behaviour.
6164 (b) When building, for example, log lines, the code previously used a
6165 macro that called string_cat() twice, in order to add two strings. This is
6166 not really sufficiently general. Furthermore, there was one instance where
6167 it was actually wrong because one of the argument was used twice, and in
6168 one call a function was used. (As it happened, calling the function twice
6169 did not affect the overall behaviour.) The macro has been replaced by a
6170 function that can join an arbitrary number of extra strings onto a growing
6173 (c) The code for expansion conditions now uses a table and a binary chop
6174 instead of a serial search (which was left over from when there were very
6175 few conditions). Also, it now recognizes conditions like "pam" even when
6176 the relevant support is not compiled in: a suitably worded error message is
6177 given if an attempt is made to use such a condition.
6179 11. Added ${time_interval:xxxxx}.
6181 12. A bug was causing one of the ddress fields not to be passed back correctly
6182 from remote delivery subprocesses. The field in question was not being
6183 subsequently used, so this caused to problems in practice.
6185 13. Added new log selectors queue_time and deliver_time.
6187 14. Might have fixed a bug in maildirsizefile handling that threw up
6188 "unexpected character" debug warnings, and recalculated the data
6189 unnecessarily. In any case, I expanded the warning message to give more
6192 15. Added the message "Restricted characters in address" to the statements in
6193 the default ACL that block characters like @ and % in local parts.
6195 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6196 Three changes have been made:
6198 (a) There was a serious bug; a negative response to MAIL caused the whole
6199 recipient domain to be cached as invalid, thereby blocking all messages
6200 to all local parts at the same domain, from all senders. This bug has
6201 been fixed. The domain is no longer cached after a negative response to
6202 MAIL if the sender used is not empty.
6204 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6207 (c) A new callout option, "use_sender" has been added for people who want
6208 the modified behaviour.
6214 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6217 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6218 indeed breaks things for older releases.
6220 3. Added additional logging to the case where there is a problem reading data
6221 from a filter that is running in a subprocess using a pipe, in order to
6222 try to track down a specific problem.
6224 4. Testing facility fudge: when running in the test harness and attempting
6225 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6226 getting "No route to host". Convert this to a timeout.
6228 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6231 6. Some OS don't have socklen_t but use size_t instead. This affects the
6232 fifth argument of getsockopt() amongst other things. This is now
6233 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6234 can be set for individual OS. I have set it for SunOS5, OSF1, and
6235 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6236 some earlier ones do not.
6238 7. Change 4.30/15 was not doing the test caselessly.
6240 8. The standard form for an IPv6 address literal was being rejected by address
6241 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6242 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6243 this, as well as the form without the "IPv6" on the front (but only when
6244 address literals are enabled, of course).
6246 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6248 10. Exim crashed if a message with an empty sender address specified by -f
6249 encountered a router with an errors_to setting. This could be provoked only
6250 by a command such as
6254 where an empty string was supplied; "<>" did not hit this bug.
6256 11. Installed PCRE release 4.5.
6258 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6259 remained set. It is now erased.
6261 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6262 times from message ids (which are base 36 rather than the normal 62).
6264 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6265 were being counted as actual protocol errors, and logged if the log
6266 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6267 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6268 and DATA following a set of rejected RCPTs do not count as protocol errors.
6269 In other words, Exim assumes they were pipelined, though this may not
6270 actually be the case. Of course, in all cases the client gets an
6271 appropriate error code.
6273 15. If a lookup fails in an ACL condition, a message about the failure may
6274 be available; it is used if testing the ACL cannot continue, because most
6275 such messages specify what the cause of the deferral is. However, some
6276 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6277 that caused an old message to be retained and used if a later statement
6278 caused a defer, replacing the real cause of the deferral.
6280 16. If an IP address had so many PTR records that the DNS lookup buffer
6281 was not large enough to hold them, Exim could crash while trying to process
6282 the truncated data. It now detects and logs this case.
6284 17. Further to 4.21/58, another change has been made: if (and only if) the
6285 first line of a message (the first header line) ends with CRLF, a bare LF
6286 in a subsequent header line has a space inserted after it, so as not to
6287 terminate the header.
6289 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6290 unnecessarily, used atoi() instead of strtol(), and didn't check the
6291 termination when getting file sizes from file names by regex.
6293 19. Completely re-implemented the support for maildirsize files, in the light
6294 of a number of problems with the previous contributed implementation
6295 (4.30/29). In particular:
6297 . If the quota is zero, the maildirsize file is maintained, but no quota is
6300 . If the maildir directory does not exist, it is created before any attempt
6301 to write a maildirsize file.
6303 . The quota value in the file is just a cache; if the quota is changed in
6304 the transport, the new value overrides.
6306 . A regular expression is available for excluding directories from the
6309 20. The autoreply transport checks the characters in options that define the
6310 message's headers; it allows continued headers, but it was checking with
6311 isspace() after an embedded newline instead of explicitly looking for a
6314 21. If all the "regular" hosts to which an address was routed had passed their
6315 expiry times, and had not reached their retry times, the address was
6316 bounced, even if fallback hosts were defined. Now Exim should go on to try
6319 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6320 equivalent code in the SMTP transport. Some hosts send humungous responses
6321 to HELO/EHLO, more than 1024 it seems.
6323 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6324 gives compiler warnings in some environments. I've now done it "properly",
6327 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6328 (because of problems with the built-in one) was declared to return uschar *
6329 instead of char *, causing compiler failure.
6331 25. Fixed a file descriptor leak when processing alias/forward files.
6333 26. Fixed a minor format string issue in dbfn.c.
6335 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6337 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6338 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6341 29. When a sender address is verified, it is cached, to save repeating the test
6342 when there is more than one recipient in a message. However, when the
6343 verification involves a callout, it is possible for different callout
6344 options to be set for different recipients. It is too complicated to keep
6345 track of this in the cache, so now Exim always runs a verification when a
6346 callout is required, relying on the callout cache for the optimization.
6347 The overhead is duplication of the address routing, but this should not be
6350 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6351 to be verified with callout=postmaster, and the main callout worked but the
6352 postmaster check failed, the verification correctly failed. However, if a
6353 subsequent RCPT command asked for sender verification *without* the
6354 postmaster check, incorrect caching caused this verification also to fail,
6357 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6358 it was not caching the DNS options (qualify_single, search_parents) that
6359 were used when the lookup failed. A subsequent lookup with different
6360 options therefore always gave the same answer, though there were cases
6361 where it should not have. (Example: a "domains = !$mx_any" option on a
6362 dnslookup router: the "domains" option is always processed without any
6363 widening, but the router might have qualify_single set.) Now Exim uses the
6364 cached value only when the same options are set.
6366 32. Added John Jetmore's "exipick" utility to the distribution.
6368 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6369 than a timeout (e.g. a certificate is required, and is not provided), an
6370 Exim server now closes the connection immediately. Previously it waited for
6371 the client to close - but if the client is SSL, it seems that they each
6372 wait for each other, leading to a delay before one of them times out.
6374 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6375 maintained 0.8.x compatibility because I don't think many are using it, and
6376 it is clearly obsolete.
6378 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6381 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6382 client certificate was expired. A simple patch fixes this, though I don't
6383 understand the full logic of why the verify callback is called multiple
6386 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6387 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6388 which causes problems with some clients (such as the Certicom SSL Plus
6389 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6390 disables the coutermeasure allowing Eudora to connect."
6392 38. Exim was not checking that a write() to a log file succeeded. This could
6393 lead to Bad Things if a log got too big, in particular if it hit a file
6394 size limit. Exim now panics and dies if it cannot write to a log file, just
6395 as it does if it cannot open a log file.
6397 39. Modified OS/Makefile-Linux so that it now contains
6399 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6401 The two -D definitions ensure that Exim is compiled with large file
6402 support, which makes it possible to handle log files that are bigger than
6405 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6406 instance) a domain was checked against a named list that involved a lookup,
6407 causing $domain_data to be set, then another domain was checked against the
6408 same list, then the first domain was re-checked, the value of $domain_data
6409 after the final check could be wrong. In particular, if the second check
6410 failed, it could be set empty. This bug probably also applied to
6413 41. The strip_trailing_dot option was not being applied to the address given
6414 with the -f command-line option.
6416 42. The code for reading a message's header from the spool was incrementing
6417 $received_count, but never initializing it. This meant that the value was
6418 incorrect (doubled) while delivering a message in the same process in which
6419 it was received. In the most common configuration of Exim, this never
6420 happens - a fresh exec is done - but it can happen when
6421 deliver_drop_privilege is set.
6423 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6424 it now includes up to 150 characters of the unexpected data in the log
6427 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6428 and building data strings. The size of both of these buffers was 10 000
6429 bytes - far larger than anybody would *ever* want, thought I. Needless to
6430 say, somebody hit the limit. I have increased the maximum line length to
6431 20 000 and the maximum data length of concatenated lines to 100 000. I have
6432 also fixed two bugs, because there was no checking on these buffers. Tsk,
6433 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6434 buffer is too small.
6436 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6437 lsearch lookups. Now it does.
6439 46. When parsing a route_list item in a manualroute router, a fixed-length
6440 buffer was used for the list of hosts. I made this 1024 bytes long,
6441 thinking that nobody would ever have a list of hosts that long. Wrong.
6442 Somebody had a whole pile of complicated expansion conditions, and the
6443 string was silently truncated, leading to an expansion error. It turns out
6444 that it is easier to change to an unlimited length (owing to other changes
6445 that have happened since this code was originally written) than to build
6446 structure for giving a limitation error. The length of the item that
6447 expands into the list of hosts is now unlimited.
6449 47. The lsearch lookup could not handle data where the length of text line was
6450 more than 4095 characters. Such lines were truncated, leading to shortened
6451 data being returned. It should now handle lines of any length.
6453 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6454 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6457 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6458 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6459 no longer happens in comment lines. A list of replacements is now placed
6460 at the head of all of the source files, except those whose only change is
6461 to replace PERL_COMMAND in the very first #! line.
6463 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6464 messages on the queue, with a bottom-up merge sort, using code contributed
6465 by Michael Haardt. This should make operations like -bp somewhat faster on
6466 large queues. It won't affect queue runners, except when queue_run_in_order
6469 51. Installed eximstats 1.31 in the distribution.
6471 52. Added support for SRV lookups to the dnslookup router.
6473 53. If an ACL referred to $message_body or $message_body_end, the value was not
6474 reset for any messages that followed in the same SMTP session.
6476 54. The store-handling optimization for building very long strings was not
6477 differentiating between the different store pools. I don't think this
6478 actually made any difference in practice, but I've tidied it.
6480 55. While running the routers to verify a sender address, $sender_address
6481 was still set to the sender address. This is wrong, because when routing to
6482 send a bounce to the sender, it would be empty. Therefore, I have changed
6483 it so that, while verifying a sender address, $sender_address is set to <>.
6484 (There is no change to what happens when verifying a recipient address.)
6486 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6487 target A or AAAA records (if not already returned) without resetting the
6488 qualify_single or search_parents options of the DNS resolver. These are
6489 inappropriate in this case because the targets of MX and SRV records must
6490 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6491 target that, when qualified, matched something in the local domain. These
6492 two options are now turned off when doing these lookups.
6494 57. It seems that at least some releases of Reiserfs (which does not have the
6495 concept of a fixed number of inodes) returns zero and not -1 for the
6496 number of available inodes. This interacted badly with check_spool_inodes,
6497 which assumed that -1 was the "no such thing" setting. What I have done is
6498 to check that the total number of inodes is greater than zero before doing
6499 the test of how many are available.
6501 58. When a "warn" ACL statement has a log_message modifier, the message is
6502 remembered, and not repeated. This is to avoid a lot of repetition when a
6503 message has many recipients that cause the same warning to be written.
6504 However, Exim was preserving the list of already written lines for an
6505 entire SMTP session, which doesn't seem right. The memory is now reset if a
6506 new message is started.
6508 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6509 debugging output unless log_rewrite was also set.
6511 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6512 of (int)(handle) when we know that handle contains (void *)(-1).
6514 61. The Exim daemon panic-logs an error return when it closes the incoming
6515 connection. However "connection reset by peer" seems to be common, and
6516 isn't really an error worthy of noting specially, so that particular error
6519 62. When Exim is trying to find all the local interfaces, it used to panic and
6520 die if the ioctl to get the interface flags failed. However, it seems that
6521 on at least one OS (Solaris 9) it is possible to have an interface that is
6522 included in the list of interfaces, but for which you get a failure error
6523 for this call. This happens when the interface is not "plumbed" into a
6524 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6525 failure of the "get flags" call assumes that the interface is down.
6527 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6528 makes life easier for people who are doing arithmetic on fields extracted
6529 from dates, where you often get leading zeros that should not be
6530 interpreted as octal.
6532 64. Added qualify_domain to the redirect router, to override the global
6535 65. If a pathologically long header line contained very many addresses (the
6536 report of this problem mentioned 10 000) and each of them was rewritten,
6537 Exim could use up a very large amount of memory. (It kept on making new
6538 copies of the header line as it rewrote, and never released the old ones.)
6539 At the expense of a bit more processing, the header rewriting function has
6540 been changed so that it no longer eats memory in this way.
6542 66. The generation of the Received: header has been moved from the time that a
6543 message starts to be received, to the time that it finishes. The timestamp
6544 in the Received: header should now be very close to that of the <= log
6545 line. There are two side-effects of this change:
6547 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6548 logged header lines no longer include the local Received: line, because
6549 it has not yet been created. The same applies to a copy of the message
6550 that is returned to a non-SMTP sender when a message is rejected.
6552 (b) When a filter file is tested using -bf, no additional Received: header
6553 is added to the test message. After some thought, I decided that this
6556 This change does not affect the value of $received_for. It is still set
6557 after address rewriting, but before local_scan() is called.
6559 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6561 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6562 gave an unhelpful panic error message, and a defer error. I have managed to
6563 change this behaviour so that it now rejects any supplied certificate,
6564 which seems right, as the list of acceptable certificates is empty.
6566 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6567 gave an unhelpful defer error. I have not managed to make this reject any
6568 supplied certificates, but the error message it gives is "no certificate
6569 supplied", which is not helpful.
6571 70. exigrep's output now also includes lines that are not associated with any
6572 message, but which match the given pattern. Implemented by a patch from
6573 Martin Sluka, which also tidied up the Perl a bit.
6575 71. Recipient callout verification, like sender verification, was using <> in
6576 the MAIL FROM command. This isn't really the right thing, since the actual
6577 sender may affect whether the remote host accepts the recipient or not. I
6578 have changed it to use the actual sender in the callout; this means that
6579 the cache record is now keyed on a recipient/sender pair, not just the
6580 recipient address. There doesn't seem to be a real danger of callout loops,
6581 since a callout by the remote host to check the sender would use <>.
6582 [SEE ABOVE: changed after hitting problems.]
6584 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6585 temporary errors. However, in the case of such a code being given after
6586 the end of a data transmission (i.e. after ".") Exim was failing to write
6587 a retry record for the message. (Yes, there was some broken host that was
6588 actually sending 8xx at this point.)
6590 73. An unknown lookup type in a host list could cause Exim to panic-die when
6591 the list was checked. (An example that provoked this was putting <; in the
6592 middle of a list instead of at the start.) If this happened during a DATA
6593 ACL check, a -D file could be left lying around. This kind of configuration
6594 error no longer causes Exim to die; instead it causes a defer error. The
6595 incident is still logged to the main and panic logs.
6597 74. Buglet left over from Exim 3 conversion. The message "too many messages
6598 in one connection" was written to the rejectlog but not the mainlog, except
6599 when address rewriting (yes!) was being logged.
6601 75. Added write_rejectlog option.
6603 76. When a system filter was run not as root (that is, when system_filter_user
6604 was set), the values of the $n variables were not being returned to the
6605 main process; thus, they were not subsequently available in the $sn
6608 77. Added +return_path_on_delivery log selector.
6610 78. A connection timeout was being treated differently from recipients deferred
6611 when testing hosts_max_try with a message that was older than the host's
6612 retry timeout. (The host should not be counted, thus allowing all hosts to
6613 be tried at least once before bouncing.) This may have been the cause of an
6614 occasionally reported bug whereby a message would remain on the queue
6615 longer than the retry timeout, but would be bounced if a delivery was
6616 forced. I say "may" because I never totally pinned down the problem;
6617 setting up timeout/retry tests is difficult. See also the next item.
6619 79. The ultimate address timeout was not being applied to errors that involved
6620 a combination of host plus message (for example, a timeout on a MAIL
6621 command). When an address resolved to a number of possible hosts, and they
6622 were not all tried for each delivery (e.g. because of hosts_max_try), a
6623 message could remain on the queue longer than the retry timeout.
6625 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6628 81. Fixed an obscure SMTP outgoing bug which required at least the following
6629 conditions: (a) there was another message waiting for the same server;
6630 (b) the server returned 5xx to all RCPT commands in the first message so
6631 that the message was not completed; (c) the server dropped the connection
6632 or gave a negative response to the RSET that Exim sends to abort the
6633 transaction. The observed case was a dropped connection after DATA that had
6634 been sent in pipelining mode. That is, the server had advertised PIPELINING
6635 but was not implementing it correctly. The effect of the bug was incorrect
6636 behaviour, such as trying another host, and this could lead to a crash.
6642 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6643 and daemon.c were passed as pointers to ints; they should have been
6644 pointers to socklen_t variables (which are typically unsigned ints).
6646 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6649 3. Fixed a really odd bug that affected only the testing scheme; patching a
6650 certain fixed string in the binary changed the value of another string that
6651 happened to be identical to the end of the original first string.
6653 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6654 name", it returns that address as the IP address. On some operating
6655 systems (e.g. Solaris), it also passes back the IP address string as the
6656 "host name". However, on others (e.g. Linux), it passes back an empty
6657 string. Exim wasn't checking for this, and was changing the host name to an
6658 empty string, assuming it had been canonicalized.
6660 5. Although rare, it is permitted to have more than one PTR record for a given
6661 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6662 all the names associated with an address, because they do in Solaris.
6663 However, it seems that they do not in Linux for data that comes from the
6664 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6665 I found this out when I moved to a new Linux workstation and tried to run
6666 the Exim test suite.
6668 To get round this problem I have changed the code so that it now does its
6669 own call to the DNS to look up PTR records when searching for a host name.
6670 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6671 addresses that are only in /etc/hosts are still found.
6673 This behaviour is, however, controlled by an option called host_lookup_
6674 order, which defaults to "bydns:byaddr". If people want to use the other
6675 order, or indeed, just use one or the other means of lookup, they can
6676 specify it in this variable.
6678 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6679 some operating systems, this comes back from gethostbyaddr() as an empty
6680 string, and this is what Exim used to test for. However, it seems that in
6681 other systems, "." is yielded. Exim now tests for this case too.
6683 7. The values of check_spool_space and check_log_space are now held internally
6684 as a number of kilobytes instead of an absolute number of bytes. If a
6685 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6686 kilobyte. This means that much larger values can be stored.
6688 8. Exim monitor: an attempt to get the action menu when not actually pointing
6689 at a message produces an empty menu entitled "No message selected". This
6690 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6691 no entries in it ("Shell widget menu has zero width and/or height"). So I
6692 have added a single, blank menu entry in this case.
6694 9. Added ${quote_local_part.
6696 10. MIME decoding is now applied to the contents of Subject: header lines when
6699 11. Now that a reference to $sender_host_address automatically causes a reverse
6700 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6701 host lookup before query-style lookups in lists that might use this
6702 variable. This has therefore been abolished, and the "net-" prefix is no
6703 longer necessary for query-style lookups.
6705 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6706 have been a typo for LFLAGS, so it has been changed.
6708 13. The install script calls Exim with "-C /dev/null" in order to find the
6709 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6710 to be output. However, since Exim outputs its version number before the
6711 error, it didn't break the script. It just looked ugly. I fixed this by
6712 always allowing "-C /dev/null" if the caller is root.
6714 14. Ignore overlarge ACL variable number when reading spool file - insurance
6715 against a later release with more variables having written the file.
6717 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6718 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6719 this, as well as the form without the "IPv6" on the front.
6721 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6722 OS/Makefile-Darwin file.
6724 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6725 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6727 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6728 variable. However, some OS use ints for inodes. I've added cast to long int
6729 to get rid of the compiler warning.
6731 19. I had forgotten to lock out "/../" in configuration file names when
6732 ALT_CONFIG_PREFIX was set.
6734 20. Routers used for verification do not need to specify transports. However,
6735 if such a router generated a host list, and callout was configured, Exim
6736 crashed, because it could not find a port number from the (non-existent)
6737 transport. It now assumes port 25 in this circumstance.
6739 21. Added the -t option to exigrep.
6741 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6742 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6743 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6744 them is set, LOOKUP_LSEARCH is forced.
6746 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6748 24. Added sender and host information to the "rejected by local_scan()" log
6749 line; previously there was no indication of these.
6751 25. Added .include_if_exists.
6753 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6754 Linux. It turns out that not all file systems support this. Apparently some
6755 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6756 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6757 not supported on the file descriptor, is now ignored when Exim is trying to
6758 sync a directory. This applies only to Linux.
6760 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6762 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6763 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6766 29. Added support for maildirsize files from supplied patch (modified a bit).
6768 30. The use of :fail: followed by an empty string could lead Exim to respond to
6769 sender verification failures with (e.g.):
6771 550 Verification failed for <xxx>
6772 550 Sender verify failed
6774 where the first response line was missing the '-' that indicates it is not
6775 the final line of the response.
6777 31. The loop for finding the name of the user that called Exim had a hardwired
6778 limit of 10; it now uses the value of finduser_retries, which is used for
6779 all other user lookups.
6781 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6784 33. Exim was neglecting to zero errno before one call of strtol() when
6785 expanding a string and expecting an integer value. On some systems this
6786 resulted in spurious "integer overflow" errors. Also, it was casting the
6787 result into an int without checking.
6789 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6790 did not work. The code looks as if it has *never* worked, though it appears
6791 to have been documented since at least release 1.62. I have made it work.
6793 35. The "timeout_DNS" error in retry rules, also documented since at least
6794 1.62, also never worked. As it isn't clear exactly what this means, and
6795 clearly it isn't a major issue, I have abolished the feature by treating it
6796 as "timeout", and writing a warning to the main and panic logs.
6798 36. The display of retry rules for -brt wasn't always showing the error code
6801 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6802 timeout_connect_A, timeout_connect_MX.
6804 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6805 to the empty sender.
6807 39. The daemon was not analysing the content of -oX till after it had closed
6808 stderr and disconnected from the controlling terminal. This meant that any
6809 syntax errors were only noted on the panic log, and the return code from
6810 the command was 0. By re-arranging the code a little, I've made the
6811 decoding happen first, so such errors now appear on stderr, and the return
6812 code is 1. However, the actual setting up of the sockets still happens in
6813 the disconnected process, so errors there are still only recorded on the
6816 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6817 connections (as happens on some IP stacks) was logged at start up time as
6818 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6819 it from "IPv6 and IPv4", which means that two separate sockets are being
6822 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6823 says whether AF_INET or AF_INET6 was passed as an argument.
6825 42. Exiwhat output was messed up when time zones were included in log
6828 43. Exiwhat now gives more information about the daemon's listening ports,
6829 and whether -tls-on-connect was used.
6831 44. The "port" option of the smtp transport is now expanded.
6833 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6834 silently ignored. Now an error message is written to the main and panic
6837 46. There's a new ACL modifier called "logwrite" which writes to a log file
6838 as soon as it is encountered.
6840 47. Added $local_user_uid and $local_user_gid at routing time.
6842 48. Exim crashed when trying to verify a sender address that was being
6845 49. Exim was recognizing only a space character after ".include". It now also
6846 recognizes a tab character.
6848 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6849 extracting the relevant information from the specification. The man page no
6850 longer contains scrambled data for the -d option, and I've added a section
6851 at the front about calling Exim under different names.
6853 51. Added "extra_headers" argument to the "mail" command in filter files.
6855 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6858 53. Installed eximstats 1.29.
6860 54. Added transport_filter_timeout as a generic transport option.
6862 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6863 Cc: header lines. This was required by RFC 822, but it not required by RFC
6866 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6867 incoming messages that did not have them. Now it does so only if the
6868 message originates locally, that is, if there is no associated remote host
6869 address. When Resent- header lines are present, this applies to the Resent-
6870 lines rather than the non-Resent- lines.
6872 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6873 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6875 58. Messages for configuration errors now include the name of the main
6876 configuration file - useful now that there may be more than one file in a
6877 list (.included file names were always shown).
6879 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6880 for those rare installations that do not start the daemon as root or run it
6881 setuid root. I've cut out the call to initgroups() if the daemon is not
6884 60. The Exim user and group can now be bound into the binary as text strings
6885 that are looked up at the start of Exim's processing.
6887 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6889 62. Added $mailstore_basename variable.
6891 63. Installed patch to sieve.c from Michael Haardt.
6893 64. When Exim failed to open the panic log after failing to open the main log,
6894 the original message it was trying to log was written to stderr and debug
6895 output, but if they were not available (the usual case in production), it
6896 was lost. Now it is written to syslog before the two lines that record the
6897 failures to open the logs.
6899 65. Users' Exim filters run in subprocesses under the user's uid. It is
6900 possible for a "deliver" command or an alias in a "personal" command to
6901 provoke an address rewrite. If logging of address rewriting is configured,
6902 this fails because the process is not running as root or exim. There may be
6903 a better way of dealing with this, but for the moment (because 4.30 needs
6904 to be released), I have disabled address rewrite logging when running a
6905 filter in a non-root, non-exim process.
6911 1. The buildconfig auxiliary program wasn't quoting the value set for
6912 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6913 not defined. This bug was masked in 4.22 by the effect that was fixed in
6916 2. Some messages that were rejected after a message id was allocated were
6917 shown as "incomplete" by exigrep. It no longer does this for messages that
6918 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6920 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6921 have allow_domain_literals set, the ID did not get logged in the <= line.
6922 Domain literals are now always recognized in Message-ID: header lines.
6924 4. The first argument for a ${extract expansion item is the key name or field
6925 number. Leading and trailing spaces in this item were not being ignored,
6926 causing some misleading effects.
6928 5. When deliver_drop_privilege was set, single queue runner processes started
6929 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6930 same command in the process it spins off) were not dropping privilege.
6932 6. When the daemon running as "exim" started a queue runner, it always
6933 re-executed Exim in the spun-off process. This is a waste of effort when
6934 deliver_drop_privilege is set. The new process now just calls the
6935 queue-runner function directly.
6941 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6944 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6945 ignored. Though the use of -oP was forcing the writing of a pid file, it
6946 was always written to the default place.
6948 3. If the message "no IP address found for host xxxx" is generated during
6949 incoming verification, it is now followed by identification of the incoming
6950 connection (so you can more easily find what provoked it).
6952 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6954 5. Added some features to "harden" Exim a bit more against certain attacks:
6956 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6957 be put in Local/Makefile. This is like the never_users runtime option,
6958 but it cannot be overridden. The default setting is "root".
6960 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6961 prefix string with which any file named in a -C command line option
6964 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6965 is retained for -C and -D only if the caller of Exim is root. Without
6966 it, the exim user may also use -C and -D and retain privilege.
6968 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6969 command line option is disabled.
6971 6. Macro names set by the -D option must start with an upper case letter, just
6972 like macro names defined in the configuration file.
6974 7. Added "dereference=" facility to LDAP.
6976 8. Two instances of the typo "uknown" in the source files are fixed.
6978 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6979 the Configure-Makefile script screwed up while processing it.
6981 10. Incorporated PCRE 4.4.
6983 11. The SMTP synchronization check was not operating right at the start of an
6984 SMTP session. For example, it could not catch a HELO sent before the client
6985 waited for the greeting. There is now a check for outstanding input at the
6986 point when the greeting is written. Because of the duplex, asynchronous
6987 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6988 way, but not yet received, when the check is performed.
6990 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6991 on TCP/IP sockets, because this apparently causes some broken clients to
6994 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6995 unchanged) from the Cygwin maintainer.
6997 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6998 was supported instead of testing for mbx. Effectively a typo.
7000 15. The spa authenticator server code was not checking that the input it
7001 received was valid base64.
7003 16. The debug output line for the "set" modifier in ACLs was not showing the
7004 name of the variable that was being set.
7006 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7008 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7009 lookup on its own. Something else was needed to trigger the lookup. For
7010 example, a match in host_lookup or the need for a host name in a host list.
7011 Now, if $sender_host_name is referenced and the host name has not yet been
7012 looked up, a lookup is performed. If the lookup fails, the variable remains
7013 empty, and $host_lookup_failed is set to "1".
7015 19. Added "eqi" as a case-independent comparison operator.
7017 20. The saslauthd authentication condition could segfault if neither service
7018 nor realm was specified.
7020 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7021 error message that was logged was misleading, and incoming SMTP
7022 connections were dropped. The message is now more accurate, and temporary
7023 errors are given to SMTP connections.
7025 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7026 (except RSET) with a 421 error, until QUIT is received. However, it was
7027 failing to send a response to QUIT.
7029 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7030 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7031 ACL is now run after the helo_try_verify_hosts code.
7033 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7034 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7035 case-independent, so other case variants are also recognized). Apparently
7036 some systems use these upper case variants.
7038 25. If more than two messages were waiting for the same host, and a transport
7039 filter was specified for the transport, Exim sent two messages over the
7040 same TCP/IP connection, and then failed with "socket operation on non-
7041 socket" when it tried to send the third.
7043 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7045 27. The extern definition of crypt16() in expand.c was not being excluded when
7046 the OS had its own crypt16() function.
7048 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7049 as a preferred synonym for return_size_limit, both as an option and as an
7052 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7054 30. Changed the default configuration ACL to relax the local part checking rule
7055 for addresses that are not in any local domains. For these addresses,
7056 slashes and pipe symbols are allowed within local parts, but the sequence
7057 /../ is explicitly forbidden.
7059 31. SPA server authentication was not clearing the challenge buffer before
7062 32. log_message in a "warn" ACL statement was writing to the reject log as
7063 well as to the main log, which contradicts the documentation and doesn't
7064 seem right (because no rejection is happening). So I have stopped it.
7066 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7067 However, I am unable to do any testing of this.
7069 34. Fixed an infelicity in the appendfile transport. When checking directories
7070 for a mailbox, to see if any needed to be created, it was accidentally
7071 using path names with one or more superfluous leading slashes; tracing
7072 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7074 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7075 contents are added to the log line that is written for every discarded
7076 recipient. (Previously a log_message setting was ignored.)
7078 36. The ${quote: operator now quotes the string if it is empty.
7080 37. The install script runs exim in order to find its version number. If for
7081 some reason other than non-existence or emptiness, which it checks, it
7082 could not run './exim', it was installing it with an empty version number,
7083 i.e. as "exim-". This error state is now caught, and the installation is
7086 38. An argument was missing from the function that creates an error message
7087 when Exim fails to connect to the socket for saslauthd authentication.
7088 This could cause Exim to crash, or give a corrupted message.
7090 39. Added isip, isip4, and isip6 to ${if conditions.
7092 40. The ACL variables $acl_xx are now saved with the message, and can be
7093 accessed later in routers, transports, and filters.
7095 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7096 strings in the file are not string-expanded.
7098 42. If a MAIL command specified a SIZE value that was too large to fit into an
7099 int variable, the check against message_size_limit failed. Such values are
7100 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7101 day this will have to be increased, but I don't think I want to be around
7102 when emails are that large.
7109 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7110 iconv() is not standard in FreeBSD.
7112 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7113 IPv6 enabled. The observed symptom was a segmentation fault on return from
7114 the function os_common_find_running_interfaces() in src/os.c.
7116 3. In the check_special_case() function in daemon.c I had used "errno" as an
7117 argument name, which causes warnings on some systems. This was basically a
7118 typo, since it was named "eno" in the comments!
7120 4. The code that waits for the clock to tick (at a resolution of some fraction
7121 of a second) so as to ensure message-id uniqueness was always waiting for
7122 at least one whole tick, when it could have waited for less. [This is
7123 almost certainly not relevant at current processor speeds, where it is
7124 unlikely to ever wait at all. But we try to future-proof.]
7126 5. The function that sleeps for a time interval that includes fractions of a
7127 second contained a race. It did not block SIGALRM between setting the
7128 timer, and suspending (a couple of lines later). If the interval was short
7129 and the sigsuspend() was delayed until after it had expired, the suspension
7130 never ended. On busy systems this could lead to processes getting stuck for
7133 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7134 process, before it forks any delivery processes. The open lookup caching
7135 mechanism meant that the open file or database connection was passed into
7136 the delivery process. The problem was that delivery processes always tidy
7137 up cached lookup data. This could cause a problem for the next delivery
7138 process started by the queue runner, because the external queue runner
7139 process does not know about the closure. So the next delivery process
7140 still has data in the lookup cache. In the case of a file lookup, there was
7141 no problem because closing a file descriptor in a subprocess doesn't affect
7142 the parent. However, if the lookup was caching a connection to a database,
7143 the connection was closed, and the second delivery process was likely to
7144 see errors such as "PGSQL: query failed: server closed the connection
7145 unexpectedly". The problem has been fixed by closing all cached lookups
7146 in a queue runner before running a delivery process.
7148 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7149 seem to have the "const" qualifier which it has on other OS. I've
7152 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7153 *of the same type* (client or server) with the same public name that an
7154 error should be diagnosed.
7156 9. When Exim looked up a host name for an IP address, but failed to find the
7157 original IP address when looking up the host name (a safety check), it
7158 output the message "<ip address> does not match any IP for NULL", which was
7159 confusing, to say the least. The bug was that the host name should have
7160 appeared instead of "NULL".
7162 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7163 user that is built into the binary, and the -C or -D options is used, root
7164 privilege is dropped before the configuration file is read. In addition,
7165 logging is switched to stderr instead of the normal log files. If the
7166 configuration then re-defines the Exim user, the unprivileged environment
7167 is probably not what is expected, so Exim logs a panic warning message (but
7170 However, if deliver_drop_privilege is set, the unprivileged state may well
7171 be exactly what is intended, so the warning has been cut out in that case,
7172 and Exim is allowed to try to write to its normal log files.
7178 1. smtp_return_error_details was not giving details for temporary sender
7179 or receiver verification errors.
7181 2. Diagnose a configuration error if two authenticators have the same public
7184 3. Exim used not to create the message log file for a message until the first
7185 delivery attempt. This could be confusing when incoming messages were held
7186 for policy or load reasons. The message log file is now created at the time
7187 the message is received, and an initial "Received" line is written to it.
7189 4. The automatically generated man page for command line options had a minor
7190 bug that caused no ill effects; however, a more serious problem was that
7191 the procedure for building the man page automatically didn't always
7192 operate. Consequently, release 4.20 contains an out-of-date version. This
7193 shouldn't happen again.
7195 5. When building Exim with embedded Perl support, the script that builds the
7196 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7197 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7199 6. The freeze_tell option was not being used for messages that were frozen on
7200 arrival, either by an ACL or by local_scan().
7202 7. Added the smtp_incomplete_transaction log selector.
7204 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7205 was accepting AUTH without a new EHLO.
7207 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7208 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7211 10. Logging of TCP/IP connections (when configured) now happens in the main
7212 daemon process instead of the child process, so that the TCP/IP connection
7213 count is more accurate (but it can never be perfect).
7215 11. The use of "drop" in a nested ACL was not being handled correctly in the
7216 outer ACL. Now, if condition failure induced by the nested "drop" causes
7217 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7218 or "require"), the connection is dropped.
7220 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7221 that yield "discard" can now be used with an "accept" or a "discard" verb,
7222 but an error is generated for any others (because I can't see a useful way
7223 to define what should happen).
7225 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7226 are now processed in the same way as lines in the Exim configuration file.
7227 In particular, continuation lines are supported.
7229 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7231 15. Added -ti meaning -t -i.
7233 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7234 domains, and warn by logging if others are found.
7236 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7237 returned from GIFCONF when Exim is trying to find the list of interfaces on
7238 a host. The code in os.c has been modified to copy each ifreq to an aligned
7239 structure in all cases.
7241 Also, in some cases, the returned ifreq's were being copied to a 'struct
7242 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7243 means the last couple of bytes of an IPv6 address could be chopped if the
7244 ifreq contained only a normal sockaddr (14 bytes storage).
7246 18. Named domain lists were not supported in the hosts_treat_as_local option.
7247 An entry such as +xxxx was not recognized, and was treated as a literal
7250 19. Ensure that header lines added by a DATA ACL are included in the reject log
7251 if the ACL subsequently rejects the message.
7253 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7254 MD5 (which is deprecated).
7256 21. When testing a filter file using -bf, Exim was writing a message when it
7257 took the sender from a "From " line in the message, but it was not doing so
7258 when it took $return_path from a Return-Path: header line. It now does.
7260 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7261 with a valid header line field name (a series of printing characters
7262 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7264 23. Changed "disc" in the source to "disk" to conform to the documentation and
7265 the book and for uniformity.
7267 24. Ignore Sendmail's -Ooption=value command line item.
7269 25. When execve() failed while trying to run a command in a pipe transport,
7270 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7271 could be confused with a return value of 69 from the command itself. This
7272 has been changed to 127, the value the shell returns if it is asked to run
7273 a non-existent command. The wording for the related log line suggests a
7274 non-existent command as the problem.
7276 26. If received_header_text expands to an empty string, do not add a Received:
7277 header line to the message. (Well, it adds a token one on the spool, but
7278 marks it "old" so that it doesn't get used or transmitted.)
7280 27. Installed eximstats 1.28 (addition of -nt option).
7282 28. There was no check for failure on the call to getsockname() in the daemon
7283 code. This can fail if there is a shortage of resources on the system, with
7284 ENOMEM, for example. A temporary error is now given on failure.
7286 29. Contrary to the C standard, it seems that in some environments, the
7287 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7288 program. Exim now does this explicitly; it affects the formatting of
7289 timestamps using strftime().
7291 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7292 complaints. I've now initialized all the variables, to avoid this.
7294 32. Header lines added by a system filter were not being "seen" during
7295 transport-time rewrites.
7297 33. The info_callback() function passed to OpenSSL is set up with type void
7298 (*)(SSL *, int, int), as described somewhere. However, when calling the
7299 function (actually a macro) that sets it up, the type void(*)() is
7300 expected. I've put in a cast to prevent warnings from picky compilers.
7302 34. If a DNS black list lookup found a CNAME record, but there were no A
7303 records associated with the domain it pointed at, Exim crashed.
7305 35. If a DNS black list lookup returned more than one A record, Exim ignored
7306 all but the first. It now scans all returned addresses if a particular IP
7307 value is being sought. In this situation, the contents of the
7308 $dnslist_value variable are a list of all the addresses, separated by a
7311 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7312 to accept a host name and all its aliases if the forward lookup for any of
7313 them yielded the IP address of the incoming connection. Now it accepts only
7314 those names whose forward lookup yields the correct IP address. Any other
7315 names are discarded. This closes a loophole whereby a rogue DNS
7316 administrator could create reverse DNS records to break through a
7317 wildcarded host restriction in an ACL.
7319 37. If a user filter or a system filter that ran in a subprocess used any of
7320 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7321 the wrong values were passed to the pipe command ($thisaddress had the
7322 value of $0, $0 had the value of $1, etc). This bug was introduced by
7323 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7325 38. Improved the line breaking for long SMTP error messages from ACLs.
7326 Previously, if there was no break point between 40 and 75 characters, Exim
7327 left the rest of the message alone. Two changes have been made: (a) I've
7328 reduced the minimum length to 35 characters; (b) if it can't find a break
7329 point between 35 and 75 characters, it looks ahead and uses the first one
7330 that it finds. This may give the occasional overlong line, but at least the
7331 remaining text gets split now.
7333 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7334 file descriptors might be low, and that setting 1000 would always raise it.
7335 It turns out that in some environments, the limit is already over 1000 and
7336 that lowering it causes trouble. So now Exim takes care not to decrease it.
7338 40. When delivering a message, the value of $return_path is set to $sender_
7339 address at the start of routing (routers may change the value). By an
7340 oversight, this default was not being set up when an address was tested by
7341 -bt or -bv, which affected the outcome if any router or filter referred to
7344 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7345 log only when "message" or "log_message" are set. However, if one of the
7346 conditions was an address verification, or a call to a nested ACL, the
7347 messages generated by the underlying test were being passed through. This
7348 no longer happens. The underlying message is available in $acl_verify_
7349 message for both "message" and "log_message" expansions, so it can be
7350 passed through if needed.
7352 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7353 new expansion $bh_ to give the encoded byte string without charset
7354 translation. Translation happens only if iconv() is available; HAVE_ICONV
7355 indicates this at build time. HEADERS_CHARSET gives the charset to
7356 translate to; headers_charset can change it in the configuration, and
7357 "headers charset" can change it in an individual filter file.
7359 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7360 that creates RFC 2047 encoded "words" labels them as that charset instead
7361 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7362 expansion operator; (ii) when Exim creates a From: line for a local
7363 message; (iii) when a header line is rewritten to include a "phrase" part.
7365 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7366 buggy, causing it to skip the first lines of messages whose message ID
7367 ended in 'D'. This would not have bitten before Exim release 4.14, because
7368 message IDs were unlikely to end in 'D' before then. The effect was to have
7369 incorrect size information for certain domains.
7371 45. #include "config.h" was missing at the start of the crypt16.c module. This
7372 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7375 46. If there was a timeout during a "random" callout check, Exim treated it as
7376 a failure of the random address, and carried on sending RSET and the real
7377 address. If the delay was just some slowness somewhere, the response to the
7378 original RCPT would be taken as a response to RSET and so on, causing
7379 mayhem of various kinds.
7381 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7382 when I implemented it. It didn't allow for the fact that some option values
7383 may legitimately be negative (e.g. size_addition), and it didn't even do
7384 the right test for positive values.
7386 48. Domain names in DNS records are case-independent. Exim always looks them up
7387 in lower case. Some resolvers return domain names in exactly the case they
7388 appear in the zone file, that is, they may contain uppercase letters. Not
7389 all resolvers do this - some return always lower case. Exim was treating a
7390 change of case by a resolver as a change of domain, similar to a widening
7391 of a domain abbreviation. This triggered its re-routing code and so it was
7392 trying to route what was effectively the same domain again. This normally
7393 caused routing to fail (because the router wouldn't handle the domain
7394 twice). Now Exim checks for this case specially, and just changes the
7395 casing of the domain that it ultimately uses when it transmits the message
7398 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7401 50. If a filter generated a file delivery with a non-absolute name (possible if
7402 no home directory exists for the router), the forbid_file option was not
7405 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7406 the existing equality matching.
7408 52. Exim was using ints instead of ino_t variables in some places where it was
7409 dealing with inode numbers.
7411 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7412 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7413 called TMPDIR, and if it finds it is different, it changes its value.
7415 54. The smtp_printf() function is now made available to local_scan() so
7416 additional output lines can be written before returning. There is also an
7417 smtp_fflush() function to enable the detection of a dropped connection.
7418 The variables smtp_input and smtp_batched_input are exported to
7421 55. Changed the default runtime configuration: the message "Unknown user"
7422 has been removed from the ACL, and instead placed on the localuser router,
7423 using the cannot_route_message feature. This means that any verification
7424 failures that generate their own messages won't get overridden. Similarly,
7425 the "Unrouteable address" message that was in the ACL for unverifiable
7426 relay addresses has also been removed.
7428 56. Added hosts_avoid_esmtp to the smtp transport.
7430 57. The exicyclog script was not checking for the esoteric option
7431 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7432 will work only if exicyclog is run under the appropriate euid.
7434 58. Following a discussion on the list, the rules by which Exim recognises line
7435 endings on incoming messages have been changed. The -dropcr and drop_cr
7436 options are now no-ops, retained only for backwards compatibility. The
7437 following line terminators are recognized: LF CRLF CR. However, special
7438 processing applies to CR:
7440 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7441 nor a local message in the state where . is a terminator.
7443 (ii) If a bare CR is encountered in a header line, an extra space is added
7444 after the line terminator so as not to end the header. The reasoning
7445 behind this is that bare CRs in header lines are most likely either
7446 to be mistakes, or people trying to play silly games.
7448 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7449 was being incorrectly given as 18 bytes larger than it should have been.
7450 This is a VOB (very old bug).
7452 60. This may never have affected anything current, but just in case it has:
7453 When the local host is found other than at the start of a list of hosts,
7454 the local host, those with the same MX, and any that follow, are discarded.
7455 When the list in question was part of a longer list of hosts, the following
7456 hosts (not currently being processed) were also being discarded. This no
7457 longer happens. I'm not sure if this situation could ever has previously
7460 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7463 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7464 header. This is something that is recommended in a new Internet Draft, and
7465 is something that is documented as being done by Sendmail. There are two
7466 possible values. For messages generated by the autoreply transport, Exim
7469 Auto-Submitted: auto-replied
7471 whereas for all other generated messages (e.g. bounces) it adds
7473 Auto-Submitted: auto-generated
7475 63. The "personal" condition in filters now includes a test for the
7476 Auto-Submitted: header. If it contains the string "auto-" the message it
7477 not considered personal.
7479 64. Added rcpt_include_affixes as a generic transport option.
7481 65. Added queue_only_override (default true).
7483 66. Added the syslog_duplication option.
7485 67. If what should have been the first header line of a message consisted of
7486 a space followed by a colon, Exim was mis-interpreting it as a header line.
7487 It isn't of course - it is syntactically invalid and should therefore be
7488 treated as the start of the message body. The misbehaviour could have
7489 caused a number of strange effects, including loss of data in subsequent
7490 header lines, and spool format errors.
7492 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7493 client host had authenticated. This control can now be exercised by an ACL
7494 for more flexibility.
7496 69. By default, callouts do not happen when testing with -bh. There is now a
7497 variant, -bhc, which does actually run the callout code, including
7498 consulting and updating the callout cache.
7500 70. Added support for saslauthd authentication, courtesy of Alexander
7503 71. If statvfs() failed on the spool or log directories while checking their
7504 size for availability, Exim confusingly gave the error "space shortage".
7505 Furthermore, in debugging mode it crashed with a floating point exception.
7506 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7507 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7508 serious problem, Exim now writes to the main and panic logs when this
7509 happens, with details of the failure. It then refuses to accept the
7510 incoming message, giving the message "spool directory problem" or "log
7511 directory problem" with a 421 code for SMTP messages.
7513 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7514 0, 1, and 2 exist, because some OS complain for execs without them (see
7515 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7516 descriptors. However, the code omitted to check that the open succeeded,
7517 causing mysterious errors if for some reason the permissions on /dev/null
7518 got screwed. Now Exim writes a message to the main and panic logs, and
7519 bombs out if it can't open /dev/null.
7521 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7522 interact so that it is all more flexible. It is supposed to remain
7523 backwards compatible. Also added extra_local_interfaces.
7525 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7526 to bomb out with an assertion failure - to the client this appears as a
7527 connection drop. This problem occurs in the part of the code that was taken
7528 from the Samba project. Fortunately, the assertion is in a very simple
7529 function, so I have fixed this by reproducing the function inline in the
7530 one place where it is called, and arranging for authentication to fail
7531 instead of killing the process with assert().
7533 75. The SPA client code was not working when the server requested OEM rather
7534 than Unicode encoding.
7536 76. Added code to make require_files with a specific uid setting more usable in
7537 the case where statting the file as root fails - usually a non-root-mounted
7538 NFS file system. When this happens and the failure is EACCES, Exim now
7539 forks a subprocess and does the per-uid checking as the relevant uid.
7541 77. Added process_log_path.
7543 78. If log_file_path was not explicitly set, a setting of check_log_space or
7544 check_log_inodes was ignored.
7546 79. If a space check for the spool or log partitions fails, the incident is now
7547 logged. Of course, in the latter case the data may get lost...
7549 80. Added the %p formatting code to string_format() so that it can be used to
7550 print addresses in debug_print(). Adjusted all the address printing in the
7551 debugging in store.c to use %p rather than %d.
7553 81. There was a concern that a line of code in smtp_in.c could overflow a
7554 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7555 initially expressed, the concern was not well-founded, because trailing
7556 spaces are removed early. However, if the trailing spaces were followed by
7557 a NULL, they did not get removed, so the overflow was possible. Two fixes
7560 (a) I re-wrote the offending code in a cleaner fashion.
7561 (b) If an incoming SMTP command contains a NULL character, it is rejected
7564 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7565 runs initgroups(), so that if the Exim user is in any additional groups,
7566 they will be used during message reception.
7572 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob