1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Avoid costly startup code when not strictly needed. This reduces time
12 for some exim process initialisations. It does mean that the logging
13 of TLS configuration problems is only done for the daemon startup.
15 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
17 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
18 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
21 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
22 DKIM signing is being done. Previously a transport_filter would always
23 disable CHUNKING, falling back to traditional DATA.
25 JH/05 Regard command-line receipients as tainted.
27 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
33 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
34 side, to decode a TLS 1.3 packet capture.
36 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
37 Previously the default library behaviour applied, sending two, each in
40 JH/03 Debug output for ACL now gives the config file name and line number for
43 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
45 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
47 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
48 buffer overrun for (non-chunking) other transports.
50 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
51 TLS1.3, means that a server rejecting a client certificate is not visible
52 to the client until the first read of encrypted data (typically the
53 response to EHLO). Add detection for that case and treat it as a failed
54 TLS connection attempt, so that the normal retry-in-clear can work (if
57 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
58 and/or domain. Found and fixed by Jason Betts.
60 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
61 configuration). If a CNAME target was not a wellformed name pattern, a
64 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
65 the OS reports them interleaved with other addresses.
67 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
68 used both for input and for a verify callout, both encrypted, SMTP
69 responses being sent by the server could be lost. This resulted in
70 dropped connections and sometimes bounces generated by a peer sending
73 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
74 string. Previously it was possible to cause undefined behaviour in a
75 library routine (usually a crash). Found by "zerons".
77 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
80 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
81 API was removed, so update to use the newer ones.
83 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
84 any timeout set, is taking a long time. Previously we would hang on to a
85 rotated logfile "forever" if the input was arriving with long gaps
86 (a previous attempt to fix addressed lack, for a long time, of initial
89 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
90 shared (NFS) environment. The length of the tempfile name is now
91 4 + 16 ("hdr.$message_exim_id") which might break on file
92 systems which restrict the file name length to lower values.
95 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
96 shared (NFS) environment.
98 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
99 did for all versions <4.90). Notably -M, -m, --invert, -I may be
102 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
103 on some platforms for bit 31.
105 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
106 to changes apparently associated with TLS1.3 handling some of the APIs
107 previously used were either nonfunctional or inappropriate. Strings
108 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
109 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
110 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
111 This affects log line X= elements, the $tls_{in,out}_cipher variables,
112 and the use of specific cipher names in the encrypted= ACL condition.
114 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
116 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
117 verification result was not updated unless hosts_require_ocsp applied.
119 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
120 queue_list_requires_admin set to false, non-admin users were denied the
123 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
124 directory-of-certs mode. Previously they were advertised despite the
127 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
128 A single TCP connection by a client will now hold a TLS connection open
129 for multiple message deliveries, by default. Previoud the default was to
132 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
133 default. If built with the facility, DANE will be used. The facility
134 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
136 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
137 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
138 must be defined and you must still, unless you define DISABLE_TLS, manage
139 the the include-dir and library-file requirements that go with that
140 choice. Non-TLS builds are still supported.
142 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
145 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
146 default. If the platform supports and has the facility enabled, it will
147 be requested on all coneections.
149 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
150 controlled by the build-time option SUPPORT_PIPE_CONNECT.
152 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
154 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
155 success-DSN messages. Previously the From: header was always the default
156 one for these; the option was ignored.
158 JH/28 Fix the timeout on smtp response to apply to the whole response.
159 Previously it was reset for every read, so a teergrubing peer sending
160 single bytes within the time limit could extend the connection for a
161 long time. Credit to Qualsys Security Advisory Team for the discovery.
163 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
164 delivery address, which leaked information of the results of local
165 forwarding. Change to the original envelope recipient address, per
168 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
169 requested. Previously not bounce was generated and a log entry of
170 error ignored was made.
172 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
174 JH/32 Introduce a general tainting mechanism for values read from the input
175 channel, and values derived from them. Refuse to expand any tainted
176 values, to catch one form of exploit.
178 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
179 was unused and the unexpanded text used for the test. Found and
180 fixed by Ruben Jenster.
182 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
183 an attempt to use a TLS library read routine dereffed a nul pointer,
186 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
187 them in our smtp response.
189 JH/36 Have the general router option retry_use_local_part default to true when
190 any of the restrictive preconditions are set (to anything). Previously it
191 was only for check_local user. The change removes one item of manual
192 configuration which is required for proper retries when a remote router
193 handles a subset of addresses for a domain.
195 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
196 link count into consideration.
198 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
199 caused the extension of big_buffer, the following lines were ignored.
201 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
202 accordance with RFC 2308. Previously there was no expiry, so a longlived
203 receive process (eg. due to ACL delays) versus a short SOA value could
206 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
208 JH/39 Promote DMARC support to mainline.
210 JH/40 Bug 2452: Add a References: header to DSNs.
212 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
213 parameters. The relevant library call is documented as "Deprecated: This
214 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
215 3.6.0, DH parameters are negotiated following RFC7919."
217 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
218 buffer was used for the filename, resulting in a trap when tainted
219 arguments (eg. $domain) were used.
221 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
222 the file-offset (which the Linux syscall does, and exim expects); this
223 resulted in an indefinite loop.
229 JH/01 Remove code calling the customisable local_scan function, unless a new
230 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
232 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
233 non-signal-safe functions being used.
235 JH/03 Bug 2269: When presented with a received message having a stupidly large
236 number of DKIM-Signature headers, disable DKIM verification to avoid
237 a resource-consumption attack. The limit is set at twenty.
239 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
240 report of oldest_pass in ${authres } in consequence, and separate out
241 some descriptions of reasons for verification fail.
243 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
244 files in the spool were present and unlocked. A queue-runner could spot
245 them, resulting in a duplicate delivery. Fix that by doing the unlock
246 after the unlink. Investigation by Tim Stewart. Take the opportunity to
247 add more error-checking on spoolfile handling while that code is being
250 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
251 No known attacks, no CVE, this is defensive hardening.
253 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
254 a queue-runner could start a delivery while other operations were ongoing.
255 Cutthrough delivery was a common victim, resulting in duplicate delivery.
256 Found and investigated by Tim Stewart. Fix by using the open message data
257 file handle rather than opening another, and not locally closing it (which
258 releases a lock) for that case, while creating the temporary .eml format
259 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
261 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
262 $sender_verify_failure/$recipient_verify_failure to "random".
264 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
267 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
268 Previously this would segfault.
270 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
273 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
274 like zero, since the resolver should be doing this for us, But we need one
275 as a CNAME but no MX presence gets the CNAME returned; we need to check
276 that doesn't point to an MX to declare it "no MX returned" rather than
277 "error, loop". A new main option is added so the older capability of
278 following some limited number of chain links is maintained.
280 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
282 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
283 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
284 now-deprecated earlier definitions used only specified the range up to TLS
285 1.2 (in the older-version library docs).
287 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
289 JH/15 Rework TLS client-side context management. Stop using a global, and
290 explicitly pass a context around. This enables future use of TLS for
291 connections to service-daemons (eg. malware scanning) while a client smtp
292 connection is using TLS; with cutthrough connections this is quite likely.
294 JH/16 Fix ARC verification to do AS checks in reverse order.
296 JH/17 Support a "tls" option on the ${readsocket } expansion item.
298 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
299 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
300 Previously the "utf8" would be re-prepended for every additional message.
302 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
303 Previously thery were accepted, resulting in issues when attempting to
304 forward messages to a non-supporting MTA.
306 PP/02 Let -n work with printing macros too, not just options.
308 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
309 one parent address was copied, and bogus data was used at delivery-logging
310 time. Either a crash (after delivery) or bogus log data could result.
311 Discovery and analysis by Tim Stewart.
313 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
314 Previously if the string ended mid-character, we did not insert the
315 promised '?' replacement.
317 PP/04 Documentation: current string operators work on bytes, not codepoints.
319 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
320 should pack well giving a smaller memory footprint so better caching and
321 therefore performance. Group the declarations where this can't be done so
322 that the byte-sized flag variables are not interspersed among pointer
323 variables, giving a better chance of good packing by the compiler.
325 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
326 non-null, to avoid issues with sites running BATV. Previously reports were
327 sent with an empty envelope sender so looked like bounces.
329 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
330 The ignore_error flag wasn't being returned from the filter subprocess so
331 was not set for later routers. Investigation and fix by Matthias Kurz.
333 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
334 and a msg:complete for the whole, when a message is manually removed using
335 -Mrm. Developement by Matthias Kurz, hacked on by JH.
337 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
338 a "Gnu special" function, asprintf() in the DB utility binary builds; I
339 hope that is portable enough.
341 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
342 requiring a known-CA anchor certificate; make it now rely entirely on the
343 TLSA as an anchor. Checking the name on the leaf cert against the name
344 on the A-record for the host is still done for TA (but not for EE mode).
346 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
347 included in delivery lines for non-proxied connections, when compiled with
348 SUPPORT_SOCKS and running with proxy logging enabled.
350 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
351 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
352 move the existing event to fire before the normal logging of message
353 failure so that custom logging is bracketed by normal logging.
355 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
356 msg:fail:internal event. Developement by Matthias Kurz.
358 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
359 far too small for todays use of crypto signatures stored there. Go all
360 the way to the max DNS message size of 64kB, even though this might be
361 overmuch for IOT constrained device use.
363 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
364 copy a string over itself. The library routine is documented as not
365 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
367 JH/32 For main options check_spool_space and check_inode_space, where the
368 platform supports 64b integers, support more than the previous 2^31 kB
369 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
370 the previous G, M, k.
372 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
373 $authenticated_fail_id variable on authentication failure. Previously
376 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
377 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
378 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
379 GNUTLS_SEC_PARAM_MEDIUM.
381 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
382 side. Previously we would continue as if no SNI had been received.
384 JH/36 Harden the handling of string-lists. When a list consisted of a sole
385 "<" character, which should be a list-separator specification, we walked
386 off past the nul-terimation.
388 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
389 causes) even when the retry time is not yet met. Previously they were
390 not, meaning that when (say) an account was over-quota and temp-rejecting,
391 and multiple senders' messages were queued, only one sender would get
392 notified on each configured delay_warning cycle.
394 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
396 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
397 cipher-suites, an error can be left on the stack even for a succeeding
398 accept; this results in impossible error messages when a later operation
401 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
402 return error codes indicating retry. Under TLS1.3 this becomes required.
404 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
405 it only wrote the new authenticators, resulting in a lack of tracking of
406 peer changes of ESMTP extensions until the next cache flush.
408 JH/41 Fix the loop reading a message header line to check for integer overflow,
409 and more-often against header_maxsize. Previously a crafted message could
410 induce a crash of the recive process; now the message is cleanly rejected.
412 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
413 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
419 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
420 When redis_servers is set to a list of > 1 element, and the Redis servers
421 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
422 case of MOVED into a DEFER case instead, thus moving the query onto the
423 next server in the list. For a cluster of N elements, all N servers must
424 be defined in redis_servers.
426 GF/02 Catch and remove uninitialized value warning in exiqsumm
427 Check for existence of @ARGV before looking at $ARGV[0]
429 JH/01 Replace the store_release() internal interface with store_newblock(),
430 which internalises the check required to safely use the old one, plus
431 the allocate and data copy operations duplicated in both (!) of the
432 extant use locations.
434 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
435 modifier. This matches the restriction on the commandline.
437 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
438 Previously only the last row was returned.
440 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
441 we assumed that tags in the header were well-formed, and parsed the
442 element content after inspecting only the first char of the tag.
443 Assumptions at that stage could crash the receive process on malformed
446 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
447 While running the DKIM ACL we operate on the Permanent memory pool so that
448 variables created with "set" persist to the DATA ACL. Also (at any time)
449 DNS lookups that fail create cache records using the Permanent pool. But
450 expansions release any allocations made on the current pool - so a dnsdb
451 lookup expansion done in the DKIM ACL releases the memory used for the
452 DNS negative-cache, and bad things result. Solution is to switch to the
453 Main pool for expansions.
454 While we're in that code, add checks on the DNS cache during store_reset,
455 active in the testsuite.
456 Problem spotted, and debugging aided, by Wolfgang Breyha.
458 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
459 When none of the hosts presented to a transport match an already-open
460 connection, close it and proceed with the list. Previously we would
461 queue the message. Spotted by Lena with Yahoo, probably involving
464 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
465 Previously a spurious "250 OK id=" response was appended to the proper
468 JH/08 The "support for" informational output now, which built with Content
469 Scanning support, has a line for the malware scanner interfaces compiled
470 in. Interface can be individually included or not at build time.
472 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
473 by the template makefile "src/EDITME". The "STREAM" support for an older
474 ClamAV interface method is removed.
476 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
477 rows affected is given instead).
479 JH/11 The runtime Berkeley DB library version is now additionally output by
480 "exim -d -bV". Previously only the compile-time version was shown.
482 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
483 SMTP connection. Previously, when one had more receipients than the
484 first, an abortive onward connection was made. Move to full support for
485 multiple onward connections in sequence, handling cutthrough connection
486 for all multi-message initiating connections.
488 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
489 routers. Previously, a multi-recipient message would fail to match the
490 onward-connection opened for the first recipient, and cause its closure.
492 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
493 a timeout on read on a GnuTLS initiating connection, resulting in the
494 initiating connection being dropped. This mattered most when the callout
495 was marked defer_ok. Fix to keep the two timeout-detection methods
498 JH/15 Relax results from ACL control request to enable cutthrough, in
499 unsupported situations, from error to silently (except under debug)
500 ignoring. This covers use with PRDR, frozen messages, queue-only and
503 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
505 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
506 metadata, resulting in a crash in free().
508 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
509 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
510 Broken also in d185889f4, with init system revamp.
512 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
513 Previously we abruptly closed the connection after reading a malware-
514 found indication; now we go on to read the "scan ok" response line,
517 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
518 ACL. Previously, a crash would result.
520 JH/19 Speed up macro lookups during configuration file read, by skipping non-
521 macro text after a replacement (previously it was only once per line) and
522 by skipping builtin macros when searching for an uppercase lead character.
524 JH/20 DANE support moved from Experimental to mainline. The Makefile control
525 for the build is renamed.
527 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
528 was allocated for every new TLS startup, meaning one per message. Fix
529 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
531 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
532 reported the original. Fix to report (as far as possible) the ACL
533 result replacing the original.
535 JH/23 Fix memory leak during multi-message connections using STARTTLS under
536 OpenSSL. Certificate information is loaded for every new TLS startup,
537 and the resources needed to be freed.
539 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
541 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
544 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
545 DATA response info to the (existing) per-recipient response info for
546 the "C=" log element. It can have useful tracking info from the
547 destination system. Patch from Simon Arlott.
549 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
550 length value. Previously this would segfault.
552 HS/02 Support Avast multiline protoocol, this allows passing flags to
553 newer versions of the scanner.
555 JH/28 Ensure that variables possibly set during message acceptance are marked
556 dead before release of memory in the daemon loop. This stops complaints
557 about them when the debug_store option is enabled. Discovered specifically
558 for sender_rate_period, but applies to a whole set of variables.
559 Do the same for the queue-runner and queue-list loops, for variables set
560 from spool message files. Do the same for the SMTP per-message loop, for
561 certain variables indirectly set in ACL operations.
563 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
564 as a multi-recipient message from a mailinglist manager). The coding had
565 an arbitrary cutoff number of characters while checking for more input;
566 enforced by writing a NUL into the buffer. This corrupted long / fast
567 input. The problem was exposed more widely when more pipelineing of SMTP
568 responses was introduced, and one Exim system was feeding another.
569 The symptom is log complaints of SMTP syntax error (NUL chars) on the
570 receiving system, and refused recipients seen by the sending system
571 (propating to people being dropped from mailing lists).
572 Discovered and pinpointed by David Carter.
574 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
575 replaced by the ${authresults } expansion.
577 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
579 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
580 allows proper process termination in container environments.
582 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
583 Previously the "final dot" had a newline after it; ensure it is CR,LF.
585 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
586 and "err_perm", deprecated since 4.83 when the RFC-defined words
587 "temperror" and "permerror" were introduced.
589 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
590 transport-filters or DKIM-signing. The restriction was lost in the
591 consolidation of verify-callout and delivery SMTP handling.
592 Extend the restriction to also cover ARC-signing.
594 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
595 in defer=pass mode supply a 450 to the initiator. Previously the message
598 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
599 tls_require_ciphers is used as before.
601 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
602 "pass_unscanned". Only tmpfails from the scanner are written to
603 the paniclog, as they may require admin intervention (permission
604 denied, license issues). Other scanner errors (like decompression
605 bombs) do not cause a paniclog entry.
607 JH/36 Fix reinitialisation of DKIM logging variable between messages.
608 Previously it was possible to log spurious information in receive log
611 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
612 triggered odd behaviour from Outlook Express clients.
614 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
617 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
618 since the IETF WG has not yet settled on that versus the original
619 "bare" representation.
621 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
622 Previously the millisecond value corrupted the output.
623 Fix also for syslog_pid=no and log_selector +pid, for which the pid
624 corrupted the output.
630 JH/01 Rework error string handling in TLS interface so that the caller in
631 more cases is responsible for logging. This permits library-sourced
632 string to be attached to addresses during delivery, and collapses
633 pairs of long lines into single ones.
635 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
636 during configuration. Wildcards are allowed and expanded.
638 JH/02 Rework error string handling in DKIM to pass more info back to callers.
639 This permits better logging.
641 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
642 do not close it down and have the child transport start it up again on
643 the passed-on TCP connection. Instead, proxy the child (and any
644 subsequent ones) for TLS via a unix-domain socket channel. Logging is
645 affected: the continued delivery log lines do not have any DNSSEC, TLS
646 Certificate or OCSP information. TLS cipher information is still logged.
648 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
649 identical IP addresses on different listening ports. Will also affect
652 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
653 add noisy ifdef guards to special-case this sillyness.
654 Patch from Bernd Kuhls.
656 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
657 than 255 are no longer allowed.
659 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
660 Disable the session-cache too, which might reduce our load. Since we
661 currrectly use a new context for every connection, both as server and
662 client, there is no benefit for these.
663 GnuTLS appears to not support tickets server-side by default (we don't
664 call gnutls_session_ticket_enable_server()) but client side is enabled
665 by default on recent versions (3.1.3 +) unless the PFS priority string
668 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
669 <https://reproducible-builds.org/specs/source-date-epoch/>.
671 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
672 the check for any unsuccessful recipients did not notice the limit, and
673 erroneously found still-pending ones.
675 JH/08 Pipeline CHUNKING command and data together, on kernels that support
676 MSG_MORE. Only in-clear (not on TLS connections).
678 JH/09 Avoid using a temporary file during transport using dkim. Unless a
679 transport-filter is involved we can buffer the headers in memory for
680 creating the signature, and read the spool data file once for the
681 signature and again for transmission.
683 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
684 4.77 as the kernel support then wasn't solid, having issues in 64bit
685 mode. Now, it's been long enough. Add support for FreeBSD also.
687 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
688 case where the routing stage had gathered several addresses to send to
689 a host before calling the transport for the first, we previously failed
690 to close down TLS in the old transport process before passing the TCP
691 connection to the new process. The new one sent a STARTTLS command
692 which naturally failed, giving a failed delivery and bloating the retry
693 database. Investigation and fix prototype from Wolfgang Breyha.
695 JH/12 Fix check on SMTP command input synchronisation. Previously there were
696 false-negatives in the check that the sender had not preempted a response
697 or prompt from Exim (running as a server), due to that code's lack of
698 awareness of the SMTP input buffering.
700 PP/04 Add commandline_checks_require_admin option.
701 Exim drops privileges sanely, various checks such as -be aren't a
702 security problem, as long as you trust local users with access to their
703 own account. When invoked by services which pass untrusted data to
704 Exim, this might be an issue. Set this option in main configuration
705 AND make fixes to the calling application, such as using `--` to stop
708 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
709 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
710 responses to those, into a single TLS record each way (this usually means
711 a single packet). As a side issue, smtp_enforce_sync now works on TLS
714 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
715 affects you only if you're dancing at the edge of the param size limits.
716 If you are, and this message makes sense to you, then: raise the
717 configured limit or use OpenSSL 1.1. Nothing we can do for older
720 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
721 cmdline element to get the documented default one. Previously it was
724 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
725 are unsuable for non-message-specific reasons - eg. connection timeout,
726 banner-time rejection.
728 JH/16 Fix logging of delivery remote port, when specified by router, under
731 PP/06 Repair manualroute's ability to take options in any order, even if one
732 is the name of a transport.
735 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
737 JH/17 Change the list-building routines interface to use the expanding-string
738 triplet model, for better allocation and copying behaviour.
740 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
741 Previously it was constructed the first time a possibly-matching string
742 was met in the configuration file input during startup; now it is done
745 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
746 compatible one, to avoid the (poorly documented) possibility of a config
747 file in the working directory redirecting the DB files, possibly correpting
748 some existing file. CVE-2017-10140 assigned for BDB.
750 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
751 cache-hot. Previously, although the result was properly cached, the
752 initial verify call returned a defer.
754 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
755 the main verify for receipient in uncached-mode.
757 JH/22 Retire historical build files to an "unsupported" subdir. These are
758 defined as "ones for which we have no current evidence of testing".
760 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
761 if present. Previously it was ignored.
763 JH/24 Start using specified-initialisers in C structure init coding. This is
764 a C99 feature (it's 2017, so now considered safe).
766 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
767 if was a fixed-sized field and bitmask ops via macros; it is now more
770 PP/07 GitHub PR 56: Apply MariaDB build fix.
771 Patch provided by Jaroslav Škarvada.
773 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
774 during Coverity cleanups [4.87 JH/47]
775 Diagnosis and fix provided by Michael Fischer v. Mollard.
777 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
778 the right size to place the terminating semicolon on its own folded
779 line, the header hash was calculated to an incorrect value thanks to
780 the (relaxed) space the fold became.
782 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
783 and confused the parent.
785 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
786 which could crash as a result. This could lead to undeliverable messages.
788 JH/28 Logging: "next input sent too soon" now shows where input was truncated
791 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
792 matters on fast-turnover and PID-randomising systems, which were getting
793 out-of-order delivery.
795 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
796 a possibly-overlapping copy. The symptom was that "Remote host closed
797 connection in response to HELO" was logged instead of the actual 4xx
800 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
801 Previously only that bufferd was discarded, resulting in SYMTP command
804 JH/32 DKIM: when a message has multiple signatures matching an identity given
805 in dkim_verify_signers, run the dkim acl once for each. Previously only
806 one run was done. Bug 2189.
808 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
809 "panic the current process" to "deliberately defer". The panic log is
810 still written with the problem list name; the mail and reject logs now
811 get a temp-reject line for the message that was being handled, saying
812 something like "domains check lookup or other defer". The SMTP 451
813 message is still "Temporary local problem".
815 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
816 A crafted sequence of BDAT commands could result in in-use memory beeing
817 freed. CVE-2017-16943.
819 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
820 from SMTP input. Previously it was always done; now only done for DATA
821 and not BDAT commands. CVE-2017-16944.
823 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
824 to the message (such as an overlong header line). Previously this was
825 not done and we did not exit BDAT mode. Followon from the previous item
826 though a different problem.
832 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
833 than -2003 did; needs libidn2 in addition to libidn.
835 JH/02 The path option on a pipe transport is now expanded before use.
837 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
838 Patch provided by "Björn", documentation fix added too.
840 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
841 missing a wire-to-host endian conversion.
843 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
844 close after a BDAT command line could be taken as a following command,
845 giving a synch failure. Fix by only checking for synch immediately
846 before acknowledging the chunk.
848 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
849 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
850 macro. Patches provided by Josh Soref.
852 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
853 Previously we did not; the RFC seems ambiguous and VRFY is not listed
854 by IANA as a service extension. However, John Klensin suggests that we
857 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
858 the dkim code may be unix-mode line endings rather than smtp wire-format
859 CRLF, so prepend a CR to any bare LF.
861 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
862 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
864 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
865 the first were themselves being wrongly included in the feed into dkim
866 processing; with most chunk sizes in use this resulted in an incorrect
867 body hash calculated value.
869 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
870 DKIM signature block, for verification. Although advised against by
871 standards it is specifically not ruled illegal.
873 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
875 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
876 missing a body hash (the bh= tag).
878 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
879 It seems that HAProxy sends the Proxy Protocol information in clear and
880 only then does a TLS startup, so do the same.
882 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
883 TCP connections (such as for Spamd) unless the daemon successfully set
884 Fast Open mode on its listening sockets. This fixes breakage seen on
885 too-old kernels or those not configured for Fast Open, at the cost of
886 requiring both directions being enabled for TFO, and TFO never being used
887 by non-daemon-related Exim processes.
889 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
890 endings, at least on the first header line. Try to canonify any that get
891 past that check, despite the cost.
893 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
894 now limited to an arbitrary five deep, while parsing addresses with the
895 strip_excess_angle_brackets option enabled.
897 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
898 instead leave the unprompted TLS handshake in socket buffer for the
899 TLS library to consume.
901 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
903 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
905 JH/16 Drop variables when they go out of scope. Memory management drops a whole
906 region in one operation, for speed, and this leaves assigned pointers
907 dangling. Add checks run only under the testsuite which checks all
908 variables at a store-reset and panics on a dangling pointer; add code
909 explicitly nulling out all the variables discovered. Fixes one known
910 bug: a transport crash, where a dangling pointer for $sending_ip_address
911 originally assigned in a verify callout, is re-used.
913 PP/06 Drop '.' from @INC in various Perl scripts.
915 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
917 PP/08 Reduce a number of compilation warnings under clang; building with
918 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
919 should be warning-free.
921 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
923 HS/01 Fix portability problems introduced by PP/08 for platforms where
924 realloc(NULL) is not equivalent to malloc() [SunOS et al].
926 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
927 chunk. This allows us to accept broken chunked messages. We need a more
928 general solution here.
930 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
931 already-broken messages in the queue.
933 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
935 JH/19 Fix reference counting bug in routing-generated-address tracking.
941 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
942 supports it and a size is available (ie. the sending peer gave us one).
944 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
945 years of being deprecated). The replacements are the ACLs
946 acl_smtp_mime and acl_not_smtp_mime.
948 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
949 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
950 or even an in-clear connection were permitted. Now, if the host lookup
951 was dnssec and dane was requested then the host is only used if the
952 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
953 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
954 if one fails this test.
955 This means that a poorly-configured remote DNS will make it incommunicado;
956 but it protects against a DNS-interception attack on it.
958 JH/04 Bug 1810: make continued-use of an open smtp transport connection
959 non-noisy when a race steals the message being considered.
961 JH/05 If main configuration option tls_certificate is unset, generate a
962 self-signed certificate for inbound TLS connections.
964 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
965 in rewrites and routers.
967 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
968 and logged a warning sing 4.83; now they are a configuration file error.
970 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
971 (lacking @domain). Apply the same qualification processing as RCPT.
973 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
975 JH/10 Support ${sha256:} applied to a string (as well as the previous
978 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
979 a cutthrough deliver is pending, as we always want to make a connection.
980 This also avoids re-routing the message when later placing the cutthrough
981 connection after a verify cache hit.
982 Do not update it with the verify result either.
984 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
985 when routing results in more than one destination address.
987 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
988 signing (which inhibits the cutthrough capability). Previously only
989 the presence of an option was tested; now an expansion evaluating as
990 empty is permissible (obviously it should depend only on data available
991 when the cutthrough connection is made).
993 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
994 the relevant preceding SMTP command did not note the pipelining mode.
996 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
997 Previously they were not counted.
999 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1000 as one having no matching records. Previously we deferred the message
1001 that needed the lookup.
1003 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1004 distinguished as "(=".
1006 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1007 for missing MX records. Previously it only worked for missing A records.
1009 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1011 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1012 after the data-go-ahead and data-ack. Patch from Jason Betts.
1014 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1015 even for a "none" policy. Patch from Tony Meyer.
1017 JH/22 Fix continued use of a connection for further deliveries. If a port was
1018 specified by a router, it must also match for the delivery to be
1021 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1022 When one of the recipients of a message was unsuitable for the connection
1023 (has no matching addresses), we lost track of needing to mark it
1024 deferred. As a result mail would be lost.
1026 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1028 JH/25 Decoding ACL controls is now done using a binary search; the source code
1029 takes up less space and should be simpler to maintain. Merge the ACL
1030 condition decode tables also, with similar effect.
1032 JH/26 Fix problem with one_time used on a redirect router which returned the
1033 parent address unchanged. A retry would see the parent address marked as
1034 delivered, so not attempt the (identical) child. As a result mail would
1037 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1038 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1039 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1042 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1043 Main config options check_{log,spool}_{inodes,space} are now
1044 100 inodes, 10MB unless set otherwise in the configuration.
1046 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1047 Previously it only applied to the main-section connection policy
1050 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1052 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1053 by me. Added RFC7919 DH primes as an alternative.
1055 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1056 are not in the system include path.
1058 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1059 GnuTLS, when a session startup failed (eg because the client disconnected)
1060 Exim did stdio operations after fclose. This was exposed by a recent
1061 change which nulled out the file handle after the fclose.
1063 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1064 signed directly by the cert-signing cert, rather than an intermediate
1065 OCSP-signing cert. This is the model used by LetsEncrypt.
1067 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1069 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1070 an incoming connection.
1072 HS/02 Bug 1802: Do not half-close the connection after sending a request
1075 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1076 fallback to "prime256v1".
1078 JH/34 SECURITY: Use proper copy of DATA command in error message.
1079 Could leak key material. Remotely exploitable. CVE-2016-9963.
1085 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1086 and 3.4.4 - once the server is enabled to respond to an OCSP request
1087 it does even when not requested, resulting in a stapling non-aware
1088 client dropping the TLS connection.
1090 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1091 support variable-length bit vectors. No functional change.
1093 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1094 The I= interface field on outgoing lines is now after the H= remote
1095 host field, same as incoming lines. There is a separate
1096 outgoing_interface log selector which allows you to disable the
1099 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1100 If not running log_selector +smtp_connection the mainlog would be held
1101 open indefinitely after a "too many connections" event, including to a
1102 deleted file after a log rotate. Leave the per net connection logging
1103 leaving it open for efficiency as that will be quickly detected by the
1104 check on the next write.
1106 HS/01 Bug 1671: Fix post transport crash.
1107 Processing the wait-<transport> messages could crash the delivery
1108 process if the message IDs didn't exist for some reason. When
1109 using 'split_spool_directory=yes' the construction of the spool
1110 file name failed already, exposing the same netto behaviour.
1112 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1113 mime_regex ACL conditions.
1115 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1116 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1117 to HELO, local diagnostic string.
1119 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1120 log line to debug. Even when configured with a tls authenticator many
1121 client connections are expected to not authenticate in this way, so
1122 an authenticate fail is not an error.
1124 HS/02 Add the Exim version string to the process info. This way exiwhat
1125 gives some more detail about the running daemon.
1127 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1128 matter for fast-change records such as DNSBLs.
1130 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1131 retry record, even if constant. There may be multiple transports with
1132 different interface settings and the retry behaviour needs to be kept
1135 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1137 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1139 JH/10 Bug 840: fix log_defer_output option of pipe transport
1141 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1142 hosts_randomize. This matters a lot when combined with mua_wrapper.
1144 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1145 ${quote_pgsql:<string>} operator.
1147 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1149 JH/14 Change header file name for freeradius-client. Relevant if compiling
1150 with Radius support; from the Gentoo tree and checked under Fedora.
1152 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1154 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1155 option transport_filter, meaning no filtering.
1157 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1159 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1160 defaults to "*" (all hosts). The variable is now available when not built
1161 with TLS, default unset, mainly to enable keeping the testsuite sane.
1162 If a server certificate is not supplied (via tls_certificate) an error is
1163 logged, and clients will find TLS connections fail on startup. Presumably
1164 they will retry in-clear.
1165 Packagers of Exim are strongly encouraged to create a server certificate
1166 at installation time.
1168 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1169 with the $config_file variable.
1171 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1172 in transport context, after the attempt, and per-recipient. The latter type
1173 is per host attempted. The event data is the error message, and the errno
1174 information encodes the lookup type (A vs. MX) used for the (first) host,
1175 and the trailing two digits of the smtp 4xx response.
1177 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1178 to write to mainlog (or rejectlog, paniclog) in the window between file
1179 creation and permissions/ownership being changed. Particularly affects
1180 installations where exicyclog is run as root, rather than exim user;
1181 result is that the running daemon panics and dies.
1183 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1185 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1186 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1187 "pri" and "weight". Note that the previous implicit priority given by the
1188 list order is no longer honoured.
1190 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1191 for DKIM processing.
1193 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1194 by defining SUPPORT_SOCKS.
1196 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1197 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1198 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1199 variables are renamed to proxy_{local,external}_{address,port}.
1201 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1202 for a build by defining SUPPORT_I18N
1204 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1205 of the query string, and make ${quote_redis:} do that quoting.
1207 JH/29 Move Events support from Experimental to mainline, enabled by default
1208 and removable for a build by defining DISABLE_EVENT.
1210 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1212 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1213 cached by the daemon.
1215 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1216 by defining LOOKUP_REDIS. The libhiredis library is required.
1218 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1219 keys are given for lookup.
1221 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1222 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1223 only supported when built with TLS support. The PolarSSL SHA routines
1224 are still used when the TLS library is too old for convenient support.
1226 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1227 openssl_options), for security. OpenSSL forces this from version 1.1.0
1228 server-side so match that on older versions.
1230 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1231 allocation for $value could be released as the expansion processing
1232 concluded, but leaving the global pointer active for it.
1234 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1235 and to use the domains and local_parts ACL conditions.
1237 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1238 incorrectly not doubled on cutthrough transmission, hence seen as a
1239 body-termination at the receiving system - resulting in truncated mails.
1240 Commonly the sender saw a TCP-level error, and retransmitted the message
1241 via the normal store-and-forward channel. This could result in duplicates
1242 received - but deduplicating mailstores were liable to retain only the
1243 initial truncated version.
1245 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1247 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1249 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1250 we're in there, support oversigning also; bug 1309.
1252 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1254 HS/04 Add support for keep_environment and add_environment options.
1256 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1257 either intentional arithmetic overflow during PRNG, or testing config-
1260 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1261 delivery resulted in actual delivery. Cancel cutthrough before DATA
1264 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1265 rejects a recipient: pass the reject to the originator.
1267 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1268 Many were false-positives and ignorable, but it's worth fixing the
1271 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1272 for the new environment-manipulation done at startup. Move the routines
1273 from being local to tls.c to being global via the os.c file.
1275 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1276 an extract embedded as result-arg for a map, the first arg for extract
1277 is unavailable so we cannot tell if this is a numbered or keyed
1278 extraction. Accept either.
1284 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1287 JH/02 The smtp transport option "multi_domain" is now expanded.
1289 JH/03 The smtp transport now requests PRDR by default, if the server offers
1292 JH/04 Certificate name checking on server certificates, when exim is a client,
1293 is now done by default. The transport option tls_verify_cert_hostnames
1294 can be used to disable this per-host. The build option
1295 EXPERIMENTAL_CERTNAMES is withdrawn.
1297 JH/05 The value of the tls_verify_certificates smtp transport and main options
1298 default to the word "system" to access the system default CA bundle.
1299 For GnuTLS, only version 3.0.20 or later.
1301 JH/06 Verification of the server certificate for a TLS connection is now tried
1302 (but not required) by default. The verification status is now logged by
1303 default, for both outbound TLS and client-certificate supplying inbound
1306 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1309 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1310 Status Notification (bounce) messages are now MIME format per RFC 3464.
1311 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1312 under the control of the dsn_advertise_hosts option, and routers may
1313 have a dsn_lasthop option.
1315 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1316 default, modifiable by a malware= option. The list separator for
1317 the options can now be changed in the usual way. Bug 68.
1319 JH/10 The smtp_receive_timeout main option is now expanded before use.
1321 JH/11 The incoming_interface log option now also enables logging of the
1322 local interface on delivery outgoing connections.
1324 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1325 if the interface and destination host and port all match.
1327 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1330 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1331 Patch from Andrew Lewis.
1333 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1334 now supports optional time-restrictions, weighting, and priority
1335 modifiers per server. Patch originally by <rommer@active.by>.
1337 JH/16 The spamd_address main option now supports a mixed list of local
1338 and remote servers. Remote servers can be IPv6 addresses, and
1339 specify a port-range.
1341 JH/17 Bug 68: The spamd_address main option now supports an optional
1342 timeout value per server.
1344 JH/18 Bug 1581: Router and transport options headers_add/remove can
1345 now have the list separator specified.
1347 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1350 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1353 JH/21 Support for the A6 type of dns record is withdrawn.
1355 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1356 rather than the verbs used.
1358 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1359 from 255 to 1024 chars.
1361 JH/24 Verification callouts now attempt to use TLS by default.
1363 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1364 are generic router options now. The defaults didn't change.
1366 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1367 Original patch from Alexander Shikoff, worked over by JH.
1369 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1370 files if ZCAT_COMMAND is not executable.
1372 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1374 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1376 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1377 Normally benign, it bites when the pair was led to by a CNAME;
1378 modern usage is to not canonicalize the domain to a CNAME target
1379 (and we were inconsistent anyway for A-only vs AAAA+A).
1381 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1383 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1384 when evaluating $sender_host_dnssec.
1386 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1387 $sender_helo_dnssec variable.
1389 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1391 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1393 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1395 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1396 documented as working, but never had. Support all but $spam_report.
1398 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1399 added for tls authenticator.
1401 HS/03 Add perl_taintmode main config option
1407 TL/01 When running the test suite, the README says that variables such as
1408 no_msglog_check are global and can be placed anywhere in a specific
1409 test's script, however it was observed that placement needed to be near
1410 the beginning for it to behave that way. Changed the runtest perl
1411 script to read through the entire script once to detect and set these
1412 variables, reset to the beginning of the script, and then run through
1413 the script parsing/test process like normal.
1415 TL/02 The BSD's have an arc4random API. One of the functions to induce
1416 adding randomness was arc4random_stir(), but it has been removed in
1417 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1418 function when detected.
1420 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1421 cause callback expansion.
1423 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1424 syntax errors in an expansion can be treated as a string instead of
1425 logging or causing an error, due to the internal use of bool_lax
1426 instead of bool when processing it.
1428 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1429 server certificates when making smtp deliveries.
1431 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1433 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1435 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1437 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1438 Merged patch from Sebastian Wiedenroth.
1440 JH/05 Fix results-pipe from transport process. Several recipients, combined
1441 with certificate use, exposed issues where response data items split
1442 over buffer boundaries were not parsed properly. This eventually
1443 resulted in duplicates being sent. This issue only became common enough
1444 to notice due to the introduction of connection certificate information,
1445 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1447 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1448 size buffer was used, resulting in syntax errors when an expansion
1451 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1452 version 3.3.6 or later.
1454 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1455 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1456 both become "event_action", the variables become $event_name, $event_data
1457 and $event_defer_errno. There is a new variable $verify_mode, usable in
1458 routers, transports and related events. The tls:cert event is now also
1459 raised for inbound connections, if the main configuration event_action
1462 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1463 early OCSP support, but no stapling (appears to be less than 1.0.0).
1465 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1466 server certificate names available under the smtp transport option
1467 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1470 JH/10 Time-related extraction expansions from certificates now use the main
1471 option "timezone" setting for output formatting, and are consistent
1472 between OpenSSL and GnuTLS compilations. Bug 1541.
1474 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1475 encoded parameter in the incoming message. Bug 1558.
1477 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1478 include certificate info, eximon was claiming there were spoolfile
1481 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1483 JH/14 Log delivery-related information more consistently, using the sequence
1484 "H=<name> [<ip>]" wherever possible.
1486 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1487 are problematic for Debian distribution, omit them from the release
1490 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1492 JH/16 Fix string representation of time values on 64bit time_t architectures.
1495 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1496 output list separator was used.
1501 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1502 checkers that were complaining about end of non-void function with no
1505 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1506 This was a regression introduced in 4.83 by another bugfix.
1508 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1510 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1511 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1517 TF/01 Correctly close the server side of TLS when forking for delivery.
1519 When a message was received over SMTP with TLS, Exim failed to clear up
1520 the incoming connection properly after forking off the child process to
1521 deliver the message. In some situations the subsequent outgoing
1522 delivery connection happened to have the same fd number as the incoming
1523 connection previously had. Exim would try to use TLS and fail, logging
1524 a "Bad file descriptor" error.
1526 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1527 utilities have not been installed.
1529 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1530 temporary space as the ACL may create new global variables.
1532 TL/01 LDAP support uses per connection or global context settings, depending
1533 upon the detected version of the libraries at build time.
1535 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1536 to extract and use the src ip:port in logging and expansions as if it
1537 were a direct connection from the outside internet. PPv2 support was
1538 updated based on HAProxy spec change in May 2014.
1540 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1542 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1543 Properly escape header and check for NULL return.
1545 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1548 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1550 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1551 characters in header names, implemented as a verify condition.
1552 Contributed by Michael Fischer v. Mollard.
1554 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1555 results permerror and temperror. Previous values are deprecated but
1556 still accepted. In a future release, err_perm and err_temp will be
1557 completely removed, which will be a backward incompatibility if the
1558 ACL tests for either of these two old results. Patch contributed by
1559 user bes-internal on the mailing list.
1561 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1563 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1564 selectors, in both main and reject logs.
1566 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1569 JH/07 Add malware type "sock" for talking to simple daemon.
1571 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1573 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1574 routers/transports under cutthrough routing.
1576 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1577 numbers. Touch up "bool" conditional to keep the same definition.
1579 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1581 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1583 JH/12 Expand items in router/transport headers_add or headers_remove lists
1584 individually rather than the list as a whole. Bug 1452.
1586 Required for reasonable handling of multiple headers_ options when
1587 they may be empty; requires that headers_remove items with embedded
1588 colons must have them doubled (or the list-separator changed).
1590 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1591 view the policy declared in the DMARC record. Currently, $dmarc_status
1592 is a combined value of both the record presence and the result of the
1595 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1597 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1598 dnslookup router and the smtp transport (applying to the forward
1601 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1602 of ldap servers used for a specific lookup. Patch provided by Heiko
1605 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1606 New variable $lookup_dnssec_authenticated for observability.
1608 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1609 Patch submitted by Lars Timman.
1611 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1613 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1614 Requires trusted mode and valid format message id, aborts otherwise.
1615 Patch contributed by Heiko Schlichting.
1617 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1618 certextract with support for various fields. Bug 1358.
1620 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1621 is requested by default, modifiable by smtp transport option
1624 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1625 operate on certificate variables to give certificate fingerprints
1626 Also new ${sha256:cert_variable}.
1628 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1630 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1631 Christian Aistleitner.
1633 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1635 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1636 file. Patch from Wolfgang Breyha.
1638 JH/25 Expand the coverage of the delivery $host and $host_address to
1639 client authenticators run in verify callout. Bug 1476.
1641 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1642 align with daemon_smtp_ports. Bug 72.
1644 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1645 support and error reporting did not work properly.
1647 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1648 and is readable. Patch from Andrew Colin Kissa.
1650 TL/14 Enhance documentation of ${run expansion and how it parses the
1651 commandline after expansion, particularly in the case when an
1652 unquoted variable expansion results in an empty value.
1654 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1656 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1657 by renaming away from it. Addresses GH issue 15, reported by
1660 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1662 TL/15 SECURITY: prevent double expansion in math comparison functions
1663 (can expand unsanitized data). Not remotely exploitable.
1670 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1672 PP/02 Make -n do something, by making it not do something.
1673 When combined with -bP, the name of an option is not output.
1675 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1678 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1679 $sender_host_name and config options to manage this, and basic check
1682 PP/05 DSCP support for outbound connections and control modifier for inbound.
1684 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1685 (Only plugin which currently uses this is kerberos4, which nobody should
1686 be using, but we should make it available and other future plugins might
1687 conceivably use it, even though it would break NAT; stuff *should* be
1688 using channel bindings instead).
1690 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1691 name; added for Sendmail compatibility; requires admin caller.
1692 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1693 just ignore it); requires trusted caller.
1694 Also parse but ignore: -Ac -Am -X<logfile>
1697 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1699 TL/02 Add +smtp_confirmation as a default logging option.
1701 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1702 Patch by Magnus Holmgren from 2007-02-20.
1704 TL/04 Bugzilla 1281 - Spec typo.
1705 Bugzilla 1283 - Spec typo.
1706 Bugzilla 1290 - Spec grammar fixes.
1708 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1710 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1712 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1713 mailing list by by Dmitry Isaikin.
1715 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1717 JH/02 Support "G" suffix to numbers in ${if comparisons.
1719 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1721 NM/01 Bugzilla 1197 - Spec typo
1722 Bugzilla 1196 - Spec examples corrections
1724 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1726 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1727 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1730 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1731 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1733 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1734 "acl {{name}{arg}...}", and optional args on acl condition
1737 JH/05 Permit multiple router/transport headers_add/remove lines.
1739 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1741 JH/07 Avoid using a waiting database for a single-message-only transport.
1742 Performance patch from Paul Fisher. Bugzilla 1262.
1744 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1747 JH/09 Add $headers_added variable, with content from use of ACL modifier
1748 add_header (but not yet added to the message). Bugzilla 199.
1750 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1751 Pulled from Bugzilla 817 by Wolfgang Breyha.
1753 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1755 (nb: this is the same fix as in Exim 4.80.1)
1757 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1760 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1763 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1765 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1766 advertise SMTP AUTH mechanism to us, instead of a generic
1767 protocol violation error. Also, make Exim more robust to bad
1768 data from the Dovecot auth socket.
1770 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1772 When a queue runner is handling a message, Exim first routes the
1773 recipient addresses, during which it prunes them based on the retry
1774 hints database. After that it attempts to deliver the message to
1775 any remaining recipients. It then updates the hints database using
1778 So if a recipient address works intermittently, it can get repeatedly
1779 deferred at routing time. The retry hints record remains fresh so the
1780 address never reaches the final cutoff time.
1782 This is a fairly common occurrence when a user is bumping up against
1783 their storage quota. Exim had some logic in its local delivery code
1784 to deal with this. However it did not apply to per-recipient defers
1785 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1787 This change adds a proper retry rule check during routing so that the
1788 final cutoff time is checked against the message's age. We only do
1789 this check if there is an address retry record and there is not a
1790 domain retry record; this implies that previous attempts to handle
1791 the address had the retry_use_local_parts option turned on. We use
1792 this as an approximation for the destination being like a local
1793 delivery, as in LMTP.
1795 I suspect this new check makes the old local delivery cutoff check
1796 redundant, but I have not verified this so I left the code in place.
1798 TF/02 Correct gecos expansion when From: is a prefix of the username.
1800 Test 0254 submits a message to Exim with the header
1804 When I ran the test suite under the user fanf2, Exim expanded
1805 the header to contain my full name, whereas it should have added
1806 a Resent-Sender: header. It erroneously treats any prefix of the
1807 username as equal to the username.
1809 This change corrects that bug.
1811 GF/01 DCC debug and logging tidyup
1812 Error conditions log to paniclog rather than rejectlog.
1813 Debug lines prefixed by "DCC: " to remove any ambiguity.
1815 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1817 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1818 Bug spotted by Jeremy Harris; was flawed since initial commit.
1819 Would have resulted in OCSP responses post-SNI triggering an Exim
1820 NULL dereference and crash.
1822 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1824 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1825 Bug detection, analysis and fix by Samuel Thibault.
1826 Bugzilla 1331, Debian bug #698092.
1828 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1830 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1831 Server implementation by Todd Lyons, client by JH.
1832 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1833 config variable "prdr_enable" controls whether the server
1834 advertises the facility. If the client requests PRDR a new
1835 acl_data_smtp_prdr ACL is called once for each recipient, after
1836 the body content is received and before the acl_smtp_data ACL.
1837 The client is controlled by both of: a hosts_try_prdr option
1838 on the smtp transport, and the server advertisement.
1839 Default client logging of deliveries and rejections involving
1840 PRDR are flagged with the string "PRDR".
1842 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1843 fclose(). Diagnosis by Todd Lyons.
1845 PP/17 Update configure.default to handle IPv6 localhost better.
1846 Patch by Alain Williams (plus minor tweaks).
1849 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1850 This is now consistent with GnuTLS, and is now documented: the
1851 previous undocumented portable approach to treating the option as
1852 unset was to force an expansion failure. That still works, and
1853 an empty string is now equivalent.
1855 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1856 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1857 not performing validation itself.
1859 PP/20 Added force_command boolean option to pipe transport.
1860 Patch from Nick Koston, of cPanel Inc.
1862 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1865 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1867 PP/21 Fix eximon continuous updating with timestamped log-files.
1868 Broken in a format-string cleanup in 4.80, missed when I repaired the
1869 other false fix of the same issue.
1870 Report and fix from Heiko Schlichting.
1873 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1874 Report from Prashanth Katuri.
1876 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1877 It's SecureTransport, so affects any MacOS clients which use the
1878 system-integrated TLS libraries, including email clients.
1880 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1881 using a MIME ACL for non-SMTP local injection.
1882 Report and assistance in diagnosis by Warren Baker.
1884 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1886 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1888 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1889 last id that failed so it may be referenced in subsequent ACL's.
1891 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1894 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1895 ldap library initialization, allowing self-signed CA's to be
1896 used. Also properly sets require_cert option later in code by
1897 using NULL (global ldap config) instead of ldap handle (per
1898 session). Bug diagnosis and testing by alxgomz.
1900 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1901 the src/util/ subdirectory.
1903 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1904 renamed to Transport Post Delivery Action by Jeremy Harris, as
1907 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1908 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1909 redis_servers = needs to be configured which will be used by the redis
1910 lookup. Patch from Warren Baker, of The Packet Hub.
1912 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1914 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1915 hostname or reverse DNS when processing a host list. Used suggestions
1916 from multiple comments on this bug.
1918 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1920 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1921 Missed a few lines, added it to make the runtest require no keyboard
1924 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1925 contains upper case chars. Make router use caseful_local_part.
1927 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1928 support when GnuTLS has been built with p11-kit.
1934 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1936 This, or similar/improved, will also be change PP/11 of 4.82.
1942 PP/01 Handle short writes when writing local log-files.
1943 In practice, only affects FreeBSD (8 onwards).
1944 Bugzilla 1053, with thanks to Dmitry Isaikin.
1946 NM/01 Bugzilla 949 - Documentation tweak
1948 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1951 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1953 PP/02 Implemented gsasl authenticator.
1955 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1957 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1958 `pkg-config foo` for cflags/libs.
1960 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1961 with rest of GSASL and with heimdal_gssapi.
1963 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1964 `pkg-config foo` for cflags/libs for the TLS implementation.
1966 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1967 properties get this fed in as external SSF. A number of robustness
1968 and debugging improvements to the cyrus_sasl authenticator.
1970 PP/08 cyrus_sasl server now expands the server_realm option.
1972 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1973 Patch by Jeremy Harris.
1975 PP/10 Added dbmjz lookup type.
1977 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1979 PP/12 MAIL args handles TAB as well as SP, for better interop with
1980 non-compliant senders.
1981 Analysis and variant patch by Todd Lyons.
1983 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1984 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1985 Patch from Dirk Mueller <dmueller@suse.com>
1987 PP/13 tls_peerdn now print-escaped for spool files.
1988 Observed some $tls_peerdn in wild which contained \n, which resulted
1989 in spool file corruption.
1991 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1992 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1993 or write after TLS renegotiation, which otherwise led to messages
1996 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1997 as a tracking header (ie: a signed header comes before the signature).
1998 Patch from Wolfgang Breyha.
2000 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2001 comma-sep list; embedded commas doubled.
2003 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2005 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2007 Report and patch from Dmitry Banschikov.
2009 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2010 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2011 failures. We appear to now support TLS1.1+ with Exim.
2013 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2014 lets Exim select keys and certificates based upon TLS SNI from client.
2015 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2016 before an outbound SMTP session. New log_selector, +tls_sni.
2018 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2019 NULL dereference. Report and patch from Alun Jones.
2021 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2022 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2023 Not seeing resolver debug output on NetBSD, but suspect this is a
2024 resolver implementation change.
2026 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2027 Left warnings. Added "eximon gdb" invocation mode.
2029 PP/21 Defaulting "accept_8bitmime" to true, not false.
2031 PP/22 Added -bw for inetd wait mode support.
2033 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2034 locate the relevant includes and libraries. Made this the default.
2036 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2037 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2039 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2040 This may cause build issues on older platforms.
2042 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2043 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2044 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2045 Added SNI support via GnuTLS too.
2046 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2048 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2050 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2051 Applied second patch from Janne, implementing suggestion to default
2052 multiple-strings-in-record handling to match SPF spec.
2054 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2056 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2057 read-only, out of scope).
2058 Patch from Wolfgang Breyha, report from Stuart Northfield.
2060 PP/29 Fix three issues highlighted by clang analyser static analysis.
2061 Only crash-plausible issue would require the Cambridge-specific
2062 iplookup router and a misconfiguration.
2063 Report from Marcin Mirosław.
2065 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2067 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2068 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2069 As part of this, removing so much warning spew let me fix some minor
2070 real issues in debug logging.
2072 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2073 assignment on my part. Fixed.
2075 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2076 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2077 Janne Snabb (who went above and beyond: thank you).
2079 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2080 string otherwise requires a connection and a bunch more work and it's
2081 relatively easy to get wrong. Should also expose TLS library linkage
2084 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2085 64-bit ${eval} (JH/03).
2087 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2088 GNU libc to support some of the 64-bit stuff, should not lead to
2089 conflicts. Defined before os.h is pulled in, so if a given platform
2090 needs to override this, it can.
2092 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2093 protection layer was required, which is not implemented.
2094 Bugzilla 1254, patch from Wolfgang Breyha.
2096 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2097 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2098 tls_dhparam take prime identifiers. Also unbreak combination of
2099 OpenSSL+DH_params+TLSSNI.
2101 PP/39 Disable SSLv2 by default in OpenSSL support.
2107 PP/01 Solaris build fix for Oracle's LDAP libraries.
2108 Bugzilla 1109, patch from Stephen Usher.
2110 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2112 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2115 TF/02 Fix a couple more cases where we did not log the error message
2116 when unlink() failed. See also change 4.74-TF/03.
2118 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2119 lock up or crash if it happened to be inside a call to libc when it
2120 got a SIGUSR1 from exiwhat.
2122 The SIGUSR1 handler appends the current process status to the process
2123 log which is later printed by exiwhat. It used to use the general
2124 purpose logging code to do this, but several functions it calls are
2125 not safe for signals.
2127 The new output code in the SIGUSR1 handler is specific to the process
2128 log, and simple enough that it's easy to inspect for signal safety.
2129 Removing some special cases also simplifies the general logging code.
2130 Removing the spurious timestamps from the process log simplifies
2133 TF/04 Improved ratelimit ACL condition.
2135 The /noupdate option has been deprecated in favour of /readonly which
2136 has clearer semantics. The /leaky, /strict, and /readonly update modes
2137 are mutually exclusive. The update mode is no longer included in the
2138 database key; it just determines when the database is updated. (This
2139 means that when you upgrade Exim will forget old rate measurements.)
2141 Exim now checks that the per_* options are used with an update mode that
2142 makes sense for the current ACL. For example, when Exim is processing a
2143 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2144 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2145 must specify per_mail/readonly. If you omit the update mode it defaults to
2146 /leaky where that makes sense (as before) or /readonly where required.
2148 The /noupdate option is now undocumented but still supported for
2149 backwards compatibility. It is equivalent to /readonly except that in
2150 ACLs where /readonly is required you may specify /leaky/noupdate or
2151 /strict/noupdate which are treated the same as /readonly.
2153 A useful new feature is the /count= option. This is a generalization
2154 of the per_byte option, so that you can measure the throughput of other
2155 aggregate values. For example, the per_byte option is now equivalent
2156 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2158 The per_rcpt option has been generalized using the /count= mechanism
2159 (though it's more complicated than the per_byte equivalence). When it is
2160 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2161 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2162 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2163 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2164 Note that using per_rcpt with a non-readonly update mode in more than
2165 one ACL will cause the recipients to be double-counted. (The per_mail
2166 and per_byte options don't have this problem.)
2168 The handling of very low rates has changed slightly. If the computed rate
2169 is less than the event's count (usually one) then this event is the first
2170 after a long gap. In this case the rate is set to the same as this event's
2171 count, so that the first message of a spam run is counted properly.
2173 The major new feature is a mechanism for counting the rate of unique
2174 events. The new per_addr option counts the number of different
2175 recipients that someone has sent messages to in the last time period. It
2176 behaves like per_rcpt if all the recipient addresses are different, but
2177 duplicate recipient addresses do not increase the measured rate. Like
2178 the /count= option this is a general mechanism, so the per_addr option
2179 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2180 example, measure the rate that a client uses different sender addresses
2181 with the options per_mail/unique=$sender_address. There are further
2182 details in the main documentation.
2184 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2186 TF/06 Removed a few PCRE remnants.
2188 TF/07 Automatically extract Exim's version number from tags in the git
2189 repository when doing development or release builds.
2191 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2192 Bugzilla 879. Patch from Paul Fisher.
2194 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2195 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2198 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2200 PP/05 Variable $av_failed, true if the AV scanner deferred.
2201 Bugzilla 1078. Patch from John Horne.
2203 PP/06 Stop make process more reliably on build failure.
2204 Bugzilla 1087. Patch from Heiko Schlittermann.
2206 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2207 Bugzilla 1089. Patch from Heiko Schlittermann.
2209 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2210 Bugzilla 1131. Patch from Holger Weiß.
2212 PP/09 Handle IPv6 addresses with SPF.
2213 Bugzilla 860. Patch from Wolfgang Breyha.
2215 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2217 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2220 PP/11 match_* no longer expand right-hand-side by default.
2221 New compile-time build option, EXPAND_LISTMATCH_RHS.
2222 New expansion conditions, "inlist", "inlisti".
2224 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2226 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2228 PP/14 fix log_write() format string regression from TF/03.
2229 Bugzilla 1152. Patch from Dmitry Isaikin.
2235 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2237 PP/02 Harmonised TLS library version reporting; only show if debugging.
2238 Layout now matches that introduced for other libraries in 4.74 PP/03.
2240 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2242 PP/04 New "dns_use_edns0" global option.
2244 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2247 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2248 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2250 TK/01 Updated PolarSSL code to 0.14.2.
2251 Bugzilla 1097. Patch from Andreas Metzler.
2253 PP/07 Catch divide-by-zero in ${eval:...}.
2254 Fixes bugzilla 1102.
2256 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2259 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2260 format-string attack -- SECURITY: remote arbitrary code execution.
2262 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2263 time unintentionally subject to list matching rules, letting the header
2264 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2265 arbitrary string expansion). This allowed for information disclosure.
2267 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2268 INT_MIN/-1 -- value coerced to INT_MAX.
2274 NM/01 Workaround for PCRE version dependency in version reporting
2277 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2278 This fixes portability to compilers other than gcc, notably
2279 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2281 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2282 makefiles for portability to HP-UX and POSIX correctness.
2284 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2285 Also via indented variable definition in the Makefile.
2286 (Debugging by Oliver Heesakkers).
2288 PP/02 Restore caching of spamd results with expanded spamd_address.
2289 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2291 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2292 Improves build reliability. Fix from: Frank Elsner
2294 NM/02 Fix wide character breakage in the rfc2047 coding
2295 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2297 NM/03 Allow underscore in dnslist lookups
2298 Fixes bug 1026. Patch from Graeme Fowler
2300 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2301 Code patches from Adam Ciarcinski of NetBSD.
2303 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2306 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2307 is logged, to avoid truncation. Patch from John Horne.
2309 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2310 Patch from Jakob Hirsch.
2312 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2313 SQL string expansion failure details.
2314 Patch from Andrey Oktyabrski.
2316 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2317 Patch from Simon Arlott.
2319 PP/09 New lookups functionality failed to compile on old gcc which rejects
2320 extern declarations in function scope.
2321 Patch from Oliver Fleischmann
2323 PP/10 Use sig_atomic_t for flags set from signal handlers.
2324 Check getgroups() return and improve debugging.
2325 Fixed developed for diagnosis in bug 927 (which turned out to be
2328 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2329 Patch from Mark Zealey.
2331 PP/12 Bugzilla 1056: Improved spamd server selection.
2332 Patch from Mark Zealey.
2334 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2335 Based on patch from Heiko Schlittermann.
2337 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2338 Patch from Uwe Doering, sign-off by Michael Haardt.
2340 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2341 variable declaration deep within a block. Bug and patch from
2344 PP/15 lookups-Makefile IRIX compatibility coercion.
2346 PP/16 Make DISABLE_DKIM build knob functional.
2348 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2349 Patch by Simon Arlott
2351 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2352 variable argument macros. Our copy now differs from upstream.
2358 TF/01 Failure to get a lock on a hints database can have serious
2359 consequences so log it to the panic log.
2361 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2362 controlled using the smtp_confirmation log selector.
2364 TF/03 Include the error message when we fail to unlink a spool file.
2366 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2367 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2368 for maintaining out-of-tree patches for some time.
2370 PP/01 Bugzilla 139: Documentation and portability issues.
2371 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2372 Handle per-OS dynamic-module compilation flags.
2374 PP/02 Let /dev/null have normal permissions.
2375 The 4.73 fixes were a little too stringent and complained about the
2376 permissions on /dev/null. Exempt it from some checks.
2377 Reported by Andreas M. Kirchwitz.
2379 PP/03 Report version information for many libraries, including
2380 Exim version information for dynamically loaded libraries. Created
2381 version.h, now support a version extension string for distributors
2382 who patch heavily. Dynamic module ABI change.
2384 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2385 privilege escalation vulnerability whereby the Exim run-time user
2386 can cause root to append content of the attacker's choosing to
2389 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2392 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2393 If dropping privileges for untrusted macros, we disabled normal logging
2394 on the basis that it would fail; for the Exim run-time user, this is not
2395 the case, and it resulted in successful deliveries going unlogged.
2396 Fixed. Reported by Andreas Metzler.
2402 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2403 only prepend for the Resent-* case. Fixes regression introduced in
2404 Exim 4.70 by NM/22 for Bugzilla 607.
2406 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2407 increasing numbers of administrators be bitten by this.
2409 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2411 PP/03 Bugzilla 994: added openssl_options main configuration option.
2413 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2415 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2417 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2419 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2420 without a peer certificate, leading to a segfault because of an
2421 assumption that peers always have certificates. Be a little more
2422 paranoid. Problem reported by Martin Tscholak.
2424 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2425 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2426 NB: ClamAV planning to remove STREAM in "middle of 2010".
2427 CL also introduces -bmalware, various -d+acl logging additions and
2428 more caution in buffer sizes.
2430 PP/09 Implemented reverse_ip expansion operator.
2432 PP/10 Bugzilla 937: provide a "debug" ACL control.
2434 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2436 PP/12 Bugzilla 973: Implement --version.
2438 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2440 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2442 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2444 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2445 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2446 ignore trailing whitespace.
2448 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2450 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2453 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2454 Notification from Dr Andrew Aitchison.
2456 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2457 ExtendedDetectionInfo response format.
2458 Notification from John Horne.
2460 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2463 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2464 XSL and documented dependency on system catalogs, with examples of how
2467 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2470 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2471 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2472 configuration file which is writeable by the Exim user or group.
2474 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2475 of configuration files to cover files specified with the -C option if
2476 they are going to be used with root privileges, not just the default
2479 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2480 option (effectively making it always true).
2482 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2483 files to be used while preserving root privileges.
2485 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2486 that rogue child processes cannot use them.
2488 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2489 run-time user, instead of root.
2491 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2492 Exim run-time user without dropping privileges.
2494 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2495 result string, instead of calling string_vformat() twice with the same
2498 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2499 for other users. Others should always drop root privileges if they use
2500 -C on the command line, even for a whitelisted configure file.
2502 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2504 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2510 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2511 $data_path, and $header_path variables; fixed documentation bugs and
2514 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2515 exipick to access non-standard spools, including the "frozen" queue
2518 NM/01 Bugzilla 965: Support mysql stored procedures.
2519 Patch from Alain Williams
2521 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2523 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2524 Patch from Andreas Metzler
2526 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2527 Patch from Kirill Miazine
2529 NM/05 Bugzilla 671: Added umask to procmail example.
2531 JJ/03 installed exipick 20100323.0, fixing doc bug
2533 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2534 directory. Notification and patch from Dan Rosenberg.
2536 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2538 TK/02 Improve log output when DKIM signing operation fails.
2540 MH/01 Treat the transport option dkim_domain as a colon separated
2541 list, not as a single string, and sign the message with each element,
2542 omitting multiple occurences of the same signer.
2544 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2545 Bugzilla 985, 986. Patch by Simon Arlott
2547 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2548 Patch by Simon Arlott
2550 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2551 MBX locking. Notification from Dan Rosenberg.
2557 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2559 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2561 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2563 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2565 NM/04 Bugzilla 915: Flag broken perl installation during build.
2571 TK/01 Added patch by Johannes Berg that expands the main option
2572 "spamd_address" if it starts with a dollar sign.
2574 TK/02 Write list of recipients to X-Envelope-Sender header when building
2575 the mbox-format spool file for content scanning (suggested by Jakob
2578 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2579 (http://www.dcc-servers.net/) support via dccifd. Activated by
2580 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2582 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2583 by Mark Daniel Reidel <mr@df.eu>.
2585 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2586 When building exim an external PCRE library is now needed -
2587 PCRE is a system library on the majority of modern systems.
2588 See entry on PCRE_LIBS in EDITME file.
2590 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2591 conversation. Added nologin parameter to request.
2592 Patch contributed by Kirill Miazine.
2594 TF/01 Do not log submission mode rewrites if they do not change the address.
2596 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2598 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2599 log files in place. Contributed by Roberto Lima.
2601 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2603 TF/03 Bugzilla 615: When checking the local_parts router precondition
2604 after a local_part_suffix or local_part_prefix option, Exim now
2605 does not use the address's named list lookup cache, since this
2606 contains cached lookups for the whole local part.
2608 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2609 Robert Millan. Documentation is in experimental-spec.txt.
2611 TF/04 Bugzilla 668: Fix parallel build (make -j).
2613 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2615 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2616 Patch provided by Jan Srzednicki.
2618 TF/05 Leading white space used to be stripped from $spam_report which
2619 wrecked the formatting. Now it is preserved.
2621 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2622 that they are available at delivery time.
2624 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2626 TF/08 TLS error reporting now respects the incoming_interface and
2627 incoming_port log selectors.
2629 TF/09 Produce a more useful error message if an SMTP transport's hosts
2630 setting expands to an empty string.
2632 NM/06 Bugzilla 744: EXPN did not work under TLS.
2633 Patch provided by Phil Pennock.
2635 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2636 Patch provided by Richard Godbee.
2638 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2639 acl_smtp_notquit, added index entry.
2641 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2642 Patch provided by Eugene Bujak.
2644 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2645 accept(). Patch provided by Maxim Dounin.
2647 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2648 Patch provided by Phil Pennock.
2650 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2652 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2653 Patch provided by Brad "anomie" Jorsch.
2655 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2656 Patch provided by Dean Brooks.
2658 TK/05 Add native DKIM support (does not depend on external libraries).
2660 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2661 Patch provided by Graeme Fowler.
2663 NM/16 Bugzilla 851: Documentation example syntax fix.
2665 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2667 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2670 NM/19 Bugzilla 745: TLS version reporting.
2671 Patch provided by Phil Pennock.
2673 NM/20 Bugzilla 167: bool: condition support.
2674 Patch provided by Phil Pennock.
2676 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2677 clients. Patch provided by Phil Pennock.
2679 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2680 Patch provided by Brad "anomie" Jorsch.
2682 NM/23 Bugzilla 687: Fix misparses in eximstats.
2683 Patch provided by Heiko Schlittermann.
2685 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2686 Patch provided by Heiko Schlittermann.
2688 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2689 plus update to original patch.
2691 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2693 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2694 Patch provided by David Brownlee.
2696 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2698 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2700 NM/30 Bugzilla 888: TLS documentation bugfixes.
2702 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2704 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2705 Unlike the original bugzilla I have changed all shell scripts in src tree.
2707 NM/33 Bugzilla 898: Transport filter timeout fix.
2708 Patch by Todd Rinaldo.
2710 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2711 Patch by Serge Demonchaux.
2713 NM/35 Bugzilla 39: Base64 decode bug fixes.
2714 Patch by Jakob Hirsch.
2716 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2718 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2720 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2722 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2728 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2729 ALT-N's libdkim that I have put here:
2730 http://duncanthrax.net/exim-experimental/
2732 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2733 were called 'true' and it seems that C99 defines that as a reserved
2734 keyword to be used with 'bool' variable types. That means you could
2735 not include C99-style headers which use bools without triggering
2736 build errors in sieve.c.
2738 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2739 as mailq or other aliases. Changed the --help handling significantly
2740 to do whats expected. exim_usage() emits usage/help information.
2742 SC/01 Added the -bylocaldomain option to eximstats.
2744 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2746 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2748 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2754 PH/01 Another patch from the Sieve maintainer.
2756 PH/02 When an IPv6 address is converted to a string for single-key lookup
2757 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2758 dots are used instead of colons so that keys in lsearch files need not
2759 contain colons. This was done some time before quoting was made available
2760 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2761 (notated using the quote facility) so as to distinguish them from IPv4
2762 keys. This meant that lookups for IP addresses in host lists did not work
2763 for iplsearch lookups.
2765 This has been fixed by arranging for IPv6 addresses to be expressed with
2766 colons if the lookup type is iplsearch. This is not incompatible, because
2767 previously such lookups could never work.
2769 The situation is now rather anomalous, since one *can* have colons in
2770 ordinary lsearch keys. However, making the change in all cases is
2771 incompatible and would probably break a number of configurations.
2773 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2776 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2777 conversion specification without a maximum field width, thereby enabling
2778 a rogue spamd server to cause a buffer overflow. While nobody in their
2779 right mind would setup Exim to query an untrusted spamd server, an
2780 attacker that gains access to a server running spamd could potentially
2781 exploit this vulnerability to run arbitrary code as the Exim user.
2783 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2784 $primary_hostname instead of what libspf2 thinks the hosts name is.
2786 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2787 a directory entry by the name of the lookup key. Previously, if a
2788 symlink pointed to a non-existing file or a file in a directory that
2789 Exim lacked permissions to read, a lookup for a key matching that
2790 symlink would fail. Now it is enough that a matching directory entry
2791 exists, symlink or not. (Bugzilla 503.)
2793 PH/03 The body_linecount and body_zerocount variables are now exported in the
2796 PH/04 Added the $dnslist_matched variable.
2798 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2799 This means they are set thereafter only if the connection becomes
2802 PH/06 Added the client_condition to authenticators so that some can be skipped
2803 by clients under certain conditions.
2805 PH/07 The error message for a badly-placed control=no_multiline_responses left
2806 "_responses" off the end of the name.
2808 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2810 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2811 (without spaces) instead of just copying the configuration text.
2813 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2815 PH/11 Added $max_received_linelength.
2817 PH/12 Added +ignore_defer and +include_defer to host lists.
2819 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2820 way in which PCRE > 7.0 is built.
2822 PH/14 Implemented queue_only_load_latch.
2824 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2825 MAIL command. The effect was to mangle the value on 64-bit systems.
2827 PH/16 Another patch from the Sieve maintainer.
2829 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2831 PH/18 If a system quota error occurred while trying to create the file for
2832 a maildir delivery, the message "Mailbox is full" was not appended to the
2833 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2834 only to a quota excession during the actual writing of the file.
2836 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2837 characters?) which causes problems in log lines. The DN values are now
2838 passed through string_printing() before being added to log lines.
2840 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2841 and InterBase are left for another time.)
2843 PH/21 Added message_body_newlines option.
2845 PH/22 Guard against possible overflow in moan_check_errorcopy().
2847 PH/23 POSIX allows open() to be a macro; guard against that.
2849 PH/24 If the recipient of an error message contained an @ in the local part
2850 (suitably quoted, of course), incorrect values were put in $domain and
2851 $local_part during the evaluation of errors_copy.
2857 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2858 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2861 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2862 issue a MAIL command.
2864 PH/02 In an ACL statement such as
2866 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2868 if a client was not listed at all, or was listed with a value other than
2869 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2870 the condition was not true (as it should be), so access was not denied.
2871 The bug was that the ! inversion was incorrectly passed on to the second
2872 item. This has been fixed.
2874 PH/03 Added additional dnslists conditions == and =& which are different from
2875 = and & when the dns lookup returns more than one IP address.
2877 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2878 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2880 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2881 FSYNC, which compiles an option called disable_fsync that allows for
2882 bypassing fsync(). The documentation is heavily laced with warnings.
2884 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2886 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2887 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2888 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2889 including adding "make clean"; (3) Added -fPIC when compiling the test
2890 dynamically loaded module, to get rid of a warning.
2892 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2893 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2894 = 0s. The bug is otherwise harmless.
2896 PH/07 There was a bug in the dovecot authenticator such that the value of
2897 $auth1 could be overwritten, and so not correctly preserved, after a
2898 successful authentication. This usually meant that the value preserved by
2899 the server_setid option was incorrect.
2901 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2903 PH/09 Installed PCRE release 7.0.
2905 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2906 run for batched SMTP input. It is now run at the start of every message
2907 in the batch. While fixing this I discovered that the process information
2908 (output by running exiwhat) was not always getting set for -bs and -bS
2909 input. This is fixed, and it now also says "batched" for BSMTP.
2911 PH/11 Added control=no_pipelining.
2913 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2914 patch, slightly modified), and move the expansion of helo_data till after
2915 the connection is made in the smtp transport (so it can use these
2918 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2920 PH/14 Added log_selector = +pid.
2922 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2924 PH/16 Add ${if forany and ${if forall.
2926 PH/17 Added dsn_from option to vary the From: line in DSNs.
2928 PH/18 Flush SMTP output before performing a callout, unless control =
2929 no_callout_flush is set.
2931 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2932 was true (the default) a successful delivery failed to delete the retry
2933 item, thus causing premature timeout of the address. The bug is now
2936 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2938 PH/21 Long custom messages for fakedefer and fakereject are now split up
2939 into multiline responses in the same way that messages for "deny" and
2940 other ACL rejections are.
2942 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2943 with slight modification.
2945 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2946 draft, changing the syntax and factoring some duplicate code".
2948 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2949 for deliveries of the second and subsequent messages over the same SMTP
2952 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2953 ${reduce, with only minor "tidies".
2955 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2957 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2958 expansion side effects.
2960 PH/27 When a message times out after an over-quota error from an Exim-imposed
2961 quota, the bounce message says "mailbox is full". This message was not
2962 being given when it was a system quota that was exceeded. It now should
2965 MH/03 Made $recipients available in local_scan(). local_scan() already has
2966 better access to the recipient list through recipients_list[], but
2967 $recipients can be useful in postmaster-provided expansion strings.
2969 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2970 in the case of a MAIL command with additional options following the
2971 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2972 were accidentally chopped off.
2974 PH/29 SMTP synchronization checks are implemented when a command is read -
2975 there is a check that no more input is waiting when there shouldn't be
2976 any. However, for some commands, a delay in an ACL can mean that it is
2977 some time before the response is written. In this time, more input might
2978 arrive, invalidly. So now there are extra checks after an ACL has run for
2979 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2980 pipelining has not been advertised.
2982 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2984 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2985 returned in the SMTP response when smtp_return_error_details was false.
2986 This has been fixed.
2988 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2989 instead of the C I/O that was originally supplied, because problems were
2990 reported on Solaris.
2992 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2993 Exim which did not show up earlier: it was assuming that a call to
2994 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2995 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2996 was a macro that became an assignment, so it seemed to work. This has
2997 changed to a proper function call with a void return, hence the compile
2998 error. Exim's code has been fixed.
3000 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3003 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3005 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3007 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3008 because it turns out to be too big or there is a timeout) there is no
3009 "Completed" line in the log. When some messages of this type were
3010 selected by exigrep, they were listed as "not completed". Others were
3011 picked up by some special patterns. I have improved the selection
3012 criteria to be more general.
3014 PH/38 The host_find_failed option in the manualroute router can now be set
3015 to "ignore", to completely ignore a host whose IP address cannot be
3016 found. If all hosts are ignored, the behaviour is controlled by the new
3017 host_all_ignored option.
3019 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3020 homing or because of multiple MX records with /mx) generated more than
3021 one IP address, and the following item turned out to be the local host,
3022 all the secondary addresses of the first item were incorrectly removed
3023 from the list, along with the local host and any following hosts (which
3024 is what is supposed to happen).
3026 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3027 whoever called Exim into the -H file. In the case of the daemon it was
3028 behaving confusingly. When first started, it used values for whoever
3029 started the daemon, but after a SIGHUP it used the Exim user (because it
3030 calls itself on a restart). I have changed the code so that it now always
3033 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3034 message are rejected with the same error (e.g. no authentication or bad
3035 sender address), and a DATA command is nevertheless sent (as can happen
3036 with PIPELINING or a stupid MUA), the error message that was given to the
3037 RCPT commands is included in the rejection of the DATA command. This is
3038 intended to be helpful for MUAs that show only the final error to their
3041 PH/42 Another patch from the Sieve maintainer.
3043 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3044 Eximstats - Fixed some broken HTML links and added missing column headers
3046 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3047 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3049 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3051 PH/43 Yet another patch from the Sieve maintainer.
3053 PH/44 I found a way to check for a TCP/IP connection going away before sending
3054 the response to the final '.' that terminates a message, but only in the
3055 case where the client has not sent further data following the '.'
3056 (unfortunately, this is allowed). However, in many cases there won't be
3057 any further data because there won't be any more messages to send. A call
3058 to select() can be used: if it shows that the input is "ready", there is
3059 either input waiting, or the socket has been closed. An attempt to read
3060 the next input character can distinguish the two cases. Previously, Exim
3061 would have sent an OK response which the client would never have see.
3062 This could lead to message repetition. This fix should cure that, at
3063 least in a lot of common cases.
3065 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3066 advertised in response to EHLO.
3072 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3073 fixed by 4.65/MH/01 (is this a record?) are fixed:
3075 (i) An empty string was always treated as zero by the numeric comparison
3076 operators. This behaviour has been restored.
3078 (ii) It is documented that the numeric comparison operators always treat
3079 their arguments as decimal numbers. This was broken in that numbers
3080 starting with 0 were being interpreted as octal.
3082 While fixing these problems I realized that there was another issue that
3083 hadn't been noticed. Values of message_size_limit (both the global option
3084 and the transport option) were treated as octal if they started with 0.
3085 The documentation was vague. These values are now always treated as
3086 decimal, and I will make that clear in the documentation.
3092 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3093 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3096 MH/01 Don't check that the operands of numeric comparison operators are
3097 integers when their expansion is in "skipping" mode (fixes bug
3098 introduced by 4.64-PH/07).
3100 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3101 child addresses, Exim now panics and dies. Previously, because the count
3102 is held in a short int, deliveries were likely to be lost. As such a
3103 large number of recipients for a single message is ridiculous
3104 (performance will be very, very poor), I have chosen to impose a limit
3105 rather than extend the field.
3111 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3112 leftover -K file (the existence of which was triggered by #402).
3113 While we were at it, introduced process PID as part of the -K
3114 filename. This should rule out race conditions when creating
3117 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3118 processing considerably. Previous code took too long for large mails,
3119 triggering a timeout which in turn triggers #401.
3121 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3122 in the DK code in transports.c. sendfile() is not really portable,
3123 hence the _LINUX specificness.
3125 TF/01 In the add_headers option to the mail command in an Exim filter,
3126 there was a bug that Exim would claim a syntax error in any
3127 header after the first one which had an odd number of characters
3130 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3131 callout verification, Exim cached a "reject" for the entire domain. This
3132 is correct for most verifications, but it is not correct for a recipient
3133 verification with use_sender or use_postmaster set, because in that case
3134 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3135 case of MAIL FROM:<> rejection from other early rejections (e.g.
3136 rejection of HELO). When verifying a recipient using a non-null MAIL
3137 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3138 Whatever the result of the callout, the value of the domain cache is
3139 left unchanged (for any other kind of callout, getting as far as trying
3140 RCPT means that the domain itself is ok).
3142 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3145 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3146 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3147 session, the daemon ignores ECONNECT errors and logs others; it now
3148 ignores EPIPE as well.
3150 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3151 (quoted-printable decoding).
3153 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3154 later the small subsequent patch to fix an introduced bug.
3156 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3158 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3160 PH/08 An error is now given if message_size_limit is specified negative.
3162 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3163 to be given (somewhat) arbitrary names.
3165 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3168 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3169 miscellaneous code fixes
3171 PH/10 Added the log_reject_target ACL modifier to specify where to log
3174 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3175 hostname. This is wrong, because it relates to the incoming message (and
3176 probably the interface on which it is arriving) and not to the outgoing
3177 callout (which could be using a different interface). This has been
3178 changed to use the value of the helo_data option from the smtp transport
3179 instead - this is what is used when a message is actually being sent. If
3180 there is no remote transport (possible with a router that sets up host
3181 addresses), $smtp_active_hostname is used.
3183 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3184 tweaks were necessary in order to get it to work (see also 21 below):
3185 (a) The code assumed that strncpy() returns a negative number on buffer
3186 overflow, which isn't the case. Replaced with Exim's string_format()
3188 (b) There were several signed/unsigned issues. I just did the minimum
3189 hacking in of casts. There is scope for a larger refactoring.
3190 (c) The code used strcasecmp() which is not a standard C function.
3191 Replaced with Exim's strcmpic() function.
3192 (d) The code set only $1; it now sets $auth1 as well.
3193 (e) A simple test gave the error "authentication client didn't specify
3194 service in request". It would seem that Dovecot has changed its
3195 interface. Fortunately there's a specification; I followed it and
3196 changed what the client sends and it appears to be working now.
3198 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3201 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3202 address A is aliased to B and C, where B exists and C does not. Without
3203 -v the output is "A verified" because verification stops after a
3204 successful redirection if more than one address is generated. However,
3205 with -v the child addresses are also verified. Exim was outputting "A
3206 failed to verify" and then showing the successful verification for C,
3207 with its parentage. It now outputs "B failed to verify", showing B's
3208 parentage before showing the successful verification of C.
3210 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3211 look up a TXT record in a specific list after matching in a combined
3214 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3215 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3216 they consult the DNS. I had assumed they would set it the way they
3217 wanted; and indeed my experiments on Linux seem to show that in some
3218 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3219 To be on the safe side, however, I have now made the interface to
3220 host_find_byname() similar to host_find_bydns(), with an argument
3221 containing the DNS resolver options. The host_find_byname() function now
3222 sets these options at its start, just as host_find_bydns() does. The smtp
3223 transport options dns_qualify_single and dns_search_parents are passed to
3224 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3225 of host_find_byname() use the default settings of RES_DEFNAMES
3226 (qualify_single) but not RES_DNSRCH (search_parents).
3228 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3229 spool_read_header() do less string testing, by means of a preliminary
3230 switch on the second character of optional "-foo" lines. (This is
3231 overdue, caused by the large number of possibilities that now exist.
3232 Originally there were few.) While I was there, I also converted the
3233 str(n)cmp tests so they don't re-test the leading "-" and the first
3234 character, in the hope this might squeeze out yet more improvement.
3236 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3237 flag allowing group syntax was set by the header_syntax check but not
3238 turned off, possible causing trouble later; (2) The flag was not being
3239 set at all for the header_verify test, causing "group"-style headers to
3240 be rejected. I have now set it in this case, and also caused header_
3241 verify to ignore an empty address taken from a group. While doing this, I
3242 came across some other cases where the code for allowing group syntax
3243 while scanning a header line wasn't quite right (mostly, not resetting
3244 the flag correctly in the right place). These bugs could have caused
3245 trouble for malformed header lines. I hope it is now all correct.
3247 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3248 with the "reply" argument non-NULL. The code, however (which originally
3249 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3250 but it didn't always do it. This confused somebody who was copying the
3251 code for some other use. I have removed all the tests.
3253 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3254 feature that was used to support insecure browsers during the U.S. crypto
3255 embargo. It requires special client support, and Exim is probably the
3256 only MTA that supported it -- and would never use it because real RSA is
3257 always available. This code has been removed, because it had the bad
3258 effect of slowing Exim down by computing (never used) parameters for the
3259 RSA_EXPORT functionality.
3261 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3262 authenticator to fail if there's a tab character in the incoming data
3263 (there should never be unless someone is messing about, as it's supposed
3264 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3265 if the connection is using TLS or if the remote IP is the same as the
3266 local IP, and the "valid-client-cert option" if a client certificate has
3269 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3270 authenticators. This can be used for authorization after authentication
3271 succeeds. (In the case of plaintext, it servers for both authentication
3274 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3275 if any retry times were supplied.
3277 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3278 connection, where there is no client IP address to check. In this
3279 situation, the verify now always succeeds.
3281 PH/25 Applied John Jetmore's -Mset patch.
3283 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3285 PH/27 In a string expansion for a processed (not raw) header when multiple
3286 headers of the same name were present, leading whitespace was being
3287 removed from all of them, but trailing whitespace was being removed only
3288 from the last one. Now trailing whitespace is removed from each header
3289 before concatenation. Completely empty headers in a concatenation (as
3290 before) are ignored.
3292 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3293 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3295 PH/29 [Removed. This was a change that I later backed out, and forgot to
3296 correct the ChangeLog entry (that I had efficiently created) before
3297 committing the later change.]
3299 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3300 address errors (4xx response to RCPT) over the same connection as other
3301 messages routed to the same hosts. Such deliveries are always "forced",
3302 so retry times are not inspected. This resulted in far too many retries
3303 for the affected addresses. The effect occurred only when there were more
3304 hosts than the hosts_max_try setting in the smtp transport when it had
3305 the 4xx errors. Those hosts that it had tried were not added to the list
3306 of hosts for which the message was waiting, so if all were tried, there
3307 was no problem. Two fixes have been applied:
3309 (i) If there are any address or message errors in an SMTP delivery, none
3310 of the hosts (tried or untried) are now added to the list of hosts
3311 for which the message is waiting, so the message should not be a
3312 candidate for sending over the same connection that was used for a
3313 successful delivery of some other message. This seems entirely
3314 reasonable: after all the message is NOT "waiting for some host".
3315 This is so "obvious" that I'm not sure why it wasn't done
3316 previously. Hope I haven't missed anything, but it can't do any
3317 harm, as the worst effect is to miss an optimization.
3319 (ii) If, despite (i), such a delivery is accidentally attempted, the
3320 routing retry time is respected, so at least it doesn't keep
3321 hammering the server.
3323 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3324 in ${readsocket because some servers need this prod.
3326 PH/32 Added some extra debug output when updating a wait-xxx database.
3328 PH/33 The hint "could be header name not terminated by colon", which has been
3329 given for certain expansion errors for a long time, was not being given
3330 for the ${if def:h_colon_omitted{... case.
3332 PH/34 The spec says: "With one important exception, whenever a domain list is
3333 being scanned, $domain contains the subject domain." There was at least
3334 one case where this was not true.
3336 PH/35 The error "getsockname() failed: connection reset by peer" was being
3337 written to the panic log as well as the main log, but it isn't really
3338 panic-worthy as it just means the connection died rather early on. I have
3339 removed the panic log writing for the ECONNRESET error when getsockname()
3342 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3343 runs only) independently of the message's sender address. This meant
3344 that, if the 4xx error was in fact related to the sender, a different
3345 message to the same recipient with a different sender could confuse
3346 things. In particular, this can happen when sending to a greylisting
3347 server, but other circumstances could also provoke similar problems.
3348 I have changed the default so that the retry time for these errors is now
3349 based a combination of the sender and recipient addresses. This change
3350 can be overridden by setting address_retry_include_sender=false in the
3353 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3354 remote server are returned as part of bounce messages. This was not
3355 happening for LMTP over a pipe (the lmtp transport), but now it is the
3356 same for both kinds of LMTP.
3358 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3359 in header lines that were in fact CNAMEs. This is no longer the case.
3361 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3362 and queue runs started by the daemon processed all messages. This has
3363 been fixed so that -R and -S can now usefully be given with -q<time>.
3365 PH/40 Import PCRE release 6.7 (fixes some bugs).
3367 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3369 PH/42 Give an error if -q is specified more than once.
3371 PH/43 Renamed the variables $interface_address and $interface_port as
3372 $received_ip_address and $received_port, to make it clear that these
3373 values apply to message reception, and not to the outgoing interface when
3374 a message is delivered. (The old names remain recognized, of course.)
3376 PH/44 There was no timeout on the connect() call when using a Unix domain
3377 socket in the ${readsocket expansion. There now is.
3379 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3380 be meaningful with "accept".
3382 SC/01 Eximstats V1.43
3383 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3385 SC/02 Eximstats V1.44
3386 Use a glob alias rather than an array ref in the generated
3387 parser. This improves both readability and performance.
3389 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3390 Collect SpamAssassin and rejection statistics.
3391 Don't display local sender or destination tables unless
3392 there is data to show.
3393 Added average volumes into the top table text output.
3395 SC/04 Eximstats V1.46
3396 Collect data on the number of addresses (recipients)
3397 as well as the number of messages.
3399 SC/05 Eximstats V1.47
3400 Added 'Message too big' to the list of mail rejection
3401 reasons (thanks to Marco Gaiarin).
3403 SC/06 Eximstats V1.48
3404 Mainlog lines which have GMT offsets and are too short to
3405 have a flag are now skipped.
3407 SC/07 Eximstats V1.49 (Alain Williams)
3408 Added the -emptyok flag.
3410 SC/08 Eximstats V1.50
3411 Fixes for obtaining the IP address from reject messages.
3413 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3414 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3415 whitespace changes from 4.64-PH/27
3417 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3420 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3421 are found, allow negative numbers in numeric criteria)
3423 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3425 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3426 to match changes made in 4.64-PH/43
3428 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3430 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3431 "domain" is actually an IP address, and force a failure. This locks out
3432 those revolvers/nameservers that support "A-for-A" lookups, in
3433 contravention of the specifications.
3435 PH/48 When a host name was looked up from an IP address, and the subsequent
3436 forward lookup of the name timed out, the host name was left in
3437 $sender_host_name, contrary to the specification.
3439 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3440 restricted to single-key lookups, Exim was not diagnosing an error if
3441 * or *@ was used with a query-style lookup.
3443 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3445 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3446 long ago, but noone interested enough thought of it. Let's just say that
3447 the "1.1" means that there are some new functions that weren't there at
3448 some point in the past.
3450 PH/51 Error processing for expansion failure of helo_data from an smtp
3451 transport during callout processing was broken.
3453 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3454 tested/used via the -bh/-bhc/-bs options.
3456 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3457 bug, fixed in subsequent PCRE releases).
3459 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3460 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3462 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3468 SC/01 Use a glob alias rather than an array ref in eximstats generated
3469 parser. This improves both readability and performance.
3471 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3472 Don't display local sender or destination tables in eximstats unless
3473 there is data to show.
3474 Added average volumes into the eximstats top table text output.
3476 SC/03 Collect data on the number of addresses (recipients) as well
3477 as the number of messages in eximstats.
3479 TF/01 Correct an error in the documentation for the redirect router. Exim
3480 does (usually) call initgroups() when daemonizing.
3482 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3483 with consistent privilege compared to when running as a daemon.
3485 TF/03 Note in the spec that $authenticated_id is not set for local
3486 submissions from trusted users.
3488 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3489 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3491 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3492 by adding some example configuration directives to the default
3493 configuration file. A little bit of work is required to uncomment the
3494 directives and define how usernames and passwords are checked, but
3495 there is now a framework to start from.
3497 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3498 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3499 without this. I don't know how relevant this is to other LDAP libraries.
3501 PH/02 Add the verb name to the "unknown ACL verb" error.
3503 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3505 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3507 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3508 directory not expanded when it should be if an expanded home directory
3509 was set for the address (which is overridden by the transport).
3511 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3514 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3515 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3516 because it is too late at that time, and has no effect.
3518 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3519 security issue with \' (bugzilla #107). I could not use the
3520 PQescapeStringConn() function, because it needs a PGconn value as one of
3523 PH/08 When testing addresses using -bt, indicate those final addresses that
3524 are duplicates that would not cause an additional delivery. At least one
3525 person was confused, thinking that -bt output corresponded to deliveries.
3526 (Suppressing duplicates isn't a good idea as you lose the information
3527 about possibly different redirections that led to the duplicates.)
3529 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3530 systems where poll() doesn't work, in particular OS X.
3532 PH/10 Added more information to debugging output for retry time not reached.
3534 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3535 operations in malware.c.
3537 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3540 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3541 syslog_duplication set false, log lines that would normally be written
3542 both the the main log and to the reject log were not written to syslog at
3545 PH/14 In the default configuration, change the use of "message" in ACL warn
3546 statements to "add_header".
3548 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3549 not followed by a command (e.g. "seen endif").
3551 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3552 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3555 PH/17 Added extra conditions to the default value of delay_warning_condition
3559 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3560 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3561 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3564 The Auto-Submitted: and various List- headers are standardised, whereas I
3565 don't think Precedence: ever was.
3567 PH/18 Refactored debugging code in route_finduser() to show more information,
3568 in particular, the error code if getpwnam() issues one.
3570 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3571 This is apparently needed in addition to the PH/07 change above to avoid
3572 any possible encoding problems.
3574 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3575 but not after initializing Perl.
3577 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3578 output them only if debugging. By default they are written stderr,
3579 apparently, which is not desirable.
3581 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3584 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3587 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3589 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3590 authenticated or an ident call has been made. Suppress the default
3591 values for $authenticated_id and $authenticated_sender (but permit -oMai
3592 and -oMas) when testing with -bh.
3594 PH/24 Re-jigged the order of the tests in the default configuration so that the
3595 tests for valid domains and recipients precede the DNS black list and CSA
3596 tests, on the grounds that those ones are more expensive.
3598 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3599 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3600 This bug exists in every version of Exim that I still have, right back to
3603 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3604 However, an attempt to turn on case-sensitivity in a regex key by
3605 including (?-i) didn't work because the subject string was already
3606 lowercased, and the effects were non-intuitive. It turns out that a
3607 one-line patch can be used to allow (?-i) to work as expected.
3613 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3614 other effects) broke the use of negated acl sub-conditions.
3616 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3619 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3620 "Deny" causes Exim to reject the incoming connection with a 554 error.
3621 Unfortunately, if there is a major crisis, such as a disk failure,
3622 tcp-wrappers gives "deny", whereas what one would like would be some
3623 kind of temporary error. A kludge has been added to help with this.
3624 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3625 554 error is used if errno is still zero or contains ENOENT (which occurs
3626 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3629 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3631 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3632 errors. Otherwise a message that provokes a temporary error (when other
3633 messages do not) can cause a whole host to time out.
3635 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3636 addresses were routed directly to files or pipes from a redirect router.
3637 File deliveries just didn't batch; pipe deliveries might have suffered
3640 PH/06 A failure to get a lock for a hints database would erroneously always say
3641 "Failed to get write lock", even when it was really a read lock.
3643 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3644 of 0600. This has been changed to use the value of the lockfile_mode
3645 option (which defaults to 0600).
3647 PH/08 Applied small patch from the Sieve maintainer.
3649 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3650 folder from quota calculations, a direct delivery into this folder messed
3651 up the contents of the maildirsize file. This was because the regex was
3652 used only to exclude .Trash (or whatever) when the size of the mailbox
3653 was calculated. There was no check that a delivery was happening into an
3654 excluded directory. This bug has been fixed by ignoring all quota
3655 processing for deliveries into excluded directories.
3657 PH/10 Added the maildirfolder_create_regex option to appendfile.
3663 PH/01 The code for finding all the local interface addresses on a FreeBSD
3664 system running IPv6 was broken. This may well have applied to all BSD
3665 systems, as well as to others that have similar system calls. The broken
3666 code found IPv4 interfaces correctly, but gave incorrect values for the
3667 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3668 that it would not match correctly against @[] and not recognize the IPv6
3671 PH/02 The ipliteral router was not recognizing addresses of the form user@
3672 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3674 PH/03 Added disable_ipv6.
3676 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3677 decoded form, because it is most often used to construct To: headers
3678 lines in autoreplies, and the decoded form may well be syntactically
3679 invalid. However, $reply_address has leading white space removed, and all
3680 newlines turned into spaces so that the autoreply transport does not
3683 PH/05 If group was specified without a user on a router, and no group or user
3684 was specified on a transport, the group from the router was ignored.
3686 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3687 for visible compile-time settings that can be used to change these
3688 numbers, for those that want even more. Backwards compatibility with old
3689 spool files has been maintained. However, going back to a previous Exim
3690 release will lost any variables that are in spool files.
3692 PH/07 Two small changes when running in the test harness: increase delay when
3693 passing a TCP/IP connection to a new process, in case the original
3694 process has to generate a bounce, and remove special handling of
3695 127.0.0.2 (sic), which is no longer necessary.
3697 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3698 be the same on different OS.
3700 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3703 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3704 whether --show-vars was specified or not
3706 JJ/02 exipick: Added support for new ACL variable spool format introduced
3709 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3710 syntactically invalid From: or Reply-to: line, and a filter used this to
3711 generate an autoreply, and therefore failed to obtain an address for the
3712 autoreply, Exim could try to deliver to a non-existent relative file
3713 name, causing unrelated and misleading errors. What now happens is that
3714 it logs this as a hard delivery error, but does not attempt to create a
3717 PH/11 The exinext utility has a -C option for testing purposes, but although
3718 the given file was scanned by exinext itself; it wasn't being passed on
3719 when Exim was called.
3721 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3722 an end-of-file indication when reading a command response.
3724 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3725 compiled. In many other places in Exim, IPv6 addresses are always
3726 recognized, so I have changed this. It also means that IPv4 domain
3727 literals of the form [IPV4:n.n.n.n] are now always recognized.
3729 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3730 used if the router is not running as root, for example, when verifying at
3731 ACL time, or when using -bh. The debugging output from this situation was
3732 non-existent - all you got was a failure to exec. I have made two
3735 (a) Failures to set uid/gid, the current directory, or a process leader
3736 in a subprocess such as that created by queryprogram now generate
3737 suitable debugging output when -d is set.
3739 (b) The queryprogram router detects when it is not running as root,
3740 outputs suitable debugging information if -d is set, and then runs
3741 the subprocess without attempting to change uid/gid.
3743 PH/15 Minor change to Makefile for building test_host (undocumented testing
3746 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3747 additional section of a DNS packet that returns MX or SRV records.
3748 Instead, it always explicitly searches for A/AAAA records. This avoids
3749 major problems that occur when a DNS server includes only records of one
3750 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3751 fixed another bug: if SRV records were looked up and the corresponding
3752 address records were *not* found in the additional section, the port
3753 values from the SRV records were lost.
3755 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3756 using the correct key (the original address) when searching the retry
3757 rules in order to find which one to use for generating the retry hint.
3759 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3760 adding the default one. Similarly, if it contains a Reply-To: header, the
3761 errors_reply_to option, if set, is not used.
3763 PH/19 When calculating a retry time, Exim used to measure the "time since
3764 failure" by looking at the "first failed" field in the retry record. Now
3765 it does not use this if it is later than than the arrival time of the
3766 message. Instead it uses the arrival time. This makes for better
3767 behaviour in cases where some deliveries succeed, thus re-setting the
3768 "first failed" field. An example is a quota failure for a huge message
3769 when small messages continue to be delivered. Without this change, the
3770 "time since failure" will always be short, possible causing more frequent
3771 delivery attempts for the huge message than are intended.
3772 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3774 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3775 $1, $2, $3) because the numerical variables can be reset during some
3776 expansion items (e.g. "match"), thereby losing the authentication data.
3778 PH/21 Make -bV show the size of off_t variables so that the test suite can
3779 decide whether to run tests for quotas > 2G.
3781 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3782 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3783 filecount value is greater than 2G or if a quota value is greater than 2G
3784 on a system where the size of off_t is not greater than 4, a panic error
3787 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3788 never match. The debug and -bh output now contains an explicit error
3789 message indicating a malformed IPv4 address or mask.
3791 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3792 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3793 PH/23 above applies.
3795 PH/25 Do not write to syslog when running in the test harness. The only
3796 occasion when this arises is a failure to open the main or panic logs
3797 (for which there is an explicit test).
3799 PH/26 Added the /no_tell option to "control=freeze".
3801 PH/27 If a host name lookup failed very early in a connection, for example, if
3802 the IP address matched host_lookup and the reverse lookup yielded a name
3803 that did not have a forward lookup, an error message of the form "no IP
3804 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3805 could be logged. Now it outputs the IP address instead of "NULL".
3807 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3808 allows the sender and the authenticated sender to be set when
3809 submitting a message from within Exim. Since child_open_exim() is
3810 documented for local_scan(), the new function should be too.
3812 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3813 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3814 results in an empty string is now treated as unset.
3816 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3818 PH/31 Added sender_verify_fail logging option.
3820 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3821 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3822 tidied the source and removed it altogether.
3824 PH/33 When a queue run was abandoned because the load average was too high, a
3825 log line was always written; now it is written only if the queue_run log
3826 selector is set. In addition, the log line for abandonment now contains
3827 information about the queue run such as the pid. This is always present
3828 in "start" and "stop" lines but was omitted from the "abandon" line.
3830 PH/34 Omit spaces between a header name and the colon in the error message that
3831 is given when verify = headers_syntax fails (if there are lots of them,
3832 the message gets confusing).
3834 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3835 names, as there are now some PTR records that contain slashes. This check
3836 is only to protect against broken name servers that fall over on strange
3837 characters, so the fact that it applies to all lookups doesn't matter.
3839 PH/36 Now that the new test suite is complete, we can remove some of the
3840 special code in Exim that was needed for the old test suite. For example,
3841 sorting DNS records because real resolvers return them in an arbitrary
3842 order. The new test suite's fake resolver always returns records in the
3845 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3846 bounces) except when queue_only is set, to avoid logging races between
3847 the different processes.
3849 PH/38 Panic-die if .include specifies a non-absolute path.
3851 PH/39 A tweak to the "H" retry rule from its user.
3853 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3854 a label. They prevented compilation on older perls.
3856 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3857 a warning to be raised on newish perls.
3859 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3860 on queue. Changes to match documented behaviour of showing count of
3861 messages matching specified criteria.
3863 PH/40 Changed the default ident timeout from 30s to 5s.
3865 PH/41 Added support for the use of login_cap features, on those BSD systems
3866 that have them, for controlling the resources used by pipe deliveries.
3868 PH/42 The content-scanning code uses fopen() to create files in which to put
3869 message data. Previously it was not paying any attention to the mode of
3870 the files. Exim runs with umask(0) because the rest of the code creates
3871 files with open(), and sets the required mode explicitly. Thus, these
3872 files were ending up world-writeable. This was not a big issue, because,
3873 being within the spool directory, they were not world-accessible. I have
3874 created a function called modefopen, which takes an additional mode
3875 argument. It sets umask(777), creates the file, chmods it to the required
3876 mode, then resets the umask. All the relevant calls to fopen() in the
3877 content scanning code have been changed to use this function.
3879 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3880 to 24 hours. This avoids potential overflow problems when processing G
3881 and H retry rules. I suspect nobody ever tinkers with this value.
3883 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3885 PH/45 When the plaintext authenticator is running as a client, the server's
3886 challenges are checked to ensure they are valid base64 strings. By
3887 default, the authentication attempt is cancelled if an invalid string is
3888 received. Setting client_ignore_invalid_base64 true ignores these errors.
3889 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3890 they are received. Thus, the responses can be made to depend on the
3891 challenges. If an invalid string is ignored, an empty string is placed in
3894 PH/46 Messages that are created by the autoreply transport now contains a
3895 References: header, in accordance with RFCs 2822 and 3834.
3897 PH/47 Added authenticated_sender_force to the smtp transport.
3899 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3901 PH/49 Installed latest patch from the Sieve maintainer.
3903 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3904 was also set, the appendfile transport was unnecessarily scanning a
3905 directory of message files (e.g. for maildir delivery) to find the count
3906 of files (along with the size), even though it did not need this
3907 information. It now does the scan only if it needs to find either the
3908 size of the count of files.
3910 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3912 PH/52 Two bugs concerned with error handling when the smtp transport is
3915 (i) Exim was not creating retry information for temporary errors given
3916 for individual recipients after the DATA command when the smtp transport
3917 was used in LMTP mode. This meant that they could be retried too
3918 frequently, and not timed out correctly.
3920 (ii) Exim was setting the flag that allows error details to be returned
3921 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3922 recipients that were returned after the DATA command.
3924 PH/53 This is related to PH/52, but is more general: for any failing address,
3925 when detailed error information was permitted to be returned to the
3926 sender, but the error was temporary, then after the final timeout, only
3927 "retry timeout exceeded" was returned. Now it returns the full error as
3928 well as "retry timeout exceeded".
3930 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3931 do this, and (what is worse) MTAs that accept it.
3933 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3934 will now be deprecated.
3936 PH/56 New os.c-cygwin from the Cygwin maintainer.
3938 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3939 formats (previously only available in exim formats via -bpr, -bpru,
3940 and -bpra. Now also available in native and exiqgrep formats)
3942 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3943 with very large, slow to parse queues
3945 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3947 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3949 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3950 responses to authentication challenges, though it was showing the
3951 challenges; (ii) I've removed the CR characters from the debug output for
3954 PH/58 Allow for the insertion of a newline as well as a space when a string
3955 is turned into more than one encoded-word during RFC 2047 encoding. The
3956 Sieve code now uses this.
3958 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3959 data_4xx, lost_connection, tls_required.
3961 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3962 message was being sent as an SMTP response.
3964 PH/61 Add -l and -k options to exicyclog.
3966 PH/62 When verifying, if an address was redirected to one new address, so that
3967 verification continued, and the new address failed or deferred after
3968 having set something in $address_data, the value of $address_data was not
3969 passed back to the ACL. This was different to the case when no
3970 redirection occurred. The value is now passed back in both cases.
3972 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3973 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3974 use login_cap.h, so on its own it isn't the distinguishing feature. The
3975 new name refers directly to the setclassresources() function.
3977 PH/65 Added configuration files for NetBSD3.
3979 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3981 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3982 is preferred over IPv4.
3984 PH/68 The bounce_return_message and bounce_return_body options were not being
3985 honoured for bounces generated during the reception of non-SMTP messages.
3986 In particular, this applied to messages rejected by the ACL. This bug has
3987 been fixed. However, if bounce_return_message is true and bounce_return_
3988 body is false, the headers that are returned for a non-SMTP message
3989 include only those that have been read before the error was detected.
3990 (In the case of an ACL rejection, they have all been read.)
3992 PH/69 The HTML version of the specification is now built in a directory called
3993 spec_html instead of spec.html, because the latter looks like a path with
3994 a MIME-type, and this confuses some software.
3996 PH/70 Catch two compiler warnings in sieve.c.
3998 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3999 function verify_get_ident() calls ip_connect() to connect a socket, but
4000 if the "connect()" function timed out, ip_connect() used to close the
4001 socket. However, verify_get_ident() also closes the socket later, and in
4002 between Exim writes to the log, which may get opened at this point. When
4003 the socket was closed in ip_connect(), the log could get the same file
4004 descriptor number as the socket. This naturally causes chaos. The fix is
4005 not to close the socket in ip_connect(); the socket should be closed by
4006 the function that creates it. There was only one place in the code where
4007 this was missing, in the iplookup router, which I don't think anybody now
4008 uses, but I've fixed it anyway.
4010 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4011 well as to direct DNS lookups. Otherwise the handling of names in host
4012 lists is inconsistent and therefore confusing.
4018 PH/01 Two changes to the default runtime configuration:
4020 (1) Move the checks for relay_from_hosts and authenticated clients from
4021 after to before the (commented out) DNS black list checks.
4023 (2) Add control=submission to the relay_from_hosts and authenticated
4024 clients checks, on the grounds that messages accepted by these
4025 statements are most likely to be submissions.
4027 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4029 (1) Generate an error if the third argument for the ${prvs expansion is
4032 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4035 (3) Reset the variables that are obtained from the first argument of
4036 ${prvscheck and used in the second argument before leaving the code,
4037 because their memory is reclaimed, so using them afterwards may do
4040 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4041 one (it's much easier than Tom thought :-).
4043 (5) Because of (4), we can now allow for the use of $prvscheck_result
4044 inside the third argument.
4046 PH/03 For some reason, the default setting of PATH when running a command from
4047 a pipe transport was just "/usr/bin". I have changed it to
4050 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4051 anything to be listed in the output from -bV.
4053 PH/05 When a filter generated an autoreply, the entire To: header line was
4054 quoted in the delivery log line, like this:
4056 => >A.N.Other <ano@some.domain> <original@ddress> ...
4058 This has been changed so that it extracts the operative address. There
4059 may be more than one such address. If so, they are comma-separated, like
4062 => >ano@some.domain,ona@other.domain <original@ddress> ...
4064 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4065 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4066 not being looked up in the rDNS to get a host name, Exim was showing the
4067 IP address twice in Received: lines, even though the IP addresses were
4068 identical. For example:
4070 Received: from [1.2.3.4] (helo=[1.2.3.4])
4072 However, if the real host name was known, it was omitting the HELO data
4073 if it matched the actual IP address. This has been tidied up so that it
4074 doesn't show the same IP address twice.
4076 PH/07 When both +timestamp and +memory debugging was on, the value given by
4077 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4078 called by the debug printing, thereby overwriting the timestamp buffer.
4079 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4081 PH/08 When the original message was included in an autoreply transport, it
4082 always said "this is a copy of the message, including all the headers",
4083 even if body_only or headers_only was set. It now gives an appropriate
4086 PH/09 Applied a patch from the Sieve maintainer which:
4088 o fixes some comments
4089 o adds the (disabled) notify extension core
4090 o adds some debug output for the result of if/elsif tests
4091 o points to the current vacation draft in the documentation
4092 and documents the missing references header update
4096 o fixes a bug in processing the envelope test (when testing
4097 multiple envelope elements, the last element determined the
4100 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4101 Electronic Mail") by including:
4103 Auto-submitted: auto-generated
4105 in the messages that it generates (bounce messages and others, such as
4106 warnings). In the case of bounce messages for non-SMTP messages, there was
4107 also a typo: it was using "Auto_submitted" (underscore instead of
4108 hyphen). Since every message generated by Exim is necessarily in response
4109 to another message, thes have all been changed to:
4111 Auto-Submitted: auto-replied
4113 in accordance with these statements in the RFC:
4115 The auto-replied keyword:
4117 - SHOULD be used on messages sent in direct response to another
4118 message by an automatic process,
4120 - MUST NOT be used on manually-generated messages,
4122 - MAY be used on Delivery Status Notifications (DSNs) and Message
4123 Disposition Notifications (MDNs),
4125 - MUST NOT be used on messages generated by automatic or periodic
4126 processes, except for messages which are automatic responses to
4129 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4130 to the default Received: header definition.
4132 PH/12 Added log selector acl_warn_skipped (default on).
4134 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4135 variables because (a) they are in the wrong storage pool and (b) even if
4136 they were copied, it wouldn't work properly because of the caching.
4138 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4139 checking when decoding. Apparently there are clients that generate
4140 overlong encoded strings. Why am I not surprised?
4142 PH/15 If the first argument of "${if match_address" was not empty, but did not
4143 contain an "@" character, Exim crashed. Now it writes a panic log message
4144 and treats the condition as false.
4146 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4148 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4149 extension "envelope-auth". The code is finished and in agreement with
4150 other implementations, but there is no documentation so far and in fact,
4151 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4152 not changing the active code.
4154 Print executed "if" and "elsif" statements when debugging is used. This
4155 helps a great deal to understand what a filter does.
4157 Document more things not specified clearly in RFC3028. I had all this
4158 sorted out, when out of a sudden new issues came to my mind. Oops."
4160 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4163 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4164 canonical form (as documented). However, after a host name lookup from
4165 the IP address, check_host() was doing a simple string comparison with
4166 addresses acquired from the DNS when checking that the found name did
4167 have the original IP as one of its addresses. Since any found IPv6
4168 addresses are likely to be in abbreviated form, the comparison could
4169 fail. Luckily, there already exists a function for doing the comparison
4170 by converting both addresses to binary, so now that is used instead of
4171 the text comparison.
4173 PH/20 There was another similar case to PH/19, when a complete host name was
4174 given in a host list; looking up its IP address could give an abbreviated
4175 form, whereas the current host's name might or might not be abbreviated.
4176 The same fix has been applied.
4182 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4183 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4186 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4187 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4189 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4191 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4192 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4193 Also, for "Auto-Submitted", treat anything other than "no" as
4194 non-personal, in accordance with RFC 3834. (Previously it treated
4195 anything starting "auto-" as non-personal.)
4197 TF/01 The control=submission/name=... option had a problem with syntax
4198 errors if the name included a slash character. The /name= option
4199 now slurps the rest of the string, so it can include any characters
4200 but it must come last in the list of options (after /sender_retain
4203 PH/05 Some modifications to the interface to the fake nameserver for the new
4211 TK/01 Added the "success_on_redirect" address verification option. See
4212 NewStuff for rationale and an example.
4214 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4216 PH/02 Patch to exigrep to allow it to work on syslog lines.
4218 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4219 fread() to skip over the body file's header line, because in Cygwin the
4220 header line is locked and is inaccessible.
4222 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4223 co-exist for some time) to make it clear that it is the Exim ID that is
4224 referenced, not the Message-ID: header line.
4226 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4227 string_format() function, because snprintf() does not exist on all
4230 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4231 ${stat: expansion item.
4233 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4234 protocol synchronization error", to keep the pedants happy.
4236 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4237 well as for IRIX systems, when gcc is being used. See the host.c source
4240 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4242 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4245 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4246 transport and to the smtp transport in LMTP mode.
4248 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4250 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4251 run a filter in a subprocess. This could lead to confusion in subsequent
4252 lookups in the parent process. There should also be a search_tidyup() at
4253 the end of the subprocess.
4255 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4256 only if the host matched helo_try_verify_hosts, which caused the
4257 verification to occur when the EHLO/HELO command was issued. The ACL just
4258 tested the remembered result. Now, if a previous verification attempt has
4259 not happened, "verify = helo" does it there and then.
4261 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4263 TK/03 Fix log output including CR from clamd.
4265 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4266 exist provoked a memory error which could cause a segfault.
4268 PH/15 Installed PCRE 6.2
4270 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4272 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4273 of the problem. Specifically, suggested +O2 rather than +O1 for the
4276 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4278 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4279 record, the port was not correctly logged when the outgoing_port log
4280 selector was set (it logged the transort's default port).
4282 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4283 fallback_hosts, and "hosts" in the smtp transport.
4285 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4286 host errors such as "Connection refused".
4288 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4289 authentication with radiusclient 0.4.9:
4291 - Error returned from rc_read_config was caught wrongly
4292 - Username/password not passed on to radius server due to wrong length.
4294 The presumption is that some radiusclient API changes for 4.51/PH/17
4295 were not taken care of correctly. The code is still untested by me (my
4296 Linux distribution still has 0.3.2 of radiusclient), but it was
4297 contributed by a Radius user.
4299 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4300 expanding the "port" option of the smtp transport.
4302 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4303 while reading a MIME header. Thanks to Tom Hughes for a patch.
4305 PH/24 Include config.h inside local_scan.h so that configuration settings are
4308 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4309 that in an ACL for RCPT (for example), you can examine exactly what was
4312 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4313 commands, but it was not correctly comparing the address with the actual
4314 client host address. Thus, it would show the EHLO address in Received:
4315 header lines when this was not necessary.
4317 PH/27 Added the % operator to ${eval:}.
4319 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4320 it should be ignoring failures (because with -C, for example, it has lost
4321 privilege). It wasn't ignoring creation failures other than "already
4324 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4327 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4328 because an input line was too long, either on its own, or by virtue of
4329 too many continuations, the temporary file was not being removed, and the
4330 return code was incorrect.
4332 PH/31 Missing "BOOL" in function definition in filtertest.c.
4334 PH/32 Applied Sieve patches from the maintainer.
4336 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4338 PH/33 Added "verify = not_blind".
4340 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4341 Local/Makefile (with some defaults set). These are used in built scripts
4342 such as exicyclog, but they have never been used in the exim_install
4343 script (though there are many overriding facilities there). I have
4344 arranged that the exim_install script now takes note of these two
4347 PH/35 Installed configuration files for Dragonfly.
4349 PH/36 When a locally submitted message by a trusted user did not contain a
4350 From: header, and the sender address was obtained from -f or from an SMTP
4351 MAIL command, and the trusted user did not use -F to supply a sender
4352 name, $originator_name was incorrectly used when constructing a From:
4353 header. Furthermore, $originator_name was used for submission mode
4354 messages from external hosts without From: headers in a similar way,
4355 which is clearly wrong.
4357 PH/37 Added control=suppress_local_fixups.
4359 PH/38 When log_selector = +received_sender was set, and the addition of the
4360 sender made the log line's construction buffer exactly full, or one byte
4361 less than full, an overflow happened when the terminating "\n" was
4364 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4365 when the result of a list match is failure because a DNS lookup failed.
4367 PH/40 RM_COMMAND is now used in the building process.
4369 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4370 the "build-* directories that it finds.
4372 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4373 address in a domain literal was a prefix of an interface address.
4375 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4376 when verifying a sender address, unless rewrite_headers is false.
4378 PH/44 Wrote a long comment about why errors_to addresses are verified as
4379 recipients, not senders.
4381 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4382 the ratelimit ACL was added.
4384 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4386 PH/46 Added extra information about PostgreSQL errors to the error string.
4388 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4389 avoiding the need to install special zones in a real server. This is
4390 backwards compatible; if it can't find the fake resolver, it drops back.
4391 Thus, both old and new test suites can be run.
4393 TF/02 Added util/ratelimit.pl
4395 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4396 clock is set back in time.
4398 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4399 Candler <B.Candler@pobox.com>.
4401 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4402 OK for addresses that are the result of redirections.
4404 PH/48 A number of further additions for the benefit of the new test suite,
4405 including a fake gethostbyname() that interfaces to the fake DNS resolver
4408 TF/06 The fix for widen_domains has also been applied to qualify_single and
4409 search_parents which are the other dnslookup options that can cause
4412 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4415 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4417 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4418 service" if it encountered an MX record with an empty target hostname.
4419 The message is now "an MX or SRV record indicated no SMTP service".
4421 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4422 if the DNS of the sending site is misconfigured. This is quite a
4423 common situation. This change restores the behaviour of treating a
4424 helo verification defer as a failure.
4426 PH/51 If self=fail was set on a router, the bounce message did not include the
4427 actual error message.
4433 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4435 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4436 command itself ended in error, the underlying message about the transport
4437 filter timeout was being overwritten with the pipe command error. Now the
4438 underlying error message should be appended to the second error message.
4440 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4442 PH/02 Reduce the amount of output that "make" produces by default. Full output
4443 can still be requested.
4445 PH/03 The warning log line about a condition test deferring for a "warn" verb
4446 was being output only once per connection, rather than after each
4447 occurrence (because it was using the same function as for successful
4448 "warn" verbs). This seems wrong, so I have changed it.
4450 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4451 it should not have, which might have caused a crash in the right
4452 circumstances, but probably never did.
4454 PH/04 Installed a modified version of Tony Finch's patch to make submission
4455 mode fix the return path as well as the Sender: header line, and to
4456 add a /name= option so that you can make the user's friendly name appear
4459 TF/03 Added the control = fakedefer ACL modifier.
4461 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4462 Mark Lowes for thorough testing.
4464 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4466 TK/03 Merged latest SRS patch from Miles Wilton.
4468 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4469 with the definition in sysexits.h (which is #included earlier).
4470 Fortunately, Exim does not actually use EX_OK. The code used to try to
4471 preserve the sysexits.h value, by assuming that macro definitions were
4472 scanned for macro replacements. I have been disabused of this notion,
4473 so now the code just undefines EX_OK before #including unistd.h.
4475 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4476 in the smtp transport. When a block could not be written in a single
4477 write() function, the timeout was being re-applied to each part-write.
4478 This seems wrong - if the receiver was accepting one byte at a time it
4479 would take for ever. The timeout is now adjusted when this happens. It
4480 doesn't have to be particularly precise.
4482 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4483 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4485 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4486 without a domain if the check to <postmaster@domain> fails.
4488 SC/01 Eximstats: added -xls and the ability to specify output files
4489 (patch written by Frank Heydlauf).
4491 SC/02 Eximstats: use FileHandles for outputting results.
4493 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4495 SC/04 Eximstats: fixed display of large numbers with -nvr option
4497 SC/05 Eximstats: fixed merging of reports with empty tables.
4499 SC/06 Eximstats: added the -include_original_destination flag
4501 SC/07 Eximstats: removed tabs and trailing whitespace.
4503 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4505 TK/06 MBOX spool code: Add real "From " MBOX separator line
4506 so the .eml file is really in mbox format (even though
4507 most programs do not really care). Patch from Alex Miller.
4509 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4510 The latter is generated from $received_to and is only set if the
4511 message has one envelope recipient. SA can use these headers,
4512 obviously out-of-the-box. Patch from Alex Miller.
4514 PH/08 The ${def test on a variable was returning false if the variable's
4515 value was "0", contrary to what the specification has always said!
4516 The result should be true unless the variable is empty.
4518 PH/09 The syntax error of a character other than { following "${if
4519 def:variable_name" (after optional whitespace) was not being diagnosed.
4520 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4521 accidental colon was present, for example, could give incorrect results.
4523 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4524 result is used (not including appendfile, where other changes are about
4527 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4528 This involved changing a lot of size variables from int to off_t. It
4529 should work with maildirs and everything.
4531 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4532 spamd dying while we are connected to it.
4534 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4537 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4538 function for BDB 4.3.
4540 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4542 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4543 the sources for PCRE have been re-arranged and more files are now
4546 PH/15 The code I had for printing potentially long long variables in PH/11
4547 above was not the best (it lost precision). The length of off_t variables
4548 is now inspected at build time, and an appropriate printing format (%ld
4549 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4550 to be "long long int" or "long int". This is needed for the internal
4551 formatting function string_vformat().
4553 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4554 the configuration file to be ":syslog", then the script "guesses" where
4555 the logs files are, rather than using the compiled in default. In our
4556 case the guess is not the same as the compiled default, so the script
4557 suddenly stopped working when I started to use syslog. The patch checks
4558 to see if log_file_path is "". If so, it attempts to read it from exim
4559 with no configuration file to get the compiled in version, before it
4560 falls back to the previous guessing code."
4562 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4563 implementing BATV in an Exim configuration. See NewStuff for the gory
4566 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4567 Makefile that are specific to HP-UX.
4569 PH/18 If the "use_postmaster" option was set for a recipient callout together
4570 with the "random" option, the postmaster address was used as the MAIL
4571 FROM address for the random test, but not for the subsequent recipient
4572 test. It is now used for both.
4574 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4575 patch removes a few documentation additions to RFC 3028, because the
4576 latest draft now contains them. It adds the new en;ascii-case comparator
4577 and a new error check for 8bit text in MIME parts. Comparator and
4578 require names are now matched exactly. I enabled the subaddress
4579 extension, but it is not well tested yet (read: it works for me)."
4581 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4582 rework some of the code of TK/09 above to avoid the hardwired use of
4583 "%lld" and "long long". Replaced the call to snprintf() with a call to
4586 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4587 records point to non-existent hosts", "retry timeout exceeded", and
4588 "retry time not reached for any host after a long failure period".
4590 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4591 experimental DomainKeys support:
4593 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4594 (2) On an error such as an illegally used "control", the wrong name for
4595 the control was given.
4597 These problems did NOT occur unless DomainKeys support was compiled.
4599 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4601 PH/24 Added ${if match_ip condition.
4603 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4604 databases so that it will be absolutely obvious if a crash occurs in the
4605 DB library. This is a regular occurrence (often caused by mis-matched
4608 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4609 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4610 picked up on a user's system that detects such things. There doesn't seem
4611 to be a gcc warning option for this - only an attribute that has to be
4612 put on the function's prototype. It seems that in Fedora Core 4 they have
4613 set this on a number of new functions. No doubt there will be more in due
4616 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4617 specify a transport. However, if an address that was verified by such a
4618 router was the subject of a callout, Exim crashed because it tried to
4619 read the rcpt_include_affixes from the non-existent transport. Now it
4620 just assumes that the setting of that option is false. This bug was
4621 introduced by 4.51/PH/31.
4623 PH/28 Changed -d+all to exclude +memory, because that information is very
4624 rarely of interest, but it makes the output a lot bigger. People tend to
4625 do -d+all out of habit.
4627 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4628 code in os-type was giving problems when libc.so lives in lib64, like on
4631 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4632 aren't the modern standard, and it seems that some systems' include files
4633 don't always have them. Exim was already checking for some of the newer
4634 ones like T_AAAA, and defining it itself. I've added checks for all the
4635 record types that Exim uses.
4637 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4638 not automatically generating a new one, as it is supposed to. This
4639 prevented TLS from working. If the file did exist, but contained invalid
4640 data, a new version was generated, as expected. It was only the case of a
4641 non-existent file that was broken.
4643 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4644 with a change in libdomainkeys > 0.64.
4646 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4647 from DNS. If the selector record carries the flag, it now has
4648 precedence over the domain-wide flag.
4650 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4652 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4653 the use of a port name that isn't defined in /etc/services) occurred, the
4654 message was deferred as in a normal delivery, and thus remained on the
4655 spool, instead of being failed because of the mua_wrapper setting. This
4656 is now fixed, and I tidied up some of the mua_wrapper messages at the
4659 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4660 the messages in a hash of arrays rather than using individual hashes.
4661 This is a bit cleaner and results in dramatic memory savings, albeit
4662 at a slight CPU cost.
4664 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4665 as requested by Marc Sherman.
4667 SC/10 Eximstats: added histograms for user specified patterns as requested
4670 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4672 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4673 fopen() in the content-scanning modules that did not already have it.
4679 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4680 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4682 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4684 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4686 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4687 to test Sieve filters that use "vacation".
4689 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4690 that changes the way the GnuTLS parameters are stored in the cache file.
4691 The new format can be generated externally. For backward compatibility,
4692 if the data in the cache doesn't make sense, Exim assumes it has read an
4693 old-format file, and it generates new data and writes a new file. This
4694 means that you can't go back to an older release without removing the
4697 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4698 work if there are any delivery delays because "one_time" forces the
4699 parent to be marked "delivered", so its unseen clone is never tried
4700 again. For this reason, Exim now forbids the simultaneous setting of
4703 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4704 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4705 entry if you want to know the details. The fix, however, neglected to
4706 consider the case when local delivery batching is involved. The test for
4707 "previously delivered" was not happening when checking to see if an
4708 address could be batched with a previous (undelivered) one; under
4709 certain circumstances this could lead to multiple deliveries to the same
4712 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4713 in its include files, and this causes problems building Exim.
4715 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4716 header_syntax) but Exim was just ignoring anything given after a slash.
4717 In particular, this caused confusion with an attempt to use "verify =
4718 reverse_host_lookup/defer_ok". An error is now given when options are
4719 supplied for verify items that do not have them. (Maybe reverse_host_
4720 lookup should have a defer_ok option, but that's a different point.)
4722 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4723 defined by RFC 821) to 2048, because there were problems with some AUTH
4724 commands, and RFC 1869 says the size should be increased for extended
4725 SMTP commands that take arguments.
4727 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4730 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4731 "unknown" error; now it says that the functionality isn't in the binary.
4733 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4734 an address' error message when a string expansion fails (syntax or
4735 whatever). Otherwise the password may appear in the log. Following change
4736 PH/42 below, there is no longer a chance of it appearing in a bounce
4739 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4741 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4742 of its addresses was ever tried. (Bugzilla bug #2.)
4744 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4745 the result incorrectly in the debug output. (It correctly added a newline
4746 to what was transported.)
4748 TF/01 Added $received_time.
4750 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4751 commented out examples of how to interface to a virus scanner and to
4752 SpamAssassin. Also added commented examples of av_scanner and
4753 spamd_address settings.
4755 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4756 and controls are allowed in which ACLs. There were a couple of minor
4757 errors. Some of the entries in the conditions table (which is a table of
4758 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4759 negation of where the condition IS allowed.
4761 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4763 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4764 header file does not have a version number, so I've had to invent a new
4765 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4766 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4767 radiusclient), but it was contributed by a Radius user.
4769 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4770 files or directories, for OpenSSL.
4772 PH/19 When an Exim process that is running as root has to create an Exim log
4773 file, it does so in a subprocess that runs as exim:exim so as to get the
4774 ownership right at creation (otherwise, other Exim processes might see
4775 the file with the wrong ownership). There was no test for failure of this
4776 fork() call, which would lead to the process getting stuck as it waited
4777 for a non-existent subprocess. Forks do occasionally fail when resources
4778 run out. I reviewed all the other calls to fork(); they all seem to check
4781 PH/20 When checking for unexpected SMTP input at connect time (before writing
4782 the banner), Exim was not dealing correctly with a non-positive return
4783 from the read() function. If the client had disconnected by this time,
4784 the result was a log entry for a synchronization error with an empty
4785 string after "input=" when read() returned zero. If read() returned -1
4786 (an event I could not check), uninitialized data bytes were printed.
4787 There were reports of junk text (parts of files, etc) appearing after
4790 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4792 PH/22 Added support for macro redefinition, and (re)definition in between
4793 driver and ACL definitions.
4795 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4796 forgetting to use the resulting value; it was using the unexpanded value.
4798 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4799 hadn't been configured. The fix is from Juergen Kreileder, who
4800 understands it better than I do:
4802 "Here's what I see happening with three configured cyrus_sasl
4803 authenticators configured (plain, login, cram-md5):
4805 On startup auth_cyrus_sasl_init() gets called for each of these.
4806 This means three calls to sasl_listmech() without a specified mech_list.
4807 => SASL tests which mechs of all available mechs actually work
4808 => three warnings about OTP not working
4809 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4811 With the patch, sasl_listmech() also gets called three times. But now
4812 SASL's mech_list option is set to the server_mech specified in the the
4813 authenticator. Or in other words, the answer from sasl_listmech()
4814 gets limited to just the mech you're testing for (which is different
4816 => the return list contains just 'plain' or 'login', 'cram-md5' or
4817 nothing depending on the value of ob->server_mech.
4819 I've just tested the patch: Authentication still works fine,
4820 unavailable mechs specified in the exim configuration are still
4821 caught, and the auth.log warnings about OTP are gone."
4823 PH/25 When debugging is enabled, the contents of the command line are added
4824 to the debugging output, even when log_selector=+arguments is not
4827 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4828 answer is "GNU", and only if the return is "GNU/something" is the answer
4831 PH/27 $acl_verify_message is now set immediately after the failure of a
4832 verification in an ACL, and so is available in subsequent modifiers. In
4833 particular, the message can be preserved by coding like this:
4835 warn !verify = sender
4836 set acl_m0 = $acl_verify_message
4838 Previously, $acl_verify_message was set only while expanding "message"
4839 and "log_message" when a very denied access.
4841 PH/28 Modified OS/os.c-Linux with
4843 -#ifndef OS_LOAD_AVERAGE
4844 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4846 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4847 nomenclature these days.)
4849 PH/29 Installed patch from the Sieve maintainer that adds the options
4850 sieve_useraddress and sieve_subaddress to the redirect router.
4852 PH/30 In these circumstances:
4853 . Two addresses routed to the same list of hosts;
4854 . First host does not offer TLS;
4855 . First host accepts first address;
4856 . First host gives temporary error to second address;
4857 . Second host offers TLS and a TLS session is established;
4858 . Second host accepts second address.
4859 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4860 and peerdn, if requested) that were in fact used only for the second
4863 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4864 attention to any local part prefix or suffix that was matched by the
4865 router that accepted the address. It now behaves in the same way as it
4866 does for delivery: the affixes are removed from the local part unless
4867 rcpt_include_affixes is set on the transport.
4869 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4870 timeout during the DATA phase of an incoming message.
4872 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4873 applied a patch sanctioned by the Sieve maintainer.
4875 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4876 the uid or gid is negative. A case of a negative gid caused this to be
4877 noticed. The fix allows for either to be negative.
4879 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4880 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4883 PH/36 The change PH/12 above was broken. Fixed it.
4885 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4886 the grounds that routing the same address twice would always produce the
4887 same answer. This might have been true once, but it is certainly no
4888 longer true now. Routing a child address may depend on the previous
4889 routing that produced that child. Some complicated redirection strategies
4890 went wrong when messages had multiple recipients, and made Exim's
4891 behaviour dependent on the order in which the addresses were given.
4893 I have moved the duplicate checking until after the routing is complete.
4894 Exim scans the addresses that are assigned to local and remote
4895 transports, and removes any duplicates. This means that more work will be
4896 done, as duplicates will always all be routed, but duplicates are
4897 presumably rare, so I don't expect this is of any significance.
4899 For deliveries to pipes, files, and autoreplies, the duplicate checking
4900 still happens during the routing process, since they are not going to be
4903 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4904 It corrects a timeout issue with spamd. This is Ian's comment: "The
4905 background is that sometimes spamd either never reads data from a
4906 connection it has accepted, or it never writes response data. The exiscan
4907 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4908 blindly assumes that writes won't block so it may never time out."
4910 PH/39 Allow G after quota size as well as K and M.
4912 PH/40 The value set for $authenticated_id in an authenticator may not contain
4913 binary zeroes or newlines because the value is written to log lines and
4914 to spool files. There was no check on this. Now the value is run through
4915 the string_printing() function so that such characters are converted to
4916 printable escape sequences.
4918 PH/41 $message_linecount is a new variable that contains the total number of
4919 lines in the message. Compare $body_linecount, which is the count for the
4922 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4923 bounce and delay warning messages, except in certain special cases, which
4926 (a) An SMTP error message from a remote host;
4927 (b) A message specified in a :fail: redirection;
4928 (c) A message specified in a "fail" command in a system filter;
4929 (d) A message specified in a FAIL return from the queryprogram router;
4930 (e) A message specified by the cannot_route_message router option.
4932 In these cases only, Exim does include the error details in bounce and
4933 warning messages. There are also a few cases where bland messages such
4934 as "unrouteable address" or "local delivery error" are given.
4936 PH/43 $value is now also set for the "else" part of a ${run expansion.
4938 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4939 being worked on, but at least Exim now implements the latest version to
4942 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4943 process to complete was treated as a delivery failure, a timeout while
4944 writing the message to the pipe was logged, but erroneously treated as a
4945 successful delivery. Such timeouts include transport filter timeouts. For
4946 consistency with the overall process timeout, these timeouts are now
4947 treated as errors, giving rise to delivery failures by default. However,
4948 there is now a new Boolean option for the pipe transport called
4949 timeout_defer, which, if set TRUE, converts the failures into defers for
4950 both kinds of timeout. A transport filter timeout is now identified in
4953 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4954 systems where "make" and "gmake" are different, calling "gmake" at top
4955 level broke things. I've arranged for the value of $(MAKE) to be passed
4956 from the Makefile to this script so that it can call the same version of
4960 A note about Exim versions 4.44 and 4.50
4961 ----------------------------------------
4963 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4964 changes of various kinds. As a consequence, a big documentation update was
4965 needed. This delayed the release for rather longer than seemed good, especially
4966 in the light of a couple of (minor) security issues. Therefore, the changes
4967 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4968 release. So 4.44 and 4.50 are in effect two different branches that both start
4971 I have left the 4.50 change log unchanged; it contains all the changes since
4972 4.43. The change log for 4.44 is below; many of its items are identical to
4973 those for 4.50. This seems to be the most sensible way to preserve the
4974 historical information.
4980 1. Minor wording change to the doc/README.SIEVE file.
4982 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4983 computation of the current number of files was incorrect.
4985 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4986 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4989 4. Give more explanation in the error message when the command for a transport
4990 filter fails to execute.
4992 5. There are several places where Exim runs a non-Exim command in a
4993 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4994 was being done only for the command run by the queryprogram router. It is
4995 now done for all such subprocesses. The other cases are: ${run, transport
4996 filters, and the commands run by the lmtp and pipe transports.
4998 6. Added CONFIGURE_GROUP build-time option.
5000 7. Some older OS have a limit of 256 on the maximum number of file
5001 descriptors. Exim was using setrlimit() to set 1000 as a large value
5002 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5003 systems. I've change it so that if it can't get 1000, it tries for 256.
5005 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5006 was an oversight, and furthermore, ever since the addition of extra
5007 controls (e.g. 4.43/32), the checks on when to allow different forms of
5008 "control" were broken. There should now be diagnostics for all cases when a
5009 control that does not make sense is encountered.
5011 9. Added the /retain_sender option to "control=submission".
5013 10. $recipients is now available in the predata ACL (oversight).
5015 11. Tidy the search cache before the fork to do a delivery from a message
5016 received from the command line. Otherwise the child will trigger a lookup
5017 failure and thereby defer the delivery if it tries to use (for example) a
5018 cached ldap connection that the parent has called unbind on.
5020 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5021 of $address_data from the recipient verification was clobbered by the
5022 sender verification.
5024 13. The value of address_data from a sender verification is now available in
5025 $sender_address_data in subsequent conditions in the ACL statement.
5027 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5029 15. Added a new option "connect=<time>" to callout options, to set a different
5032 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5033 was its contents. (It was OK if the option was not defined at all.)
5035 17. A "Completed" log line is now written for messages that are removed from
5036 the spool by the -Mrm option.
5038 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5039 information about exactly what failed.
5041 19. Added -dd to debug only the daemon process.
5043 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5044 handles timeouts, both on the server side and network timeouts. Renamed the
5045 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5047 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5048 It is now set to "smtps".
5050 22. $host_address is now set to the target address during the checking of
5051 ignore_target_hosts.
5053 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5054 being passed; this would have caused $sender_host_name to have been used if
5055 matching the list had actually called for a host name (not very likely,
5056 since this list is usually IP addresses). A host name is now passed as
5059 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5060 code that specifies a non-restarting handler (typically sigaction() in
5061 modern systems) in an attempt to fix a rare and obscure crash bug.
5063 25. Narrowed the window for a race in the daemon that could cause it to ignore
5064 SIGCHLD signals. This is not a major problem, because they are used only to
5065 wake it up if nothing else does.
5067 26. A malformed maildirsize file could cause Exim to calculate negative values
5068 for the mailbox size or file count. Odd effects could occur as a result.
5069 The maildirsize information is now recalculated if the size or filecount
5072 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5073 support for a long time. Removed HAVE_SYS_VFS_H.
5075 28. Installed the latest version of exipick from John Jetmore.
5077 29. In an address list, if the pattern was not a regular expression, an empty
5078 subject address (from a bounce message) matched only if the pattern was an
5079 empty string. Non-empty patterns were not even tested. This was the wrong
5080 because it is perfectly reasonable to use an empty address as part of a
5081 database query. An empty address is now tested by patterns that are
5082 lookups. However, all the other forms of pattern expect the subject to
5083 contain a local part and a domain, and therefore, for them, an empty
5084 address still always fails if the pattern is not itself empty.
5086 30. Exim went into a mad DNS loop when attempting to do a callout where the
5087 host was specified on an smtp transport, and looking it up yielded more
5088 than one IP address.
5090 31. Re-factored the code for checking spool and log partition space into a
5091 function that finds that data and another that does the check. The former
5092 is then used to implement four new variables: $spool_space, $log_space,
5093 $spool_inodes, and $log_inodes.
5095 32. The RFC2047 encoding function was originally intended for short strings
5096 such as real names; it was not keeping to the 75-character limit for
5097 encoded words that the RFC imposes. It now respects the limit, and
5098 generates multiple encoded words if necessary. To be on the safe side, I
5099 have increased the buffer size for the ${rfc2047: expansion operator from
5102 33. It is now permitted to omit both strings after an "if" condition; if the
5103 condition is true, the result is "true". As before, when the second string
5104 is omitted, a false condition yields an empty string. This makes it less
5105 cumbersome to write custom ACL and router conditions.
5107 34. Failure to deliver a bounce message always caused it to be frozen, even if
5108 there was an errors_to setting on the router. The errors_to setting is now
5111 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5112 canonical form (fully expanded) before being placed in
5113 $sender_host_address.
5115 36. The table in the code that translates DNS record types into text (T_A to
5116 "A" for instance) was missing entries for NS and CNAME. It is just possible
5117 that this could have caused confusion if both these types were looked up
5118 for the same domain, because the text type is used as part of Exim's
5119 per-process caching. But the chance of anyone hitting this buglet seems
5122 37. The dnsdb lookup has been extended in a number of ways.
5124 (1) There is a new type, "zns", which walks up the domain tree until it
5125 finds some nameserver records. It should be used with care.
5127 (2) There is a new type, "mxh", which is like "mx" except that it returns
5128 just the host names, not the priorities.
5130 (3) It is now possible to give a list of domains (or IP addresses) to be
5131 looked up. The behaviour when one of the lookups defers can be
5132 controlled by a keyword.
5134 (4) It is now possible to specify the separator character for use when
5135 multiple records are returned.
5137 38. The dnslists ACL condition has been extended: it is now possible to supply
5138 a list of IP addresses and/or domains to be looked up in a particular DNS
5141 39. Added log_selector=+queue_time_overall.
5143 40. When running the queue in the test harness, wait just a tad after forking a
5144 delivery process, to get repeatability of debugging output.
5146 41. Include certificate and key file names in error message when GnuTLS fails
5147 to set them up, because the GnuTLS error message doesn't include the name
5148 of the failing file when there is a problem reading it.
5150 42. Allow both -bf and -bF in the same test run.
5152 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5154 44. The "Exiscan patch" is now merged into the mainline Exim source.
5156 45. Sometimes the final signoff response after QUIT could fail to get
5157 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5158 < 0 before doing a fflush(). This bug looks as though it goes back to the
5159 introduction of TLS in release 3.20, but "sometimes" must have been rare
5160 because the tests only now provoked it.
5162 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5163 (this can affect the format of dates).
5165 47. exim_tidydb, when checking for the continued existence of a message for
5166 which it has found a message-specific retry record, was not finding
5167 messages that were in split spool directories. Consequently, it was
5168 deleting retry records that should have stayed in existence.
5170 48. Steve fixed some bugs in eximstats.
5172 49. The SPA authentication driver was not abandoning authentication and moving
5173 on to the next authenticator when an expansion was forced to fail,
5174 contradicting the general specification for all authenticators. Instead it
5175 was generating a temporary error. It now behaves as specified.
5177 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5178 (the order specifies the preference for clients). The order is now AES256,
5179 AES128, 3DES, ARCFOUR128.
5181 51. Small patch to Sieve code - explicitly set From: when generating an
5184 52. Exim crashed if a remote delivery caused a very long error message to be
5185 recorded - for instance if somebody sent an entire SpamAssassin report back
5186 as a large number of 550 error lines. This bug was coincidentally fixed by
5187 increasing the size of one of Exim's internal buffers (big_buffer) that
5188 happened as part of the Exiscan merge. However, to be on the safe side, I
5189 have made the code more robust (and fixed the comments that describe what
5192 53. Now that there can be additional text after "Completed" in log lines (if
5193 the queue_time_overall log selector is set), a one-byte patch to exigrep
5194 was needed to allow it to recognize "Completed" as not the last thing in
5197 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5198 patch that reportedly fixes this has been added. I am not expert enough to
5199 create a test for it. This is what the patch creator wrote:
5201 "I found a little strange behaviour of ldap code when working with
5202 Windows 2003 AD Domain, where users was placed in more than one
5203 Organization Units. When I tried to give exim partial DN, the exit code
5204 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5205 But simultaneously result of request was absolutely normal ldap result,
5206 so I produce this patch..."
5208 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5209 REFERENCE, so I have modified the code to exclude the patch when that macro
5212 55. Some experimental protocols are using DNS PTR records for new purposes. The
5213 keys for these records are domain names, not reversed IP addresses. The
5214 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5215 leaves it alone. Component reversal etc. now happens only for IP addresses.
5218 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5220 57. Double the size of the debug message buffer (to 2048) so that more of very
5221 long debug lines gets shown.
5223 58. The exicyclog utility now does better if the number of log files to keep
5224 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5226 59. Two changes related to the smtp_active_hostname option:
5228 (1) $smtp_active_hostname is now available as a variable.
5229 (2) The default for smtp_banner uses $smtp_active_hostname instead
5230 of $primary_hostname.
5232 60. The host_aton() function is supposed to be passed a string that is known
5233 to be a valid IP address. However, in the case of IPv6 addresses, it was
5234 not checking this. This is a hostage to fortune. Exim now panics and dies
5235 if the condition is not met. A case was found where this could be provoked
5236 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5237 components; fortuitously, this particular loophole had already been fixed
5238 by change 4.50/55 above.
5240 If there are any other similar loopholes, the new check in host_aton()
5241 itself should stop them being exploited. The report I received stated that
5242 data on the command line could provoke the exploit when Exim was running as
5243 exim, but did not say which command line option was involved. All I could
5244 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5245 running as the user.
5248 61. There was a buffer overflow vulnerability in the SPA authentication code
5249 (which came originally from the Samba project). I have added a test to the
5250 spa_base64_to_bits() function which I hope fixes it.
5253 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5254 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5256 63. The daemon start-up calls getloadavg() while still root for those OS that
5257 need the first call to be done as root, but it missed one case: when
5258 deliver_queue_load_max is set with deliver_drop_privilege. This is
5259 necessary for the benefit of the queue runner, because there is no re-exec
5260 when deliver_drop_privilege is set.
5262 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5263 This has been fixed.
5265 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5266 was in use, was not putting the data itself into the right store pool;
5267 consequently, it could be overwritten for a subsequent message in the same
5268 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5271 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5273 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5274 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5275 the function were treating the return as a boolean value, which happened to
5276 work because 0=false and not-0=true, but is not correct code.
5278 68. The host_aton() function was not handling scoped IPv6 addresses (those
5279 with, for example, "%eth0" on the end) correctly.
5281 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5282 negated items (that is, ~something) in unsigned ints. Some compilers
5283 apparently mutter when there is no cast.
5285 70. If an address verification called from an ACL failed, and did not produce a
5286 user-specific message (i.e. there was only a "system" message), nothing was
5287 put in $acl_verify_message. In this situation, it now puts the system
5290 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5291 session; change 4.31/43 added the unwanted input to the log line - except
5292 that it did not do this in the start of session case. It now does.
5294 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5295 This is wrong and can cause the other end to generate a synchronization
5296 error if it is another Exim or anything else that does the synchronization
5297 check. A QUIT command is no longer sent after a timeout.
5299 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5300 during host lookups.
5302 74. The defer_ok option of callout verification was not working if it was used
5303 when verifying addresses in header lines, that is, for this case:
5305 verify = header_sender/callout=defer_ok
5307 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5308 those file descriptors could be used for SMTP connections. If anything
5309 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5310 could be sent to the SMTP client, causing chaos. The daemon now opens
5311 stdin, stdout, and stderr to /dev/null when it puts itself into the
5314 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5315 log by default. The user can override this with suitable Perl magic.
5317 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5318 the log message when discard triggers, was not working for the DATA ACL or
5319 for the non-SMTP ACL.
5321 78. Error message wording change in sieve.c.
5323 79. If smtp_accept_max_per_host was set, the number of connections could be
5324 restricted to fewer than expected, because the daemon was trying to set up
5325 a new connection before checking whether the processes handling previous
5326 connections had finished. The check for completed processes is now done
5327 earlier. On busy systems, this bug wouldn't be noticed because something
5328 else would have woken the daemon, and it would have reaped the completed
5331 80. If a message was submitted locally by a user whose login name contained one
5332 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5333 It caused a spool format error. I have fixed the spool reading code. A
5334 related problem was that the "from" clause in the Received: line became
5335 illegal because of the space(s). It is now covered by ${quote_local_part.
5337 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5340 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5342 83. Patch from Sieve maintainer to fix unterminated string problem in
5343 "vacation" handling.
5345 84. Some minor changes to the Linux configuration files to help with other
5346 OS variants using glibc.
5348 85. One more patch for Sieve to update vacation handling to latest spec.
5351 ----------------------------------------------------
5352 See the note above about the 4.44 and 4.50 releases.
5353 ----------------------------------------------------
5359 1. Change 4.43/35 introduced a bug that caused file counts to be
5360 incorrectly computed when quota_filecount was set in an appendfile
5363 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5364 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5367 3. Give more explanation in the error message when the command for a transport
5368 filter fails to execute.
5370 4. There are several places where Exim runs a non-Exim command in a
5371 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5372 was being done only for the command run by the queryprogram router. It is
5373 now done for all such subprocesses. The other cases are: ${run, transport
5374 filters, and the commands run by the lmtp and pipe transports.
5376 5. Some older OS have a limit of 256 on the maximum number of file
5377 descriptors. Exim was using setrlimit() to set 1000 as a large value
5378 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5379 systems. I've change it so that if it can't get 1000, it tries for 256.
5381 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5382 was an oversight, and furthermore, ever since the addition of extra
5383 controls (e.g. 4.43/32), the checks on when to allow different forms of
5384 "control" were broken. There should now be diagnostics for all cases when a
5385 control that does not make sense is encountered.
5387 7. $recipients is now available in the predata ACL (oversight).
5389 8. Tidy the search cache before the fork to do a delivery from a message
5390 received from the command line. Otherwise the child will trigger a lookup
5391 failure and thereby defer the delivery if it tries to use (for example) a
5392 cached ldap connection that the parent has called unbind on.
5394 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5395 of $address_data from the recipient verification was clobbered by the
5396 sender verification.
5398 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5399 was its contents. (It was OK if the option was not defined at all.)
5401 11. A "Completed" log line is now written for messages that are removed from
5402 the spool by the -Mrm option.
5404 12. $host_address is now set to the target address during the checking of
5405 ignore_target_hosts.
5407 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5408 being passed; this would have caused $sender_host_name to have been used if
5409 matching the list had actually called for a host name (not very likely,
5410 since this list is usually IP addresses). A host name is now passed as
5413 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5414 code that specifies a non-restarting handler (typically sigaction() in
5415 modern systems) in an attempt to fix a rare and obscure crash bug.
5417 15. Narrowed the window for a race in the daemon that could cause it to ignore
5418 SIGCHLD signals. This is not a major problem, because they are used only to
5419 wake it up if nothing else does.
5421 16. A malformed maildirsize file could cause Exim to calculate negative values
5422 for the mailbox size or file count. Odd effects could occur as a result.
5423 The maildirsize information is now recalculated if the size or filecount
5426 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5427 support for a long time. Removed HAVE_SYS_VFS_H.
5429 18. Updated exipick to current release from John Jetmore.
5431 19. Allow an empty sender to be matched against a lookup in an address list.
5432 Previously the only cases considered were a regular expression, or an
5435 20. Exim went into a mad DNS lookup loop when doing a callout where the
5436 host was specified on the transport, if the DNS lookup yielded more than
5439 21. The RFC2047 encoding function was originally intended for short strings
5440 such as real names; it was not keeping to the 75-character limit for
5441 encoded words that the RFC imposes. It now respects the limit, and
5442 generates multiple encoded words if necessary. To be on the safe side, I
5443 have increased the buffer size for the ${rfc2047: expansion operator from
5446 22. Failure to deliver a bounce message always caused it to be frozen, even if
5447 there was an errors_to setting on the router. The errors_to setting is now
5450 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5451 canonical form (fully expanded) before being placed in
5452 $sender_host_address.
5454 24. Updated eximstats to version 1.33
5456 25. Include certificate and key file names in error message when GnuTLS fails
5457 to set them up, because the GnuTLS error message doesn't include the name
5458 of the failing file when there is a problem reading it.
5460 26. Expand error message when OpenSSL has problems setting up cert/key files.
5463 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5464 (this can affect the format of dates).
5466 28. exim_tidydb, when checking for the continued existence of a message for
5467 which it has found a message-specific retry record, was not finding
5468 messages that were in split spool directories. Consequently, it was
5469 deleting retry records that should have stayed in existence.
5471 29. eximstats updated to version 1.35
5472 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5473 1.35 - bugfix such that pie charts by volume are generated correctly
5475 30. The SPA authentication driver was not abandoning authentication and moving
5476 on to the next authenticator when an expansion was forced to fail,
5477 contradicting the general specification for all authenticators. Instead it
5478 was generating a temporary error. It now behaves as specified.
5480 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5481 (the order specifies the preference for clients). The order is now AES256,
5482 AES128, 3DES, ARCFOUR128.
5484 31. Small patch to Sieve code - explicitly set From: when generating an
5487 32. Exim crashed if a remote delivery caused a very long error message to be
5488 recorded - for instance if somebody sent an entire SpamAssassin report back
5489 as a large number of 550 error lines. This bug was coincidentally fixed by
5490 increasing the size of one of Exim's internal buffers (big_buffer) that
5491 happened as part of the Exiscan merge. However, to be on the safe side, I
5492 have made the code more robust (and fixed the comments that describe what
5495 33. Some experimental protocols are using DNS PTR records for new purposes. The
5496 keys for these records are domain names, not reversed IP addresses. The
5497 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5498 leaves it alone. Component reversal etc. now happens only for IP addresses.
5501 34. The host_aton() function is supposed to be passed a string that is known
5502 to be a valid IP address. However, in the case of IPv6 addresses, it was
5503 not checking this. This is a hostage to fortune. Exim now panics and dies
5504 if the condition is not met. A case was found where this could be provoked
5505 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5506 components; fortuitously, this particular loophole had already been fixed
5507 by change 4.50/55 or 4.44/33 above.
5509 If there are any other similar loopholes, the new check in host_aton()
5510 itself should stop them being exploited. The report I received stated that
5511 data on the command line could provoke the exploit when Exim was running as
5512 exim, but did not say which command line option was involved. All I could
5513 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5514 running as the user.
5517 35. There was a buffer overflow vulnerability in the SPA authentication code
5518 (which came originally from the Samba project). I have added a test to the
5519 spa_base64_to_bits() function which I hope fixes it.
5522 36. The daemon start-up calls getloadavg() while still root for those OS that
5523 need the first call to be done as root, but it missed one case: when
5524 deliver_queue_load_max is set with deliver_drop_privilege. This is
5525 necessary for the benefit of the queue runner, because there is no re-exec
5526 when deliver_drop_privilege is set.
5528 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5529 was in use, was not putting the data itself into the right store pool;
5530 consequently, it could be overwritten for a subsequent message in the same
5531 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5534 38. Sometimes the final signoff response after QUIT could fail to get
5535 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5536 < 0 before doing a fflush(). This bug looks as though it goes back to the
5537 introduction of TLS in release 3.20, but "sometimes" must have been rare
5538 because the tests only now provoked it.
5544 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5545 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5546 Now it returns an integer. A number of calls were still expecting a T/F
5547 return. Fortuitously, in all cases, the tests worked in OK situations,
5548 which is the norm. However, things would have gone wrong on any write
5549 failures on the smtp file descriptor. This function is used when sending
5550 messages over SMTP and also when doing verify callouts.
5552 2. When Exim is called to do synchronous delivery of a locally submitted
5553 message (the -odf or -odi options), it no longer closes stderr before doing
5556 3. Implemented the mua_wrapper option.
5558 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5560 5. Implemented the functions header_remove(), header_testname(),
5561 header_add_at_position(), and receive_remove_recipient(), and exported them
5564 6. If an ACL "warn" statement specified the addition of headers, Exim already
5565 inserted X-ACL-Warn: at the start if there was no header name. However, it
5566 was not making this test for the second and subsequent header lines if
5567 there were newlines in the string. This meant that an invalid header could
5568 be inserted if Exim was badly configured.
5570 7. Allow an ACL "warn" statement to add header lines at the start or after all
5571 the Received: headers, as well as at the end.
5573 8. Added the rcpt_4xx retry error code.
5575 9. Added postmaster_mailfrom=xxx to callout verification option.
5577 10. Added mailfrom=xxxx to the callout verification option, for verify=
5580 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5581 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5583 12. Inserted some casts to stop certain compilers warning when using pointer
5584 differences as field lengths or precisions in printf-type calls (mostly
5585 affecting debugging statements).
5587 13. Added optional readline() support for -be (dynamically loaded).
5589 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5590 same clock tick as a message's arrival, so that its received time was the
5591 same as the "first fail" time on the retry record, and that message
5592 remained on the queue past the ultimate address timeout, every queue runner
5593 would try a delivery (because it was past the ultimate address timeout) but
5594 after another failure, the ultimate address timeout, which should have then
5595 bounced the address, did not kick in. This was a "< instead of <=" error;
5596 in most cases the first failure would have been in the next clock tick
5597 after the received time, and all would be well.
5599 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5600 being recognized when the domain list was tested by the match_domain
5601 condition in an expansion string.
5603 16. Added the ${str2b64: operator.
5605 17. Exim was always calling setrlimit() to set a large limit for the number of
5606 processes, without checking whether the existing limit was already
5607 adequate. (It did check for the limit on file descriptors.) Furthermore,
5608 errors from getrlimit() and setrlimit() were being ignored. Now they are
5609 logged to the main and panic logs, but Exim does carry on, to try to do its
5610 job under whatever limits there are.
5612 18. Imported PCRE 5.0.
5614 19. Trivial typo in log message " temporarily refused connection" (the leading
5617 20. If the log selector return_path_on_delivery was set and an address was
5618 redirected to /dev/null, the delivery process crashed because it assumed
5619 that a return path would always be set for a "successful" delivery. In this
5620 case, the whole delivery is bypassed as an optimization, and therefore no
5623 21. Internal re-arrangement: the function for sending a challenge and reading
5624 a response while authentication was assuming a zero-terminated challenge
5625 string. It's now changed to take a pointer and a length, to allow for
5626 binary data in such strings.
5628 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5630 23. Exim was not respecting finduser_retries when seeking the login of the
5631 uid under which it was called; it was always trying 10 times. (The default
5632 setting of finduser_retries is zero.) Also, it was sleeping after the final
5633 failure, which is pointless.
5635 24. Implemented tls_on_connect_ports.
5637 25. Implemented acl_smtp_predata.
5639 26. If the domain in control=submission is set empty, Exim assumes that the
5640 authenticated id is a complete email address when it generates From: or
5641 Sender: header lines.
5643 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5644 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5645 chown and chgrp in /bin and hostname in /usr/bin.
5647 28. Exim was keeping the "process log" file open after each use, just as it
5648 does for the main log. This opens the possibility of it remaining open for
5649 long periods when the USR1 signal hits a daemon. Occasional processlog
5650 errors were reported, that could have been caused by this. Anyway, it seems
5651 much more sensible not to leave this file open at all, so that is what now
5654 29. The long-running daemon process does not normally write to the log once it
5655 has entered its main loop, and it closes the log before doing so. This is
5656 so that log files can straightforwardly be renamed and moved. However,
5657 there are a couple of unusual error situations where the daemon does write
5658 log entries, and I had neglected to close the log afterwards.
5660 30. The text of an SMTP error response that was received during a remote
5661 delivery was being truncated at 512 bytes. This is too short for some of
5662 the long messages that one sometimes sees. I've increased the limit to
5665 31. It is now possible to make retry rules that apply only when a message has a
5666 specific sender, in particular, an empty sender.
5668 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5669 it possible to be selective about when SMTP synchronization is enforced.
5671 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5673 32. Implemented hosts_connection_nolog.
5675 33. Added an ACL for QUIT.
5677 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5680 35. Added mailbox_size and mailbox_filecount to appendfile.
5682 36. Added control = no_multiline_responses to ACLs.
5684 37. There was a bug in the logic of the code that waits for the clock to tick
5685 in the case where the clock went backwards by a substantial amount such
5686 that the microsecond fraction of "now" was more than the microsecond
5687 fraction of "then" (but the whole seconds number was less).
5689 38. Added support for the libradius Radius client library this is found on
5690 FreeBSD (previously only the radiusclient library was supported).
5696 1. When certain lookups returned multiple values in the form name=value, the
5697 quoting of the values was not always being done properly. Specifically:
5698 (a) If the value started with a double quote, but contained no whitespace,
5700 (b) If the value contained whitespace other than a space character (i.e.
5701 tabs or newlines or carriage returns) it was not quoted.
5702 This fix has been applied to the mysql and pgsql lookups by writing a
5703 separate quoting function and calling it from the lookup code. The fix
5704 should probably also be applied to nisplus, ibase and oracle lookups, but
5705 since I cannot test any of those, I have not disturbed their existing code.
5707 2. A hit in the callout cache for a specific address caused a log line with no
5708 reason for rejecting RCPT. Now it says "Previous (cached) callout
5709 verification failure".
5711 3. There was an off-by-one bug in the queryprogram router. An over-long
5712 return line was truncated at 256 instead of 255 characters, thereby
5713 overflowing its buffer with the terminating zero. As well as fixing this, I
5714 have increased the buffer size to 1024 (and made a note to document this).
5716 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5717 when Exim is waiting for an SMTP response from a remote server, Exim
5718 restarts its select() call on the socket, thereby resetting its timeout.
5719 This is not a problem when such interrupts are rare. Somebody set up a cron
5720 job to run exiwhat every 2 minutes, which is less than the normal select()
5721 timeout (5 or 10 minutes). This meant that the select() timeout never
5722 kicked in because it was always reset. I have fixed this by comparing the
5723 time when an interrupt arrives with the time at the start of the first call
5724 to select(). If more time than the timeout has elapsed, the interrupt is
5725 treated as a timeout.
5727 5. Some internal re-factoring in preparation for the addition of Sieve
5728 extensions (by MH). In particular, the "personal" test is moved to a
5729 separate function, and given an option for scanning Cc: and Bcc: (which is
5730 not set for Exim filters).
5732 6. When Exim created an email address using the login of the caller as the
5733 local part (e.g. when creating a From: or Sender: header line), it was not
5734 quoting the local part when it contained special characters such as @.
5736 7. Installed new OpenBSD configuration files.
5738 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5739 try to make them clearer.
5741 9. Callout options, other than the timeout value, were being ignored when
5742 verifying sender addresses in header lines. For example, when using
5744 verify = header_sender/callout=no_cache
5746 the cache was (incorrectly) being used.
5748 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5749 only the Cygwin environment.
5751 11. When return_path_on_delivery was set as a log selector, if different remote
5752 addresses in the same message used different return paths and parallel
5753 remote delivery occurred, the wrong values would sometimes be logged.
5754 (Whenever a remote delivery process finished, the return path value from
5755 the most recently started remote delivery process was logged.)
5757 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5758 lines when AUTH and/or TLS are in use. This is the "received protocol"
5759 field. Exim used to use "asmtp" for authenticated SMTP, without any
5760 indication (in the protocol name) for TLS use. Now it follows the RFC and
5761 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5762 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5763 names appear in log lines as well as in Received: header lines.
5765 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5766 extensions, and comparison tests, and to fix some bugs.
5768 14. Changes to the "personal" filter test:
5770 (1) The test was buggy in that it was just doing the equivalent of
5771 "contains" tests on header lines. For example, if a user's address was
5772 anne@some.where, the "personal" test would incorrectly be true for
5774 To: susanne@some.where
5776 This test is now done by extracting each address from the header in turn,
5777 and checking the entire address. Other tests that are part of "personal"
5778 are now done using regular expressions (for example, to check local parts
5779 of addresses in From: header lines).
5781 (2) The list of non-personal local parts in From: addresses has been
5782 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5783 taken from the Sieve specification recommendations.
5785 (3) If the message contains any header line starting with "List-" it is
5786 treated as non-personal.
5788 (4) The test for "circular" in the Subject: header line has been removed
5789 because it now seems ill-conceived.
5791 15. Minor typos in src/EDITME comments corrected.
5793 16. Installed latest exipick from John Jetmore.
5795 17. If headers_add on a router specified a text string that was too long for
5796 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5797 of string_sprintf() is now avoided.
5799 18. $message_body_size was not set (it was always zero) when running the DATA
5800 ACL and the local_scan() function.
5802 19. For the "mail" command in an Exim filter, no default was being set for
5803 the once_repeat time, causing a random time value to be used if "once" was
5804 specified. (If the value happened to be <= 0, no repeat happened.) The
5805 default is now 0s, meaning "never repeat". The "vacation" command was OK
5806 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5807 (I found it when inspecting the code).
5809 20. There is now an overall timeout for performing a callout verification. It
5810 defaults to 4 times the callout timeout, which applies to individual SMTP
5811 commands during the callout. The overall timeout applies when there is more
5812 than one host that can be tried. The timeout is checked before trying the
5813 next host. This prevents very long delays if there are a large number of
5814 hosts and all are timing out (e.g. when the network connections are timing
5815 out). The value of the overall timeout can be changed by specifying an
5816 additional sub-option for "callout", called "maxwait". For example:
5818 verify = sender/callout=5s,maxwait=20s
5820 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5821 to the end before writing, but this should make it even safer).
5823 22. Exim was forgetting that it had advertised PIPELINING for the second and
5824 subsequent messages on an SMTP connection. It was also not resetting its
5825 memory on STARTTLS and an internal HELO.
5827 23. When Exim logs an SMTP synchronization error within a session, it now
5828 records whether PIPELINING has been advertised or not.
5830 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5831 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5832 rather than long int.
5834 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5836 26. Added the never_mail option to autoreply.
5842 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5843 crash if the getsockname() call failed; this can happen if a connection is
5844 closed very soon after it is established. The problem was simply in the
5845 order in which certain operations were done, causing Exim to try to write
5846 to the SMTP stream before it had set up the file descriptor. The bug has
5847 been fixed by making things happen in the correct order.
5853 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5854 before the connection was closed, thus losing the rejection response.
5856 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5857 some early Solaris releases, but causes trouble in current releases where
5858 socklen_t is defined.
5860 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5863 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5866 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5867 that releasing the top of it at the end releases what was used for sub-
5868 expansions (unless the block got too big). However, discard this block if
5869 the first thing is a variable or header, so that we can use its block when
5870 it is dynamic (useful for very large $message_headers, for example).
5872 6. Lookups now cache *every* query, not just the most recent. A new, separate
5873 store pool is used for this. It can be recovered when all lookup caches are
5874 flushed. Lookups now release memory at the end of their result strings.
5875 This has involved some general refactoring of the lookup sources.
5877 7. Some code has been added to the store_xxx() functions to reduce the amount
5878 of flapping under certain conditions.
5880 8. log_incoming_interface used to affect only the <= reception log lines. Now
5881 it causes the local interface and port to be added to several more SMTP log
5882 lines, for example "SMTP connection from", and rejection lines.
5884 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5886 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5888 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5889 could be overwritten at the end of the current message (or the start of a
5890 new message if it was set in a HELO ACL). The value is now preserved for
5891 the duration of the SMTP connection.
5893 12. If a transport had a headers_rewrite setting, and a matching header line
5894 contained an unqualified address, that address was qualified, even if it
5895 did not match any rewriting rules. The underlying bug was that the values
5896 of the flags that permit the existence of unqualified sender and recipient
5897 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5898 non-local messages, and by -bnq for local messages) were not being
5899 preserved with the message after it was received.
5901 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5902 "next input=" as part of the text comprising the host identity instead of
5903 the correct text. The code was using the same buffer for two different
5904 strings. However, depending on which order the printing function evaluated
5905 its arguments, the bug did not always show up. Under Linux, for example, my
5906 test suite worked just fine.
5908 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5909 doesn't work with some older versions of Perl. It has been changed to "my",
5910 which in any case is probably the better facility to use.
5912 15. A really picky compiler found some instances of statements for creating
5913 error messages that either had too many or two few arguments for the format
5916 16. The size of the buffer for calls to the DNS resolver has been increased
5917 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5918 for addresses that have a lot of PTR records. This alleviates a problem; it
5919 does not fully solve it.
5921 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5922 buffer now truncates the list and logs the incident, which is the same
5923 action as happens when Exim is looking up a host name and its aliases.
5924 Previously in this situation something unpredictable would happen;
5925 sometimes it was "internal error: store_reset failed".
5927 18. If a server dropped the connection unexpectedly when an Exim client was
5928 using GnuTLS and trying to read a response, the client delivery process
5929 crashed while trying to generate an error log message.
5931 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5932 string, for example:
5934 warn message = H1: something\nH2: something
5936 the text was added as a single header line from Exim's point of view
5937 though it ended up OK in the delivered message. However, searching for the
5938 second and subsequent header lines using $h_h2: did not work. This has been
5939 fixed. Similarly, if a system filter added multiple headers in this way,
5940 the routers could not see them.
5942 20. Expanded the error message when iplsearch is called with an invalid key to
5943 suggest using net-iplsearch in a host list.
5945 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5946 ACLs are no longer actually imposed (and a message to that effect is
5949 22. If a "gecos" field in a passwd entry contained escaped characters, in
5950 particular, if it contained a \" sequence, Exim got it wrong when building
5951 a From: or a Sender: header from that name. A second bug also caused
5952 incorrect handling when an unquoted " was present following a character
5953 that needed quoting.
5955 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5956 was not being matched caselessly.
5958 24. Arranged for all hyphens in the exim.8 source to be escaped with
5961 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5962 properly. Recipient callout cache records were still being keyed to include
5963 the sender, even when use_sender was set false. This led to far more
5964 callouts that were necessary. The sender is no longer included in the key
5965 when use_sender is false.
5967 26. Added "control = submission" modifier to ACLs.
5969 27. Added the ${base62d: operator to decode base 62 numbers.
5971 28. dnsdb lookups can now access SRV records.
5973 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5974 the configuration file.
5976 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5977 (-v) mode. This makes the output for a verbose queue run more intelligible.
5979 31. Added a use_postmaster feature to recipient callouts.
5981 32. Added the $body_zerocount variable, containing the number of binary zero
5982 bytes in the message body.
5984 33. The time of last modification of the "new" subdirectory is now used as the
5985 "mailbox time last read" when there is a quota error for a maildir
5988 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5990 35. Added +ignore_unknown as a special item in host lists.
5992 36. Code for decoding IPv6 addresses in host lists is now included, even if
5993 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5994 address was recognized as an IP address, but was then not correctly decoded
5995 into binary, causing unexpected and incorrect effects when compared with
6002 1. Very minor rewording of debugging text in manualroute to say "list of
6003 hosts" instead of "hostlist".
6005 2. If verify=header_syntax was set, and a header line with an unqualified
6006 address (no domain) and a large number of spaces between the end of the
6007 name and the colon was received, the reception process suffered a buffer
6008 overflow, and (when I tested it) crashed. This was caused by some obsolete
6009 code that should have been removed. The fix is to remove it!
6011 3. When running in the test harness, delay a bit after writing a bounce
6012 message to get a bit more predictability in the log output.
6014 4. Added a call to search_tidyup() just before forking a reception process. In
6015 theory, someone could use a lookup in the expansion of smtp_accept_max_
6016 per_host which, without the tidyup, could leave open a database connection.
6018 5. Added the variables $recipient_data and $sender_data which get set from a
6019 lookup success in an ACL "recipients" or "senders" condition, or a router
6020 "senders" option, similar to $domain_data and $local_part_data.
6022 6. Moved the writing of debug_print from before to after the "senders" test
6025 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6026 problems for message scanning, either using a data ACL, or using
6027 local_scan() because the Received: header was not generated till after they
6028 were called (in order to set the time as the time of reception completion).
6029 I have revised the way this works. The header is now generated after the
6030 body is received, but before the ACL or local_scan() are called. After they
6031 are run, the timestamp in the header is updated.
6037 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6038 before starting a queue runner without re-exec. This happened only when
6039 deliver_drop_privilege was set or when the Exim user was set to root. The
6040 effect of the bug was that timeouts during subsequent deliveries caused
6041 crashes instead of being properly handled. The handler is now left at its
6042 default (and expected) setting.
6044 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6045 message, again when deliver_drop_privilege is set or Exim is run as root.
6046 The bug described in (1) was not present in this case, but the tidying up
6047 of the other signals was missing. I have made the two cases consistent.
6049 3. The ignore_target_hosts setting on a manualroute router was being ignored
6050 for hosts that were looked up using the /MX notation.
6052 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6055 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6056 operated on the sender address. After changing the $sender_address to <>
6057 for the sender address verify, Exim was re-instated it as the original
6058 (before rewriting) address, but remembering that it had rewritten it, so it
6059 wasn't rewriting it again. This bug also had the effect of breaking the
6060 sender address verification caching when the sender address was rewritten.
6062 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6063 This has been changed so that if the ip literal address matches
6064 ignore_target_hosts, the router declines.
6066 7. Added expansion conditions match_domain, match_address, and match_local_
6067 part (NOT match_host).
6069 8. The placeholder for the Received: header didn't have a length field set.
6071 9. Added code to Exim itself and to exim_lock to test for a specific race
6072 condition that could lead to file corruption when using MBX delivery. The
6073 issue is with the lockfile that is created in /tmp. If this file is removed
6074 after a process has opened it but before that process has acquired a lock,
6075 there is the potential for a second process to recreate the file and also
6076 acquire a lock. This could lead to two Exim processes writing to the file
6077 at the same time. The added code performs the same test as UW imapd; it
6078 checks after acquiring the lock that its file descriptor still refers to
6079 the same named file.
6081 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6082 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6085 11. Added the smtp_active_hostname option. If used, this will typically be made
6086 to depend on the incoming interface address. Because $interface_address is
6087 not set up until the daemon has forked a reception process, error responses
6088 that can happen earlier (such as "too many connections") no longer contain
6091 12. If an expansion in a condition on a "warn" statement fails because a lookup
6092 defers, the "warn" statement is abandoned, and the next ACL statement is
6093 processed. Previously this caused the whole ACL to be aborted.
6095 13. Added the iplsearch lookup type.
6097 14. Added ident_timeout as a log selector.
6099 15. Added tls_certificate_verified as a log selector.
6101 16. Added a global option tls_require_ciphers (compare the smtp transport
6102 option of the same name). This controls incoming TLS connections.
6104 17. I finally figured out how to make tls_require_ciphers do a similar thing
6105 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6106 before starting the TLS session.
6108 18. Tabs are now shown as \t in -bP output.
6110 19. If the log selector return_path_on_delivery was set, Exim crashed when
6111 bouncing a message because it had too many Received: header lines.
6113 20. If two routers both had headers_remove settings, and the first one included
6114 a superfluous trailing colon, the final name in the first list and the
6115 first name in the second list were incorrectly joined into one item (with a
6116 colon in the middle).
6122 1. Added -C and -D options to the exinext utility, mainly to make it easier
6123 to include in the automated testing, but these could be helpful when
6124 multiple configurations are in use.
6126 2. The exinext utility was not formatting the output nicely when there was
6127 an alternate port involved in the retry record key, nor when there was a
6128 message id as well (for retries that were specific to a specific message
6129 and a specific host). It was also confused by IPv6 addresses, because of
6130 the additional colons they contain. I have fixed the IPv4 problem, and
6131 patched it up to do a reasonable job for IPv6.
6133 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6134 delivery, the log line now contains "pipelined" if PIPELINING was used.
6136 4. An SMTP transport process used to panic and die if the bind() call to set
6137 an explicit outgoing interface failed. This has been changed; it is now
6138 treated in the same way as a connect() failure.
6140 5. A reference to $sender_host_name in the part of a conditional expansion
6141 that was being skipped was still causing a DNS lookup. This no longer
6144 6. The def: expansion condition was not recognizing references to header lines
6145 that used bh_ and bheader_.
6147 7. Added the _cache feature to named lists.
6149 8. The code for checking quota_filecount in the appendfile transport was
6150 allowing one more file than it should have been.
6152 9. For compatibility with Sendmail, the command line option
6160 and sets the incoming protocol and host name (for trusted callers). The
6161 host name and its colon can be omitted when only the protocol is to be set.
6162 Note the Exim already has two private options, -pd and -ps, that refer to
6163 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6164 "s", but I don't think that's a major issue.
6166 10. A number of refactoring changes to the code, none of which should affect
6169 (a) The number of logging options was getting close to filling up the
6170 32-bit word that was used as a bit map. I have split them into two classes:
6171 those that are passed in the argument to log_write(), and those that are
6172 only ever tested independently outside of that function. These are now in
6173 separate 32-bit words, so there is plenty of room for expansion again.
6174 There is no change in the user interface or the logging behaviour.
6176 (b) When building, for example, log lines, the code previously used a
6177 macro that called string_cat() twice, in order to add two strings. This is
6178 not really sufficiently general. Furthermore, there was one instance where
6179 it was actually wrong because one of the argument was used twice, and in
6180 one call a function was used. (As it happened, calling the function twice
6181 did not affect the overall behaviour.) The macro has been replaced by a
6182 function that can join an arbitrary number of extra strings onto a growing
6185 (c) The code for expansion conditions now uses a table and a binary chop
6186 instead of a serial search (which was left over from when there were very
6187 few conditions). Also, it now recognizes conditions like "pam" even when
6188 the relevant support is not compiled in: a suitably worded error message is
6189 given if an attempt is made to use such a condition.
6191 11. Added ${time_interval:xxxxx}.
6193 12. A bug was causing one of the ddress fields not to be passed back correctly
6194 from remote delivery subprocesses. The field in question was not being
6195 subsequently used, so this caused to problems in practice.
6197 13. Added new log selectors queue_time and deliver_time.
6199 14. Might have fixed a bug in maildirsizefile handling that threw up
6200 "unexpected character" debug warnings, and recalculated the data
6201 unnecessarily. In any case, I expanded the warning message to give more
6204 15. Added the message "Restricted characters in address" to the statements in
6205 the default ACL that block characters like @ and % in local parts.
6207 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6208 Three changes have been made:
6210 (a) There was a serious bug; a negative response to MAIL caused the whole
6211 recipient domain to be cached as invalid, thereby blocking all messages
6212 to all local parts at the same domain, from all senders. This bug has
6213 been fixed. The domain is no longer cached after a negative response to
6214 MAIL if the sender used is not empty.
6216 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6219 (c) A new callout option, "use_sender" has been added for people who want
6220 the modified behaviour.
6226 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6229 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6230 indeed breaks things for older releases.
6232 3. Added additional logging to the case where there is a problem reading data
6233 from a filter that is running in a subprocess using a pipe, in order to
6234 try to track down a specific problem.
6236 4. Testing facility fudge: when running in the test harness and attempting
6237 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6238 getting "No route to host". Convert this to a timeout.
6240 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6243 6. Some OS don't have socklen_t but use size_t instead. This affects the
6244 fifth argument of getsockopt() amongst other things. This is now
6245 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6246 can be set for individual OS. I have set it for SunOS5, OSF1, and
6247 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6248 some earlier ones do not.
6250 7. Change 4.30/15 was not doing the test caselessly.
6252 8. The standard form for an IPv6 address literal was being rejected by address
6253 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6254 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6255 this, as well as the form without the "IPv6" on the front (but only when
6256 address literals are enabled, of course).
6258 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6260 10. Exim crashed if a message with an empty sender address specified by -f
6261 encountered a router with an errors_to setting. This could be provoked only
6262 by a command such as
6266 where an empty string was supplied; "<>" did not hit this bug.
6268 11. Installed PCRE release 4.5.
6270 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6271 remained set. It is now erased.
6273 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6274 times from message ids (which are base 36 rather than the normal 62).
6276 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6277 were being counted as actual protocol errors, and logged if the log
6278 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6279 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6280 and DATA following a set of rejected RCPTs do not count as protocol errors.
6281 In other words, Exim assumes they were pipelined, though this may not
6282 actually be the case. Of course, in all cases the client gets an
6283 appropriate error code.
6285 15. If a lookup fails in an ACL condition, a message about the failure may
6286 be available; it is used if testing the ACL cannot continue, because most
6287 such messages specify what the cause of the deferral is. However, some
6288 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6289 that caused an old message to be retained and used if a later statement
6290 caused a defer, replacing the real cause of the deferral.
6292 16. If an IP address had so many PTR records that the DNS lookup buffer
6293 was not large enough to hold them, Exim could crash while trying to process
6294 the truncated data. It now detects and logs this case.
6296 17. Further to 4.21/58, another change has been made: if (and only if) the
6297 first line of a message (the first header line) ends with CRLF, a bare LF
6298 in a subsequent header line has a space inserted after it, so as not to
6299 terminate the header.
6301 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6302 unnecessarily, used atoi() instead of strtol(), and didn't check the
6303 termination when getting file sizes from file names by regex.
6305 19. Completely re-implemented the support for maildirsize files, in the light
6306 of a number of problems with the previous contributed implementation
6307 (4.30/29). In particular:
6309 . If the quota is zero, the maildirsize file is maintained, but no quota is
6312 . If the maildir directory does not exist, it is created before any attempt
6313 to write a maildirsize file.
6315 . The quota value in the file is just a cache; if the quota is changed in
6316 the transport, the new value overrides.
6318 . A regular expression is available for excluding directories from the
6321 20. The autoreply transport checks the characters in options that define the
6322 message's headers; it allows continued headers, but it was checking with
6323 isspace() after an embedded newline instead of explicitly looking for a
6326 21. If all the "regular" hosts to which an address was routed had passed their
6327 expiry times, and had not reached their retry times, the address was
6328 bounced, even if fallback hosts were defined. Now Exim should go on to try
6331 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6332 equivalent code in the SMTP transport. Some hosts send humungous responses
6333 to HELO/EHLO, more than 1024 it seems.
6335 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6336 gives compiler warnings in some environments. I've now done it "properly",
6339 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6340 (because of problems with the built-in one) was declared to return uschar *
6341 instead of char *, causing compiler failure.
6343 25. Fixed a file descriptor leak when processing alias/forward files.
6345 26. Fixed a minor format string issue in dbfn.c.
6347 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6349 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6350 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6353 29. When a sender address is verified, it is cached, to save repeating the test
6354 when there is more than one recipient in a message. However, when the
6355 verification involves a callout, it is possible for different callout
6356 options to be set for different recipients. It is too complicated to keep
6357 track of this in the cache, so now Exim always runs a verification when a
6358 callout is required, relying on the callout cache for the optimization.
6359 The overhead is duplication of the address routing, but this should not be
6362 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6363 to be verified with callout=postmaster, and the main callout worked but the
6364 postmaster check failed, the verification correctly failed. However, if a
6365 subsequent RCPT command asked for sender verification *without* the
6366 postmaster check, incorrect caching caused this verification also to fail,
6369 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6370 it was not caching the DNS options (qualify_single, search_parents) that
6371 were used when the lookup failed. A subsequent lookup with different
6372 options therefore always gave the same answer, though there were cases
6373 where it should not have. (Example: a "domains = !$mx_any" option on a
6374 dnslookup router: the "domains" option is always processed without any
6375 widening, but the router might have qualify_single set.) Now Exim uses the
6376 cached value only when the same options are set.
6378 32. Added John Jetmore's "exipick" utility to the distribution.
6380 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6381 than a timeout (e.g. a certificate is required, and is not provided), an
6382 Exim server now closes the connection immediately. Previously it waited for
6383 the client to close - but if the client is SSL, it seems that they each
6384 wait for each other, leading to a delay before one of them times out.
6386 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6387 maintained 0.8.x compatibility because I don't think many are using it, and
6388 it is clearly obsolete.
6390 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6393 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6394 client certificate was expired. A simple patch fixes this, though I don't
6395 understand the full logic of why the verify callback is called multiple
6398 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6399 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6400 which causes problems with some clients (such as the Certicom SSL Plus
6401 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6402 disables the coutermeasure allowing Eudora to connect."
6404 38. Exim was not checking that a write() to a log file succeeded. This could
6405 lead to Bad Things if a log got too big, in particular if it hit a file
6406 size limit. Exim now panics and dies if it cannot write to a log file, just
6407 as it does if it cannot open a log file.
6409 39. Modified OS/Makefile-Linux so that it now contains
6411 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6413 The two -D definitions ensure that Exim is compiled with large file
6414 support, which makes it possible to handle log files that are bigger than
6417 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6418 instance) a domain was checked against a named list that involved a lookup,
6419 causing $domain_data to be set, then another domain was checked against the
6420 same list, then the first domain was re-checked, the value of $domain_data
6421 after the final check could be wrong. In particular, if the second check
6422 failed, it could be set empty. This bug probably also applied to
6425 41. The strip_trailing_dot option was not being applied to the address given
6426 with the -f command-line option.
6428 42. The code for reading a message's header from the spool was incrementing
6429 $received_count, but never initializing it. This meant that the value was
6430 incorrect (doubled) while delivering a message in the same process in which
6431 it was received. In the most common configuration of Exim, this never
6432 happens - a fresh exec is done - but it can happen when
6433 deliver_drop_privilege is set.
6435 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6436 it now includes up to 150 characters of the unexpected data in the log
6439 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6440 and building data strings. The size of both of these buffers was 10 000
6441 bytes - far larger than anybody would *ever* want, thought I. Needless to
6442 say, somebody hit the limit. I have increased the maximum line length to
6443 20 000 and the maximum data length of concatenated lines to 100 000. I have
6444 also fixed two bugs, because there was no checking on these buffers. Tsk,
6445 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6446 buffer is too small.
6448 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6449 lsearch lookups. Now it does.
6451 46. When parsing a route_list item in a manualroute router, a fixed-length
6452 buffer was used for the list of hosts. I made this 1024 bytes long,
6453 thinking that nobody would ever have a list of hosts that long. Wrong.
6454 Somebody had a whole pile of complicated expansion conditions, and the
6455 string was silently truncated, leading to an expansion error. It turns out
6456 that it is easier to change to an unlimited length (owing to other changes
6457 that have happened since this code was originally written) than to build
6458 structure for giving a limitation error. The length of the item that
6459 expands into the list of hosts is now unlimited.
6461 47. The lsearch lookup could not handle data where the length of text line was
6462 more than 4095 characters. Such lines were truncated, leading to shortened
6463 data being returned. It should now handle lines of any length.
6465 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6466 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6469 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6470 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6471 no longer happens in comment lines. A list of replacements is now placed
6472 at the head of all of the source files, except those whose only change is
6473 to replace PERL_COMMAND in the very first #! line.
6475 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6476 messages on the queue, with a bottom-up merge sort, using code contributed
6477 by Michael Haardt. This should make operations like -bp somewhat faster on
6478 large queues. It won't affect queue runners, except when queue_run_in_order
6481 51. Installed eximstats 1.31 in the distribution.
6483 52. Added support for SRV lookups to the dnslookup router.
6485 53. If an ACL referred to $message_body or $message_body_end, the value was not
6486 reset for any messages that followed in the same SMTP session.
6488 54. The store-handling optimization for building very long strings was not
6489 differentiating between the different store pools. I don't think this
6490 actually made any difference in practice, but I've tidied it.
6492 55. While running the routers to verify a sender address, $sender_address
6493 was still set to the sender address. This is wrong, because when routing to
6494 send a bounce to the sender, it would be empty. Therefore, I have changed
6495 it so that, while verifying a sender address, $sender_address is set to <>.
6496 (There is no change to what happens when verifying a recipient address.)
6498 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6499 target A or AAAA records (if not already returned) without resetting the
6500 qualify_single or search_parents options of the DNS resolver. These are
6501 inappropriate in this case because the targets of MX and SRV records must
6502 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6503 target that, when qualified, matched something in the local domain. These
6504 two options are now turned off when doing these lookups.
6506 57. It seems that at least some releases of Reiserfs (which does not have the
6507 concept of a fixed number of inodes) returns zero and not -1 for the
6508 number of available inodes. This interacted badly with check_spool_inodes,
6509 which assumed that -1 was the "no such thing" setting. What I have done is
6510 to check that the total number of inodes is greater than zero before doing
6511 the test of how many are available.
6513 58. When a "warn" ACL statement has a log_message modifier, the message is
6514 remembered, and not repeated. This is to avoid a lot of repetition when a
6515 message has many recipients that cause the same warning to be written.
6516 However, Exim was preserving the list of already written lines for an
6517 entire SMTP session, which doesn't seem right. The memory is now reset if a
6518 new message is started.
6520 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6521 debugging output unless log_rewrite was also set.
6523 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6524 of (int)(handle) when we know that handle contains (void *)(-1).
6526 61. The Exim daemon panic-logs an error return when it closes the incoming
6527 connection. However "connection reset by peer" seems to be common, and
6528 isn't really an error worthy of noting specially, so that particular error
6531 62. When Exim is trying to find all the local interfaces, it used to panic and
6532 die if the ioctl to get the interface flags failed. However, it seems that
6533 on at least one OS (Solaris 9) it is possible to have an interface that is
6534 included in the list of interfaces, but for which you get a failure error
6535 for this call. This happens when the interface is not "plumbed" into a
6536 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6537 failure of the "get flags" call assumes that the interface is down.
6539 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6540 makes life easier for people who are doing arithmetic on fields extracted
6541 from dates, where you often get leading zeros that should not be
6542 interpreted as octal.
6544 64. Added qualify_domain to the redirect router, to override the global
6547 65. If a pathologically long header line contained very many addresses (the
6548 report of this problem mentioned 10 000) and each of them was rewritten,
6549 Exim could use up a very large amount of memory. (It kept on making new
6550 copies of the header line as it rewrote, and never released the old ones.)
6551 At the expense of a bit more processing, the header rewriting function has
6552 been changed so that it no longer eats memory in this way.
6554 66. The generation of the Received: header has been moved from the time that a
6555 message starts to be received, to the time that it finishes. The timestamp
6556 in the Received: header should now be very close to that of the <= log
6557 line. There are two side-effects of this change:
6559 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6560 logged header lines no longer include the local Received: line, because
6561 it has not yet been created. The same applies to a copy of the message
6562 that is returned to a non-SMTP sender when a message is rejected.
6564 (b) When a filter file is tested using -bf, no additional Received: header
6565 is added to the test message. After some thought, I decided that this
6568 This change does not affect the value of $received_for. It is still set
6569 after address rewriting, but before local_scan() is called.
6571 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6573 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6574 gave an unhelpful panic error message, and a defer error. I have managed to
6575 change this behaviour so that it now rejects any supplied certificate,
6576 which seems right, as the list of acceptable certificates is empty.
6578 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6579 gave an unhelpful defer error. I have not managed to make this reject any
6580 supplied certificates, but the error message it gives is "no certificate
6581 supplied", which is not helpful.
6583 70. exigrep's output now also includes lines that are not associated with any
6584 message, but which match the given pattern. Implemented by a patch from
6585 Martin Sluka, which also tidied up the Perl a bit.
6587 71. Recipient callout verification, like sender verification, was using <> in
6588 the MAIL FROM command. This isn't really the right thing, since the actual
6589 sender may affect whether the remote host accepts the recipient or not. I
6590 have changed it to use the actual sender in the callout; this means that
6591 the cache record is now keyed on a recipient/sender pair, not just the
6592 recipient address. There doesn't seem to be a real danger of callout loops,
6593 since a callout by the remote host to check the sender would use <>.
6594 [SEE ABOVE: changed after hitting problems.]
6596 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6597 temporary errors. However, in the case of such a code being given after
6598 the end of a data transmission (i.e. after ".") Exim was failing to write
6599 a retry record for the message. (Yes, there was some broken host that was
6600 actually sending 8xx at this point.)
6602 73. An unknown lookup type in a host list could cause Exim to panic-die when
6603 the list was checked. (An example that provoked this was putting <; in the
6604 middle of a list instead of at the start.) If this happened during a DATA
6605 ACL check, a -D file could be left lying around. This kind of configuration
6606 error no longer causes Exim to die; instead it causes a defer error. The
6607 incident is still logged to the main and panic logs.
6609 74. Buglet left over from Exim 3 conversion. The message "too many messages
6610 in one connection" was written to the rejectlog but not the mainlog, except
6611 when address rewriting (yes!) was being logged.
6613 75. Added write_rejectlog option.
6615 76. When a system filter was run not as root (that is, when system_filter_user
6616 was set), the values of the $n variables were not being returned to the
6617 main process; thus, they were not subsequently available in the $sn
6620 77. Added +return_path_on_delivery log selector.
6622 78. A connection timeout was being treated differently from recipients deferred
6623 when testing hosts_max_try with a message that was older than the host's
6624 retry timeout. (The host should not be counted, thus allowing all hosts to
6625 be tried at least once before bouncing.) This may have been the cause of an
6626 occasionally reported bug whereby a message would remain on the queue
6627 longer than the retry timeout, but would be bounced if a delivery was
6628 forced. I say "may" because I never totally pinned down the problem;
6629 setting up timeout/retry tests is difficult. See also the next item.
6631 79. The ultimate address timeout was not being applied to errors that involved
6632 a combination of host plus message (for example, a timeout on a MAIL
6633 command). When an address resolved to a number of possible hosts, and they
6634 were not all tried for each delivery (e.g. because of hosts_max_try), a
6635 message could remain on the queue longer than the retry timeout.
6637 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6640 81. Fixed an obscure SMTP outgoing bug which required at least the following
6641 conditions: (a) there was another message waiting for the same server;
6642 (b) the server returned 5xx to all RCPT commands in the first message so
6643 that the message was not completed; (c) the server dropped the connection
6644 or gave a negative response to the RSET that Exim sends to abort the
6645 transaction. The observed case was a dropped connection after DATA that had
6646 been sent in pipelining mode. That is, the server had advertised PIPELINING
6647 but was not implementing it correctly. The effect of the bug was incorrect
6648 behaviour, such as trying another host, and this could lead to a crash.
6654 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6655 and daemon.c were passed as pointers to ints; they should have been
6656 pointers to socklen_t variables (which are typically unsigned ints).
6658 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6661 3. Fixed a really odd bug that affected only the testing scheme; patching a
6662 certain fixed string in the binary changed the value of another string that
6663 happened to be identical to the end of the original first string.
6665 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6666 name", it returns that address as the IP address. On some operating
6667 systems (e.g. Solaris), it also passes back the IP address string as the
6668 "host name". However, on others (e.g. Linux), it passes back an empty
6669 string. Exim wasn't checking for this, and was changing the host name to an
6670 empty string, assuming it had been canonicalized.
6672 5. Although rare, it is permitted to have more than one PTR record for a given
6673 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6674 all the names associated with an address, because they do in Solaris.
6675 However, it seems that they do not in Linux for data that comes from the
6676 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6677 I found this out when I moved to a new Linux workstation and tried to run
6678 the Exim test suite.
6680 To get round this problem I have changed the code so that it now does its
6681 own call to the DNS to look up PTR records when searching for a host name.
6682 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6683 addresses that are only in /etc/hosts are still found.
6685 This behaviour is, however, controlled by an option called host_lookup_
6686 order, which defaults to "bydns:byaddr". If people want to use the other
6687 order, or indeed, just use one or the other means of lookup, they can
6688 specify it in this variable.
6690 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6691 some operating systems, this comes back from gethostbyaddr() as an empty
6692 string, and this is what Exim used to test for. However, it seems that in
6693 other systems, "." is yielded. Exim now tests for this case too.
6695 7. The values of check_spool_space and check_log_space are now held internally
6696 as a number of kilobytes instead of an absolute number of bytes. If a
6697 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6698 kilobyte. This means that much larger values can be stored.
6700 8. Exim monitor: an attempt to get the action menu when not actually pointing
6701 at a message produces an empty menu entitled "No message selected". This
6702 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6703 no entries in it ("Shell widget menu has zero width and/or height"). So I
6704 have added a single, blank menu entry in this case.
6706 9. Added ${quote_local_part.
6708 10. MIME decoding is now applied to the contents of Subject: header lines when
6711 11. Now that a reference to $sender_host_address automatically causes a reverse
6712 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6713 host lookup before query-style lookups in lists that might use this
6714 variable. This has therefore been abolished, and the "net-" prefix is no
6715 longer necessary for query-style lookups.
6717 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6718 have been a typo for LFLAGS, so it has been changed.
6720 13. The install script calls Exim with "-C /dev/null" in order to find the
6721 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6722 to be output. However, since Exim outputs its version number before the
6723 error, it didn't break the script. It just looked ugly. I fixed this by
6724 always allowing "-C /dev/null" if the caller is root.
6726 14. Ignore overlarge ACL variable number when reading spool file - insurance
6727 against a later release with more variables having written the file.
6729 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6730 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6731 this, as well as the form without the "IPv6" on the front.
6733 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6734 OS/Makefile-Darwin file.
6736 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6737 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6739 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6740 variable. However, some OS use ints for inodes. I've added cast to long int
6741 to get rid of the compiler warning.
6743 19. I had forgotten to lock out "/../" in configuration file names when
6744 ALT_CONFIG_PREFIX was set.
6746 20. Routers used for verification do not need to specify transports. However,
6747 if such a router generated a host list, and callout was configured, Exim
6748 crashed, because it could not find a port number from the (non-existent)
6749 transport. It now assumes port 25 in this circumstance.
6751 21. Added the -t option to exigrep.
6753 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6754 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6755 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6756 them is set, LOOKUP_LSEARCH is forced.
6758 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6760 24. Added sender and host information to the "rejected by local_scan()" log
6761 line; previously there was no indication of these.
6763 25. Added .include_if_exists.
6765 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6766 Linux. It turns out that not all file systems support this. Apparently some
6767 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6768 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6769 not supported on the file descriptor, is now ignored when Exim is trying to
6770 sync a directory. This applies only to Linux.
6772 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6774 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6775 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6778 29. Added support for maildirsize files from supplied patch (modified a bit).
6780 30. The use of :fail: followed by an empty string could lead Exim to respond to
6781 sender verification failures with (e.g.):
6783 550 Verification failed for <xxx>
6784 550 Sender verify failed
6786 where the first response line was missing the '-' that indicates it is not
6787 the final line of the response.
6789 31. The loop for finding the name of the user that called Exim had a hardwired
6790 limit of 10; it now uses the value of finduser_retries, which is used for
6791 all other user lookups.
6793 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6796 33. Exim was neglecting to zero errno before one call of strtol() when
6797 expanding a string and expecting an integer value. On some systems this
6798 resulted in spurious "integer overflow" errors. Also, it was casting the
6799 result into an int without checking.
6801 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6802 did not work. The code looks as if it has *never* worked, though it appears
6803 to have been documented since at least release 1.62. I have made it work.
6805 35. The "timeout_DNS" error in retry rules, also documented since at least
6806 1.62, also never worked. As it isn't clear exactly what this means, and
6807 clearly it isn't a major issue, I have abolished the feature by treating it
6808 as "timeout", and writing a warning to the main and panic logs.
6810 36. The display of retry rules for -brt wasn't always showing the error code
6813 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6814 timeout_connect_A, timeout_connect_MX.
6816 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6817 to the empty sender.
6819 39. The daemon was not analysing the content of -oX till after it had closed
6820 stderr and disconnected from the controlling terminal. This meant that any
6821 syntax errors were only noted on the panic log, and the return code from
6822 the command was 0. By re-arranging the code a little, I've made the
6823 decoding happen first, so such errors now appear on stderr, and the return
6824 code is 1. However, the actual setting up of the sockets still happens in
6825 the disconnected process, so errors there are still only recorded on the
6828 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6829 connections (as happens on some IP stacks) was logged at start up time as
6830 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6831 it from "IPv6 and IPv4", which means that two separate sockets are being
6834 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6835 says whether AF_INET or AF_INET6 was passed as an argument.
6837 42. Exiwhat output was messed up when time zones were included in log
6840 43. Exiwhat now gives more information about the daemon's listening ports,
6841 and whether -tls-on-connect was used.
6843 44. The "port" option of the smtp transport is now expanded.
6845 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6846 silently ignored. Now an error message is written to the main and panic
6849 46. There's a new ACL modifier called "logwrite" which writes to a log file
6850 as soon as it is encountered.
6852 47. Added $local_user_uid and $local_user_gid at routing time.
6854 48. Exim crashed when trying to verify a sender address that was being
6857 49. Exim was recognizing only a space character after ".include". It now also
6858 recognizes a tab character.
6860 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6861 extracting the relevant information from the specification. The man page no
6862 longer contains scrambled data for the -d option, and I've added a section
6863 at the front about calling Exim under different names.
6865 51. Added "extra_headers" argument to the "mail" command in filter files.
6867 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6870 53. Installed eximstats 1.29.
6872 54. Added transport_filter_timeout as a generic transport option.
6874 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6875 Cc: header lines. This was required by RFC 822, but it not required by RFC
6878 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6879 incoming messages that did not have them. Now it does so only if the
6880 message originates locally, that is, if there is no associated remote host
6881 address. When Resent- header lines are present, this applies to the Resent-
6882 lines rather than the non-Resent- lines.
6884 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6885 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6887 58. Messages for configuration errors now include the name of the main
6888 configuration file - useful now that there may be more than one file in a
6889 list (.included file names were always shown).
6891 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6892 for those rare installations that do not start the daemon as root or run it
6893 setuid root. I've cut out the call to initgroups() if the daemon is not
6896 60. The Exim user and group can now be bound into the binary as text strings
6897 that are looked up at the start of Exim's processing.
6899 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6901 62. Added $mailstore_basename variable.
6903 63. Installed patch to sieve.c from Michael Haardt.
6905 64. When Exim failed to open the panic log after failing to open the main log,
6906 the original message it was trying to log was written to stderr and debug
6907 output, but if they were not available (the usual case in production), it
6908 was lost. Now it is written to syslog before the two lines that record the
6909 failures to open the logs.
6911 65. Users' Exim filters run in subprocesses under the user's uid. It is
6912 possible for a "deliver" command or an alias in a "personal" command to
6913 provoke an address rewrite. If logging of address rewriting is configured,
6914 this fails because the process is not running as root or exim. There may be
6915 a better way of dealing with this, but for the moment (because 4.30 needs
6916 to be released), I have disabled address rewrite logging when running a
6917 filter in a non-root, non-exim process.
6923 1. The buildconfig auxiliary program wasn't quoting the value set for
6924 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6925 not defined. This bug was masked in 4.22 by the effect that was fixed in
6928 2. Some messages that were rejected after a message id was allocated were
6929 shown as "incomplete" by exigrep. It no longer does this for messages that
6930 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6932 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6933 have allow_domain_literals set, the ID did not get logged in the <= line.
6934 Domain literals are now always recognized in Message-ID: header lines.
6936 4. The first argument for a ${extract expansion item is the key name or field
6937 number. Leading and trailing spaces in this item were not being ignored,
6938 causing some misleading effects.
6940 5. When deliver_drop_privilege was set, single queue runner processes started
6941 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6942 same command in the process it spins off) were not dropping privilege.
6944 6. When the daemon running as "exim" started a queue runner, it always
6945 re-executed Exim in the spun-off process. This is a waste of effort when
6946 deliver_drop_privilege is set. The new process now just calls the
6947 queue-runner function directly.
6953 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6956 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6957 ignored. Though the use of -oP was forcing the writing of a pid file, it
6958 was always written to the default place.
6960 3. If the message "no IP address found for host xxxx" is generated during
6961 incoming verification, it is now followed by identification of the incoming
6962 connection (so you can more easily find what provoked it).
6964 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6966 5. Added some features to "harden" Exim a bit more against certain attacks:
6968 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6969 be put in Local/Makefile. This is like the never_users runtime option,
6970 but it cannot be overridden. The default setting is "root".
6972 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6973 prefix string with which any file named in a -C command line option
6976 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6977 is retained for -C and -D only if the caller of Exim is root. Without
6978 it, the exim user may also use -C and -D and retain privilege.
6980 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6981 command line option is disabled.
6983 6. Macro names set by the -D option must start with an upper case letter, just
6984 like macro names defined in the configuration file.
6986 7. Added "dereference=" facility to LDAP.
6988 8. Two instances of the typo "uknown" in the source files are fixed.
6990 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6991 the Configure-Makefile script screwed up while processing it.
6993 10. Incorporated PCRE 4.4.
6995 11. The SMTP synchronization check was not operating right at the start of an
6996 SMTP session. For example, it could not catch a HELO sent before the client
6997 waited for the greeting. There is now a check for outstanding input at the
6998 point when the greeting is written. Because of the duplex, asynchronous
6999 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7000 way, but not yet received, when the check is performed.
7002 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7003 on TCP/IP sockets, because this apparently causes some broken clients to
7006 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7007 unchanged) from the Cygwin maintainer.
7009 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7010 was supported instead of testing for mbx. Effectively a typo.
7012 15. The spa authenticator server code was not checking that the input it
7013 received was valid base64.
7015 16. The debug output line for the "set" modifier in ACLs was not showing the
7016 name of the variable that was being set.
7018 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7020 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7021 lookup on its own. Something else was needed to trigger the lookup. For
7022 example, a match in host_lookup or the need for a host name in a host list.
7023 Now, if $sender_host_name is referenced and the host name has not yet been
7024 looked up, a lookup is performed. If the lookup fails, the variable remains
7025 empty, and $host_lookup_failed is set to "1".
7027 19. Added "eqi" as a case-independent comparison operator.
7029 20. The saslauthd authentication condition could segfault if neither service
7030 nor realm was specified.
7032 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7033 error message that was logged was misleading, and incoming SMTP
7034 connections were dropped. The message is now more accurate, and temporary
7035 errors are given to SMTP connections.
7037 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7038 (except RSET) with a 421 error, until QUIT is received. However, it was
7039 failing to send a response to QUIT.
7041 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7042 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7043 ACL is now run after the helo_try_verify_hosts code.
7045 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7046 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7047 case-independent, so other case variants are also recognized). Apparently
7048 some systems use these upper case variants.
7050 25. If more than two messages were waiting for the same host, and a transport
7051 filter was specified for the transport, Exim sent two messages over the
7052 same TCP/IP connection, and then failed with "socket operation on non-
7053 socket" when it tried to send the third.
7055 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7057 27. The extern definition of crypt16() in expand.c was not being excluded when
7058 the OS had its own crypt16() function.
7060 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7061 as a preferred synonym for return_size_limit, both as an option and as an
7064 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7066 30. Changed the default configuration ACL to relax the local part checking rule
7067 for addresses that are not in any local domains. For these addresses,
7068 slashes and pipe symbols are allowed within local parts, but the sequence
7069 /../ is explicitly forbidden.
7071 31. SPA server authentication was not clearing the challenge buffer before
7074 32. log_message in a "warn" ACL statement was writing to the reject log as
7075 well as to the main log, which contradicts the documentation and doesn't
7076 seem right (because no rejection is happening). So I have stopped it.
7078 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7079 However, I am unable to do any testing of this.
7081 34. Fixed an infelicity in the appendfile transport. When checking directories
7082 for a mailbox, to see if any needed to be created, it was accidentally
7083 using path names with one or more superfluous leading slashes; tracing
7084 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7086 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7087 contents are added to the log line that is written for every discarded
7088 recipient. (Previously a log_message setting was ignored.)
7090 36. The ${quote: operator now quotes the string if it is empty.
7092 37. The install script runs exim in order to find its version number. If for
7093 some reason other than non-existence or emptiness, which it checks, it
7094 could not run './exim', it was installing it with an empty version number,
7095 i.e. as "exim-". This error state is now caught, and the installation is
7098 38. An argument was missing from the function that creates an error message
7099 when Exim fails to connect to the socket for saslauthd authentication.
7100 This could cause Exim to crash, or give a corrupted message.
7102 39. Added isip, isip4, and isip6 to ${if conditions.
7104 40. The ACL variables $acl_xx are now saved with the message, and can be
7105 accessed later in routers, transports, and filters.
7107 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7108 strings in the file are not string-expanded.
7110 42. If a MAIL command specified a SIZE value that was too large to fit into an
7111 int variable, the check against message_size_limit failed. Such values are
7112 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7113 day this will have to be increased, but I don't think I want to be around
7114 when emails are that large.
7121 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7122 iconv() is not standard in FreeBSD.
7124 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7125 IPv6 enabled. The observed symptom was a segmentation fault on return from
7126 the function os_common_find_running_interfaces() in src/os.c.
7128 3. In the check_special_case() function in daemon.c I had used "errno" as an
7129 argument name, which causes warnings on some systems. This was basically a
7130 typo, since it was named "eno" in the comments!
7132 4. The code that waits for the clock to tick (at a resolution of some fraction
7133 of a second) so as to ensure message-id uniqueness was always waiting for
7134 at least one whole tick, when it could have waited for less. [This is
7135 almost certainly not relevant at current processor speeds, where it is
7136 unlikely to ever wait at all. But we try to future-proof.]
7138 5. The function that sleeps for a time interval that includes fractions of a
7139 second contained a race. It did not block SIGALRM between setting the
7140 timer, and suspending (a couple of lines later). If the interval was short
7141 and the sigsuspend() was delayed until after it had expired, the suspension
7142 never ended. On busy systems this could lead to processes getting stuck for
7145 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7146 process, before it forks any delivery processes. The open lookup caching
7147 mechanism meant that the open file or database connection was passed into
7148 the delivery process. The problem was that delivery processes always tidy
7149 up cached lookup data. This could cause a problem for the next delivery
7150 process started by the queue runner, because the external queue runner
7151 process does not know about the closure. So the next delivery process
7152 still has data in the lookup cache. In the case of a file lookup, there was
7153 no problem because closing a file descriptor in a subprocess doesn't affect
7154 the parent. However, if the lookup was caching a connection to a database,
7155 the connection was closed, and the second delivery process was likely to
7156 see errors such as "PGSQL: query failed: server closed the connection
7157 unexpectedly". The problem has been fixed by closing all cached lookups
7158 in a queue runner before running a delivery process.
7160 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7161 seem to have the "const" qualifier which it has on other OS. I've
7164 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7165 *of the same type* (client or server) with the same public name that an
7166 error should be diagnosed.
7168 9. When Exim looked up a host name for an IP address, but failed to find the
7169 original IP address when looking up the host name (a safety check), it
7170 output the message "<ip address> does not match any IP for NULL", which was
7171 confusing, to say the least. The bug was that the host name should have
7172 appeared instead of "NULL".
7174 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7175 user that is built into the binary, and the -C or -D options is used, root
7176 privilege is dropped before the configuration file is read. In addition,
7177 logging is switched to stderr instead of the normal log files. If the
7178 configuration then re-defines the Exim user, the unprivileged environment
7179 is probably not what is expected, so Exim logs a panic warning message (but
7182 However, if deliver_drop_privilege is set, the unprivileged state may well
7183 be exactly what is intended, so the warning has been cut out in that case,
7184 and Exim is allowed to try to write to its normal log files.
7190 1. smtp_return_error_details was not giving details for temporary sender
7191 or receiver verification errors.
7193 2. Diagnose a configuration error if two authenticators have the same public
7196 3. Exim used not to create the message log file for a message until the first
7197 delivery attempt. This could be confusing when incoming messages were held
7198 for policy or load reasons. The message log file is now created at the time
7199 the message is received, and an initial "Received" line is written to it.
7201 4. The automatically generated man page for command line options had a minor
7202 bug that caused no ill effects; however, a more serious problem was that
7203 the procedure for building the man page automatically didn't always
7204 operate. Consequently, release 4.20 contains an out-of-date version. This
7205 shouldn't happen again.
7207 5. When building Exim with embedded Perl support, the script that builds the
7208 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7209 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7211 6. The freeze_tell option was not being used for messages that were frozen on
7212 arrival, either by an ACL or by local_scan().
7214 7. Added the smtp_incomplete_transaction log selector.
7216 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7217 was accepting AUTH without a new EHLO.
7219 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7220 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7223 10. Logging of TCP/IP connections (when configured) now happens in the main
7224 daemon process instead of the child process, so that the TCP/IP connection
7225 count is more accurate (but it can never be perfect).
7227 11. The use of "drop" in a nested ACL was not being handled correctly in the
7228 outer ACL. Now, if condition failure induced by the nested "drop" causes
7229 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7230 or "require"), the connection is dropped.
7232 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7233 that yield "discard" can now be used with an "accept" or a "discard" verb,
7234 but an error is generated for any others (because I can't see a useful way
7235 to define what should happen).
7237 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7238 are now processed in the same way as lines in the Exim configuration file.
7239 In particular, continuation lines are supported.
7241 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7243 15. Added -ti meaning -t -i.
7245 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7246 domains, and warn by logging if others are found.
7248 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7249 returned from GIFCONF when Exim is trying to find the list of interfaces on
7250 a host. The code in os.c has been modified to copy each ifreq to an aligned
7251 structure in all cases.
7253 Also, in some cases, the returned ifreq's were being copied to a 'struct
7254 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7255 means the last couple of bytes of an IPv6 address could be chopped if the
7256 ifreq contained only a normal sockaddr (14 bytes storage).
7258 18. Named domain lists were not supported in the hosts_treat_as_local option.
7259 An entry such as +xxxx was not recognized, and was treated as a literal
7262 19. Ensure that header lines added by a DATA ACL are included in the reject log
7263 if the ACL subsequently rejects the message.
7265 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7266 MD5 (which is deprecated).
7268 21. When testing a filter file using -bf, Exim was writing a message when it
7269 took the sender from a "From " line in the message, but it was not doing so
7270 when it took $return_path from a Return-Path: header line. It now does.
7272 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7273 with a valid header line field name (a series of printing characters
7274 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7276 23. Changed "disc" in the source to "disk" to conform to the documentation and
7277 the book and for uniformity.
7279 24. Ignore Sendmail's -Ooption=value command line item.
7281 25. When execve() failed while trying to run a command in a pipe transport,
7282 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7283 could be confused with a return value of 69 from the command itself. This
7284 has been changed to 127, the value the shell returns if it is asked to run
7285 a non-existent command. The wording for the related log line suggests a
7286 non-existent command as the problem.
7288 26. If received_header_text expands to an empty string, do not add a Received:
7289 header line to the message. (Well, it adds a token one on the spool, but
7290 marks it "old" so that it doesn't get used or transmitted.)
7292 27. Installed eximstats 1.28 (addition of -nt option).
7294 28. There was no check for failure on the call to getsockname() in the daemon
7295 code. This can fail if there is a shortage of resources on the system, with
7296 ENOMEM, for example. A temporary error is now given on failure.
7298 29. Contrary to the C standard, it seems that in some environments, the
7299 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7300 program. Exim now does this explicitly; it affects the formatting of
7301 timestamps using strftime().
7303 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7304 complaints. I've now initialized all the variables, to avoid this.
7306 32. Header lines added by a system filter were not being "seen" during
7307 transport-time rewrites.
7309 33. The info_callback() function passed to OpenSSL is set up with type void
7310 (*)(SSL *, int, int), as described somewhere. However, when calling the
7311 function (actually a macro) that sets it up, the type void(*)() is
7312 expected. I've put in a cast to prevent warnings from picky compilers.
7314 34. If a DNS black list lookup found a CNAME record, but there were no A
7315 records associated with the domain it pointed at, Exim crashed.
7317 35. If a DNS black list lookup returned more than one A record, Exim ignored
7318 all but the first. It now scans all returned addresses if a particular IP
7319 value is being sought. In this situation, the contents of the
7320 $dnslist_value variable are a list of all the addresses, separated by a
7323 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7324 to accept a host name and all its aliases if the forward lookup for any of
7325 them yielded the IP address of the incoming connection. Now it accepts only
7326 those names whose forward lookup yields the correct IP address. Any other
7327 names are discarded. This closes a loophole whereby a rogue DNS
7328 administrator could create reverse DNS records to break through a
7329 wildcarded host restriction in an ACL.
7331 37. If a user filter or a system filter that ran in a subprocess used any of
7332 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7333 the wrong values were passed to the pipe command ($thisaddress had the
7334 value of $0, $0 had the value of $1, etc). This bug was introduced by
7335 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7337 38. Improved the line breaking for long SMTP error messages from ACLs.
7338 Previously, if there was no break point between 40 and 75 characters, Exim
7339 left the rest of the message alone. Two changes have been made: (a) I've
7340 reduced the minimum length to 35 characters; (b) if it can't find a break
7341 point between 35 and 75 characters, it looks ahead and uses the first one
7342 that it finds. This may give the occasional overlong line, but at least the
7343 remaining text gets split now.
7345 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7346 file descriptors might be low, and that setting 1000 would always raise it.
7347 It turns out that in some environments, the limit is already over 1000 and
7348 that lowering it causes trouble. So now Exim takes care not to decrease it.
7350 40. When delivering a message, the value of $return_path is set to $sender_
7351 address at the start of routing (routers may change the value). By an
7352 oversight, this default was not being set up when an address was tested by
7353 -bt or -bv, which affected the outcome if any router or filter referred to
7356 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7357 log only when "message" or "log_message" are set. However, if one of the
7358 conditions was an address verification, or a call to a nested ACL, the
7359 messages generated by the underlying test were being passed through. This
7360 no longer happens. The underlying message is available in $acl_verify_
7361 message for both "message" and "log_message" expansions, so it can be
7362 passed through if needed.
7364 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7365 new expansion $bh_ to give the encoded byte string without charset
7366 translation. Translation happens only if iconv() is available; HAVE_ICONV
7367 indicates this at build time. HEADERS_CHARSET gives the charset to
7368 translate to; headers_charset can change it in the configuration, and
7369 "headers charset" can change it in an individual filter file.
7371 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7372 that creates RFC 2047 encoded "words" labels them as that charset instead
7373 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7374 expansion operator; (ii) when Exim creates a From: line for a local
7375 message; (iii) when a header line is rewritten to include a "phrase" part.
7377 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7378 buggy, causing it to skip the first lines of messages whose message ID
7379 ended in 'D'. This would not have bitten before Exim release 4.14, because
7380 message IDs were unlikely to end in 'D' before then. The effect was to have
7381 incorrect size information for certain domains.
7383 45. #include "config.h" was missing at the start of the crypt16.c module. This
7384 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7387 46. If there was a timeout during a "random" callout check, Exim treated it as
7388 a failure of the random address, and carried on sending RSET and the real
7389 address. If the delay was just some slowness somewhere, the response to the
7390 original RCPT would be taken as a response to RSET and so on, causing
7391 mayhem of various kinds.
7393 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7394 when I implemented it. It didn't allow for the fact that some option values
7395 may legitimately be negative (e.g. size_addition), and it didn't even do
7396 the right test for positive values.
7398 48. Domain names in DNS records are case-independent. Exim always looks them up
7399 in lower case. Some resolvers return domain names in exactly the case they
7400 appear in the zone file, that is, they may contain uppercase letters. Not
7401 all resolvers do this - some return always lower case. Exim was treating a
7402 change of case by a resolver as a change of domain, similar to a widening
7403 of a domain abbreviation. This triggered its re-routing code and so it was
7404 trying to route what was effectively the same domain again. This normally
7405 caused routing to fail (because the router wouldn't handle the domain
7406 twice). Now Exim checks for this case specially, and just changes the
7407 casing of the domain that it ultimately uses when it transmits the message
7410 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7413 50. If a filter generated a file delivery with a non-absolute name (possible if
7414 no home directory exists for the router), the forbid_file option was not
7417 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7418 the existing equality matching.
7420 52. Exim was using ints instead of ino_t variables in some places where it was
7421 dealing with inode numbers.
7423 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7424 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7425 called TMPDIR, and if it finds it is different, it changes its value.
7427 54. The smtp_printf() function is now made available to local_scan() so
7428 additional output lines can be written before returning. There is also an
7429 smtp_fflush() function to enable the detection of a dropped connection.
7430 The variables smtp_input and smtp_batched_input are exported to
7433 55. Changed the default runtime configuration: the message "Unknown user"
7434 has been removed from the ACL, and instead placed on the localuser router,
7435 using the cannot_route_message feature. This means that any verification
7436 failures that generate their own messages won't get overridden. Similarly,
7437 the "Unrouteable address" message that was in the ACL for unverifiable
7438 relay addresses has also been removed.
7440 56. Added hosts_avoid_esmtp to the smtp transport.
7442 57. The exicyclog script was not checking for the esoteric option
7443 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7444 will work only if exicyclog is run under the appropriate euid.
7446 58. Following a discussion on the list, the rules by which Exim recognises line
7447 endings on incoming messages have been changed. The -dropcr and drop_cr
7448 options are now no-ops, retained only for backwards compatibility. The
7449 following line terminators are recognized: LF CRLF CR. However, special
7450 processing applies to CR:
7452 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7453 nor a local message in the state where . is a terminator.
7455 (ii) If a bare CR is encountered in a header line, an extra space is added
7456 after the line terminator so as not to end the header. The reasoning
7457 behind this is that bare CRs in header lines are most likely either
7458 to be mistakes, or people trying to play silly games.
7460 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7461 was being incorrectly given as 18 bytes larger than it should have been.
7462 This is a VOB (very old bug).
7464 60. This may never have affected anything current, but just in case it has:
7465 When the local host is found other than at the start of a list of hosts,
7466 the local host, those with the same MX, and any that follow, are discarded.
7467 When the list in question was part of a longer list of hosts, the following
7468 hosts (not currently being processed) were also being discarded. This no
7469 longer happens. I'm not sure if this situation could ever has previously
7472 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7475 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7476 header. This is something that is recommended in a new Internet Draft, and
7477 is something that is documented as being done by Sendmail. There are two
7478 possible values. For messages generated by the autoreply transport, Exim
7481 Auto-Submitted: auto-replied
7483 whereas for all other generated messages (e.g. bounces) it adds
7485 Auto-Submitted: auto-generated
7487 63. The "personal" condition in filters now includes a test for the
7488 Auto-Submitted: header. If it contains the string "auto-" the message it
7489 not considered personal.
7491 64. Added rcpt_include_affixes as a generic transport option.
7493 65. Added queue_only_override (default true).
7495 66. Added the syslog_duplication option.
7497 67. If what should have been the first header line of a message consisted of
7498 a space followed by a colon, Exim was mis-interpreting it as a header line.
7499 It isn't of course - it is syntactically invalid and should therefore be
7500 treated as the start of the message body. The misbehaviour could have
7501 caused a number of strange effects, including loss of data in subsequent
7502 header lines, and spool format errors.
7504 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7505 client host had authenticated. This control can now be exercised by an ACL
7506 for more flexibility.
7508 69. By default, callouts do not happen when testing with -bh. There is now a
7509 variant, -bhc, which does actually run the callout code, including
7510 consulting and updating the callout cache.
7512 70. Added support for saslauthd authentication, courtesy of Alexander
7515 71. If statvfs() failed on the spool or log directories while checking their
7516 size for availability, Exim confusingly gave the error "space shortage".
7517 Furthermore, in debugging mode it crashed with a floating point exception.
7518 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7519 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7520 serious problem, Exim now writes to the main and panic logs when this
7521 happens, with details of the failure. It then refuses to accept the
7522 incoming message, giving the message "spool directory problem" or "log
7523 directory problem" with a 421 code for SMTP messages.
7525 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7526 0, 1, and 2 exist, because some OS complain for execs without them (see
7527 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7528 descriptors. However, the code omitted to check that the open succeeded,
7529 causing mysterious errors if for some reason the permissions on /dev/null
7530 got screwed. Now Exim writes a message to the main and panic logs, and
7531 bombs out if it can't open /dev/null.
7533 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7534 interact so that it is all more flexible. It is supposed to remain
7535 backwards compatible. Also added extra_local_interfaces.
7537 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7538 to bomb out with an assertion failure - to the client this appears as a
7539 connection drop. This problem occurs in the part of the code that was taken
7540 from the Samba project. Fortunately, the assertion is in a very simple
7541 function, so I have fixed this by reproducing the function inline in the
7542 one place where it is called, and arranging for authentication to fail
7543 instead of killing the process with assert().
7545 75. The SPA client code was not working when the server requested OEM rather
7546 than Unicode encoding.
7548 76. Added code to make require_files with a specific uid setting more usable in
7549 the case where statting the file as root fails - usually a non-root-mounted
7550 NFS file system. When this happens and the failure is EACCES, Exim now
7551 forks a subprocess and does the per-uid checking as the relevant uid.
7553 77. Added process_log_path.
7555 78. If log_file_path was not explicitly set, a setting of check_log_space or
7556 check_log_inodes was ignored.
7558 79. If a space check for the spool or log partitions fails, the incident is now
7559 logged. Of course, in the latter case the data may get lost...
7561 80. Added the %p formatting code to string_format() so that it can be used to
7562 print addresses in debug_print(). Adjusted all the address printing in the
7563 debugging in store.c to use %p rather than %d.
7565 81. There was a concern that a line of code in smtp_in.c could overflow a
7566 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7567 initially expressed, the concern was not well-founded, because trailing
7568 spaces are removed early. However, if the trailing spaces were followed by
7569 a NULL, they did not get removed, so the overflow was possible. Two fixes
7572 (a) I re-wrote the offending code in a cleaner fashion.
7573 (b) If an incoming SMTP command contains a NULL character, it is rejected
7576 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7577 runs initgroups(), so that if the Exim user is in any additional groups,
7578 they will be used during message reception.
7584 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob