1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
9 JH/01 Avoid costly startup code when not strictly needed. This reduces time
10 for some exim process initialisations. It does mean that the logging
11 of TLS configuration problems is only done for the daemon startup.
13 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
15 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
16 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
19 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
20 DKIM signing is being done. Previously a transport_filter would always
21 disable CHUNKING, falling back to traditional DATA.
23 JH/05 Regard command-line receipients as tainted.
25 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
27 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
28 PAM library frees one of the arguments given to it, despite the
29 documentation. Therefore a plain malloc must be used.
31 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
32 on-stack buffers were used, resulting in a taint trap when DSN information
33 copied from a received message was written into the buffer.
35 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
36 the ordering of its ARC headers. This caused a crash.
38 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
39 a new record was being constructed with information from the peer, a trap
42 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
43 installation would get error messages from DMARC verify, when it hit the
44 nonexistent file indicated by the default. Distros wanting DMARC enabled
45 should both provide the file and set the option.
46 Also enforce no DMARC verification for command-line sourced messages.
48 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
49 could, depending on the platform, hang at the STARTTLS response.
51 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
52 message on a connection. Previously if one message had ARC headers and
53 the following one did not, a crash could result when adding an
54 Authentication-Results: header.
56 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
57 Exim main code and Exim-related utities. The introduction of taint
58 tracking also did many adjustments to string handling. Since then, eximon
59 frequently terminated with an assert failure.
61 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
62 check for 452 responses. This slightly helps the inefficieny of doing
63 a large alias-expansion into a recipient-limited target. The max_rcpt
64 transport option still applies (and at the current default, will override
65 the new feature). The check is done for either cause of synch, and forces
66 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
67 connection. The new facility is not tunable at this time.
69 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
70 library live data was being used, so the results became garbage. Make
71 copies while it is still usable.
77 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
78 side, to decode a TLS 1.3 packet capture.
80 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
81 Previously the default library behaviour applied, sending two, each in
84 JH/03 Debug output for ACL now gives the config file name and line number for
87 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
89 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
91 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
92 buffer overrun for (non-chunking) other transports.
94 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
95 TLS1.3, means that a server rejecting a client certificate is not visible
96 to the client until the first read of encrypted data (typically the
97 response to EHLO). Add detection for that case and treat it as a failed
98 TLS connection attempt, so that the normal retry-in-clear can work (if
101 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
102 and/or domain. Found and fixed by Jason Betts.
104 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
105 configuration). If a CNAME target was not a wellformed name pattern, a
108 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
109 the OS reports them interleaved with other addresses.
111 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
112 used both for input and for a verify callout, both encrypted, SMTP
113 responses being sent by the server could be lost. This resulted in
114 dropped connections and sometimes bounces generated by a peer sending
117 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
118 string. Previously it was possible to cause undefined behaviour in a
119 library routine (usually a crash). Found by "zerons".
121 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
124 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
125 API was removed, so update to use the newer ones.
127 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
128 any timeout set, is taking a long time. Previously we would hang on to a
129 rotated logfile "forever" if the input was arriving with long gaps
130 (a previous attempt to fix addressed lack, for a long time, of initial
133 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
134 shared (NFS) environment. The length of the tempfile name is now
135 4 + 16 ("hdr.$message_exim_id") which might break on file
136 systems which restrict the file name length to lower values.
139 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
140 shared (NFS) environment.
142 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
143 did for all versions <4.90). Notably -M, -m, --invert, -I may be
146 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
147 on some platforms for bit 31.
149 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
150 to changes apparently associated with TLS1.3 handling some of the APIs
151 previously used were either nonfunctional or inappropriate. Strings
152 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
153 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
154 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
155 This affects log line X= elements, the $tls_{in,out}_cipher variables,
156 and the use of specific cipher names in the encrypted= ACL condition.
158 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
160 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
161 verification result was not updated unless hosts_require_ocsp applied.
163 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
164 queue_list_requires_admin set to false, non-admin users were denied the
167 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
168 directory-of-certs mode. Previously they were advertised despite the
171 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
172 A single TCP connection by a client will now hold a TLS connection open
173 for multiple message deliveries, by default. Previoud the default was to
176 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
177 default. If built with the facility, DANE will be used. The facility
178 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
180 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
181 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
182 must be defined and you must still, unless you define DISABLE_TLS, manage
183 the the include-dir and library-file requirements that go with that
184 choice. Non-TLS builds are still supported.
186 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
189 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
190 default. If the platform supports and has the facility enabled, it will
191 be requested on all coneections.
193 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
194 controlled by the build-time option SUPPORT_PIPE_CONNECT.
196 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
198 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
199 success-DSN messages. Previously the From: header was always the default
200 one for these; the option was ignored.
202 JH/28 Fix the timeout on smtp response to apply to the whole response.
203 Previously it was reset for every read, so a teergrubing peer sending
204 single bytes within the time limit could extend the connection for a
205 long time. Credit to Qualsys Security Advisory Team for the discovery.
207 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
208 delivery address, which leaked information of the results of local
209 forwarding. Change to the original envelope recipient address, per
212 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
213 requested. Previously not bounce was generated and a log entry of
214 error ignored was made.
216 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
218 JH/32 Introduce a general tainting mechanism for values read from the input
219 channel, and values derived from them. Refuse to expand any tainted
220 values, to catch one form of exploit.
222 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
223 was unused and the unexpanded text used for the test. Found and
224 fixed by Ruben Jenster.
226 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
227 an attempt to use a TLS library read routine dereffed a nul pointer,
230 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
231 them in our smtp response.
233 JH/36 Have the general router option retry_use_local_part default to true when
234 any of the restrictive preconditions are set (to anything). Previously it
235 was only for check_local user. The change removes one item of manual
236 configuration which is required for proper retries when a remote router
237 handles a subset of addresses for a domain.
239 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
240 link count into consideration.
242 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
243 caused the extension of big_buffer, the following lines were ignored.
245 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
246 accordance with RFC 2308. Previously there was no expiry, so a longlived
247 receive process (eg. due to ACL delays) versus a short SOA value could
250 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
252 JH/39 Promote DMARC support to mainline.
254 JH/40 Bug 2452: Add a References: header to DSNs.
256 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
257 parameters. The relevant library call is documented as "Deprecated: This
258 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
259 3.6.0, DH parameters are negotiated following RFC7919."
261 HS/06 Change the default of dnssec_request_domains to "*"
263 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
264 carried on and emitted a BDAT command, even when PIPELINING was not
267 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
268 buffer was used for the filename, resulting in a trap when tainted
269 arguments (eg. $domain) were used.
271 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
272 recommended to avoid a possible server-load attack. The feature can be
273 re-enabled via the openssl_options main cofiguration option.
275 JH/45 local_scan API: documented the current smtp_printf() call. This changed
276 for version 4.90 - adding a "more data" boolean to the arguments.
277 Bumped the ABI version number also, this having been missed previously;
278 release versions 4.90 to 4.92.3 inclusive were effectively broken in
279 respect of usage of smtp_printf() by either local_scan code or libraries
280 accessed via the ${dlfunc } expansion item. Both will need coding
281 adjustment for any calls to smtp_printf() to match the new function
282 signature; a FALSE value for the new argument is always safe.
284 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
285 the file-offset (which the Linux syscall does, and exim expects); this
286 resulted in an indefinite loop.
288 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
289 to do ARC verification. The Authentication-Results: header line added
290 by the configuration then had no ARC item.
296 JH/01 Remove code calling the customisable local_scan function, unless a new
297 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
299 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
300 non-signal-safe functions being used.
302 JH/03 Bug 2269: When presented with a received message having a stupidly large
303 number of DKIM-Signature headers, disable DKIM verification to avoid
304 a resource-consumption attack. The limit is set at twenty.
306 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
307 report of oldest_pass in ${authres } in consequence, and separate out
308 some descriptions of reasons for verification fail.
310 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
311 files in the spool were present and unlocked. A queue-runner could spot
312 them, resulting in a duplicate delivery. Fix that by doing the unlock
313 after the unlink. Investigation by Tim Stewart. Take the opportunity to
314 add more error-checking on spoolfile handling while that code is being
317 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
318 No known attacks, no CVE, this is defensive hardening.
320 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
321 a queue-runner could start a delivery while other operations were ongoing.
322 Cutthrough delivery was a common victim, resulting in duplicate delivery.
323 Found and investigated by Tim Stewart. Fix by using the open message data
324 file handle rather than opening another, and not locally closing it (which
325 releases a lock) for that case, while creating the temporary .eml format
326 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
328 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
329 $sender_verify_failure/$recipient_verify_failure to "random".
331 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
334 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
335 Previously this would segfault.
337 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
340 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
341 like zero, since the resolver should be doing this for us, But we need one
342 as a CNAME but no MX presence gets the CNAME returned; we need to check
343 that doesn't point to an MX to declare it "no MX returned" rather than
344 "error, loop". A new main option is added so the older capability of
345 following some limited number of chain links is maintained.
347 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
349 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
350 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
351 now-deprecated earlier definitions used only specified the range up to TLS
352 1.2 (in the older-version library docs).
354 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
356 JH/15 Rework TLS client-side context management. Stop using a global, and
357 explicitly pass a context around. This enables future use of TLS for
358 connections to service-daemons (eg. malware scanning) while a client smtp
359 connection is using TLS; with cutthrough connections this is quite likely.
361 JH/16 Fix ARC verification to do AS checks in reverse order.
363 JH/17 Support a "tls" option on the ${readsocket } expansion item.
365 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
366 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
367 Previously the "utf8" would be re-prepended for every additional message.
369 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
370 Previously thery were accepted, resulting in issues when attempting to
371 forward messages to a non-supporting MTA.
373 PP/02 Let -n work with printing macros too, not just options.
375 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
376 one parent address was copied, and bogus data was used at delivery-logging
377 time. Either a crash (after delivery) or bogus log data could result.
378 Discovery and analysis by Tim Stewart.
380 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
381 Previously if the string ended mid-character, we did not insert the
382 promised '?' replacement.
384 PP/04 Documentation: current string operators work on bytes, not codepoints.
386 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
387 should pack well giving a smaller memory footprint so better caching and
388 therefore performance. Group the declarations where this can't be done so
389 that the byte-sized flag variables are not interspersed among pointer
390 variables, giving a better chance of good packing by the compiler.
392 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
393 non-null, to avoid issues with sites running BATV. Previously reports were
394 sent with an empty envelope sender so looked like bounces.
396 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
397 The ignore_error flag wasn't being returned from the filter subprocess so
398 was not set for later routers. Investigation and fix by Matthias Kurz.
400 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
401 and a msg:complete for the whole, when a message is manually removed using
402 -Mrm. Developement by Matthias Kurz, hacked on by JH.
404 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
405 a "Gnu special" function, asprintf() in the DB utility binary builds; I
406 hope that is portable enough.
408 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
409 requiring a known-CA anchor certificate; make it now rely entirely on the
410 TLSA as an anchor. Checking the name on the leaf cert against the name
411 on the A-record for the host is still done for TA (but not for EE mode).
413 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
414 included in delivery lines for non-proxied connections, when compiled with
415 SUPPORT_SOCKS and running with proxy logging enabled.
417 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
418 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
419 move the existing event to fire before the normal logging of message
420 failure so that custom logging is bracketed by normal logging.
422 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
423 msg:fail:internal event. Developement by Matthias Kurz.
425 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
426 far too small for todays use of crypto signatures stored there. Go all
427 the way to the max DNS message size of 64kB, even though this might be
428 overmuch for IOT constrained device use.
430 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
431 copy a string over itself. The library routine is documented as not
432 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
434 JH/32 For main options check_spool_space and check_inode_space, where the
435 platform supports 64b integers, support more than the previous 2^31 kB
436 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
437 the previous G, M, k.
439 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
440 $authenticated_fail_id variable on authentication failure. Previously
443 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
444 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
445 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
446 GNUTLS_SEC_PARAM_MEDIUM.
448 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
449 side. Previously we would continue as if no SNI had been received.
451 JH/36 Harden the handling of string-lists. When a list consisted of a sole
452 "<" character, which should be a list-separator specification, we walked
453 off past the nul-terimation.
455 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
456 causes) even when the retry time is not yet met. Previously they were
457 not, meaning that when (say) an account was over-quota and temp-rejecting,
458 and multiple senders' messages were queued, only one sender would get
459 notified on each configured delay_warning cycle.
461 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
463 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
464 cipher-suites, an error can be left on the stack even for a succeeding
465 accept; this results in impossible error messages when a later operation
468 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
469 return error codes indicating retry. Under TLS1.3 this becomes required.
471 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
472 it only wrote the new authenticators, resulting in a lack of tracking of
473 peer changes of ESMTP extensions until the next cache flush.
475 JH/41 Fix the loop reading a message header line to check for integer overflow,
476 and more-often against header_maxsize. Previously a crafted message could
477 induce a crash of the recive process; now the message is cleanly rejected.
479 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
480 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
486 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
487 When redis_servers is set to a list of > 1 element, and the Redis servers
488 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
489 case of MOVED into a DEFER case instead, thus moving the query onto the
490 next server in the list. For a cluster of N elements, all N servers must
491 be defined in redis_servers.
493 GF/02 Catch and remove uninitialized value warning in exiqsumm
494 Check for existence of @ARGV before looking at $ARGV[0]
496 JH/01 Replace the store_release() internal interface with store_newblock(),
497 which internalises the check required to safely use the old one, plus
498 the allocate and data copy operations duplicated in both (!) of the
499 extant use locations.
501 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
502 modifier. This matches the restriction on the commandline.
504 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
505 Previously only the last row was returned.
507 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
508 we assumed that tags in the header were well-formed, and parsed the
509 element content after inspecting only the first char of the tag.
510 Assumptions at that stage could crash the receive process on malformed
513 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
514 While running the DKIM ACL we operate on the Permanent memory pool so that
515 variables created with "set" persist to the DATA ACL. Also (at any time)
516 DNS lookups that fail create cache records using the Permanent pool. But
517 expansions release any allocations made on the current pool - so a dnsdb
518 lookup expansion done in the DKIM ACL releases the memory used for the
519 DNS negative-cache, and bad things result. Solution is to switch to the
520 Main pool for expansions.
521 While we're in that code, add checks on the DNS cache during store_reset,
522 active in the testsuite.
523 Problem spotted, and debugging aided, by Wolfgang Breyha.
525 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
526 When none of the hosts presented to a transport match an already-open
527 connection, close it and proceed with the list. Previously we would
528 queue the message. Spotted by Lena with Yahoo, probably involving
531 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
532 Previously a spurious "250 OK id=" response was appended to the proper
535 JH/08 The "support for" informational output now, which built with Content
536 Scanning support, has a line for the malware scanner interfaces compiled
537 in. Interface can be individually included or not at build time.
539 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
540 by the template makefile "src/EDITME". The "STREAM" support for an older
541 ClamAV interface method is removed.
543 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
544 rows affected is given instead).
546 JH/11 The runtime Berkeley DB library version is now additionally output by
547 "exim -d -bV". Previously only the compile-time version was shown.
549 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
550 SMTP connection. Previously, when one had more receipients than the
551 first, an abortive onward connection was made. Move to full support for
552 multiple onward connections in sequence, handling cutthrough connection
553 for all multi-message initiating connections.
555 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
556 routers. Previously, a multi-recipient message would fail to match the
557 onward-connection opened for the first recipient, and cause its closure.
559 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
560 a timeout on read on a GnuTLS initiating connection, resulting in the
561 initiating connection being dropped. This mattered most when the callout
562 was marked defer_ok. Fix to keep the two timeout-detection methods
565 JH/15 Relax results from ACL control request to enable cutthrough, in
566 unsupported situations, from error to silently (except under debug)
567 ignoring. This covers use with PRDR, frozen messages, queue-only and
570 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
572 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
573 metadata, resulting in a crash in free().
575 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
576 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
577 Broken also in d185889f4, with init system revamp.
579 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
580 Previously we abruptly closed the connection after reading a malware-
581 found indication; now we go on to read the "scan ok" response line,
584 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
585 ACL. Previously, a crash would result.
587 JH/19 Speed up macro lookups during configuration file read, by skipping non-
588 macro text after a replacement (previously it was only once per line) and
589 by skipping builtin macros when searching for an uppercase lead character.
591 JH/20 DANE support moved from Experimental to mainline. The Makefile control
592 for the build is renamed.
594 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
595 was allocated for every new TLS startup, meaning one per message. Fix
596 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
598 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
599 reported the original. Fix to report (as far as possible) the ACL
600 result replacing the original.
602 JH/23 Fix memory leak during multi-message connections using STARTTLS under
603 OpenSSL. Certificate information is loaded for every new TLS startup,
604 and the resources needed to be freed.
606 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
608 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
611 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
612 DATA response info to the (existing) per-recipient response info for
613 the "C=" log element. It can have useful tracking info from the
614 destination system. Patch from Simon Arlott.
616 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
617 length value. Previously this would segfault.
619 HS/02 Support Avast multiline protoocol, this allows passing flags to
620 newer versions of the scanner.
622 JH/28 Ensure that variables possibly set during message acceptance are marked
623 dead before release of memory in the daemon loop. This stops complaints
624 about them when the debug_store option is enabled. Discovered specifically
625 for sender_rate_period, but applies to a whole set of variables.
626 Do the same for the queue-runner and queue-list loops, for variables set
627 from spool message files. Do the same for the SMTP per-message loop, for
628 certain variables indirectly set in ACL operations.
630 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
631 as a multi-recipient message from a mailinglist manager). The coding had
632 an arbitrary cutoff number of characters while checking for more input;
633 enforced by writing a NUL into the buffer. This corrupted long / fast
634 input. The problem was exposed more widely when more pipelineing of SMTP
635 responses was introduced, and one Exim system was feeding another.
636 The symptom is log complaints of SMTP syntax error (NUL chars) on the
637 receiving system, and refused recipients seen by the sending system
638 (propating to people being dropped from mailing lists).
639 Discovered and pinpointed by David Carter.
641 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
642 replaced by the ${authresults } expansion.
644 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
646 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
647 allows proper process termination in container environments.
649 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
650 Previously the "final dot" had a newline after it; ensure it is CR,LF.
652 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
653 and "err_perm", deprecated since 4.83 when the RFC-defined words
654 "temperror" and "permerror" were introduced.
656 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
657 transport-filters or DKIM-signing. The restriction was lost in the
658 consolidation of verify-callout and delivery SMTP handling.
659 Extend the restriction to also cover ARC-signing.
661 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
662 in defer=pass mode supply a 450 to the initiator. Previously the message
665 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
666 tls_require_ciphers is used as before.
668 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
669 "pass_unscanned". Only tmpfails from the scanner are written to
670 the paniclog, as they may require admin intervention (permission
671 denied, license issues). Other scanner errors (like decompression
672 bombs) do not cause a paniclog entry.
674 JH/36 Fix reinitialisation of DKIM logging variable between messages.
675 Previously it was possible to log spurious information in receive log
678 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
679 triggered odd behaviour from Outlook Express clients.
681 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
684 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
685 since the IETF WG has not yet settled on that versus the original
686 "bare" representation.
688 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
689 Previously the millisecond value corrupted the output.
690 Fix also for syslog_pid=no and log_selector +pid, for which the pid
691 corrupted the output.
697 JH/01 Rework error string handling in TLS interface so that the caller in
698 more cases is responsible for logging. This permits library-sourced
699 string to be attached to addresses during delivery, and collapses
700 pairs of long lines into single ones.
702 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
703 during configuration. Wildcards are allowed and expanded.
705 JH/02 Rework error string handling in DKIM to pass more info back to callers.
706 This permits better logging.
708 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
709 do not close it down and have the child transport start it up again on
710 the passed-on TCP connection. Instead, proxy the child (and any
711 subsequent ones) for TLS via a unix-domain socket channel. Logging is
712 affected: the continued delivery log lines do not have any DNSSEC, TLS
713 Certificate or OCSP information. TLS cipher information is still logged.
715 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
716 identical IP addresses on different listening ports. Will also affect
719 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
720 add noisy ifdef guards to special-case this sillyness.
721 Patch from Bernd Kuhls.
723 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
724 than 255 are no longer allowed.
726 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
727 Disable the session-cache too, which might reduce our load. Since we
728 currrectly use a new context for every connection, both as server and
729 client, there is no benefit for these.
730 GnuTLS appears to not support tickets server-side by default (we don't
731 call gnutls_session_ticket_enable_server()) but client side is enabled
732 by default on recent versions (3.1.3 +) unless the PFS priority string
735 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
736 <https://reproducible-builds.org/specs/source-date-epoch/>.
738 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
739 the check for any unsuccessful recipients did not notice the limit, and
740 erroneously found still-pending ones.
742 JH/08 Pipeline CHUNKING command and data together, on kernels that support
743 MSG_MORE. Only in-clear (not on TLS connections).
745 JH/09 Avoid using a temporary file during transport using dkim. Unless a
746 transport-filter is involved we can buffer the headers in memory for
747 creating the signature, and read the spool data file once for the
748 signature and again for transmission.
750 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
751 4.77 as the kernel support then wasn't solid, having issues in 64bit
752 mode. Now, it's been long enough. Add support for FreeBSD also.
754 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
755 case where the routing stage had gathered several addresses to send to
756 a host before calling the transport for the first, we previously failed
757 to close down TLS in the old transport process before passing the TCP
758 connection to the new process. The new one sent a STARTTLS command
759 which naturally failed, giving a failed delivery and bloating the retry
760 database. Investigation and fix prototype from Wolfgang Breyha.
762 JH/12 Fix check on SMTP command input synchronisation. Previously there were
763 false-negatives in the check that the sender had not preempted a response
764 or prompt from Exim (running as a server), due to that code's lack of
765 awareness of the SMTP input buffering.
767 PP/04 Add commandline_checks_require_admin option.
768 Exim drops privileges sanely, various checks such as -be aren't a
769 security problem, as long as you trust local users with access to their
770 own account. When invoked by services which pass untrusted data to
771 Exim, this might be an issue. Set this option in main configuration
772 AND make fixes to the calling application, such as using `--` to stop
775 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
776 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
777 responses to those, into a single TLS record each way (this usually means
778 a single packet). As a side issue, smtp_enforce_sync now works on TLS
781 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
782 affects you only if you're dancing at the edge of the param size limits.
783 If you are, and this message makes sense to you, then: raise the
784 configured limit or use OpenSSL 1.1. Nothing we can do for older
787 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
788 cmdline element to get the documented default one. Previously it was
791 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
792 are unsuable for non-message-specific reasons - eg. connection timeout,
793 banner-time rejection.
795 JH/16 Fix logging of delivery remote port, when specified by router, under
798 PP/06 Repair manualroute's ability to take options in any order, even if one
799 is the name of a transport.
802 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
804 JH/17 Change the list-building routines interface to use the expanding-string
805 triplet model, for better allocation and copying behaviour.
807 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
808 Previously it was constructed the first time a possibly-matching string
809 was met in the configuration file input during startup; now it is done
812 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
813 compatible one, to avoid the (poorly documented) possibility of a config
814 file in the working directory redirecting the DB files, possibly correpting
815 some existing file. CVE-2017-10140 assigned for BDB.
817 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
818 cache-hot. Previously, although the result was properly cached, the
819 initial verify call returned a defer.
821 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
822 the main verify for receipient in uncached-mode.
824 JH/22 Retire historical build files to an "unsupported" subdir. These are
825 defined as "ones for which we have no current evidence of testing".
827 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
828 if present. Previously it was ignored.
830 JH/24 Start using specified-initialisers in C structure init coding. This is
831 a C99 feature (it's 2017, so now considered safe).
833 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
834 if was a fixed-sized field and bitmask ops via macros; it is now more
837 PP/07 GitHub PR 56: Apply MariaDB build fix.
838 Patch provided by Jaroslav Škarvada.
840 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
841 during Coverity cleanups [4.87 JH/47]
842 Diagnosis and fix provided by Michael Fischer v. Mollard.
844 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
845 the right size to place the terminating semicolon on its own folded
846 line, the header hash was calculated to an incorrect value thanks to
847 the (relaxed) space the fold became.
849 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
850 and confused the parent.
852 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
853 which could crash as a result. This could lead to undeliverable messages.
855 JH/28 Logging: "next input sent too soon" now shows where input was truncated
858 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
859 matters on fast-turnover and PID-randomising systems, which were getting
860 out-of-order delivery.
862 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
863 a possibly-overlapping copy. The symptom was that "Remote host closed
864 connection in response to HELO" was logged instead of the actual 4xx
867 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
868 Previously only that bufferd was discarded, resulting in SYMTP command
871 JH/32 DKIM: when a message has multiple signatures matching an identity given
872 in dkim_verify_signers, run the dkim acl once for each. Previously only
873 one run was done. Bug 2189.
875 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
876 "panic the current process" to "deliberately defer". The panic log is
877 still written with the problem list name; the mail and reject logs now
878 get a temp-reject line for the message that was being handled, saying
879 something like "domains check lookup or other defer". The SMTP 451
880 message is still "Temporary local problem".
882 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
883 A crafted sequence of BDAT commands could result in in-use memory beeing
884 freed. CVE-2017-16943.
886 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
887 from SMTP input. Previously it was always done; now only done for DATA
888 and not BDAT commands. CVE-2017-16944.
890 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
891 to the message (such as an overlong header line). Previously this was
892 not done and we did not exit BDAT mode. Followon from the previous item
893 though a different problem.
899 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
900 than -2003 did; needs libidn2 in addition to libidn.
902 JH/02 The path option on a pipe transport is now expanded before use.
904 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
905 Patch provided by "Björn", documentation fix added too.
907 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
908 missing a wire-to-host endian conversion.
910 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
911 close after a BDAT command line could be taken as a following command,
912 giving a synch failure. Fix by only checking for synch immediately
913 before acknowledging the chunk.
915 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
916 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
917 macro. Patches provided by Josh Soref.
919 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
920 Previously we did not; the RFC seems ambiguous and VRFY is not listed
921 by IANA as a service extension. However, John Klensin suggests that we
924 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
925 the dkim code may be unix-mode line endings rather than smtp wire-format
926 CRLF, so prepend a CR to any bare LF.
928 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
929 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
931 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
932 the first were themselves being wrongly included in the feed into dkim
933 processing; with most chunk sizes in use this resulted in an incorrect
934 body hash calculated value.
936 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
937 DKIM signature block, for verification. Although advised against by
938 standards it is specifically not ruled illegal.
940 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
942 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
943 missing a body hash (the bh= tag).
945 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
946 It seems that HAProxy sends the Proxy Protocol information in clear and
947 only then does a TLS startup, so do the same.
949 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
950 TCP connections (such as for Spamd) unless the daemon successfully set
951 Fast Open mode on its listening sockets. This fixes breakage seen on
952 too-old kernels or those not configured for Fast Open, at the cost of
953 requiring both directions being enabled for TFO, and TFO never being used
954 by non-daemon-related Exim processes.
956 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
957 endings, at least on the first header line. Try to canonify any that get
958 past that check, despite the cost.
960 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
961 now limited to an arbitrary five deep, while parsing addresses with the
962 strip_excess_angle_brackets option enabled.
964 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
965 instead leave the unprompted TLS handshake in socket buffer for the
966 TLS library to consume.
968 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
970 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
972 JH/16 Drop variables when they go out of scope. Memory management drops a whole
973 region in one operation, for speed, and this leaves assigned pointers
974 dangling. Add checks run only under the testsuite which checks all
975 variables at a store-reset and panics on a dangling pointer; add code
976 explicitly nulling out all the variables discovered. Fixes one known
977 bug: a transport crash, where a dangling pointer for $sending_ip_address
978 originally assigned in a verify callout, is re-used.
980 PP/06 Drop '.' from @INC in various Perl scripts.
982 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
984 PP/08 Reduce a number of compilation warnings under clang; building with
985 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
986 should be warning-free.
988 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
990 HS/01 Fix portability problems introduced by PP/08 for platforms where
991 realloc(NULL) is not equivalent to malloc() [SunOS et al].
993 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
994 chunk. This allows us to accept broken chunked messages. We need a more
995 general solution here.
997 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
998 already-broken messages in the queue.
1000 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1002 JH/19 Fix reference counting bug in routing-generated-address tracking.
1008 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1009 supports it and a size is available (ie. the sending peer gave us one).
1011 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1012 years of being deprecated). The replacements are the ACLs
1013 acl_smtp_mime and acl_not_smtp_mime.
1015 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1016 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1017 or even an in-clear connection were permitted. Now, if the host lookup
1018 was dnssec and dane was requested then the host is only used if the
1019 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1020 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1021 if one fails this test.
1022 This means that a poorly-configured remote DNS will make it incommunicado;
1023 but it protects against a DNS-interception attack on it.
1025 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1026 non-noisy when a race steals the message being considered.
1028 JH/05 If main configuration option tls_certificate is unset, generate a
1029 self-signed certificate for inbound TLS connections.
1031 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1032 in rewrites and routers.
1034 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1035 and logged a warning sing 4.83; now they are a configuration file error.
1037 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1038 (lacking @domain). Apply the same qualification processing as RCPT.
1040 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1042 JH/10 Support ${sha256:} applied to a string (as well as the previous
1045 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1046 a cutthrough deliver is pending, as we always want to make a connection.
1047 This also avoids re-routing the message when later placing the cutthrough
1048 connection after a verify cache hit.
1049 Do not update it with the verify result either.
1051 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1052 when routing results in more than one destination address.
1054 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1055 signing (which inhibits the cutthrough capability). Previously only
1056 the presence of an option was tested; now an expansion evaluating as
1057 empty is permissible (obviously it should depend only on data available
1058 when the cutthrough connection is made).
1060 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1061 the relevant preceding SMTP command did not note the pipelining mode.
1063 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1064 Previously they were not counted.
1066 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1067 as one having no matching records. Previously we deferred the message
1068 that needed the lookup.
1070 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1071 distinguished as "(=".
1073 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1074 for missing MX records. Previously it only worked for missing A records.
1076 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1078 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1079 after the data-go-ahead and data-ack. Patch from Jason Betts.
1081 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1082 even for a "none" policy. Patch from Tony Meyer.
1084 JH/22 Fix continued use of a connection for further deliveries. If a port was
1085 specified by a router, it must also match for the delivery to be
1088 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1089 When one of the recipients of a message was unsuitable for the connection
1090 (has no matching addresses), we lost track of needing to mark it
1091 deferred. As a result mail would be lost.
1093 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1095 JH/25 Decoding ACL controls is now done using a binary search; the source code
1096 takes up less space and should be simpler to maintain. Merge the ACL
1097 condition decode tables also, with similar effect.
1099 JH/26 Fix problem with one_time used on a redirect router which returned the
1100 parent address unchanged. A retry would see the parent address marked as
1101 delivered, so not attempt the (identical) child. As a result mail would
1104 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1105 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1106 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1109 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1110 Main config options check_{log,spool}_{inodes,space} are now
1111 100 inodes, 10MB unless set otherwise in the configuration.
1113 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1114 Previously it only applied to the main-section connection policy
1117 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1119 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1120 by me. Added RFC7919 DH primes as an alternative.
1122 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1123 are not in the system include path.
1125 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1126 GnuTLS, when a session startup failed (eg because the client disconnected)
1127 Exim did stdio operations after fclose. This was exposed by a recent
1128 change which nulled out the file handle after the fclose.
1130 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1131 signed directly by the cert-signing cert, rather than an intermediate
1132 OCSP-signing cert. This is the model used by LetsEncrypt.
1134 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1136 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1137 an incoming connection.
1139 HS/02 Bug 1802: Do not half-close the connection after sending a request
1142 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1143 fallback to "prime256v1".
1145 JH/34 SECURITY: Use proper copy of DATA command in error message.
1146 Could leak key material. Remotely exploitable. CVE-2016-9963.
1152 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1153 and 3.4.4 - once the server is enabled to respond to an OCSP request
1154 it does even when not requested, resulting in a stapling non-aware
1155 client dropping the TLS connection.
1157 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1158 support variable-length bit vectors. No functional change.
1160 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1161 The I= interface field on outgoing lines is now after the H= remote
1162 host field, same as incoming lines. There is a separate
1163 outgoing_interface log selector which allows you to disable the
1166 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1167 If not running log_selector +smtp_connection the mainlog would be held
1168 open indefinitely after a "too many connections" event, including to a
1169 deleted file after a log rotate. Leave the per net connection logging
1170 leaving it open for efficiency as that will be quickly detected by the
1171 check on the next write.
1173 HS/01 Bug 1671: Fix post transport crash.
1174 Processing the wait-<transport> messages could crash the delivery
1175 process if the message IDs didn't exist for some reason. When
1176 using 'split_spool_directory=yes' the construction of the spool
1177 file name failed already, exposing the same netto behaviour.
1179 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1180 mime_regex ACL conditions.
1182 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1183 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1184 to HELO, local diagnostic string.
1186 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1187 log line to debug. Even when configured with a tls authenticator many
1188 client connections are expected to not authenticate in this way, so
1189 an authenticate fail is not an error.
1191 HS/02 Add the Exim version string to the process info. This way exiwhat
1192 gives some more detail about the running daemon.
1194 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1195 matter for fast-change records such as DNSBLs.
1197 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1198 retry record, even if constant. There may be multiple transports with
1199 different interface settings and the retry behaviour needs to be kept
1202 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1204 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1206 JH/10 Bug 840: fix log_defer_output option of pipe transport
1208 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1209 hosts_randomize. This matters a lot when combined with mua_wrapper.
1211 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1212 ${quote_pgsql:<string>} operator.
1214 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1216 JH/14 Change header file name for freeradius-client. Relevant if compiling
1217 with Radius support; from the Gentoo tree and checked under Fedora.
1219 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1221 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1222 option transport_filter, meaning no filtering.
1224 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1226 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1227 defaults to "*" (all hosts). The variable is now available when not built
1228 with TLS, default unset, mainly to enable keeping the testsuite sane.
1229 If a server certificate is not supplied (via tls_certificate) an error is
1230 logged, and clients will find TLS connections fail on startup. Presumably
1231 they will retry in-clear.
1232 Packagers of Exim are strongly encouraged to create a server certificate
1233 at installation time.
1235 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1236 with the $config_file variable.
1238 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1239 in transport context, after the attempt, and per-recipient. The latter type
1240 is per host attempted. The event data is the error message, and the errno
1241 information encodes the lookup type (A vs. MX) used for the (first) host,
1242 and the trailing two digits of the smtp 4xx response.
1244 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1245 to write to mainlog (or rejectlog, paniclog) in the window between file
1246 creation and permissions/ownership being changed. Particularly affects
1247 installations where exicyclog is run as root, rather than exim user;
1248 result is that the running daemon panics and dies.
1250 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1252 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1253 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1254 "pri" and "weight". Note that the previous implicit priority given by the
1255 list order is no longer honoured.
1257 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1258 for DKIM processing.
1260 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1261 by defining SUPPORT_SOCKS.
1263 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1264 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1265 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1266 variables are renamed to proxy_{local,external}_{address,port}.
1268 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1269 for a build by defining SUPPORT_I18N
1271 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1272 of the query string, and make ${quote_redis:} do that quoting.
1274 JH/29 Move Events support from Experimental to mainline, enabled by default
1275 and removable for a build by defining DISABLE_EVENT.
1277 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1279 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1280 cached by the daemon.
1282 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1283 by defining LOOKUP_REDIS. The libhiredis library is required.
1285 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1286 keys are given for lookup.
1288 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1289 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1290 only supported when built with TLS support. The PolarSSL SHA routines
1291 are still used when the TLS library is too old for convenient support.
1293 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1294 openssl_options), for security. OpenSSL forces this from version 1.1.0
1295 server-side so match that on older versions.
1297 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1298 allocation for $value could be released as the expansion processing
1299 concluded, but leaving the global pointer active for it.
1301 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1302 and to use the domains and local_parts ACL conditions.
1304 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1305 incorrectly not doubled on cutthrough transmission, hence seen as a
1306 body-termination at the receiving system - resulting in truncated mails.
1307 Commonly the sender saw a TCP-level error, and retransmitted the message
1308 via the normal store-and-forward channel. This could result in duplicates
1309 received - but deduplicating mailstores were liable to retain only the
1310 initial truncated version.
1312 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1314 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1316 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1317 we're in there, support oversigning also; bug 1309.
1319 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1321 HS/04 Add support for keep_environment and add_environment options.
1323 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1324 either intentional arithmetic overflow during PRNG, or testing config-
1327 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1328 delivery resulted in actual delivery. Cancel cutthrough before DATA
1331 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1332 rejects a recipient: pass the reject to the originator.
1334 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1335 Many were false-positives and ignorable, but it's worth fixing the
1338 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1339 for the new environment-manipulation done at startup. Move the routines
1340 from being local to tls.c to being global via the os.c file.
1342 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1343 an extract embedded as result-arg for a map, the first arg for extract
1344 is unavailable so we cannot tell if this is a numbered or keyed
1345 extraction. Accept either.
1351 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1354 JH/02 The smtp transport option "multi_domain" is now expanded.
1356 JH/03 The smtp transport now requests PRDR by default, if the server offers
1359 JH/04 Certificate name checking on server certificates, when exim is a client,
1360 is now done by default. The transport option tls_verify_cert_hostnames
1361 can be used to disable this per-host. The build option
1362 EXPERIMENTAL_CERTNAMES is withdrawn.
1364 JH/05 The value of the tls_verify_certificates smtp transport and main options
1365 default to the word "system" to access the system default CA bundle.
1366 For GnuTLS, only version 3.0.20 or later.
1368 JH/06 Verification of the server certificate for a TLS connection is now tried
1369 (but not required) by default. The verification status is now logged by
1370 default, for both outbound TLS and client-certificate supplying inbound
1373 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1376 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1377 Status Notification (bounce) messages are now MIME format per RFC 3464.
1378 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1379 under the control of the dsn_advertise_hosts option, and routers may
1380 have a dsn_lasthop option.
1382 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1383 default, modifiable by a malware= option. The list separator for
1384 the options can now be changed in the usual way. Bug 68.
1386 JH/10 The smtp_receive_timeout main option is now expanded before use.
1388 JH/11 The incoming_interface log option now also enables logging of the
1389 local interface on delivery outgoing connections.
1391 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1392 if the interface and destination host and port all match.
1394 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1397 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1398 Patch from Andrew Lewis.
1400 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1401 now supports optional time-restrictions, weighting, and priority
1402 modifiers per server. Patch originally by <rommer@active.by>.
1404 JH/16 The spamd_address main option now supports a mixed list of local
1405 and remote servers. Remote servers can be IPv6 addresses, and
1406 specify a port-range.
1408 JH/17 Bug 68: The spamd_address main option now supports an optional
1409 timeout value per server.
1411 JH/18 Bug 1581: Router and transport options headers_add/remove can
1412 now have the list separator specified.
1414 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1417 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1420 JH/21 Support for the A6 type of dns record is withdrawn.
1422 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1423 rather than the verbs used.
1425 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1426 from 255 to 1024 chars.
1428 JH/24 Verification callouts now attempt to use TLS by default.
1430 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1431 are generic router options now. The defaults didn't change.
1433 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1434 Original patch from Alexander Shikoff, worked over by JH.
1436 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1437 files if ZCAT_COMMAND is not executable.
1439 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1441 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1443 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1444 Normally benign, it bites when the pair was led to by a CNAME;
1445 modern usage is to not canonicalize the domain to a CNAME target
1446 (and we were inconsistent anyway for A-only vs AAAA+A).
1448 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1450 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1451 when evaluating $sender_host_dnssec.
1453 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1454 $sender_helo_dnssec variable.
1456 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1458 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1460 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1462 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1463 documented as working, but never had. Support all but $spam_report.
1465 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1466 added for tls authenticator.
1468 HS/03 Add perl_taintmode main config option
1474 TL/01 When running the test suite, the README says that variables such as
1475 no_msglog_check are global and can be placed anywhere in a specific
1476 test's script, however it was observed that placement needed to be near
1477 the beginning for it to behave that way. Changed the runtest perl
1478 script to read through the entire script once to detect and set these
1479 variables, reset to the beginning of the script, and then run through
1480 the script parsing/test process like normal.
1482 TL/02 The BSD's have an arc4random API. One of the functions to induce
1483 adding randomness was arc4random_stir(), but it has been removed in
1484 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1485 function when detected.
1487 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1488 cause callback expansion.
1490 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1491 syntax errors in an expansion can be treated as a string instead of
1492 logging or causing an error, due to the internal use of bool_lax
1493 instead of bool when processing it.
1495 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1496 server certificates when making smtp deliveries.
1498 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1500 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1502 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1504 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1505 Merged patch from Sebastian Wiedenroth.
1507 JH/05 Fix results-pipe from transport process. Several recipients, combined
1508 with certificate use, exposed issues where response data items split
1509 over buffer boundaries were not parsed properly. This eventually
1510 resulted in duplicates being sent. This issue only became common enough
1511 to notice due to the introduction of connection certificate information,
1512 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1514 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1515 size buffer was used, resulting in syntax errors when an expansion
1518 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1519 version 3.3.6 or later.
1521 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1522 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1523 both become "event_action", the variables become $event_name, $event_data
1524 and $event_defer_errno. There is a new variable $verify_mode, usable in
1525 routers, transports and related events. The tls:cert event is now also
1526 raised for inbound connections, if the main configuration event_action
1529 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1530 early OCSP support, but no stapling (appears to be less than 1.0.0).
1532 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1533 server certificate names available under the smtp transport option
1534 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1537 JH/10 Time-related extraction expansions from certificates now use the main
1538 option "timezone" setting for output formatting, and are consistent
1539 between OpenSSL and GnuTLS compilations. Bug 1541.
1541 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1542 encoded parameter in the incoming message. Bug 1558.
1544 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1545 include certificate info, eximon was claiming there were spoolfile
1548 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1550 JH/14 Log delivery-related information more consistently, using the sequence
1551 "H=<name> [<ip>]" wherever possible.
1553 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1554 are problematic for Debian distribution, omit them from the release
1557 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1559 JH/16 Fix string representation of time values on 64bit time_t architectures.
1562 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1563 output list separator was used.
1568 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1569 checkers that were complaining about end of non-void function with no
1572 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1573 This was a regression introduced in 4.83 by another bugfix.
1575 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1577 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1578 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1584 TF/01 Correctly close the server side of TLS when forking for delivery.
1586 When a message was received over SMTP with TLS, Exim failed to clear up
1587 the incoming connection properly after forking off the child process to
1588 deliver the message. In some situations the subsequent outgoing
1589 delivery connection happened to have the same fd number as the incoming
1590 connection previously had. Exim would try to use TLS and fail, logging
1591 a "Bad file descriptor" error.
1593 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1594 utilities have not been installed.
1596 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1597 temporary space as the ACL may create new global variables.
1599 TL/01 LDAP support uses per connection or global context settings, depending
1600 upon the detected version of the libraries at build time.
1602 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1603 to extract and use the src ip:port in logging and expansions as if it
1604 were a direct connection from the outside internet. PPv2 support was
1605 updated based on HAProxy spec change in May 2014.
1607 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1609 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1610 Properly escape header and check for NULL return.
1612 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1615 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1617 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1618 characters in header names, implemented as a verify condition.
1619 Contributed by Michael Fischer v. Mollard.
1621 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1622 results permerror and temperror. Previous values are deprecated but
1623 still accepted. In a future release, err_perm and err_temp will be
1624 completely removed, which will be a backward incompatibility if the
1625 ACL tests for either of these two old results. Patch contributed by
1626 user bes-internal on the mailing list.
1628 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1630 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1631 selectors, in both main and reject logs.
1633 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1636 JH/07 Add malware type "sock" for talking to simple daemon.
1638 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1640 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1641 routers/transports under cutthrough routing.
1643 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1644 numbers. Touch up "bool" conditional to keep the same definition.
1646 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1648 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1650 JH/12 Expand items in router/transport headers_add or headers_remove lists
1651 individually rather than the list as a whole. Bug 1452.
1653 Required for reasonable handling of multiple headers_ options when
1654 they may be empty; requires that headers_remove items with embedded
1655 colons must have them doubled (or the list-separator changed).
1657 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1658 view the policy declared in the DMARC record. Currently, $dmarc_status
1659 is a combined value of both the record presence and the result of the
1662 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1664 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1665 dnslookup router and the smtp transport (applying to the forward
1668 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1669 of ldap servers used for a specific lookup. Patch provided by Heiko
1672 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1673 New variable $lookup_dnssec_authenticated for observability.
1675 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1676 Patch submitted by Lars Timman.
1678 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1680 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1681 Requires trusted mode and valid format message id, aborts otherwise.
1682 Patch contributed by Heiko Schlichting.
1684 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1685 certextract with support for various fields. Bug 1358.
1687 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1688 is requested by default, modifiable by smtp transport option
1691 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1692 operate on certificate variables to give certificate fingerprints
1693 Also new ${sha256:cert_variable}.
1695 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1697 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1698 Christian Aistleitner.
1700 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1702 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1703 file. Patch from Wolfgang Breyha.
1705 JH/25 Expand the coverage of the delivery $host and $host_address to
1706 client authenticators run in verify callout. Bug 1476.
1708 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1709 align with daemon_smtp_ports. Bug 72.
1711 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1712 support and error reporting did not work properly.
1714 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1715 and is readable. Patch from Andrew Colin Kissa.
1717 TL/14 Enhance documentation of ${run expansion and how it parses the
1718 commandline after expansion, particularly in the case when an
1719 unquoted variable expansion results in an empty value.
1721 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1723 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1724 by renaming away from it. Addresses GH issue 15, reported by
1727 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1729 TL/15 SECURITY: prevent double expansion in math comparison functions
1730 (can expand unsanitized data). Not remotely exploitable.
1737 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1739 PP/02 Make -n do something, by making it not do something.
1740 When combined with -bP, the name of an option is not output.
1742 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1745 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1746 $sender_host_name and config options to manage this, and basic check
1749 PP/05 DSCP support for outbound connections and control modifier for inbound.
1751 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1752 (Only plugin which currently uses this is kerberos4, which nobody should
1753 be using, but we should make it available and other future plugins might
1754 conceivably use it, even though it would break NAT; stuff *should* be
1755 using channel bindings instead).
1757 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1758 name; added for Sendmail compatibility; requires admin caller.
1759 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1760 just ignore it); requires trusted caller.
1761 Also parse but ignore: -Ac -Am -X<logfile>
1764 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1766 TL/02 Add +smtp_confirmation as a default logging option.
1768 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1769 Patch by Magnus Holmgren from 2007-02-20.
1771 TL/04 Bugzilla 1281 - Spec typo.
1772 Bugzilla 1283 - Spec typo.
1773 Bugzilla 1290 - Spec grammar fixes.
1775 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1777 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1779 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1780 mailing list by by Dmitry Isaikin.
1782 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1784 JH/02 Support "G" suffix to numbers in ${if comparisons.
1786 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1788 NM/01 Bugzilla 1197 - Spec typo
1789 Bugzilla 1196 - Spec examples corrections
1791 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1793 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1794 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1797 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1798 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1800 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1801 "acl {{name}{arg}...}", and optional args on acl condition
1804 JH/05 Permit multiple router/transport headers_add/remove lines.
1806 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1808 JH/07 Avoid using a waiting database for a single-message-only transport.
1809 Performance patch from Paul Fisher. Bugzilla 1262.
1811 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1814 JH/09 Add $headers_added variable, with content from use of ACL modifier
1815 add_header (but not yet added to the message). Bugzilla 199.
1817 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1818 Pulled from Bugzilla 817 by Wolfgang Breyha.
1820 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1822 (nb: this is the same fix as in Exim 4.80.1)
1824 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1827 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1830 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1832 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1833 advertise SMTP AUTH mechanism to us, instead of a generic
1834 protocol violation error. Also, make Exim more robust to bad
1835 data from the Dovecot auth socket.
1837 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1839 When a queue runner is handling a message, Exim first routes the
1840 recipient addresses, during which it prunes them based on the retry
1841 hints database. After that it attempts to deliver the message to
1842 any remaining recipients. It then updates the hints database using
1845 So if a recipient address works intermittently, it can get repeatedly
1846 deferred at routing time. The retry hints record remains fresh so the
1847 address never reaches the final cutoff time.
1849 This is a fairly common occurrence when a user is bumping up against
1850 their storage quota. Exim had some logic in its local delivery code
1851 to deal with this. However it did not apply to per-recipient defers
1852 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1854 This change adds a proper retry rule check during routing so that the
1855 final cutoff time is checked against the message's age. We only do
1856 this check if there is an address retry record and there is not a
1857 domain retry record; this implies that previous attempts to handle
1858 the address had the retry_use_local_parts option turned on. We use
1859 this as an approximation for the destination being like a local
1860 delivery, as in LMTP.
1862 I suspect this new check makes the old local delivery cutoff check
1863 redundant, but I have not verified this so I left the code in place.
1865 TF/02 Correct gecos expansion when From: is a prefix of the username.
1867 Test 0254 submits a message to Exim with the header
1871 When I ran the test suite under the user fanf2, Exim expanded
1872 the header to contain my full name, whereas it should have added
1873 a Resent-Sender: header. It erroneously treats any prefix of the
1874 username as equal to the username.
1876 This change corrects that bug.
1878 GF/01 DCC debug and logging tidyup
1879 Error conditions log to paniclog rather than rejectlog.
1880 Debug lines prefixed by "DCC: " to remove any ambiguity.
1882 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1884 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1885 Bug spotted by Jeremy Harris; was flawed since initial commit.
1886 Would have resulted in OCSP responses post-SNI triggering an Exim
1887 NULL dereference and crash.
1889 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1891 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1892 Bug detection, analysis and fix by Samuel Thibault.
1893 Bugzilla 1331, Debian bug #698092.
1895 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1897 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1898 Server implementation by Todd Lyons, client by JH.
1899 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1900 config variable "prdr_enable" controls whether the server
1901 advertises the facility. If the client requests PRDR a new
1902 acl_data_smtp_prdr ACL is called once for each recipient, after
1903 the body content is received and before the acl_smtp_data ACL.
1904 The client is controlled by both of: a hosts_try_prdr option
1905 on the smtp transport, and the server advertisement.
1906 Default client logging of deliveries and rejections involving
1907 PRDR are flagged with the string "PRDR".
1909 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1910 fclose(). Diagnosis by Todd Lyons.
1912 PP/17 Update configure.default to handle IPv6 localhost better.
1913 Patch by Alain Williams (plus minor tweaks).
1916 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1917 This is now consistent with GnuTLS, and is now documented: the
1918 previous undocumented portable approach to treating the option as
1919 unset was to force an expansion failure. That still works, and
1920 an empty string is now equivalent.
1922 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1923 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1924 not performing validation itself.
1926 PP/20 Added force_command boolean option to pipe transport.
1927 Patch from Nick Koston, of cPanel Inc.
1929 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1932 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1934 PP/21 Fix eximon continuous updating with timestamped log-files.
1935 Broken in a format-string cleanup in 4.80, missed when I repaired the
1936 other false fix of the same issue.
1937 Report and fix from Heiko Schlichting.
1940 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1941 Report from Prashanth Katuri.
1943 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1944 It's SecureTransport, so affects any MacOS clients which use the
1945 system-integrated TLS libraries, including email clients.
1947 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1948 using a MIME ACL for non-SMTP local injection.
1949 Report and assistance in diagnosis by Warren Baker.
1951 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1953 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1955 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1956 last id that failed so it may be referenced in subsequent ACL's.
1958 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1961 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1962 ldap library initialization, allowing self-signed CA's to be
1963 used. Also properly sets require_cert option later in code by
1964 using NULL (global ldap config) instead of ldap handle (per
1965 session). Bug diagnosis and testing by alxgomz.
1967 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1968 the src/util/ subdirectory.
1970 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1971 renamed to Transport Post Delivery Action by Jeremy Harris, as
1974 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1975 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1976 redis_servers = needs to be configured which will be used by the redis
1977 lookup. Patch from Warren Baker, of The Packet Hub.
1979 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1981 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1982 hostname or reverse DNS when processing a host list. Used suggestions
1983 from multiple comments on this bug.
1985 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1987 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1988 Missed a few lines, added it to make the runtest require no keyboard
1991 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1992 contains upper case chars. Make router use caseful_local_part.
1994 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1995 support when GnuTLS has been built with p11-kit.
2001 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2003 This, or similar/improved, will also be change PP/11 of 4.82.
2009 PP/01 Handle short writes when writing local log-files.
2010 In practice, only affects FreeBSD (8 onwards).
2011 Bugzilla 1053, with thanks to Dmitry Isaikin.
2013 NM/01 Bugzilla 949 - Documentation tweak
2015 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2018 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2020 PP/02 Implemented gsasl authenticator.
2022 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2024 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2025 `pkg-config foo` for cflags/libs.
2027 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2028 with rest of GSASL and with heimdal_gssapi.
2030 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2031 `pkg-config foo` for cflags/libs for the TLS implementation.
2033 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2034 properties get this fed in as external SSF. A number of robustness
2035 and debugging improvements to the cyrus_sasl authenticator.
2037 PP/08 cyrus_sasl server now expands the server_realm option.
2039 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2040 Patch by Jeremy Harris.
2042 PP/10 Added dbmjz lookup type.
2044 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2046 PP/12 MAIL args handles TAB as well as SP, for better interop with
2047 non-compliant senders.
2048 Analysis and variant patch by Todd Lyons.
2050 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2051 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2052 Patch from Dirk Mueller <dmueller@suse.com>
2054 PP/13 tls_peerdn now print-escaped for spool files.
2055 Observed some $tls_peerdn in wild which contained \n, which resulted
2056 in spool file corruption.
2058 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2059 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2060 or write after TLS renegotiation, which otherwise led to messages
2063 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2064 as a tracking header (ie: a signed header comes before the signature).
2065 Patch from Wolfgang Breyha.
2067 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2068 comma-sep list; embedded commas doubled.
2070 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2072 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2074 Report and patch from Dmitry Banschikov.
2076 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2077 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2078 failures. We appear to now support TLS1.1+ with Exim.
2080 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2081 lets Exim select keys and certificates based upon TLS SNI from client.
2082 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2083 before an outbound SMTP session. New log_selector, +tls_sni.
2085 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2086 NULL dereference. Report and patch from Alun Jones.
2088 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2089 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2090 Not seeing resolver debug output on NetBSD, but suspect this is a
2091 resolver implementation change.
2093 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2094 Left warnings. Added "eximon gdb" invocation mode.
2096 PP/21 Defaulting "accept_8bitmime" to true, not false.
2098 PP/22 Added -bw for inetd wait mode support.
2100 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2101 locate the relevant includes and libraries. Made this the default.
2103 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2104 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2106 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2107 This may cause build issues on older platforms.
2109 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2110 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2111 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2112 Added SNI support via GnuTLS too.
2113 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2115 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2117 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2118 Applied second patch from Janne, implementing suggestion to default
2119 multiple-strings-in-record handling to match SPF spec.
2121 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2123 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2124 read-only, out of scope).
2125 Patch from Wolfgang Breyha, report from Stuart Northfield.
2127 PP/29 Fix three issues highlighted by clang analyser static analysis.
2128 Only crash-plausible issue would require the Cambridge-specific
2129 iplookup router and a misconfiguration.
2130 Report from Marcin Mirosław.
2132 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2134 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2135 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2136 As part of this, removing so much warning spew let me fix some minor
2137 real issues in debug logging.
2139 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2140 assignment on my part. Fixed.
2142 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2143 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2144 Janne Snabb (who went above and beyond: thank you).
2146 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2147 string otherwise requires a connection and a bunch more work and it's
2148 relatively easy to get wrong. Should also expose TLS library linkage
2151 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2152 64-bit ${eval} (JH/03).
2154 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2155 GNU libc to support some of the 64-bit stuff, should not lead to
2156 conflicts. Defined before os.h is pulled in, so if a given platform
2157 needs to override this, it can.
2159 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2160 protection layer was required, which is not implemented.
2161 Bugzilla 1254, patch from Wolfgang Breyha.
2163 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2164 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2165 tls_dhparam take prime identifiers. Also unbreak combination of
2166 OpenSSL+DH_params+TLSSNI.
2168 PP/39 Disable SSLv2 by default in OpenSSL support.
2174 PP/01 Solaris build fix for Oracle's LDAP libraries.
2175 Bugzilla 1109, patch from Stephen Usher.
2177 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2179 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2182 TF/02 Fix a couple more cases where we did not log the error message
2183 when unlink() failed. See also change 4.74-TF/03.
2185 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2186 lock up or crash if it happened to be inside a call to libc when it
2187 got a SIGUSR1 from exiwhat.
2189 The SIGUSR1 handler appends the current process status to the process
2190 log which is later printed by exiwhat. It used to use the general
2191 purpose logging code to do this, but several functions it calls are
2192 not safe for signals.
2194 The new output code in the SIGUSR1 handler is specific to the process
2195 log, and simple enough that it's easy to inspect for signal safety.
2196 Removing some special cases also simplifies the general logging code.
2197 Removing the spurious timestamps from the process log simplifies
2200 TF/04 Improved ratelimit ACL condition.
2202 The /noupdate option has been deprecated in favour of /readonly which
2203 has clearer semantics. The /leaky, /strict, and /readonly update modes
2204 are mutually exclusive. The update mode is no longer included in the
2205 database key; it just determines when the database is updated. (This
2206 means that when you upgrade Exim will forget old rate measurements.)
2208 Exim now checks that the per_* options are used with an update mode that
2209 makes sense for the current ACL. For example, when Exim is processing a
2210 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2211 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2212 must specify per_mail/readonly. If you omit the update mode it defaults to
2213 /leaky where that makes sense (as before) or /readonly where required.
2215 The /noupdate option is now undocumented but still supported for
2216 backwards compatibility. It is equivalent to /readonly except that in
2217 ACLs where /readonly is required you may specify /leaky/noupdate or
2218 /strict/noupdate which are treated the same as /readonly.
2220 A useful new feature is the /count= option. This is a generalization
2221 of the per_byte option, so that you can measure the throughput of other
2222 aggregate values. For example, the per_byte option is now equivalent
2223 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2225 The per_rcpt option has been generalized using the /count= mechanism
2226 (though it's more complicated than the per_byte equivalence). When it is
2227 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2228 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2229 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2230 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2231 Note that using per_rcpt with a non-readonly update mode in more than
2232 one ACL will cause the recipients to be double-counted. (The per_mail
2233 and per_byte options don't have this problem.)
2235 The handling of very low rates has changed slightly. If the computed rate
2236 is less than the event's count (usually one) then this event is the first
2237 after a long gap. In this case the rate is set to the same as this event's
2238 count, so that the first message of a spam run is counted properly.
2240 The major new feature is a mechanism for counting the rate of unique
2241 events. The new per_addr option counts the number of different
2242 recipients that someone has sent messages to in the last time period. It
2243 behaves like per_rcpt if all the recipient addresses are different, but
2244 duplicate recipient addresses do not increase the measured rate. Like
2245 the /count= option this is a general mechanism, so the per_addr option
2246 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2247 example, measure the rate that a client uses different sender addresses
2248 with the options per_mail/unique=$sender_address. There are further
2249 details in the main documentation.
2251 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2253 TF/06 Removed a few PCRE remnants.
2255 TF/07 Automatically extract Exim's version number from tags in the git
2256 repository when doing development or release builds.
2258 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2259 Bugzilla 879. Patch from Paul Fisher.
2261 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2262 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2265 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2267 PP/05 Variable $av_failed, true if the AV scanner deferred.
2268 Bugzilla 1078. Patch from John Horne.
2270 PP/06 Stop make process more reliably on build failure.
2271 Bugzilla 1087. Patch from Heiko Schlittermann.
2273 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2274 Bugzilla 1089. Patch from Heiko Schlittermann.
2276 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2277 Bugzilla 1131. Patch from Holger Weiß.
2279 PP/09 Handle IPv6 addresses with SPF.
2280 Bugzilla 860. Patch from Wolfgang Breyha.
2282 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2284 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2287 PP/11 match_* no longer expand right-hand-side by default.
2288 New compile-time build option, EXPAND_LISTMATCH_RHS.
2289 New expansion conditions, "inlist", "inlisti".
2291 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2293 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2295 PP/14 fix log_write() format string regression from TF/03.
2296 Bugzilla 1152. Patch from Dmitry Isaikin.
2302 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2304 PP/02 Harmonised TLS library version reporting; only show if debugging.
2305 Layout now matches that introduced for other libraries in 4.74 PP/03.
2307 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2309 PP/04 New "dns_use_edns0" global option.
2311 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2314 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2315 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2317 TK/01 Updated PolarSSL code to 0.14.2.
2318 Bugzilla 1097. Patch from Andreas Metzler.
2320 PP/07 Catch divide-by-zero in ${eval:...}.
2321 Fixes bugzilla 1102.
2323 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2326 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2327 format-string attack -- SECURITY: remote arbitrary code execution.
2329 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2330 time unintentionally subject to list matching rules, letting the header
2331 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2332 arbitrary string expansion). This allowed for information disclosure.
2334 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2335 INT_MIN/-1 -- value coerced to INT_MAX.
2341 NM/01 Workaround for PCRE version dependency in version reporting
2344 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2345 This fixes portability to compilers other than gcc, notably
2346 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2348 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2349 makefiles for portability to HP-UX and POSIX correctness.
2351 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2352 Also via indented variable definition in the Makefile.
2353 (Debugging by Oliver Heesakkers).
2355 PP/02 Restore caching of spamd results with expanded spamd_address.
2356 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2358 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2359 Improves build reliability. Fix from: Frank Elsner
2361 NM/02 Fix wide character breakage in the rfc2047 coding
2362 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2364 NM/03 Allow underscore in dnslist lookups
2365 Fixes bug 1026. Patch from Graeme Fowler
2367 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2368 Code patches from Adam Ciarcinski of NetBSD.
2370 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2373 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2374 is logged, to avoid truncation. Patch from John Horne.
2376 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2377 Patch from Jakob Hirsch.
2379 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2380 SQL string expansion failure details.
2381 Patch from Andrey Oktyabrski.
2383 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2384 Patch from Simon Arlott.
2386 PP/09 New lookups functionality failed to compile on old gcc which rejects
2387 extern declarations in function scope.
2388 Patch from Oliver Fleischmann
2390 PP/10 Use sig_atomic_t for flags set from signal handlers.
2391 Check getgroups() return and improve debugging.
2392 Fixed developed for diagnosis in bug 927 (which turned out to be
2395 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2396 Patch from Mark Zealey.
2398 PP/12 Bugzilla 1056: Improved spamd server selection.
2399 Patch from Mark Zealey.
2401 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2402 Based on patch from Heiko Schlittermann.
2404 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2405 Patch from Uwe Doering, sign-off by Michael Haardt.
2407 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2408 variable declaration deep within a block. Bug and patch from
2411 PP/15 lookups-Makefile IRIX compatibility coercion.
2413 PP/16 Make DISABLE_DKIM build knob functional.
2415 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2416 Patch by Simon Arlott
2418 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2419 variable argument macros. Our copy now differs from upstream.
2425 TF/01 Failure to get a lock on a hints database can have serious
2426 consequences so log it to the panic log.
2428 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2429 controlled using the smtp_confirmation log selector.
2431 TF/03 Include the error message when we fail to unlink a spool file.
2433 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2434 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2435 for maintaining out-of-tree patches for some time.
2437 PP/01 Bugzilla 139: Documentation and portability issues.
2438 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2439 Handle per-OS dynamic-module compilation flags.
2441 PP/02 Let /dev/null have normal permissions.
2442 The 4.73 fixes were a little too stringent and complained about the
2443 permissions on /dev/null. Exempt it from some checks.
2444 Reported by Andreas M. Kirchwitz.
2446 PP/03 Report version information for many libraries, including
2447 Exim version information for dynamically loaded libraries. Created
2448 version.h, now support a version extension string for distributors
2449 who patch heavily. Dynamic module ABI change.
2451 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2452 privilege escalation vulnerability whereby the Exim run-time user
2453 can cause root to append content of the attacker's choosing to
2456 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2459 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2460 If dropping privileges for untrusted macros, we disabled normal logging
2461 on the basis that it would fail; for the Exim run-time user, this is not
2462 the case, and it resulted in successful deliveries going unlogged.
2463 Fixed. Reported by Andreas Metzler.
2469 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2470 only prepend for the Resent-* case. Fixes regression introduced in
2471 Exim 4.70 by NM/22 for Bugzilla 607.
2473 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2474 increasing numbers of administrators be bitten by this.
2476 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2478 PP/03 Bugzilla 994: added openssl_options main configuration option.
2480 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2482 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2484 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2486 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2487 without a peer certificate, leading to a segfault because of an
2488 assumption that peers always have certificates. Be a little more
2489 paranoid. Problem reported by Martin Tscholak.
2491 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2492 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2493 NB: ClamAV planning to remove STREAM in "middle of 2010".
2494 CL also introduces -bmalware, various -d+acl logging additions and
2495 more caution in buffer sizes.
2497 PP/09 Implemented reverse_ip expansion operator.
2499 PP/10 Bugzilla 937: provide a "debug" ACL control.
2501 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2503 PP/12 Bugzilla 973: Implement --version.
2505 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2507 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2509 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2511 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2512 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2513 ignore trailing whitespace.
2515 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2517 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2520 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2521 Notification from Dr Andrew Aitchison.
2523 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2524 ExtendedDetectionInfo response format.
2525 Notification from John Horne.
2527 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2530 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2531 XSL and documented dependency on system catalogs, with examples of how
2534 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2537 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2538 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2539 configuration file which is writeable by the Exim user or group.
2541 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2542 of configuration files to cover files specified with the -C option if
2543 they are going to be used with root privileges, not just the default
2546 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2547 option (effectively making it always true).
2549 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2550 files to be used while preserving root privileges.
2552 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2553 that rogue child processes cannot use them.
2555 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2556 run-time user, instead of root.
2558 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2559 Exim run-time user without dropping privileges.
2561 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2562 result string, instead of calling string_vformat() twice with the same
2565 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2566 for other users. Others should always drop root privileges if they use
2567 -C on the command line, even for a whitelisted configure file.
2569 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2571 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2577 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2578 $data_path, and $header_path variables; fixed documentation bugs and
2581 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2582 exipick to access non-standard spools, including the "frozen" queue
2585 NM/01 Bugzilla 965: Support mysql stored procedures.
2586 Patch from Alain Williams
2588 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2590 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2591 Patch from Andreas Metzler
2593 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2594 Patch from Kirill Miazine
2596 NM/05 Bugzilla 671: Added umask to procmail example.
2598 JJ/03 installed exipick 20100323.0, fixing doc bug
2600 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2601 directory. Notification and patch from Dan Rosenberg.
2603 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2605 TK/02 Improve log output when DKIM signing operation fails.
2607 MH/01 Treat the transport option dkim_domain as a colon separated
2608 list, not as a single string, and sign the message with each element,
2609 omitting multiple occurences of the same signer.
2611 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2612 Bugzilla 985, 986. Patch by Simon Arlott
2614 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2615 Patch by Simon Arlott
2617 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2618 MBX locking. Notification from Dan Rosenberg.
2624 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2626 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2628 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2630 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2632 NM/04 Bugzilla 915: Flag broken perl installation during build.
2638 TK/01 Added patch by Johannes Berg that expands the main option
2639 "spamd_address" if it starts with a dollar sign.
2641 TK/02 Write list of recipients to X-Envelope-Sender header when building
2642 the mbox-format spool file for content scanning (suggested by Jakob
2645 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2646 (http://www.dcc-servers.net/) support via dccifd. Activated by
2647 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2649 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2650 by Mark Daniel Reidel <mr@df.eu>.
2652 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2653 When building exim an external PCRE library is now needed -
2654 PCRE is a system library on the majority of modern systems.
2655 See entry on PCRE_LIBS in EDITME file.
2657 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2658 conversation. Added nologin parameter to request.
2659 Patch contributed by Kirill Miazine.
2661 TF/01 Do not log submission mode rewrites if they do not change the address.
2663 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2665 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2666 log files in place. Contributed by Roberto Lima.
2668 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2670 TF/03 Bugzilla 615: When checking the local_parts router precondition
2671 after a local_part_suffix or local_part_prefix option, Exim now
2672 does not use the address's named list lookup cache, since this
2673 contains cached lookups for the whole local part.
2675 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2676 Robert Millan. Documentation is in experimental-spec.txt.
2678 TF/04 Bugzilla 668: Fix parallel build (make -j).
2680 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2682 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2683 Patch provided by Jan Srzednicki.
2685 TF/05 Leading white space used to be stripped from $spam_report which
2686 wrecked the formatting. Now it is preserved.
2688 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2689 that they are available at delivery time.
2691 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2693 TF/08 TLS error reporting now respects the incoming_interface and
2694 incoming_port log selectors.
2696 TF/09 Produce a more useful error message if an SMTP transport's hosts
2697 setting expands to an empty string.
2699 NM/06 Bugzilla 744: EXPN did not work under TLS.
2700 Patch provided by Phil Pennock.
2702 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2703 Patch provided by Richard Godbee.
2705 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2706 acl_smtp_notquit, added index entry.
2708 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2709 Patch provided by Eugene Bujak.
2711 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2712 accept(). Patch provided by Maxim Dounin.
2714 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2715 Patch provided by Phil Pennock.
2717 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2719 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2720 Patch provided by Brad "anomie" Jorsch.
2722 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2723 Patch provided by Dean Brooks.
2725 TK/05 Add native DKIM support (does not depend on external libraries).
2727 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2728 Patch provided by Graeme Fowler.
2730 NM/16 Bugzilla 851: Documentation example syntax fix.
2732 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2734 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2737 NM/19 Bugzilla 745: TLS version reporting.
2738 Patch provided by Phil Pennock.
2740 NM/20 Bugzilla 167: bool: condition support.
2741 Patch provided by Phil Pennock.
2743 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2744 clients. Patch provided by Phil Pennock.
2746 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2747 Patch provided by Brad "anomie" Jorsch.
2749 NM/23 Bugzilla 687: Fix misparses in eximstats.
2750 Patch provided by Heiko Schlittermann.
2752 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2753 Patch provided by Heiko Schlittermann.
2755 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2756 plus update to original patch.
2758 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2760 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2761 Patch provided by David Brownlee.
2763 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2765 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2767 NM/30 Bugzilla 888: TLS documentation bugfixes.
2769 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2771 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2772 Unlike the original bugzilla I have changed all shell scripts in src tree.
2774 NM/33 Bugzilla 898: Transport filter timeout fix.
2775 Patch by Todd Rinaldo.
2777 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2778 Patch by Serge Demonchaux.
2780 NM/35 Bugzilla 39: Base64 decode bug fixes.
2781 Patch by Jakob Hirsch.
2783 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2785 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2787 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2789 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2795 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2796 ALT-N's libdkim that I have put here:
2797 http://duncanthrax.net/exim-experimental/
2799 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2800 were called 'true' and it seems that C99 defines that as a reserved
2801 keyword to be used with 'bool' variable types. That means you could
2802 not include C99-style headers which use bools without triggering
2803 build errors in sieve.c.
2805 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2806 as mailq or other aliases. Changed the --help handling significantly
2807 to do whats expected. exim_usage() emits usage/help information.
2809 SC/01 Added the -bylocaldomain option to eximstats.
2811 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2813 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2815 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2821 PH/01 Another patch from the Sieve maintainer.
2823 PH/02 When an IPv6 address is converted to a string for single-key lookup
2824 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2825 dots are used instead of colons so that keys in lsearch files need not
2826 contain colons. This was done some time before quoting was made available
2827 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2828 (notated using the quote facility) so as to distinguish them from IPv4
2829 keys. This meant that lookups for IP addresses in host lists did not work
2830 for iplsearch lookups.
2832 This has been fixed by arranging for IPv6 addresses to be expressed with
2833 colons if the lookup type is iplsearch. This is not incompatible, because
2834 previously such lookups could never work.
2836 The situation is now rather anomalous, since one *can* have colons in
2837 ordinary lsearch keys. However, making the change in all cases is
2838 incompatible and would probably break a number of configurations.
2840 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2843 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2844 conversion specification without a maximum field width, thereby enabling
2845 a rogue spamd server to cause a buffer overflow. While nobody in their
2846 right mind would setup Exim to query an untrusted spamd server, an
2847 attacker that gains access to a server running spamd could potentially
2848 exploit this vulnerability to run arbitrary code as the Exim user.
2850 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2851 $primary_hostname instead of what libspf2 thinks the hosts name is.
2853 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2854 a directory entry by the name of the lookup key. Previously, if a
2855 symlink pointed to a non-existing file or a file in a directory that
2856 Exim lacked permissions to read, a lookup for a key matching that
2857 symlink would fail. Now it is enough that a matching directory entry
2858 exists, symlink or not. (Bugzilla 503.)
2860 PH/03 The body_linecount and body_zerocount variables are now exported in the
2863 PH/04 Added the $dnslist_matched variable.
2865 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2866 This means they are set thereafter only if the connection becomes
2869 PH/06 Added the client_condition to authenticators so that some can be skipped
2870 by clients under certain conditions.
2872 PH/07 The error message for a badly-placed control=no_multiline_responses left
2873 "_responses" off the end of the name.
2875 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2877 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2878 (without spaces) instead of just copying the configuration text.
2880 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2882 PH/11 Added $max_received_linelength.
2884 PH/12 Added +ignore_defer and +include_defer to host lists.
2886 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2887 way in which PCRE > 7.0 is built.
2889 PH/14 Implemented queue_only_load_latch.
2891 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2892 MAIL command. The effect was to mangle the value on 64-bit systems.
2894 PH/16 Another patch from the Sieve maintainer.
2896 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2898 PH/18 If a system quota error occurred while trying to create the file for
2899 a maildir delivery, the message "Mailbox is full" was not appended to the
2900 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2901 only to a quota excession during the actual writing of the file.
2903 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2904 characters?) which causes problems in log lines. The DN values are now
2905 passed through string_printing() before being added to log lines.
2907 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2908 and InterBase are left for another time.)
2910 PH/21 Added message_body_newlines option.
2912 PH/22 Guard against possible overflow in moan_check_errorcopy().
2914 PH/23 POSIX allows open() to be a macro; guard against that.
2916 PH/24 If the recipient of an error message contained an @ in the local part
2917 (suitably quoted, of course), incorrect values were put in $domain and
2918 $local_part during the evaluation of errors_copy.
2924 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2925 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2928 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2929 issue a MAIL command.
2931 PH/02 In an ACL statement such as
2933 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2935 if a client was not listed at all, or was listed with a value other than
2936 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2937 the condition was not true (as it should be), so access was not denied.
2938 The bug was that the ! inversion was incorrectly passed on to the second
2939 item. This has been fixed.
2941 PH/03 Added additional dnslists conditions == and =& which are different from
2942 = and & when the dns lookup returns more than one IP address.
2944 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2945 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2947 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2948 FSYNC, which compiles an option called disable_fsync that allows for
2949 bypassing fsync(). The documentation is heavily laced with warnings.
2951 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2953 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2954 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2955 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2956 including adding "make clean"; (3) Added -fPIC when compiling the test
2957 dynamically loaded module, to get rid of a warning.
2959 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2960 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2961 = 0s. The bug is otherwise harmless.
2963 PH/07 There was a bug in the dovecot authenticator such that the value of
2964 $auth1 could be overwritten, and so not correctly preserved, after a
2965 successful authentication. This usually meant that the value preserved by
2966 the server_setid option was incorrect.
2968 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2970 PH/09 Installed PCRE release 7.0.
2972 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2973 run for batched SMTP input. It is now run at the start of every message
2974 in the batch. While fixing this I discovered that the process information
2975 (output by running exiwhat) was not always getting set for -bs and -bS
2976 input. This is fixed, and it now also says "batched" for BSMTP.
2978 PH/11 Added control=no_pipelining.
2980 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2981 patch, slightly modified), and move the expansion of helo_data till after
2982 the connection is made in the smtp transport (so it can use these
2985 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2987 PH/14 Added log_selector = +pid.
2989 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2991 PH/16 Add ${if forany and ${if forall.
2993 PH/17 Added dsn_from option to vary the From: line in DSNs.
2995 PH/18 Flush SMTP output before performing a callout, unless control =
2996 no_callout_flush is set.
2998 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2999 was true (the default) a successful delivery failed to delete the retry
3000 item, thus causing premature timeout of the address. The bug is now
3003 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3005 PH/21 Long custom messages for fakedefer and fakereject are now split up
3006 into multiline responses in the same way that messages for "deny" and
3007 other ACL rejections are.
3009 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3010 with slight modification.
3012 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3013 draft, changing the syntax and factoring some duplicate code".
3015 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3016 for deliveries of the second and subsequent messages over the same SMTP
3019 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3020 ${reduce, with only minor "tidies".
3022 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3024 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3025 expansion side effects.
3027 PH/27 When a message times out after an over-quota error from an Exim-imposed
3028 quota, the bounce message says "mailbox is full". This message was not
3029 being given when it was a system quota that was exceeded. It now should
3032 MH/03 Made $recipients available in local_scan(). local_scan() already has
3033 better access to the recipient list through recipients_list[], but
3034 $recipients can be useful in postmaster-provided expansion strings.
3036 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3037 in the case of a MAIL command with additional options following the
3038 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3039 were accidentally chopped off.
3041 PH/29 SMTP synchronization checks are implemented when a command is read -
3042 there is a check that no more input is waiting when there shouldn't be
3043 any. However, for some commands, a delay in an ACL can mean that it is
3044 some time before the response is written. In this time, more input might
3045 arrive, invalidly. So now there are extra checks after an ACL has run for
3046 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3047 pipelining has not been advertised.
3049 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3051 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3052 returned in the SMTP response when smtp_return_error_details was false.
3053 This has been fixed.
3055 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3056 instead of the C I/O that was originally supplied, because problems were
3057 reported on Solaris.
3059 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3060 Exim which did not show up earlier: it was assuming that a call to
3061 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3062 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3063 was a macro that became an assignment, so it seemed to work. This has
3064 changed to a proper function call with a void return, hence the compile
3065 error. Exim's code has been fixed.
3067 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3070 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3072 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3074 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3075 because it turns out to be too big or there is a timeout) there is no
3076 "Completed" line in the log. When some messages of this type were
3077 selected by exigrep, they were listed as "not completed". Others were
3078 picked up by some special patterns. I have improved the selection
3079 criteria to be more general.
3081 PH/38 The host_find_failed option in the manualroute router can now be set
3082 to "ignore", to completely ignore a host whose IP address cannot be
3083 found. If all hosts are ignored, the behaviour is controlled by the new
3084 host_all_ignored option.
3086 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3087 homing or because of multiple MX records with /mx) generated more than
3088 one IP address, and the following item turned out to be the local host,
3089 all the secondary addresses of the first item were incorrectly removed
3090 from the list, along with the local host and any following hosts (which
3091 is what is supposed to happen).
3093 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3094 whoever called Exim into the -H file. In the case of the daemon it was
3095 behaving confusingly. When first started, it used values for whoever
3096 started the daemon, but after a SIGHUP it used the Exim user (because it
3097 calls itself on a restart). I have changed the code so that it now always
3100 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3101 message are rejected with the same error (e.g. no authentication or bad
3102 sender address), and a DATA command is nevertheless sent (as can happen
3103 with PIPELINING or a stupid MUA), the error message that was given to the
3104 RCPT commands is included in the rejection of the DATA command. This is
3105 intended to be helpful for MUAs that show only the final error to their
3108 PH/42 Another patch from the Sieve maintainer.
3110 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3111 Eximstats - Fixed some broken HTML links and added missing column headers
3113 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3114 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3116 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3118 PH/43 Yet another patch from the Sieve maintainer.
3120 PH/44 I found a way to check for a TCP/IP connection going away before sending
3121 the response to the final '.' that terminates a message, but only in the
3122 case where the client has not sent further data following the '.'
3123 (unfortunately, this is allowed). However, in many cases there won't be
3124 any further data because there won't be any more messages to send. A call
3125 to select() can be used: if it shows that the input is "ready", there is
3126 either input waiting, or the socket has been closed. An attempt to read
3127 the next input character can distinguish the two cases. Previously, Exim
3128 would have sent an OK response which the client would never have see.
3129 This could lead to message repetition. This fix should cure that, at
3130 least in a lot of common cases.
3132 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3133 advertised in response to EHLO.
3139 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3140 fixed by 4.65/MH/01 (is this a record?) are fixed:
3142 (i) An empty string was always treated as zero by the numeric comparison
3143 operators. This behaviour has been restored.
3145 (ii) It is documented that the numeric comparison operators always treat
3146 their arguments as decimal numbers. This was broken in that numbers
3147 starting with 0 were being interpreted as octal.
3149 While fixing these problems I realized that there was another issue that
3150 hadn't been noticed. Values of message_size_limit (both the global option
3151 and the transport option) were treated as octal if they started with 0.
3152 The documentation was vague. These values are now always treated as
3153 decimal, and I will make that clear in the documentation.
3159 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3160 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3163 MH/01 Don't check that the operands of numeric comparison operators are
3164 integers when their expansion is in "skipping" mode (fixes bug
3165 introduced by 4.64-PH/07).
3167 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3168 child addresses, Exim now panics and dies. Previously, because the count
3169 is held in a short int, deliveries were likely to be lost. As such a
3170 large number of recipients for a single message is ridiculous
3171 (performance will be very, very poor), I have chosen to impose a limit
3172 rather than extend the field.
3178 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3179 leftover -K file (the existence of which was triggered by #402).
3180 While we were at it, introduced process PID as part of the -K
3181 filename. This should rule out race conditions when creating
3184 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3185 processing considerably. Previous code took too long for large mails,
3186 triggering a timeout which in turn triggers #401.
3188 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3189 in the DK code in transports.c. sendfile() is not really portable,
3190 hence the _LINUX specificness.
3192 TF/01 In the add_headers option to the mail command in an Exim filter,
3193 there was a bug that Exim would claim a syntax error in any
3194 header after the first one which had an odd number of characters
3197 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3198 callout verification, Exim cached a "reject" for the entire domain. This
3199 is correct for most verifications, but it is not correct for a recipient
3200 verification with use_sender or use_postmaster set, because in that case
3201 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3202 case of MAIL FROM:<> rejection from other early rejections (e.g.
3203 rejection of HELO). When verifying a recipient using a non-null MAIL
3204 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3205 Whatever the result of the callout, the value of the domain cache is
3206 left unchanged (for any other kind of callout, getting as far as trying
3207 RCPT means that the domain itself is ok).
3209 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3212 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3213 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3214 session, the daemon ignores ECONNECT errors and logs others; it now
3215 ignores EPIPE as well.
3217 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3218 (quoted-printable decoding).
3220 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3221 later the small subsequent patch to fix an introduced bug.
3223 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3225 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3227 PH/08 An error is now given if message_size_limit is specified negative.
3229 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3230 to be given (somewhat) arbitrary names.
3232 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3235 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3236 miscellaneous code fixes
3238 PH/10 Added the log_reject_target ACL modifier to specify where to log
3241 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3242 hostname. This is wrong, because it relates to the incoming message (and
3243 probably the interface on which it is arriving) and not to the outgoing
3244 callout (which could be using a different interface). This has been
3245 changed to use the value of the helo_data option from the smtp transport
3246 instead - this is what is used when a message is actually being sent. If
3247 there is no remote transport (possible with a router that sets up host
3248 addresses), $smtp_active_hostname is used.
3250 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3251 tweaks were necessary in order to get it to work (see also 21 below):
3252 (a) The code assumed that strncpy() returns a negative number on buffer
3253 overflow, which isn't the case. Replaced with Exim's string_format()
3255 (b) There were several signed/unsigned issues. I just did the minimum
3256 hacking in of casts. There is scope for a larger refactoring.
3257 (c) The code used strcasecmp() which is not a standard C function.
3258 Replaced with Exim's strcmpic() function.
3259 (d) The code set only $1; it now sets $auth1 as well.
3260 (e) A simple test gave the error "authentication client didn't specify
3261 service in request". It would seem that Dovecot has changed its
3262 interface. Fortunately there's a specification; I followed it and
3263 changed what the client sends and it appears to be working now.
3265 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3268 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3269 address A is aliased to B and C, where B exists and C does not. Without
3270 -v the output is "A verified" because verification stops after a
3271 successful redirection if more than one address is generated. However,
3272 with -v the child addresses are also verified. Exim was outputting "A
3273 failed to verify" and then showing the successful verification for C,
3274 with its parentage. It now outputs "B failed to verify", showing B's
3275 parentage before showing the successful verification of C.
3277 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3278 look up a TXT record in a specific list after matching in a combined
3281 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3282 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3283 they consult the DNS. I had assumed they would set it the way they
3284 wanted; and indeed my experiments on Linux seem to show that in some
3285 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3286 To be on the safe side, however, I have now made the interface to
3287 host_find_byname() similar to host_find_bydns(), with an argument
3288 containing the DNS resolver options. The host_find_byname() function now
3289 sets these options at its start, just as host_find_bydns() does. The smtp
3290 transport options dns_qualify_single and dns_search_parents are passed to
3291 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3292 of host_find_byname() use the default settings of RES_DEFNAMES
3293 (qualify_single) but not RES_DNSRCH (search_parents).
3295 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3296 spool_read_header() do less string testing, by means of a preliminary
3297 switch on the second character of optional "-foo" lines. (This is
3298 overdue, caused by the large number of possibilities that now exist.
3299 Originally there were few.) While I was there, I also converted the
3300 str(n)cmp tests so they don't re-test the leading "-" and the first
3301 character, in the hope this might squeeze out yet more improvement.
3303 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3304 flag allowing group syntax was set by the header_syntax check but not
3305 turned off, possible causing trouble later; (2) The flag was not being
3306 set at all for the header_verify test, causing "group"-style headers to
3307 be rejected. I have now set it in this case, and also caused header_
3308 verify to ignore an empty address taken from a group. While doing this, I
3309 came across some other cases where the code for allowing group syntax
3310 while scanning a header line wasn't quite right (mostly, not resetting
3311 the flag correctly in the right place). These bugs could have caused
3312 trouble for malformed header lines. I hope it is now all correct.
3314 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3315 with the "reply" argument non-NULL. The code, however (which originally
3316 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3317 but it didn't always do it. This confused somebody who was copying the
3318 code for some other use. I have removed all the tests.
3320 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3321 feature that was used to support insecure browsers during the U.S. crypto
3322 embargo. It requires special client support, and Exim is probably the
3323 only MTA that supported it -- and would never use it because real RSA is
3324 always available. This code has been removed, because it had the bad
3325 effect of slowing Exim down by computing (never used) parameters for the
3326 RSA_EXPORT functionality.
3328 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3329 authenticator to fail if there's a tab character in the incoming data
3330 (there should never be unless someone is messing about, as it's supposed
3331 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3332 if the connection is using TLS or if the remote IP is the same as the
3333 local IP, and the "valid-client-cert option" if a client certificate has
3336 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3337 authenticators. This can be used for authorization after authentication
3338 succeeds. (In the case of plaintext, it servers for both authentication
3341 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3342 if any retry times were supplied.
3344 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3345 connection, where there is no client IP address to check. In this
3346 situation, the verify now always succeeds.
3348 PH/25 Applied John Jetmore's -Mset patch.
3350 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3352 PH/27 In a string expansion for a processed (not raw) header when multiple
3353 headers of the same name were present, leading whitespace was being
3354 removed from all of them, but trailing whitespace was being removed only
3355 from the last one. Now trailing whitespace is removed from each header
3356 before concatenation. Completely empty headers in a concatenation (as
3357 before) are ignored.
3359 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3360 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3362 PH/29 [Removed. This was a change that I later backed out, and forgot to
3363 correct the ChangeLog entry (that I had efficiently created) before
3364 committing the later change.]
3366 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3367 address errors (4xx response to RCPT) over the same connection as other
3368 messages routed to the same hosts. Such deliveries are always "forced",
3369 so retry times are not inspected. This resulted in far too many retries
3370 for the affected addresses. The effect occurred only when there were more
3371 hosts than the hosts_max_try setting in the smtp transport when it had
3372 the 4xx errors. Those hosts that it had tried were not added to the list
3373 of hosts for which the message was waiting, so if all were tried, there
3374 was no problem. Two fixes have been applied:
3376 (i) If there are any address or message errors in an SMTP delivery, none
3377 of the hosts (tried or untried) are now added to the list of hosts
3378 for which the message is waiting, so the message should not be a
3379 candidate for sending over the same connection that was used for a
3380 successful delivery of some other message. This seems entirely
3381 reasonable: after all the message is NOT "waiting for some host".
3382 This is so "obvious" that I'm not sure why it wasn't done
3383 previously. Hope I haven't missed anything, but it can't do any
3384 harm, as the worst effect is to miss an optimization.
3386 (ii) If, despite (i), such a delivery is accidentally attempted, the
3387 routing retry time is respected, so at least it doesn't keep
3388 hammering the server.
3390 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3391 in ${readsocket because some servers need this prod.
3393 PH/32 Added some extra debug output when updating a wait-xxx database.
3395 PH/33 The hint "could be header name not terminated by colon", which has been
3396 given for certain expansion errors for a long time, was not being given
3397 for the ${if def:h_colon_omitted{... case.
3399 PH/34 The spec says: "With one important exception, whenever a domain list is
3400 being scanned, $domain contains the subject domain." There was at least
3401 one case where this was not true.
3403 PH/35 The error "getsockname() failed: connection reset by peer" was being
3404 written to the panic log as well as the main log, but it isn't really
3405 panic-worthy as it just means the connection died rather early on. I have
3406 removed the panic log writing for the ECONNRESET error when getsockname()
3409 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3410 runs only) independently of the message's sender address. This meant
3411 that, if the 4xx error was in fact related to the sender, a different
3412 message to the same recipient with a different sender could confuse
3413 things. In particular, this can happen when sending to a greylisting
3414 server, but other circumstances could also provoke similar problems.
3415 I have changed the default so that the retry time for these errors is now
3416 based a combination of the sender and recipient addresses. This change
3417 can be overridden by setting address_retry_include_sender=false in the
3420 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3421 remote server are returned as part of bounce messages. This was not
3422 happening for LMTP over a pipe (the lmtp transport), but now it is the
3423 same for both kinds of LMTP.
3425 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3426 in header lines that were in fact CNAMEs. This is no longer the case.
3428 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3429 and queue runs started by the daemon processed all messages. This has
3430 been fixed so that -R and -S can now usefully be given with -q<time>.
3432 PH/40 Import PCRE release 6.7 (fixes some bugs).
3434 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3436 PH/42 Give an error if -q is specified more than once.
3438 PH/43 Renamed the variables $interface_address and $interface_port as
3439 $received_ip_address and $received_port, to make it clear that these
3440 values apply to message reception, and not to the outgoing interface when
3441 a message is delivered. (The old names remain recognized, of course.)
3443 PH/44 There was no timeout on the connect() call when using a Unix domain
3444 socket in the ${readsocket expansion. There now is.
3446 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3447 be meaningful with "accept".
3449 SC/01 Eximstats V1.43
3450 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3452 SC/02 Eximstats V1.44
3453 Use a glob alias rather than an array ref in the generated
3454 parser. This improves both readability and performance.
3456 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3457 Collect SpamAssassin and rejection statistics.
3458 Don't display local sender or destination tables unless
3459 there is data to show.
3460 Added average volumes into the top table text output.
3462 SC/04 Eximstats V1.46
3463 Collect data on the number of addresses (recipients)
3464 as well as the number of messages.
3466 SC/05 Eximstats V1.47
3467 Added 'Message too big' to the list of mail rejection
3468 reasons (thanks to Marco Gaiarin).
3470 SC/06 Eximstats V1.48
3471 Mainlog lines which have GMT offsets and are too short to
3472 have a flag are now skipped.
3474 SC/07 Eximstats V1.49 (Alain Williams)
3475 Added the -emptyok flag.
3477 SC/08 Eximstats V1.50
3478 Fixes for obtaining the IP address from reject messages.
3480 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3481 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3482 whitespace changes from 4.64-PH/27
3484 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3487 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3488 are found, allow negative numbers in numeric criteria)
3490 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3492 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3493 to match changes made in 4.64-PH/43
3495 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3497 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3498 "domain" is actually an IP address, and force a failure. This locks out
3499 those revolvers/nameservers that support "A-for-A" lookups, in
3500 contravention of the specifications.
3502 PH/48 When a host name was looked up from an IP address, and the subsequent
3503 forward lookup of the name timed out, the host name was left in
3504 $sender_host_name, contrary to the specification.
3506 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3507 restricted to single-key lookups, Exim was not diagnosing an error if
3508 * or *@ was used with a query-style lookup.
3510 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3512 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3513 long ago, but noone interested enough thought of it. Let's just say that
3514 the "1.1" means that there are some new functions that weren't there at
3515 some point in the past.
3517 PH/51 Error processing for expansion failure of helo_data from an smtp
3518 transport during callout processing was broken.
3520 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3521 tested/used via the -bh/-bhc/-bs options.
3523 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3524 bug, fixed in subsequent PCRE releases).
3526 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3527 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3529 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3535 SC/01 Use a glob alias rather than an array ref in eximstats generated
3536 parser. This improves both readability and performance.
3538 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3539 Don't display local sender or destination tables in eximstats unless
3540 there is data to show.
3541 Added average volumes into the eximstats top table text output.
3543 SC/03 Collect data on the number of addresses (recipients) as well
3544 as the number of messages in eximstats.
3546 TF/01 Correct an error in the documentation for the redirect router. Exim
3547 does (usually) call initgroups() when daemonizing.
3549 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3550 with consistent privilege compared to when running as a daemon.
3552 TF/03 Note in the spec that $authenticated_id is not set for local
3553 submissions from trusted users.
3555 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3556 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3558 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3559 by adding some example configuration directives to the default
3560 configuration file. A little bit of work is required to uncomment the
3561 directives and define how usernames and passwords are checked, but
3562 there is now a framework to start from.
3564 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3565 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3566 without this. I don't know how relevant this is to other LDAP libraries.
3568 PH/02 Add the verb name to the "unknown ACL verb" error.
3570 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3572 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3574 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3575 directory not expanded when it should be if an expanded home directory
3576 was set for the address (which is overridden by the transport).
3578 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3581 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3582 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3583 because it is too late at that time, and has no effect.
3585 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3586 security issue with \' (bugzilla #107). I could not use the
3587 PQescapeStringConn() function, because it needs a PGconn value as one of
3590 PH/08 When testing addresses using -bt, indicate those final addresses that
3591 are duplicates that would not cause an additional delivery. At least one
3592 person was confused, thinking that -bt output corresponded to deliveries.
3593 (Suppressing duplicates isn't a good idea as you lose the information
3594 about possibly different redirections that led to the duplicates.)
3596 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3597 systems where poll() doesn't work, in particular OS X.
3599 PH/10 Added more information to debugging output for retry time not reached.
3601 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3602 operations in malware.c.
3604 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3607 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3608 syslog_duplication set false, log lines that would normally be written
3609 both the the main log and to the reject log were not written to syslog at
3612 PH/14 In the default configuration, change the use of "message" in ACL warn
3613 statements to "add_header".
3615 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3616 not followed by a command (e.g. "seen endif").
3618 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3619 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3622 PH/17 Added extra conditions to the default value of delay_warning_condition
3626 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3627 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3628 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3631 The Auto-Submitted: and various List- headers are standardised, whereas I
3632 don't think Precedence: ever was.
3634 PH/18 Refactored debugging code in route_finduser() to show more information,
3635 in particular, the error code if getpwnam() issues one.
3637 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3638 This is apparently needed in addition to the PH/07 change above to avoid
3639 any possible encoding problems.
3641 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3642 but not after initializing Perl.
3644 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3645 output them only if debugging. By default they are written stderr,
3646 apparently, which is not desirable.
3648 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3651 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3654 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3656 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3657 authenticated or an ident call has been made. Suppress the default
3658 values for $authenticated_id and $authenticated_sender (but permit -oMai
3659 and -oMas) when testing with -bh.
3661 PH/24 Re-jigged the order of the tests in the default configuration so that the
3662 tests for valid domains and recipients precede the DNS black list and CSA
3663 tests, on the grounds that those ones are more expensive.
3665 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3666 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3667 This bug exists in every version of Exim that I still have, right back to
3670 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3671 However, an attempt to turn on case-sensitivity in a regex key by
3672 including (?-i) didn't work because the subject string was already
3673 lowercased, and the effects were non-intuitive. It turns out that a
3674 one-line patch can be used to allow (?-i) to work as expected.
3680 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3681 other effects) broke the use of negated acl sub-conditions.
3683 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3686 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3687 "Deny" causes Exim to reject the incoming connection with a 554 error.
3688 Unfortunately, if there is a major crisis, such as a disk failure,
3689 tcp-wrappers gives "deny", whereas what one would like would be some
3690 kind of temporary error. A kludge has been added to help with this.
3691 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3692 554 error is used if errno is still zero or contains ENOENT (which occurs
3693 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3696 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3698 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3699 errors. Otherwise a message that provokes a temporary error (when other
3700 messages do not) can cause a whole host to time out.
3702 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3703 addresses were routed directly to files or pipes from a redirect router.
3704 File deliveries just didn't batch; pipe deliveries might have suffered
3707 PH/06 A failure to get a lock for a hints database would erroneously always say
3708 "Failed to get write lock", even when it was really a read lock.
3710 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3711 of 0600. This has been changed to use the value of the lockfile_mode
3712 option (which defaults to 0600).
3714 PH/08 Applied small patch from the Sieve maintainer.
3716 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3717 folder from quota calculations, a direct delivery into this folder messed
3718 up the contents of the maildirsize file. This was because the regex was
3719 used only to exclude .Trash (or whatever) when the size of the mailbox
3720 was calculated. There was no check that a delivery was happening into an
3721 excluded directory. This bug has been fixed by ignoring all quota
3722 processing for deliveries into excluded directories.
3724 PH/10 Added the maildirfolder_create_regex option to appendfile.
3730 PH/01 The code for finding all the local interface addresses on a FreeBSD
3731 system running IPv6 was broken. This may well have applied to all BSD
3732 systems, as well as to others that have similar system calls. The broken
3733 code found IPv4 interfaces correctly, but gave incorrect values for the
3734 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3735 that it would not match correctly against @[] and not recognize the IPv6
3738 PH/02 The ipliteral router was not recognizing addresses of the form user@
3739 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3741 PH/03 Added disable_ipv6.
3743 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3744 decoded form, because it is most often used to construct To: headers
3745 lines in autoreplies, and the decoded form may well be syntactically
3746 invalid. However, $reply_address has leading white space removed, and all
3747 newlines turned into spaces so that the autoreply transport does not
3750 PH/05 If group was specified without a user on a router, and no group or user
3751 was specified on a transport, the group from the router was ignored.
3753 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3754 for visible compile-time settings that can be used to change these
3755 numbers, for those that want even more. Backwards compatibility with old
3756 spool files has been maintained. However, going back to a previous Exim
3757 release will lost any variables that are in spool files.
3759 PH/07 Two small changes when running in the test harness: increase delay when
3760 passing a TCP/IP connection to a new process, in case the original
3761 process has to generate a bounce, and remove special handling of
3762 127.0.0.2 (sic), which is no longer necessary.
3764 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3765 be the same on different OS.
3767 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3770 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3771 whether --show-vars was specified or not
3773 JJ/02 exipick: Added support for new ACL variable spool format introduced
3776 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3777 syntactically invalid From: or Reply-to: line, and a filter used this to
3778 generate an autoreply, and therefore failed to obtain an address for the
3779 autoreply, Exim could try to deliver to a non-existent relative file
3780 name, causing unrelated and misleading errors. What now happens is that
3781 it logs this as a hard delivery error, but does not attempt to create a
3784 PH/11 The exinext utility has a -C option for testing purposes, but although
3785 the given file was scanned by exinext itself; it wasn't being passed on
3786 when Exim was called.
3788 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3789 an end-of-file indication when reading a command response.
3791 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3792 compiled. In many other places in Exim, IPv6 addresses are always
3793 recognized, so I have changed this. It also means that IPv4 domain
3794 literals of the form [IPV4:n.n.n.n] are now always recognized.
3796 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3797 used if the router is not running as root, for example, when verifying at
3798 ACL time, or when using -bh. The debugging output from this situation was
3799 non-existent - all you got was a failure to exec. I have made two
3802 (a) Failures to set uid/gid, the current directory, or a process leader
3803 in a subprocess such as that created by queryprogram now generate
3804 suitable debugging output when -d is set.
3806 (b) The queryprogram router detects when it is not running as root,
3807 outputs suitable debugging information if -d is set, and then runs
3808 the subprocess without attempting to change uid/gid.
3810 PH/15 Minor change to Makefile for building test_host (undocumented testing
3813 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3814 additional section of a DNS packet that returns MX or SRV records.
3815 Instead, it always explicitly searches for A/AAAA records. This avoids
3816 major problems that occur when a DNS server includes only records of one
3817 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3818 fixed another bug: if SRV records were looked up and the corresponding
3819 address records were *not* found in the additional section, the port
3820 values from the SRV records were lost.
3822 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3823 using the correct key (the original address) when searching the retry
3824 rules in order to find which one to use for generating the retry hint.
3826 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3827 adding the default one. Similarly, if it contains a Reply-To: header, the
3828 errors_reply_to option, if set, is not used.
3830 PH/19 When calculating a retry time, Exim used to measure the "time since
3831 failure" by looking at the "first failed" field in the retry record. Now
3832 it does not use this if it is later than than the arrival time of the
3833 message. Instead it uses the arrival time. This makes for better
3834 behaviour in cases where some deliveries succeed, thus re-setting the
3835 "first failed" field. An example is a quota failure for a huge message
3836 when small messages continue to be delivered. Without this change, the
3837 "time since failure" will always be short, possible causing more frequent
3838 delivery attempts for the huge message than are intended.
3839 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3841 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3842 $1, $2, $3) because the numerical variables can be reset during some
3843 expansion items (e.g. "match"), thereby losing the authentication data.
3845 PH/21 Make -bV show the size of off_t variables so that the test suite can
3846 decide whether to run tests for quotas > 2G.
3848 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3849 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3850 filecount value is greater than 2G or if a quota value is greater than 2G
3851 on a system where the size of off_t is not greater than 4, a panic error
3854 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3855 never match. The debug and -bh output now contains an explicit error
3856 message indicating a malformed IPv4 address or mask.
3858 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3859 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3860 PH/23 above applies.
3862 PH/25 Do not write to syslog when running in the test harness. The only
3863 occasion when this arises is a failure to open the main or panic logs
3864 (for which there is an explicit test).
3866 PH/26 Added the /no_tell option to "control=freeze".
3868 PH/27 If a host name lookup failed very early in a connection, for example, if
3869 the IP address matched host_lookup and the reverse lookup yielded a name
3870 that did not have a forward lookup, an error message of the form "no IP
3871 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3872 could be logged. Now it outputs the IP address instead of "NULL".
3874 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3875 allows the sender and the authenticated sender to be set when
3876 submitting a message from within Exim. Since child_open_exim() is
3877 documented for local_scan(), the new function should be too.
3879 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3880 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3881 results in an empty string is now treated as unset.
3883 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3885 PH/31 Added sender_verify_fail logging option.
3887 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3888 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3889 tidied the source and removed it altogether.
3891 PH/33 When a queue run was abandoned because the load average was too high, a
3892 log line was always written; now it is written only if the queue_run log
3893 selector is set. In addition, the log line for abandonment now contains
3894 information about the queue run such as the pid. This is always present
3895 in "start" and "stop" lines but was omitted from the "abandon" line.
3897 PH/34 Omit spaces between a header name and the colon in the error message that
3898 is given when verify = headers_syntax fails (if there are lots of them,
3899 the message gets confusing).
3901 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3902 names, as there are now some PTR records that contain slashes. This check
3903 is only to protect against broken name servers that fall over on strange
3904 characters, so the fact that it applies to all lookups doesn't matter.
3906 PH/36 Now that the new test suite is complete, we can remove some of the
3907 special code in Exim that was needed for the old test suite. For example,
3908 sorting DNS records because real resolvers return them in an arbitrary
3909 order. The new test suite's fake resolver always returns records in the
3912 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3913 bounces) except when queue_only is set, to avoid logging races between
3914 the different processes.
3916 PH/38 Panic-die if .include specifies a non-absolute path.
3918 PH/39 A tweak to the "H" retry rule from its user.
3920 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3921 a label. They prevented compilation on older perls.
3923 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3924 a warning to be raised on newish perls.
3926 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3927 on queue. Changes to match documented behaviour of showing count of
3928 messages matching specified criteria.
3930 PH/40 Changed the default ident timeout from 30s to 5s.
3932 PH/41 Added support for the use of login_cap features, on those BSD systems
3933 that have them, for controlling the resources used by pipe deliveries.
3935 PH/42 The content-scanning code uses fopen() to create files in which to put
3936 message data. Previously it was not paying any attention to the mode of
3937 the files. Exim runs with umask(0) because the rest of the code creates
3938 files with open(), and sets the required mode explicitly. Thus, these
3939 files were ending up world-writeable. This was not a big issue, because,
3940 being within the spool directory, they were not world-accessible. I have
3941 created a function called modefopen, which takes an additional mode
3942 argument. It sets umask(777), creates the file, chmods it to the required
3943 mode, then resets the umask. All the relevant calls to fopen() in the
3944 content scanning code have been changed to use this function.
3946 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3947 to 24 hours. This avoids potential overflow problems when processing G
3948 and H retry rules. I suspect nobody ever tinkers with this value.
3950 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3952 PH/45 When the plaintext authenticator is running as a client, the server's
3953 challenges are checked to ensure they are valid base64 strings. By
3954 default, the authentication attempt is cancelled if an invalid string is
3955 received. Setting client_ignore_invalid_base64 true ignores these errors.
3956 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3957 they are received. Thus, the responses can be made to depend on the
3958 challenges. If an invalid string is ignored, an empty string is placed in
3961 PH/46 Messages that are created by the autoreply transport now contains a
3962 References: header, in accordance with RFCs 2822 and 3834.
3964 PH/47 Added authenticated_sender_force to the smtp transport.
3966 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3968 PH/49 Installed latest patch from the Sieve maintainer.
3970 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3971 was also set, the appendfile transport was unnecessarily scanning a
3972 directory of message files (e.g. for maildir delivery) to find the count
3973 of files (along with the size), even though it did not need this
3974 information. It now does the scan only if it needs to find either the
3975 size of the count of files.
3977 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3979 PH/52 Two bugs concerned with error handling when the smtp transport is
3982 (i) Exim was not creating retry information for temporary errors given
3983 for individual recipients after the DATA command when the smtp transport
3984 was used in LMTP mode. This meant that they could be retried too
3985 frequently, and not timed out correctly.
3987 (ii) Exim was setting the flag that allows error details to be returned
3988 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3989 recipients that were returned after the DATA command.
3991 PH/53 This is related to PH/52, but is more general: for any failing address,
3992 when detailed error information was permitted to be returned to the
3993 sender, but the error was temporary, then after the final timeout, only
3994 "retry timeout exceeded" was returned. Now it returns the full error as
3995 well as "retry timeout exceeded".
3997 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3998 do this, and (what is worse) MTAs that accept it.
4000 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4001 will now be deprecated.
4003 PH/56 New os.c-cygwin from the Cygwin maintainer.
4005 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4006 formats (previously only available in exim formats via -bpr, -bpru,
4007 and -bpra. Now also available in native and exiqgrep formats)
4009 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4010 with very large, slow to parse queues
4012 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4014 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4016 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4017 responses to authentication challenges, though it was showing the
4018 challenges; (ii) I've removed the CR characters from the debug output for
4021 PH/58 Allow for the insertion of a newline as well as a space when a string
4022 is turned into more than one encoded-word during RFC 2047 encoding. The
4023 Sieve code now uses this.
4025 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4026 data_4xx, lost_connection, tls_required.
4028 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4029 message was being sent as an SMTP response.
4031 PH/61 Add -l and -k options to exicyclog.
4033 PH/62 When verifying, if an address was redirected to one new address, so that
4034 verification continued, and the new address failed or deferred after
4035 having set something in $address_data, the value of $address_data was not
4036 passed back to the ACL. This was different to the case when no
4037 redirection occurred. The value is now passed back in both cases.
4039 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4040 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4041 use login_cap.h, so on its own it isn't the distinguishing feature. The
4042 new name refers directly to the setclassresources() function.
4044 PH/65 Added configuration files for NetBSD3.
4046 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4048 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4049 is preferred over IPv4.
4051 PH/68 The bounce_return_message and bounce_return_body options were not being
4052 honoured for bounces generated during the reception of non-SMTP messages.
4053 In particular, this applied to messages rejected by the ACL. This bug has
4054 been fixed. However, if bounce_return_message is true and bounce_return_
4055 body is false, the headers that are returned for a non-SMTP message
4056 include only those that have been read before the error was detected.
4057 (In the case of an ACL rejection, they have all been read.)
4059 PH/69 The HTML version of the specification is now built in a directory called
4060 spec_html instead of spec.html, because the latter looks like a path with
4061 a MIME-type, and this confuses some software.
4063 PH/70 Catch two compiler warnings in sieve.c.
4065 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4066 function verify_get_ident() calls ip_connect() to connect a socket, but
4067 if the "connect()" function timed out, ip_connect() used to close the
4068 socket. However, verify_get_ident() also closes the socket later, and in
4069 between Exim writes to the log, which may get opened at this point. When
4070 the socket was closed in ip_connect(), the log could get the same file
4071 descriptor number as the socket. This naturally causes chaos. The fix is
4072 not to close the socket in ip_connect(); the socket should be closed by
4073 the function that creates it. There was only one place in the code where
4074 this was missing, in the iplookup router, which I don't think anybody now
4075 uses, but I've fixed it anyway.
4077 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4078 well as to direct DNS lookups. Otherwise the handling of names in host
4079 lists is inconsistent and therefore confusing.
4085 PH/01 Two changes to the default runtime configuration:
4087 (1) Move the checks for relay_from_hosts and authenticated clients from
4088 after to before the (commented out) DNS black list checks.
4090 (2) Add control=submission to the relay_from_hosts and authenticated
4091 clients checks, on the grounds that messages accepted by these
4092 statements are most likely to be submissions.
4094 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4096 (1) Generate an error if the third argument for the ${prvs expansion is
4099 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4102 (3) Reset the variables that are obtained from the first argument of
4103 ${prvscheck and used in the second argument before leaving the code,
4104 because their memory is reclaimed, so using them afterwards may do
4107 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4108 one (it's much easier than Tom thought :-).
4110 (5) Because of (4), we can now allow for the use of $prvscheck_result
4111 inside the third argument.
4113 PH/03 For some reason, the default setting of PATH when running a command from
4114 a pipe transport was just "/usr/bin". I have changed it to
4117 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4118 anything to be listed in the output from -bV.
4120 PH/05 When a filter generated an autoreply, the entire To: header line was
4121 quoted in the delivery log line, like this:
4123 => >A.N.Other <ano@some.domain> <original@ddress> ...
4125 This has been changed so that it extracts the operative address. There
4126 may be more than one such address. If so, they are comma-separated, like
4129 => >ano@some.domain,ona@other.domain <original@ddress> ...
4131 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4132 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4133 not being looked up in the rDNS to get a host name, Exim was showing the
4134 IP address twice in Received: lines, even though the IP addresses were
4135 identical. For example:
4137 Received: from [1.2.3.4] (helo=[1.2.3.4])
4139 However, if the real host name was known, it was omitting the HELO data
4140 if it matched the actual IP address. This has been tidied up so that it
4141 doesn't show the same IP address twice.
4143 PH/07 When both +timestamp and +memory debugging was on, the value given by
4144 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4145 called by the debug printing, thereby overwriting the timestamp buffer.
4146 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4148 PH/08 When the original message was included in an autoreply transport, it
4149 always said "this is a copy of the message, including all the headers",
4150 even if body_only or headers_only was set. It now gives an appropriate
4153 PH/09 Applied a patch from the Sieve maintainer which:
4155 o fixes some comments
4156 o adds the (disabled) notify extension core
4157 o adds some debug output for the result of if/elsif tests
4158 o points to the current vacation draft in the documentation
4159 and documents the missing references header update
4163 o fixes a bug in processing the envelope test (when testing
4164 multiple envelope elements, the last element determined the
4167 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4168 Electronic Mail") by including:
4170 Auto-submitted: auto-generated
4172 in the messages that it generates (bounce messages and others, such as
4173 warnings). In the case of bounce messages for non-SMTP messages, there was
4174 also a typo: it was using "Auto_submitted" (underscore instead of
4175 hyphen). Since every message generated by Exim is necessarily in response
4176 to another message, thes have all been changed to:
4178 Auto-Submitted: auto-replied
4180 in accordance with these statements in the RFC:
4182 The auto-replied keyword:
4184 - SHOULD be used on messages sent in direct response to another
4185 message by an automatic process,
4187 - MUST NOT be used on manually-generated messages,
4189 - MAY be used on Delivery Status Notifications (DSNs) and Message
4190 Disposition Notifications (MDNs),
4192 - MUST NOT be used on messages generated by automatic or periodic
4193 processes, except for messages which are automatic responses to
4196 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4197 to the default Received: header definition.
4199 PH/12 Added log selector acl_warn_skipped (default on).
4201 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4202 variables because (a) they are in the wrong storage pool and (b) even if
4203 they were copied, it wouldn't work properly because of the caching.
4205 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4206 checking when decoding. Apparently there are clients that generate
4207 overlong encoded strings. Why am I not surprised?
4209 PH/15 If the first argument of "${if match_address" was not empty, but did not
4210 contain an "@" character, Exim crashed. Now it writes a panic log message
4211 and treats the condition as false.
4213 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4215 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4216 extension "envelope-auth". The code is finished and in agreement with
4217 other implementations, but there is no documentation so far and in fact,
4218 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4219 not changing the active code.
4221 Print executed "if" and "elsif" statements when debugging is used. This
4222 helps a great deal to understand what a filter does.
4224 Document more things not specified clearly in RFC3028. I had all this
4225 sorted out, when out of a sudden new issues came to my mind. Oops."
4227 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4230 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4231 canonical form (as documented). However, after a host name lookup from
4232 the IP address, check_host() was doing a simple string comparison with
4233 addresses acquired from the DNS when checking that the found name did
4234 have the original IP as one of its addresses. Since any found IPv6
4235 addresses are likely to be in abbreviated form, the comparison could
4236 fail. Luckily, there already exists a function for doing the comparison
4237 by converting both addresses to binary, so now that is used instead of
4238 the text comparison.
4240 PH/20 There was another similar case to PH/19, when a complete host name was
4241 given in a host list; looking up its IP address could give an abbreviated
4242 form, whereas the current host's name might or might not be abbreviated.
4243 The same fix has been applied.
4249 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4250 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4253 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4254 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4256 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4258 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4259 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4260 Also, for "Auto-Submitted", treat anything other than "no" as
4261 non-personal, in accordance with RFC 3834. (Previously it treated
4262 anything starting "auto-" as non-personal.)
4264 TF/01 The control=submission/name=... option had a problem with syntax
4265 errors if the name included a slash character. The /name= option
4266 now slurps the rest of the string, so it can include any characters
4267 but it must come last in the list of options (after /sender_retain
4270 PH/05 Some modifications to the interface to the fake nameserver for the new
4278 TK/01 Added the "success_on_redirect" address verification option. See
4279 NewStuff for rationale and an example.
4281 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4283 PH/02 Patch to exigrep to allow it to work on syslog lines.
4285 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4286 fread() to skip over the body file's header line, because in Cygwin the
4287 header line is locked and is inaccessible.
4289 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4290 co-exist for some time) to make it clear that it is the Exim ID that is
4291 referenced, not the Message-ID: header line.
4293 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4294 string_format() function, because snprintf() does not exist on all
4297 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4298 ${stat: expansion item.
4300 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4301 protocol synchronization error", to keep the pedants happy.
4303 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4304 well as for IRIX systems, when gcc is being used. See the host.c source
4307 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4309 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4312 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4313 transport and to the smtp transport in LMTP mode.
4315 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4317 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4318 run a filter in a subprocess. This could lead to confusion in subsequent
4319 lookups in the parent process. There should also be a search_tidyup() at
4320 the end of the subprocess.
4322 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4323 only if the host matched helo_try_verify_hosts, which caused the
4324 verification to occur when the EHLO/HELO command was issued. The ACL just
4325 tested the remembered result. Now, if a previous verification attempt has
4326 not happened, "verify = helo" does it there and then.
4328 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4330 TK/03 Fix log output including CR from clamd.
4332 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4333 exist provoked a memory error which could cause a segfault.
4335 PH/15 Installed PCRE 6.2
4337 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4339 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4340 of the problem. Specifically, suggested +O2 rather than +O1 for the
4343 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4345 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4346 record, the port was not correctly logged when the outgoing_port log
4347 selector was set (it logged the transort's default port).
4349 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4350 fallback_hosts, and "hosts" in the smtp transport.
4352 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4353 host errors such as "Connection refused".
4355 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4356 authentication with radiusclient 0.4.9:
4358 - Error returned from rc_read_config was caught wrongly
4359 - Username/password not passed on to radius server due to wrong length.
4361 The presumption is that some radiusclient API changes for 4.51/PH/17
4362 were not taken care of correctly. The code is still untested by me (my
4363 Linux distribution still has 0.3.2 of radiusclient), but it was
4364 contributed by a Radius user.
4366 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4367 expanding the "port" option of the smtp transport.
4369 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4370 while reading a MIME header. Thanks to Tom Hughes for a patch.
4372 PH/24 Include config.h inside local_scan.h so that configuration settings are
4375 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4376 that in an ACL for RCPT (for example), you can examine exactly what was
4379 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4380 commands, but it was not correctly comparing the address with the actual
4381 client host address. Thus, it would show the EHLO address in Received:
4382 header lines when this was not necessary.
4384 PH/27 Added the % operator to ${eval:}.
4386 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4387 it should be ignoring failures (because with -C, for example, it has lost
4388 privilege). It wasn't ignoring creation failures other than "already
4391 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4394 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4395 because an input line was too long, either on its own, or by virtue of
4396 too many continuations, the temporary file was not being removed, and the
4397 return code was incorrect.
4399 PH/31 Missing "BOOL" in function definition in filtertest.c.
4401 PH/32 Applied Sieve patches from the maintainer.
4403 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4405 PH/33 Added "verify = not_blind".
4407 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4408 Local/Makefile (with some defaults set). These are used in built scripts
4409 such as exicyclog, but they have never been used in the exim_install
4410 script (though there are many overriding facilities there). I have
4411 arranged that the exim_install script now takes note of these two
4414 PH/35 Installed configuration files for Dragonfly.
4416 PH/36 When a locally submitted message by a trusted user did not contain a
4417 From: header, and the sender address was obtained from -f or from an SMTP
4418 MAIL command, and the trusted user did not use -F to supply a sender
4419 name, $originator_name was incorrectly used when constructing a From:
4420 header. Furthermore, $originator_name was used for submission mode
4421 messages from external hosts without From: headers in a similar way,
4422 which is clearly wrong.
4424 PH/37 Added control=suppress_local_fixups.
4426 PH/38 When log_selector = +received_sender was set, and the addition of the
4427 sender made the log line's construction buffer exactly full, or one byte
4428 less than full, an overflow happened when the terminating "\n" was
4431 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4432 when the result of a list match is failure because a DNS lookup failed.
4434 PH/40 RM_COMMAND is now used in the building process.
4436 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4437 the "build-* directories that it finds.
4439 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4440 address in a domain literal was a prefix of an interface address.
4442 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4443 when verifying a sender address, unless rewrite_headers is false.
4445 PH/44 Wrote a long comment about why errors_to addresses are verified as
4446 recipients, not senders.
4448 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4449 the ratelimit ACL was added.
4451 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4453 PH/46 Added extra information about PostgreSQL errors to the error string.
4455 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4456 avoiding the need to install special zones in a real server. This is
4457 backwards compatible; if it can't find the fake resolver, it drops back.
4458 Thus, both old and new test suites can be run.
4460 TF/02 Added util/ratelimit.pl
4462 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4463 clock is set back in time.
4465 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4466 Candler <B.Candler@pobox.com>.
4468 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4469 OK for addresses that are the result of redirections.
4471 PH/48 A number of further additions for the benefit of the new test suite,
4472 including a fake gethostbyname() that interfaces to the fake DNS resolver
4475 TF/06 The fix for widen_domains has also been applied to qualify_single and
4476 search_parents which are the other dnslookup options that can cause
4479 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4482 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4484 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4485 service" if it encountered an MX record with an empty target hostname.
4486 The message is now "an MX or SRV record indicated no SMTP service".
4488 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4489 if the DNS of the sending site is misconfigured. This is quite a
4490 common situation. This change restores the behaviour of treating a
4491 helo verification defer as a failure.
4493 PH/51 If self=fail was set on a router, the bounce message did not include the
4494 actual error message.
4500 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4502 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4503 command itself ended in error, the underlying message about the transport
4504 filter timeout was being overwritten with the pipe command error. Now the
4505 underlying error message should be appended to the second error message.
4507 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4509 PH/02 Reduce the amount of output that "make" produces by default. Full output
4510 can still be requested.
4512 PH/03 The warning log line about a condition test deferring for a "warn" verb
4513 was being output only once per connection, rather than after each
4514 occurrence (because it was using the same function as for successful
4515 "warn" verbs). This seems wrong, so I have changed it.
4517 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4518 it should not have, which might have caused a crash in the right
4519 circumstances, but probably never did.
4521 PH/04 Installed a modified version of Tony Finch's patch to make submission
4522 mode fix the return path as well as the Sender: header line, and to
4523 add a /name= option so that you can make the user's friendly name appear
4526 TF/03 Added the control = fakedefer ACL modifier.
4528 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4529 Mark Lowes for thorough testing.
4531 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4533 TK/03 Merged latest SRS patch from Miles Wilton.
4535 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4536 with the definition in sysexits.h (which is #included earlier).
4537 Fortunately, Exim does not actually use EX_OK. The code used to try to
4538 preserve the sysexits.h value, by assuming that macro definitions were
4539 scanned for macro replacements. I have been disabused of this notion,
4540 so now the code just undefines EX_OK before #including unistd.h.
4542 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4543 in the smtp transport. When a block could not be written in a single
4544 write() function, the timeout was being re-applied to each part-write.
4545 This seems wrong - if the receiver was accepting one byte at a time it
4546 would take for ever. The timeout is now adjusted when this happens. It
4547 doesn't have to be particularly precise.
4549 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4550 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4552 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4553 without a domain if the check to <postmaster@domain> fails.
4555 SC/01 Eximstats: added -xls and the ability to specify output files
4556 (patch written by Frank Heydlauf).
4558 SC/02 Eximstats: use FileHandles for outputting results.
4560 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4562 SC/04 Eximstats: fixed display of large numbers with -nvr option
4564 SC/05 Eximstats: fixed merging of reports with empty tables.
4566 SC/06 Eximstats: added the -include_original_destination flag
4568 SC/07 Eximstats: removed tabs and trailing whitespace.
4570 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4572 TK/06 MBOX spool code: Add real "From " MBOX separator line
4573 so the .eml file is really in mbox format (even though
4574 most programs do not really care). Patch from Alex Miller.
4576 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4577 The latter is generated from $received_to and is only set if the
4578 message has one envelope recipient. SA can use these headers,
4579 obviously out-of-the-box. Patch from Alex Miller.
4581 PH/08 The ${def test on a variable was returning false if the variable's
4582 value was "0", contrary to what the specification has always said!
4583 The result should be true unless the variable is empty.
4585 PH/09 The syntax error of a character other than { following "${if
4586 def:variable_name" (after optional whitespace) was not being diagnosed.
4587 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4588 accidental colon was present, for example, could give incorrect results.
4590 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4591 result is used (not including appendfile, where other changes are about
4594 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4595 This involved changing a lot of size variables from int to off_t. It
4596 should work with maildirs and everything.
4598 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4599 spamd dying while we are connected to it.
4601 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4604 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4605 function for BDB 4.3.
4607 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4609 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4610 the sources for PCRE have been re-arranged and more files are now
4613 PH/15 The code I had for printing potentially long long variables in PH/11
4614 above was not the best (it lost precision). The length of off_t variables
4615 is now inspected at build time, and an appropriate printing format (%ld
4616 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4617 to be "long long int" or "long int". This is needed for the internal
4618 formatting function string_vformat().
4620 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4621 the configuration file to be ":syslog", then the script "guesses" where
4622 the logs files are, rather than using the compiled in default. In our
4623 case the guess is not the same as the compiled default, so the script
4624 suddenly stopped working when I started to use syslog. The patch checks
4625 to see if log_file_path is "". If so, it attempts to read it from exim
4626 with no configuration file to get the compiled in version, before it
4627 falls back to the previous guessing code."
4629 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4630 implementing BATV in an Exim configuration. See NewStuff for the gory
4633 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4634 Makefile that are specific to HP-UX.
4636 PH/18 If the "use_postmaster" option was set for a recipient callout together
4637 with the "random" option, the postmaster address was used as the MAIL
4638 FROM address for the random test, but not for the subsequent recipient
4639 test. It is now used for both.
4641 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4642 patch removes a few documentation additions to RFC 3028, because the
4643 latest draft now contains them. It adds the new en;ascii-case comparator
4644 and a new error check for 8bit text in MIME parts. Comparator and
4645 require names are now matched exactly. I enabled the subaddress
4646 extension, but it is not well tested yet (read: it works for me)."
4648 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4649 rework some of the code of TK/09 above to avoid the hardwired use of
4650 "%lld" and "long long". Replaced the call to snprintf() with a call to
4653 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4654 records point to non-existent hosts", "retry timeout exceeded", and
4655 "retry time not reached for any host after a long failure period".
4657 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4658 experimental DomainKeys support:
4660 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4661 (2) On an error such as an illegally used "control", the wrong name for
4662 the control was given.
4664 These problems did NOT occur unless DomainKeys support was compiled.
4666 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4668 PH/24 Added ${if match_ip condition.
4670 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4671 databases so that it will be absolutely obvious if a crash occurs in the
4672 DB library. This is a regular occurrence (often caused by mis-matched
4675 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4676 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4677 picked up on a user's system that detects such things. There doesn't seem
4678 to be a gcc warning option for this - only an attribute that has to be
4679 put on the function's prototype. It seems that in Fedora Core 4 they have
4680 set this on a number of new functions. No doubt there will be more in due
4683 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4684 specify a transport. However, if an address that was verified by such a
4685 router was the subject of a callout, Exim crashed because it tried to
4686 read the rcpt_include_affixes from the non-existent transport. Now it
4687 just assumes that the setting of that option is false. This bug was
4688 introduced by 4.51/PH/31.
4690 PH/28 Changed -d+all to exclude +memory, because that information is very
4691 rarely of interest, but it makes the output a lot bigger. People tend to
4692 do -d+all out of habit.
4694 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4695 code in os-type was giving problems when libc.so lives in lib64, like on
4698 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4699 aren't the modern standard, and it seems that some systems' include files
4700 don't always have them. Exim was already checking for some of the newer
4701 ones like T_AAAA, and defining it itself. I've added checks for all the
4702 record types that Exim uses.
4704 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4705 not automatically generating a new one, as it is supposed to. This
4706 prevented TLS from working. If the file did exist, but contained invalid
4707 data, a new version was generated, as expected. It was only the case of a
4708 non-existent file that was broken.
4710 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4711 with a change in libdomainkeys > 0.64.
4713 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4714 from DNS. If the selector record carries the flag, it now has
4715 precedence over the domain-wide flag.
4717 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4719 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4720 the use of a port name that isn't defined in /etc/services) occurred, the
4721 message was deferred as in a normal delivery, and thus remained on the
4722 spool, instead of being failed because of the mua_wrapper setting. This
4723 is now fixed, and I tidied up some of the mua_wrapper messages at the
4726 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4727 the messages in a hash of arrays rather than using individual hashes.
4728 This is a bit cleaner and results in dramatic memory savings, albeit
4729 at a slight CPU cost.
4731 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4732 as requested by Marc Sherman.
4734 SC/10 Eximstats: added histograms for user specified patterns as requested
4737 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4739 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4740 fopen() in the content-scanning modules that did not already have it.
4746 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4747 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4749 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4751 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4753 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4754 to test Sieve filters that use "vacation".
4756 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4757 that changes the way the GnuTLS parameters are stored in the cache file.
4758 The new format can be generated externally. For backward compatibility,
4759 if the data in the cache doesn't make sense, Exim assumes it has read an
4760 old-format file, and it generates new data and writes a new file. This
4761 means that you can't go back to an older release without removing the
4764 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4765 work if there are any delivery delays because "one_time" forces the
4766 parent to be marked "delivered", so its unseen clone is never tried
4767 again. For this reason, Exim now forbids the simultaneous setting of
4770 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4771 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4772 entry if you want to know the details. The fix, however, neglected to
4773 consider the case when local delivery batching is involved. The test for
4774 "previously delivered" was not happening when checking to see if an
4775 address could be batched with a previous (undelivered) one; under
4776 certain circumstances this could lead to multiple deliveries to the same
4779 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4780 in its include files, and this causes problems building Exim.
4782 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4783 header_syntax) but Exim was just ignoring anything given after a slash.
4784 In particular, this caused confusion with an attempt to use "verify =
4785 reverse_host_lookup/defer_ok". An error is now given when options are
4786 supplied for verify items that do not have them. (Maybe reverse_host_
4787 lookup should have a defer_ok option, but that's a different point.)
4789 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4790 defined by RFC 821) to 2048, because there were problems with some AUTH
4791 commands, and RFC 1869 says the size should be increased for extended
4792 SMTP commands that take arguments.
4794 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4797 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4798 "unknown" error; now it says that the functionality isn't in the binary.
4800 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4801 an address' error message when a string expansion fails (syntax or
4802 whatever). Otherwise the password may appear in the log. Following change
4803 PH/42 below, there is no longer a chance of it appearing in a bounce
4806 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4808 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4809 of its addresses was ever tried. (Bugzilla bug #2.)
4811 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4812 the result incorrectly in the debug output. (It correctly added a newline
4813 to what was transported.)
4815 TF/01 Added $received_time.
4817 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4818 commented out examples of how to interface to a virus scanner and to
4819 SpamAssassin. Also added commented examples of av_scanner and
4820 spamd_address settings.
4822 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4823 and controls are allowed in which ACLs. There were a couple of minor
4824 errors. Some of the entries in the conditions table (which is a table of
4825 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4826 negation of where the condition IS allowed.
4828 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4830 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4831 header file does not have a version number, so I've had to invent a new
4832 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4833 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4834 radiusclient), but it was contributed by a Radius user.
4836 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4837 files or directories, for OpenSSL.
4839 PH/19 When an Exim process that is running as root has to create an Exim log
4840 file, it does so in a subprocess that runs as exim:exim so as to get the
4841 ownership right at creation (otherwise, other Exim processes might see
4842 the file with the wrong ownership). There was no test for failure of this
4843 fork() call, which would lead to the process getting stuck as it waited
4844 for a non-existent subprocess. Forks do occasionally fail when resources
4845 run out. I reviewed all the other calls to fork(); they all seem to check
4848 PH/20 When checking for unexpected SMTP input at connect time (before writing
4849 the banner), Exim was not dealing correctly with a non-positive return
4850 from the read() function. If the client had disconnected by this time,
4851 the result was a log entry for a synchronization error with an empty
4852 string after "input=" when read() returned zero. If read() returned -1
4853 (an event I could not check), uninitialized data bytes were printed.
4854 There were reports of junk text (parts of files, etc) appearing after
4857 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4859 PH/22 Added support for macro redefinition, and (re)definition in between
4860 driver and ACL definitions.
4862 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4863 forgetting to use the resulting value; it was using the unexpanded value.
4865 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4866 hadn't been configured. The fix is from Juergen Kreileder, who
4867 understands it better than I do:
4869 "Here's what I see happening with three configured cyrus_sasl
4870 authenticators configured (plain, login, cram-md5):
4872 On startup auth_cyrus_sasl_init() gets called for each of these.
4873 This means three calls to sasl_listmech() without a specified mech_list.
4874 => SASL tests which mechs of all available mechs actually work
4875 => three warnings about OTP not working
4876 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4878 With the patch, sasl_listmech() also gets called three times. But now
4879 SASL's mech_list option is set to the server_mech specified in the the
4880 authenticator. Or in other words, the answer from sasl_listmech()
4881 gets limited to just the mech you're testing for (which is different
4883 => the return list contains just 'plain' or 'login', 'cram-md5' or
4884 nothing depending on the value of ob->server_mech.
4886 I've just tested the patch: Authentication still works fine,
4887 unavailable mechs specified in the exim configuration are still
4888 caught, and the auth.log warnings about OTP are gone."
4890 PH/25 When debugging is enabled, the contents of the command line are added
4891 to the debugging output, even when log_selector=+arguments is not
4894 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4895 answer is "GNU", and only if the return is "GNU/something" is the answer
4898 PH/27 $acl_verify_message is now set immediately after the failure of a
4899 verification in an ACL, and so is available in subsequent modifiers. In
4900 particular, the message can be preserved by coding like this:
4902 warn !verify = sender
4903 set acl_m0 = $acl_verify_message
4905 Previously, $acl_verify_message was set only while expanding "message"
4906 and "log_message" when a very denied access.
4908 PH/28 Modified OS/os.c-Linux with
4910 -#ifndef OS_LOAD_AVERAGE
4911 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4913 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4914 nomenclature these days.)
4916 PH/29 Installed patch from the Sieve maintainer that adds the options
4917 sieve_useraddress and sieve_subaddress to the redirect router.
4919 PH/30 In these circumstances:
4920 . Two addresses routed to the same list of hosts;
4921 . First host does not offer TLS;
4922 . First host accepts first address;
4923 . First host gives temporary error to second address;
4924 . Second host offers TLS and a TLS session is established;
4925 . Second host accepts second address.
4926 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4927 and peerdn, if requested) that were in fact used only for the second
4930 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4931 attention to any local part prefix or suffix that was matched by the
4932 router that accepted the address. It now behaves in the same way as it
4933 does for delivery: the affixes are removed from the local part unless
4934 rcpt_include_affixes is set on the transport.
4936 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4937 timeout during the DATA phase of an incoming message.
4939 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4940 applied a patch sanctioned by the Sieve maintainer.
4942 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4943 the uid or gid is negative. A case of a negative gid caused this to be
4944 noticed. The fix allows for either to be negative.
4946 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4947 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4950 PH/36 The change PH/12 above was broken. Fixed it.
4952 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4953 the grounds that routing the same address twice would always produce the
4954 same answer. This might have been true once, but it is certainly no
4955 longer true now. Routing a child address may depend on the previous
4956 routing that produced that child. Some complicated redirection strategies
4957 went wrong when messages had multiple recipients, and made Exim's
4958 behaviour dependent on the order in which the addresses were given.
4960 I have moved the duplicate checking until after the routing is complete.
4961 Exim scans the addresses that are assigned to local and remote
4962 transports, and removes any duplicates. This means that more work will be
4963 done, as duplicates will always all be routed, but duplicates are
4964 presumably rare, so I don't expect this is of any significance.
4966 For deliveries to pipes, files, and autoreplies, the duplicate checking
4967 still happens during the routing process, since they are not going to be
4970 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4971 It corrects a timeout issue with spamd. This is Ian's comment: "The
4972 background is that sometimes spamd either never reads data from a
4973 connection it has accepted, or it never writes response data. The exiscan
4974 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4975 blindly assumes that writes won't block so it may never time out."
4977 PH/39 Allow G after quota size as well as K and M.
4979 PH/40 The value set for $authenticated_id in an authenticator may not contain
4980 binary zeroes or newlines because the value is written to log lines and
4981 to spool files. There was no check on this. Now the value is run through
4982 the string_printing() function so that such characters are converted to
4983 printable escape sequences.
4985 PH/41 $message_linecount is a new variable that contains the total number of
4986 lines in the message. Compare $body_linecount, which is the count for the
4989 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4990 bounce and delay warning messages, except in certain special cases, which
4993 (a) An SMTP error message from a remote host;
4994 (b) A message specified in a :fail: redirection;
4995 (c) A message specified in a "fail" command in a system filter;
4996 (d) A message specified in a FAIL return from the queryprogram router;
4997 (e) A message specified by the cannot_route_message router option.
4999 In these cases only, Exim does include the error details in bounce and
5000 warning messages. There are also a few cases where bland messages such
5001 as "unrouteable address" or "local delivery error" are given.
5003 PH/43 $value is now also set for the "else" part of a ${run expansion.
5005 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5006 being worked on, but at least Exim now implements the latest version to
5009 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5010 process to complete was treated as a delivery failure, a timeout while
5011 writing the message to the pipe was logged, but erroneously treated as a
5012 successful delivery. Such timeouts include transport filter timeouts. For
5013 consistency with the overall process timeout, these timeouts are now
5014 treated as errors, giving rise to delivery failures by default. However,
5015 there is now a new Boolean option for the pipe transport called
5016 timeout_defer, which, if set TRUE, converts the failures into defers for
5017 both kinds of timeout. A transport filter timeout is now identified in
5020 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5021 systems where "make" and "gmake" are different, calling "gmake" at top
5022 level broke things. I've arranged for the value of $(MAKE) to be passed
5023 from the Makefile to this script so that it can call the same version of
5027 A note about Exim versions 4.44 and 4.50
5028 ----------------------------------------
5030 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5031 changes of various kinds. As a consequence, a big documentation update was
5032 needed. This delayed the release for rather longer than seemed good, especially
5033 in the light of a couple of (minor) security issues. Therefore, the changes
5034 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5035 release. So 4.44 and 4.50 are in effect two different branches that both start
5038 I have left the 4.50 change log unchanged; it contains all the changes since
5039 4.43. The change log for 4.44 is below; many of its items are identical to
5040 those for 4.50. This seems to be the most sensible way to preserve the
5041 historical information.
5047 1. Minor wording change to the doc/README.SIEVE file.
5049 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5050 computation of the current number of files was incorrect.
5052 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5053 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5056 4. Give more explanation in the error message when the command for a transport
5057 filter fails to execute.
5059 5. There are several places where Exim runs a non-Exim command in a
5060 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5061 was being done only for the command run by the queryprogram router. It is
5062 now done for all such subprocesses. The other cases are: ${run, transport
5063 filters, and the commands run by the lmtp and pipe transports.
5065 6. Added CONFIGURE_GROUP build-time option.
5067 7. Some older OS have a limit of 256 on the maximum number of file
5068 descriptors. Exim was using setrlimit() to set 1000 as a large value
5069 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5070 systems. I've change it so that if it can't get 1000, it tries for 256.
5072 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5073 was an oversight, and furthermore, ever since the addition of extra
5074 controls (e.g. 4.43/32), the checks on when to allow different forms of
5075 "control" were broken. There should now be diagnostics for all cases when a
5076 control that does not make sense is encountered.
5078 9. Added the /retain_sender option to "control=submission".
5080 10. $recipients is now available in the predata ACL (oversight).
5082 11. Tidy the search cache before the fork to do a delivery from a message
5083 received from the command line. Otherwise the child will trigger a lookup
5084 failure and thereby defer the delivery if it tries to use (for example) a
5085 cached ldap connection that the parent has called unbind on.
5087 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5088 of $address_data from the recipient verification was clobbered by the
5089 sender verification.
5091 13. The value of address_data from a sender verification is now available in
5092 $sender_address_data in subsequent conditions in the ACL statement.
5094 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5096 15. Added a new option "connect=<time>" to callout options, to set a different
5099 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5100 was its contents. (It was OK if the option was not defined at all.)
5102 17. A "Completed" log line is now written for messages that are removed from
5103 the spool by the -Mrm option.
5105 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5106 information about exactly what failed.
5108 19. Added -dd to debug only the daemon process.
5110 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5111 handles timeouts, both on the server side and network timeouts. Renamed the
5112 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5114 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5115 It is now set to "smtps".
5117 22. $host_address is now set to the target address during the checking of
5118 ignore_target_hosts.
5120 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5121 being passed; this would have caused $sender_host_name to have been used if
5122 matching the list had actually called for a host name (not very likely,
5123 since this list is usually IP addresses). A host name is now passed as
5126 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5127 code that specifies a non-restarting handler (typically sigaction() in
5128 modern systems) in an attempt to fix a rare and obscure crash bug.
5130 25. Narrowed the window for a race in the daemon that could cause it to ignore
5131 SIGCHLD signals. This is not a major problem, because they are used only to
5132 wake it up if nothing else does.
5134 26. A malformed maildirsize file could cause Exim to calculate negative values
5135 for the mailbox size or file count. Odd effects could occur as a result.
5136 The maildirsize information is now recalculated if the size or filecount
5139 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5140 support for a long time. Removed HAVE_SYS_VFS_H.
5142 28. Installed the latest version of exipick from John Jetmore.
5144 29. In an address list, if the pattern was not a regular expression, an empty
5145 subject address (from a bounce message) matched only if the pattern was an
5146 empty string. Non-empty patterns were not even tested. This was the wrong
5147 because it is perfectly reasonable to use an empty address as part of a
5148 database query. An empty address is now tested by patterns that are
5149 lookups. However, all the other forms of pattern expect the subject to
5150 contain a local part and a domain, and therefore, for them, an empty
5151 address still always fails if the pattern is not itself empty.
5153 30. Exim went into a mad DNS loop when attempting to do a callout where the
5154 host was specified on an smtp transport, and looking it up yielded more
5155 than one IP address.
5157 31. Re-factored the code for checking spool and log partition space into a
5158 function that finds that data and another that does the check. The former
5159 is then used to implement four new variables: $spool_space, $log_space,
5160 $spool_inodes, and $log_inodes.
5162 32. The RFC2047 encoding function was originally intended for short strings
5163 such as real names; it was not keeping to the 75-character limit for
5164 encoded words that the RFC imposes. It now respects the limit, and
5165 generates multiple encoded words if necessary. To be on the safe side, I
5166 have increased the buffer size for the ${rfc2047: expansion operator from
5169 33. It is now permitted to omit both strings after an "if" condition; if the
5170 condition is true, the result is "true". As before, when the second string
5171 is omitted, a false condition yields an empty string. This makes it less
5172 cumbersome to write custom ACL and router conditions.
5174 34. Failure to deliver a bounce message always caused it to be frozen, even if
5175 there was an errors_to setting on the router. The errors_to setting is now
5178 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5179 canonical form (fully expanded) before being placed in
5180 $sender_host_address.
5182 36. The table in the code that translates DNS record types into text (T_A to
5183 "A" for instance) was missing entries for NS and CNAME. It is just possible
5184 that this could have caused confusion if both these types were looked up
5185 for the same domain, because the text type is used as part of Exim's
5186 per-process caching. But the chance of anyone hitting this buglet seems
5189 37. The dnsdb lookup has been extended in a number of ways.
5191 (1) There is a new type, "zns", which walks up the domain tree until it
5192 finds some nameserver records. It should be used with care.
5194 (2) There is a new type, "mxh", which is like "mx" except that it returns
5195 just the host names, not the priorities.
5197 (3) It is now possible to give a list of domains (or IP addresses) to be
5198 looked up. The behaviour when one of the lookups defers can be
5199 controlled by a keyword.
5201 (4) It is now possible to specify the separator character for use when
5202 multiple records are returned.
5204 38. The dnslists ACL condition has been extended: it is now possible to supply
5205 a list of IP addresses and/or domains to be looked up in a particular DNS
5208 39. Added log_selector=+queue_time_overall.
5210 40. When running the queue in the test harness, wait just a tad after forking a
5211 delivery process, to get repeatability of debugging output.
5213 41. Include certificate and key file names in error message when GnuTLS fails
5214 to set them up, because the GnuTLS error message doesn't include the name
5215 of the failing file when there is a problem reading it.
5217 42. Allow both -bf and -bF in the same test run.
5219 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5221 44. The "Exiscan patch" is now merged into the mainline Exim source.
5223 45. Sometimes the final signoff response after QUIT could fail to get
5224 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5225 < 0 before doing a fflush(). This bug looks as though it goes back to the
5226 introduction of TLS in release 3.20, but "sometimes" must have been rare
5227 because the tests only now provoked it.
5229 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5230 (this can affect the format of dates).
5232 47. exim_tidydb, when checking for the continued existence of a message for
5233 which it has found a message-specific retry record, was not finding
5234 messages that were in split spool directories. Consequently, it was
5235 deleting retry records that should have stayed in existence.
5237 48. Steve fixed some bugs in eximstats.
5239 49. The SPA authentication driver was not abandoning authentication and moving
5240 on to the next authenticator when an expansion was forced to fail,
5241 contradicting the general specification for all authenticators. Instead it
5242 was generating a temporary error. It now behaves as specified.
5244 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5245 (the order specifies the preference for clients). The order is now AES256,
5246 AES128, 3DES, ARCFOUR128.
5248 51. Small patch to Sieve code - explicitly set From: when generating an
5251 52. Exim crashed if a remote delivery caused a very long error message to be
5252 recorded - for instance if somebody sent an entire SpamAssassin report back
5253 as a large number of 550 error lines. This bug was coincidentally fixed by
5254 increasing the size of one of Exim's internal buffers (big_buffer) that
5255 happened as part of the Exiscan merge. However, to be on the safe side, I
5256 have made the code more robust (and fixed the comments that describe what
5259 53. Now that there can be additional text after "Completed" in log lines (if
5260 the queue_time_overall log selector is set), a one-byte patch to exigrep
5261 was needed to allow it to recognize "Completed" as not the last thing in
5264 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5265 patch that reportedly fixes this has been added. I am not expert enough to
5266 create a test for it. This is what the patch creator wrote:
5268 "I found a little strange behaviour of ldap code when working with
5269 Windows 2003 AD Domain, where users was placed in more than one
5270 Organization Units. When I tried to give exim partial DN, the exit code
5271 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5272 But simultaneously result of request was absolutely normal ldap result,
5273 so I produce this patch..."
5275 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5276 REFERENCE, so I have modified the code to exclude the patch when that macro
5279 55. Some experimental protocols are using DNS PTR records for new purposes. The
5280 keys for these records are domain names, not reversed IP addresses. The
5281 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5282 leaves it alone. Component reversal etc. now happens only for IP addresses.
5285 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5287 57. Double the size of the debug message buffer (to 2048) so that more of very
5288 long debug lines gets shown.
5290 58. The exicyclog utility now does better if the number of log files to keep
5291 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5293 59. Two changes related to the smtp_active_hostname option:
5295 (1) $smtp_active_hostname is now available as a variable.
5296 (2) The default for smtp_banner uses $smtp_active_hostname instead
5297 of $primary_hostname.
5299 60. The host_aton() function is supposed to be passed a string that is known
5300 to be a valid IP address. However, in the case of IPv6 addresses, it was
5301 not checking this. This is a hostage to fortune. Exim now panics and dies
5302 if the condition is not met. A case was found where this could be provoked
5303 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5304 components; fortuitously, this particular loophole had already been fixed
5305 by change 4.50/55 above.
5307 If there are any other similar loopholes, the new check in host_aton()
5308 itself should stop them being exploited. The report I received stated that
5309 data on the command line could provoke the exploit when Exim was running as
5310 exim, but did not say which command line option was involved. All I could
5311 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5312 running as the user.
5315 61. There was a buffer overflow vulnerability in the SPA authentication code
5316 (which came originally from the Samba project). I have added a test to the
5317 spa_base64_to_bits() function which I hope fixes it.
5320 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5321 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5323 63. The daemon start-up calls getloadavg() while still root for those OS that
5324 need the first call to be done as root, but it missed one case: when
5325 deliver_queue_load_max is set with deliver_drop_privilege. This is
5326 necessary for the benefit of the queue runner, because there is no re-exec
5327 when deliver_drop_privilege is set.
5329 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5330 This has been fixed.
5332 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5333 was in use, was not putting the data itself into the right store pool;
5334 consequently, it could be overwritten for a subsequent message in the same
5335 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5338 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5340 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5341 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5342 the function were treating the return as a boolean value, which happened to
5343 work because 0=false and not-0=true, but is not correct code.
5345 68. The host_aton() function was not handling scoped IPv6 addresses (those
5346 with, for example, "%eth0" on the end) correctly.
5348 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5349 negated items (that is, ~something) in unsigned ints. Some compilers
5350 apparently mutter when there is no cast.
5352 70. If an address verification called from an ACL failed, and did not produce a
5353 user-specific message (i.e. there was only a "system" message), nothing was
5354 put in $acl_verify_message. In this situation, it now puts the system
5357 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5358 session; change 4.31/43 added the unwanted input to the log line - except
5359 that it did not do this in the start of session case. It now does.
5361 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5362 This is wrong and can cause the other end to generate a synchronization
5363 error if it is another Exim or anything else that does the synchronization
5364 check. A QUIT command is no longer sent after a timeout.
5366 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5367 during host lookups.
5369 74. The defer_ok option of callout verification was not working if it was used
5370 when verifying addresses in header lines, that is, for this case:
5372 verify = header_sender/callout=defer_ok
5374 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5375 those file descriptors could be used for SMTP connections. If anything
5376 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5377 could be sent to the SMTP client, causing chaos. The daemon now opens
5378 stdin, stdout, and stderr to /dev/null when it puts itself into the
5381 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5382 log by default. The user can override this with suitable Perl magic.
5384 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5385 the log message when discard triggers, was not working for the DATA ACL or
5386 for the non-SMTP ACL.
5388 78. Error message wording change in sieve.c.
5390 79. If smtp_accept_max_per_host was set, the number of connections could be
5391 restricted to fewer than expected, because the daemon was trying to set up
5392 a new connection before checking whether the processes handling previous
5393 connections had finished. The check for completed processes is now done
5394 earlier. On busy systems, this bug wouldn't be noticed because something
5395 else would have woken the daemon, and it would have reaped the completed
5398 80. If a message was submitted locally by a user whose login name contained one
5399 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5400 It caused a spool format error. I have fixed the spool reading code. A
5401 related problem was that the "from" clause in the Received: line became
5402 illegal because of the space(s). It is now covered by ${quote_local_part.
5404 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5407 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5409 83. Patch from Sieve maintainer to fix unterminated string problem in
5410 "vacation" handling.
5412 84. Some minor changes to the Linux configuration files to help with other
5413 OS variants using glibc.
5415 85. One more patch for Sieve to update vacation handling to latest spec.
5418 ----------------------------------------------------
5419 See the note above about the 4.44 and 4.50 releases.
5420 ----------------------------------------------------
5426 1. Change 4.43/35 introduced a bug that caused file counts to be
5427 incorrectly computed when quota_filecount was set in an appendfile
5430 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5431 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5434 3. Give more explanation in the error message when the command for a transport
5435 filter fails to execute.
5437 4. There are several places where Exim runs a non-Exim command in a
5438 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5439 was being done only for the command run by the queryprogram router. It is
5440 now done for all such subprocesses. The other cases are: ${run, transport
5441 filters, and the commands run by the lmtp and pipe transports.
5443 5. Some older OS have a limit of 256 on the maximum number of file
5444 descriptors. Exim was using setrlimit() to set 1000 as a large value
5445 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5446 systems. I've change it so that if it can't get 1000, it tries for 256.
5448 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5449 was an oversight, and furthermore, ever since the addition of extra
5450 controls (e.g. 4.43/32), the checks on when to allow different forms of
5451 "control" were broken. There should now be diagnostics for all cases when a
5452 control that does not make sense is encountered.
5454 7. $recipients is now available in the predata ACL (oversight).
5456 8. Tidy the search cache before the fork to do a delivery from a message
5457 received from the command line. Otherwise the child will trigger a lookup
5458 failure and thereby defer the delivery if it tries to use (for example) a
5459 cached ldap connection that the parent has called unbind on.
5461 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5462 of $address_data from the recipient verification was clobbered by the
5463 sender verification.
5465 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5466 was its contents. (It was OK if the option was not defined at all.)
5468 11. A "Completed" log line is now written for messages that are removed from
5469 the spool by the -Mrm option.
5471 12. $host_address is now set to the target address during the checking of
5472 ignore_target_hosts.
5474 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5475 being passed; this would have caused $sender_host_name to have been used if
5476 matching the list had actually called for a host name (not very likely,
5477 since this list is usually IP addresses). A host name is now passed as
5480 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5481 code that specifies a non-restarting handler (typically sigaction() in
5482 modern systems) in an attempt to fix a rare and obscure crash bug.
5484 15. Narrowed the window for a race in the daemon that could cause it to ignore
5485 SIGCHLD signals. This is not a major problem, because they are used only to
5486 wake it up if nothing else does.
5488 16. A malformed maildirsize file could cause Exim to calculate negative values
5489 for the mailbox size or file count. Odd effects could occur as a result.
5490 The maildirsize information is now recalculated if the size or filecount
5493 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5494 support for a long time. Removed HAVE_SYS_VFS_H.
5496 18. Updated exipick to current release from John Jetmore.
5498 19. Allow an empty sender to be matched against a lookup in an address list.
5499 Previously the only cases considered were a regular expression, or an
5502 20. Exim went into a mad DNS lookup loop when doing a callout where the
5503 host was specified on the transport, if the DNS lookup yielded more than
5506 21. The RFC2047 encoding function was originally intended for short strings
5507 such as real names; it was not keeping to the 75-character limit for
5508 encoded words that the RFC imposes. It now respects the limit, and
5509 generates multiple encoded words if necessary. To be on the safe side, I
5510 have increased the buffer size for the ${rfc2047: expansion operator from
5513 22. Failure to deliver a bounce message always caused it to be frozen, even if
5514 there was an errors_to setting on the router. The errors_to setting is now
5517 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5518 canonical form (fully expanded) before being placed in
5519 $sender_host_address.
5521 24. Updated eximstats to version 1.33
5523 25. Include certificate and key file names in error message when GnuTLS fails
5524 to set them up, because the GnuTLS error message doesn't include the name
5525 of the failing file when there is a problem reading it.
5527 26. Expand error message when OpenSSL has problems setting up cert/key files.
5530 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5531 (this can affect the format of dates).
5533 28. exim_tidydb, when checking for the continued existence of a message for
5534 which it has found a message-specific retry record, was not finding
5535 messages that were in split spool directories. Consequently, it was
5536 deleting retry records that should have stayed in existence.
5538 29. eximstats updated to version 1.35
5539 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5540 1.35 - bugfix such that pie charts by volume are generated correctly
5542 30. The SPA authentication driver was not abandoning authentication and moving
5543 on to the next authenticator when an expansion was forced to fail,
5544 contradicting the general specification for all authenticators. Instead it
5545 was generating a temporary error. It now behaves as specified.
5547 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5548 (the order specifies the preference for clients). The order is now AES256,
5549 AES128, 3DES, ARCFOUR128.
5551 31. Small patch to Sieve code - explicitly set From: when generating an
5554 32. Exim crashed if a remote delivery caused a very long error message to be
5555 recorded - for instance if somebody sent an entire SpamAssassin report back
5556 as a large number of 550 error lines. This bug was coincidentally fixed by
5557 increasing the size of one of Exim's internal buffers (big_buffer) that
5558 happened as part of the Exiscan merge. However, to be on the safe side, I
5559 have made the code more robust (and fixed the comments that describe what
5562 33. Some experimental protocols are using DNS PTR records for new purposes. The
5563 keys for these records are domain names, not reversed IP addresses. The
5564 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5565 leaves it alone. Component reversal etc. now happens only for IP addresses.
5568 34. The host_aton() function is supposed to be passed a string that is known
5569 to be a valid IP address. However, in the case of IPv6 addresses, it was
5570 not checking this. This is a hostage to fortune. Exim now panics and dies
5571 if the condition is not met. A case was found where this could be provoked
5572 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5573 components; fortuitously, this particular loophole had already been fixed
5574 by change 4.50/55 or 4.44/33 above.
5576 If there are any other similar loopholes, the new check in host_aton()
5577 itself should stop them being exploited. The report I received stated that
5578 data on the command line could provoke the exploit when Exim was running as
5579 exim, but did not say which command line option was involved. All I could
5580 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5581 running as the user.
5584 35. There was a buffer overflow vulnerability in the SPA authentication code
5585 (which came originally from the Samba project). I have added a test to the
5586 spa_base64_to_bits() function which I hope fixes it.
5589 36. The daemon start-up calls getloadavg() while still root for those OS that
5590 need the first call to be done as root, but it missed one case: when
5591 deliver_queue_load_max is set with deliver_drop_privilege. This is
5592 necessary for the benefit of the queue runner, because there is no re-exec
5593 when deliver_drop_privilege is set.
5595 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5596 was in use, was not putting the data itself into the right store pool;
5597 consequently, it could be overwritten for a subsequent message in the same
5598 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5601 38. Sometimes the final signoff response after QUIT could fail to get
5602 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5603 < 0 before doing a fflush(). This bug looks as though it goes back to the
5604 introduction of TLS in release 3.20, but "sometimes" must have been rare
5605 because the tests only now provoked it.
5611 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5612 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5613 Now it returns an integer. A number of calls were still expecting a T/F
5614 return. Fortuitously, in all cases, the tests worked in OK situations,
5615 which is the norm. However, things would have gone wrong on any write
5616 failures on the smtp file descriptor. This function is used when sending
5617 messages over SMTP and also when doing verify callouts.
5619 2. When Exim is called to do synchronous delivery of a locally submitted
5620 message (the -odf or -odi options), it no longer closes stderr before doing
5623 3. Implemented the mua_wrapper option.
5625 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5627 5. Implemented the functions header_remove(), header_testname(),
5628 header_add_at_position(), and receive_remove_recipient(), and exported them
5631 6. If an ACL "warn" statement specified the addition of headers, Exim already
5632 inserted X-ACL-Warn: at the start if there was no header name. However, it
5633 was not making this test for the second and subsequent header lines if
5634 there were newlines in the string. This meant that an invalid header could
5635 be inserted if Exim was badly configured.
5637 7. Allow an ACL "warn" statement to add header lines at the start or after all
5638 the Received: headers, as well as at the end.
5640 8. Added the rcpt_4xx retry error code.
5642 9. Added postmaster_mailfrom=xxx to callout verification option.
5644 10. Added mailfrom=xxxx to the callout verification option, for verify=
5647 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5648 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5650 12. Inserted some casts to stop certain compilers warning when using pointer
5651 differences as field lengths or precisions in printf-type calls (mostly
5652 affecting debugging statements).
5654 13. Added optional readline() support for -be (dynamically loaded).
5656 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5657 same clock tick as a message's arrival, so that its received time was the
5658 same as the "first fail" time on the retry record, and that message
5659 remained on the queue past the ultimate address timeout, every queue runner
5660 would try a delivery (because it was past the ultimate address timeout) but
5661 after another failure, the ultimate address timeout, which should have then
5662 bounced the address, did not kick in. This was a "< instead of <=" error;
5663 in most cases the first failure would have been in the next clock tick
5664 after the received time, and all would be well.
5666 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5667 being recognized when the domain list was tested by the match_domain
5668 condition in an expansion string.
5670 16. Added the ${str2b64: operator.
5672 17. Exim was always calling setrlimit() to set a large limit for the number of
5673 processes, without checking whether the existing limit was already
5674 adequate. (It did check for the limit on file descriptors.) Furthermore,
5675 errors from getrlimit() and setrlimit() were being ignored. Now they are
5676 logged to the main and panic logs, but Exim does carry on, to try to do its
5677 job under whatever limits there are.
5679 18. Imported PCRE 5.0.
5681 19. Trivial typo in log message " temporarily refused connection" (the leading
5684 20. If the log selector return_path_on_delivery was set and an address was
5685 redirected to /dev/null, the delivery process crashed because it assumed
5686 that a return path would always be set for a "successful" delivery. In this
5687 case, the whole delivery is bypassed as an optimization, and therefore no
5690 21. Internal re-arrangement: the function for sending a challenge and reading
5691 a response while authentication was assuming a zero-terminated challenge
5692 string. It's now changed to take a pointer and a length, to allow for
5693 binary data in such strings.
5695 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5697 23. Exim was not respecting finduser_retries when seeking the login of the
5698 uid under which it was called; it was always trying 10 times. (The default
5699 setting of finduser_retries is zero.) Also, it was sleeping after the final
5700 failure, which is pointless.
5702 24. Implemented tls_on_connect_ports.
5704 25. Implemented acl_smtp_predata.
5706 26. If the domain in control=submission is set empty, Exim assumes that the
5707 authenticated id is a complete email address when it generates From: or
5708 Sender: header lines.
5710 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5711 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5712 chown and chgrp in /bin and hostname in /usr/bin.
5714 28. Exim was keeping the "process log" file open after each use, just as it
5715 does for the main log. This opens the possibility of it remaining open for
5716 long periods when the USR1 signal hits a daemon. Occasional processlog
5717 errors were reported, that could have been caused by this. Anyway, it seems
5718 much more sensible not to leave this file open at all, so that is what now
5721 29. The long-running daemon process does not normally write to the log once it
5722 has entered its main loop, and it closes the log before doing so. This is
5723 so that log files can straightforwardly be renamed and moved. However,
5724 there are a couple of unusual error situations where the daemon does write
5725 log entries, and I had neglected to close the log afterwards.
5727 30. The text of an SMTP error response that was received during a remote
5728 delivery was being truncated at 512 bytes. This is too short for some of
5729 the long messages that one sometimes sees. I've increased the limit to
5732 31. It is now possible to make retry rules that apply only when a message has a
5733 specific sender, in particular, an empty sender.
5735 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5736 it possible to be selective about when SMTP synchronization is enforced.
5738 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5740 32. Implemented hosts_connection_nolog.
5742 33. Added an ACL for QUIT.
5744 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5747 35. Added mailbox_size and mailbox_filecount to appendfile.
5749 36. Added control = no_multiline_responses to ACLs.
5751 37. There was a bug in the logic of the code that waits for the clock to tick
5752 in the case where the clock went backwards by a substantial amount such
5753 that the microsecond fraction of "now" was more than the microsecond
5754 fraction of "then" (but the whole seconds number was less).
5756 38. Added support for the libradius Radius client library this is found on
5757 FreeBSD (previously only the radiusclient library was supported).
5763 1. When certain lookups returned multiple values in the form name=value, the
5764 quoting of the values was not always being done properly. Specifically:
5765 (a) If the value started with a double quote, but contained no whitespace,
5767 (b) If the value contained whitespace other than a space character (i.e.
5768 tabs or newlines or carriage returns) it was not quoted.
5769 This fix has been applied to the mysql and pgsql lookups by writing a
5770 separate quoting function and calling it from the lookup code. The fix
5771 should probably also be applied to nisplus, ibase and oracle lookups, but
5772 since I cannot test any of those, I have not disturbed their existing code.
5774 2. A hit in the callout cache for a specific address caused a log line with no
5775 reason for rejecting RCPT. Now it says "Previous (cached) callout
5776 verification failure".
5778 3. There was an off-by-one bug in the queryprogram router. An over-long
5779 return line was truncated at 256 instead of 255 characters, thereby
5780 overflowing its buffer with the terminating zero. As well as fixing this, I
5781 have increased the buffer size to 1024 (and made a note to document this).
5783 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5784 when Exim is waiting for an SMTP response from a remote server, Exim
5785 restarts its select() call on the socket, thereby resetting its timeout.
5786 This is not a problem when such interrupts are rare. Somebody set up a cron
5787 job to run exiwhat every 2 minutes, which is less than the normal select()
5788 timeout (5 or 10 minutes). This meant that the select() timeout never
5789 kicked in because it was always reset. I have fixed this by comparing the
5790 time when an interrupt arrives with the time at the start of the first call
5791 to select(). If more time than the timeout has elapsed, the interrupt is
5792 treated as a timeout.
5794 5. Some internal re-factoring in preparation for the addition of Sieve
5795 extensions (by MH). In particular, the "personal" test is moved to a
5796 separate function, and given an option for scanning Cc: and Bcc: (which is
5797 not set for Exim filters).
5799 6. When Exim created an email address using the login of the caller as the
5800 local part (e.g. when creating a From: or Sender: header line), it was not
5801 quoting the local part when it contained special characters such as @.
5803 7. Installed new OpenBSD configuration files.
5805 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5806 try to make them clearer.
5808 9. Callout options, other than the timeout value, were being ignored when
5809 verifying sender addresses in header lines. For example, when using
5811 verify = header_sender/callout=no_cache
5813 the cache was (incorrectly) being used.
5815 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5816 only the Cygwin environment.
5818 11. When return_path_on_delivery was set as a log selector, if different remote
5819 addresses in the same message used different return paths and parallel
5820 remote delivery occurred, the wrong values would sometimes be logged.
5821 (Whenever a remote delivery process finished, the return path value from
5822 the most recently started remote delivery process was logged.)
5824 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5825 lines when AUTH and/or TLS are in use. This is the "received protocol"
5826 field. Exim used to use "asmtp" for authenticated SMTP, without any
5827 indication (in the protocol name) for TLS use. Now it follows the RFC and
5828 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5829 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5830 names appear in log lines as well as in Received: header lines.
5832 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5833 extensions, and comparison tests, and to fix some bugs.
5835 14. Changes to the "personal" filter test:
5837 (1) The test was buggy in that it was just doing the equivalent of
5838 "contains" tests on header lines. For example, if a user's address was
5839 anne@some.where, the "personal" test would incorrectly be true for
5841 To: susanne@some.where
5843 This test is now done by extracting each address from the header in turn,
5844 and checking the entire address. Other tests that are part of "personal"
5845 are now done using regular expressions (for example, to check local parts
5846 of addresses in From: header lines).
5848 (2) The list of non-personal local parts in From: addresses has been
5849 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5850 taken from the Sieve specification recommendations.
5852 (3) If the message contains any header line starting with "List-" it is
5853 treated as non-personal.
5855 (4) The test for "circular" in the Subject: header line has been removed
5856 because it now seems ill-conceived.
5858 15. Minor typos in src/EDITME comments corrected.
5860 16. Installed latest exipick from John Jetmore.
5862 17. If headers_add on a router specified a text string that was too long for
5863 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5864 of string_sprintf() is now avoided.
5866 18. $message_body_size was not set (it was always zero) when running the DATA
5867 ACL and the local_scan() function.
5869 19. For the "mail" command in an Exim filter, no default was being set for
5870 the once_repeat time, causing a random time value to be used if "once" was
5871 specified. (If the value happened to be <= 0, no repeat happened.) The
5872 default is now 0s, meaning "never repeat". The "vacation" command was OK
5873 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5874 (I found it when inspecting the code).
5876 20. There is now an overall timeout for performing a callout verification. It
5877 defaults to 4 times the callout timeout, which applies to individual SMTP
5878 commands during the callout. The overall timeout applies when there is more
5879 than one host that can be tried. The timeout is checked before trying the
5880 next host. This prevents very long delays if there are a large number of
5881 hosts and all are timing out (e.g. when the network connections are timing
5882 out). The value of the overall timeout can be changed by specifying an
5883 additional sub-option for "callout", called "maxwait". For example:
5885 verify = sender/callout=5s,maxwait=20s
5887 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5888 to the end before writing, but this should make it even safer).
5890 22. Exim was forgetting that it had advertised PIPELINING for the second and
5891 subsequent messages on an SMTP connection. It was also not resetting its
5892 memory on STARTTLS and an internal HELO.
5894 23. When Exim logs an SMTP synchronization error within a session, it now
5895 records whether PIPELINING has been advertised or not.
5897 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5898 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5899 rather than long int.
5901 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5903 26. Added the never_mail option to autoreply.
5909 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5910 crash if the getsockname() call failed; this can happen if a connection is
5911 closed very soon after it is established. The problem was simply in the
5912 order in which certain operations were done, causing Exim to try to write
5913 to the SMTP stream before it had set up the file descriptor. The bug has
5914 been fixed by making things happen in the correct order.
5920 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5921 before the connection was closed, thus losing the rejection response.
5923 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5924 some early Solaris releases, but causes trouble in current releases where
5925 socklen_t is defined.
5927 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5930 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5933 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5934 that releasing the top of it at the end releases what was used for sub-
5935 expansions (unless the block got too big). However, discard this block if
5936 the first thing is a variable or header, so that we can use its block when
5937 it is dynamic (useful for very large $message_headers, for example).
5939 6. Lookups now cache *every* query, not just the most recent. A new, separate
5940 store pool is used for this. It can be recovered when all lookup caches are
5941 flushed. Lookups now release memory at the end of their result strings.
5942 This has involved some general refactoring of the lookup sources.
5944 7. Some code has been added to the store_xxx() functions to reduce the amount
5945 of flapping under certain conditions.
5947 8. log_incoming_interface used to affect only the <= reception log lines. Now
5948 it causes the local interface and port to be added to several more SMTP log
5949 lines, for example "SMTP connection from", and rejection lines.
5951 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5953 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5955 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5956 could be overwritten at the end of the current message (or the start of a
5957 new message if it was set in a HELO ACL). The value is now preserved for
5958 the duration of the SMTP connection.
5960 12. If a transport had a headers_rewrite setting, and a matching header line
5961 contained an unqualified address, that address was qualified, even if it
5962 did not match any rewriting rules. The underlying bug was that the values
5963 of the flags that permit the existence of unqualified sender and recipient
5964 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5965 non-local messages, and by -bnq for local messages) were not being
5966 preserved with the message after it was received.
5968 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5969 "next input=" as part of the text comprising the host identity instead of
5970 the correct text. The code was using the same buffer for two different
5971 strings. However, depending on which order the printing function evaluated
5972 its arguments, the bug did not always show up. Under Linux, for example, my
5973 test suite worked just fine.
5975 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5976 doesn't work with some older versions of Perl. It has been changed to "my",
5977 which in any case is probably the better facility to use.
5979 15. A really picky compiler found some instances of statements for creating
5980 error messages that either had too many or two few arguments for the format
5983 16. The size of the buffer for calls to the DNS resolver has been increased
5984 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5985 for addresses that have a lot of PTR records. This alleviates a problem; it
5986 does not fully solve it.
5988 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5989 buffer now truncates the list and logs the incident, which is the same
5990 action as happens when Exim is looking up a host name and its aliases.
5991 Previously in this situation something unpredictable would happen;
5992 sometimes it was "internal error: store_reset failed".
5994 18. If a server dropped the connection unexpectedly when an Exim client was
5995 using GnuTLS and trying to read a response, the client delivery process
5996 crashed while trying to generate an error log message.
5998 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5999 string, for example:
6001 warn message = H1: something\nH2: something
6003 the text was added as a single header line from Exim's point of view
6004 though it ended up OK in the delivered message. However, searching for the
6005 second and subsequent header lines using $h_h2: did not work. This has been
6006 fixed. Similarly, if a system filter added multiple headers in this way,
6007 the routers could not see them.
6009 20. Expanded the error message when iplsearch is called with an invalid key to
6010 suggest using net-iplsearch in a host list.
6012 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6013 ACLs are no longer actually imposed (and a message to that effect is
6016 22. If a "gecos" field in a passwd entry contained escaped characters, in
6017 particular, if it contained a \" sequence, Exim got it wrong when building
6018 a From: or a Sender: header from that name. A second bug also caused
6019 incorrect handling when an unquoted " was present following a character
6020 that needed quoting.
6022 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6023 was not being matched caselessly.
6025 24. Arranged for all hyphens in the exim.8 source to be escaped with
6028 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6029 properly. Recipient callout cache records were still being keyed to include
6030 the sender, even when use_sender was set false. This led to far more
6031 callouts that were necessary. The sender is no longer included in the key
6032 when use_sender is false.
6034 26. Added "control = submission" modifier to ACLs.
6036 27. Added the ${base62d: operator to decode base 62 numbers.
6038 28. dnsdb lookups can now access SRV records.
6040 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6041 the configuration file.
6043 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6044 (-v) mode. This makes the output for a verbose queue run more intelligible.
6046 31. Added a use_postmaster feature to recipient callouts.
6048 32. Added the $body_zerocount variable, containing the number of binary zero
6049 bytes in the message body.
6051 33. The time of last modification of the "new" subdirectory is now used as the
6052 "mailbox time last read" when there is a quota error for a maildir
6055 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6057 35. Added +ignore_unknown as a special item in host lists.
6059 36. Code for decoding IPv6 addresses in host lists is now included, even if
6060 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6061 address was recognized as an IP address, but was then not correctly decoded
6062 into binary, causing unexpected and incorrect effects when compared with
6069 1. Very minor rewording of debugging text in manualroute to say "list of
6070 hosts" instead of "hostlist".
6072 2. If verify=header_syntax was set, and a header line with an unqualified
6073 address (no domain) and a large number of spaces between the end of the
6074 name and the colon was received, the reception process suffered a buffer
6075 overflow, and (when I tested it) crashed. This was caused by some obsolete
6076 code that should have been removed. The fix is to remove it!
6078 3. When running in the test harness, delay a bit after writing a bounce
6079 message to get a bit more predictability in the log output.
6081 4. Added a call to search_tidyup() just before forking a reception process. In
6082 theory, someone could use a lookup in the expansion of smtp_accept_max_
6083 per_host which, without the tidyup, could leave open a database connection.
6085 5. Added the variables $recipient_data and $sender_data which get set from a
6086 lookup success in an ACL "recipients" or "senders" condition, or a router
6087 "senders" option, similar to $domain_data and $local_part_data.
6089 6. Moved the writing of debug_print from before to after the "senders" test
6092 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6093 problems for message scanning, either using a data ACL, or using
6094 local_scan() because the Received: header was not generated till after they
6095 were called (in order to set the time as the time of reception completion).
6096 I have revised the way this works. The header is now generated after the
6097 body is received, but before the ACL or local_scan() are called. After they
6098 are run, the timestamp in the header is updated.
6104 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6105 before starting a queue runner without re-exec. This happened only when
6106 deliver_drop_privilege was set or when the Exim user was set to root. The
6107 effect of the bug was that timeouts during subsequent deliveries caused
6108 crashes instead of being properly handled. The handler is now left at its
6109 default (and expected) setting.
6111 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6112 message, again when deliver_drop_privilege is set or Exim is run as root.
6113 The bug described in (1) was not present in this case, but the tidying up
6114 of the other signals was missing. I have made the two cases consistent.
6116 3. The ignore_target_hosts setting on a manualroute router was being ignored
6117 for hosts that were looked up using the /MX notation.
6119 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6122 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6123 operated on the sender address. After changing the $sender_address to <>
6124 for the sender address verify, Exim was re-instated it as the original
6125 (before rewriting) address, but remembering that it had rewritten it, so it
6126 wasn't rewriting it again. This bug also had the effect of breaking the
6127 sender address verification caching when the sender address was rewritten.
6129 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6130 This has been changed so that if the ip literal address matches
6131 ignore_target_hosts, the router declines.
6133 7. Added expansion conditions match_domain, match_address, and match_local_
6134 part (NOT match_host).
6136 8. The placeholder for the Received: header didn't have a length field set.
6138 9. Added code to Exim itself and to exim_lock to test for a specific race
6139 condition that could lead to file corruption when using MBX delivery. The
6140 issue is with the lockfile that is created in /tmp. If this file is removed
6141 after a process has opened it but before that process has acquired a lock,
6142 there is the potential for a second process to recreate the file and also
6143 acquire a lock. This could lead to two Exim processes writing to the file
6144 at the same time. The added code performs the same test as UW imapd; it
6145 checks after acquiring the lock that its file descriptor still refers to
6146 the same named file.
6148 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6149 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6152 11. Added the smtp_active_hostname option. If used, this will typically be made
6153 to depend on the incoming interface address. Because $interface_address is
6154 not set up until the daemon has forked a reception process, error responses
6155 that can happen earlier (such as "too many connections") no longer contain
6158 12. If an expansion in a condition on a "warn" statement fails because a lookup
6159 defers, the "warn" statement is abandoned, and the next ACL statement is
6160 processed. Previously this caused the whole ACL to be aborted.
6162 13. Added the iplsearch lookup type.
6164 14. Added ident_timeout as a log selector.
6166 15. Added tls_certificate_verified as a log selector.
6168 16. Added a global option tls_require_ciphers (compare the smtp transport
6169 option of the same name). This controls incoming TLS connections.
6171 17. I finally figured out how to make tls_require_ciphers do a similar thing
6172 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6173 before starting the TLS session.
6175 18. Tabs are now shown as \t in -bP output.
6177 19. If the log selector return_path_on_delivery was set, Exim crashed when
6178 bouncing a message because it had too many Received: header lines.
6180 20. If two routers both had headers_remove settings, and the first one included
6181 a superfluous trailing colon, the final name in the first list and the
6182 first name in the second list were incorrectly joined into one item (with a
6183 colon in the middle).
6189 1. Added -C and -D options to the exinext utility, mainly to make it easier
6190 to include in the automated testing, but these could be helpful when
6191 multiple configurations are in use.
6193 2. The exinext utility was not formatting the output nicely when there was
6194 an alternate port involved in the retry record key, nor when there was a
6195 message id as well (for retries that were specific to a specific message
6196 and a specific host). It was also confused by IPv6 addresses, because of
6197 the additional colons they contain. I have fixed the IPv4 problem, and
6198 patched it up to do a reasonable job for IPv6.
6200 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6201 delivery, the log line now contains "pipelined" if PIPELINING was used.
6203 4. An SMTP transport process used to panic and die if the bind() call to set
6204 an explicit outgoing interface failed. This has been changed; it is now
6205 treated in the same way as a connect() failure.
6207 5. A reference to $sender_host_name in the part of a conditional expansion
6208 that was being skipped was still causing a DNS lookup. This no longer
6211 6. The def: expansion condition was not recognizing references to header lines
6212 that used bh_ and bheader_.
6214 7. Added the _cache feature to named lists.
6216 8. The code for checking quota_filecount in the appendfile transport was
6217 allowing one more file than it should have been.
6219 9. For compatibility with Sendmail, the command line option
6227 and sets the incoming protocol and host name (for trusted callers). The
6228 host name and its colon can be omitted when only the protocol is to be set.
6229 Note the Exim already has two private options, -pd and -ps, that refer to
6230 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6231 "s", but I don't think that's a major issue.
6233 10. A number of refactoring changes to the code, none of which should affect
6236 (a) The number of logging options was getting close to filling up the
6237 32-bit word that was used as a bit map. I have split them into two classes:
6238 those that are passed in the argument to log_write(), and those that are
6239 only ever tested independently outside of that function. These are now in
6240 separate 32-bit words, so there is plenty of room for expansion again.
6241 There is no change in the user interface or the logging behaviour.
6243 (b) When building, for example, log lines, the code previously used a
6244 macro that called string_cat() twice, in order to add two strings. This is
6245 not really sufficiently general. Furthermore, there was one instance where
6246 it was actually wrong because one of the argument was used twice, and in
6247 one call a function was used. (As it happened, calling the function twice
6248 did not affect the overall behaviour.) The macro has been replaced by a
6249 function that can join an arbitrary number of extra strings onto a growing
6252 (c) The code for expansion conditions now uses a table and a binary chop
6253 instead of a serial search (which was left over from when there were very
6254 few conditions). Also, it now recognizes conditions like "pam" even when
6255 the relevant support is not compiled in: a suitably worded error message is
6256 given if an attempt is made to use such a condition.
6258 11. Added ${time_interval:xxxxx}.
6260 12. A bug was causing one of the ddress fields not to be passed back correctly
6261 from remote delivery subprocesses. The field in question was not being
6262 subsequently used, so this caused to problems in practice.
6264 13. Added new log selectors queue_time and deliver_time.
6266 14. Might have fixed a bug in maildirsizefile handling that threw up
6267 "unexpected character" debug warnings, and recalculated the data
6268 unnecessarily. In any case, I expanded the warning message to give more
6271 15. Added the message "Restricted characters in address" to the statements in
6272 the default ACL that block characters like @ and % in local parts.
6274 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6275 Three changes have been made:
6277 (a) There was a serious bug; a negative response to MAIL caused the whole
6278 recipient domain to be cached as invalid, thereby blocking all messages
6279 to all local parts at the same domain, from all senders. This bug has
6280 been fixed. The domain is no longer cached after a negative response to
6281 MAIL if the sender used is not empty.
6283 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6286 (c) A new callout option, "use_sender" has been added for people who want
6287 the modified behaviour.
6293 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6296 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6297 indeed breaks things for older releases.
6299 3. Added additional logging to the case where there is a problem reading data
6300 from a filter that is running in a subprocess using a pipe, in order to
6301 try to track down a specific problem.
6303 4. Testing facility fudge: when running in the test harness and attempting
6304 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6305 getting "No route to host". Convert this to a timeout.
6307 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6310 6. Some OS don't have socklen_t but use size_t instead. This affects the
6311 fifth argument of getsockopt() amongst other things. This is now
6312 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6313 can be set for individual OS. I have set it for SunOS5, OSF1, and
6314 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6315 some earlier ones do not.
6317 7. Change 4.30/15 was not doing the test caselessly.
6319 8. The standard form for an IPv6 address literal was being rejected by address
6320 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6321 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6322 this, as well as the form without the "IPv6" on the front (but only when
6323 address literals are enabled, of course).
6325 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6327 10. Exim crashed if a message with an empty sender address specified by -f
6328 encountered a router with an errors_to setting. This could be provoked only
6329 by a command such as
6333 where an empty string was supplied; "<>" did not hit this bug.
6335 11. Installed PCRE release 4.5.
6337 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6338 remained set. It is now erased.
6340 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6341 times from message ids (which are base 36 rather than the normal 62).
6343 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6344 were being counted as actual protocol errors, and logged if the log
6345 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6346 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6347 and DATA following a set of rejected RCPTs do not count as protocol errors.
6348 In other words, Exim assumes they were pipelined, though this may not
6349 actually be the case. Of course, in all cases the client gets an
6350 appropriate error code.
6352 15. If a lookup fails in an ACL condition, a message about the failure may
6353 be available; it is used if testing the ACL cannot continue, because most
6354 such messages specify what the cause of the deferral is. However, some
6355 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6356 that caused an old message to be retained and used if a later statement
6357 caused a defer, replacing the real cause of the deferral.
6359 16. If an IP address had so many PTR records that the DNS lookup buffer
6360 was not large enough to hold them, Exim could crash while trying to process
6361 the truncated data. It now detects and logs this case.
6363 17. Further to 4.21/58, another change has been made: if (and only if) the
6364 first line of a message (the first header line) ends with CRLF, a bare LF
6365 in a subsequent header line has a space inserted after it, so as not to
6366 terminate the header.
6368 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6369 unnecessarily, used atoi() instead of strtol(), and didn't check the
6370 termination when getting file sizes from file names by regex.
6372 19. Completely re-implemented the support for maildirsize files, in the light
6373 of a number of problems with the previous contributed implementation
6374 (4.30/29). In particular:
6376 . If the quota is zero, the maildirsize file is maintained, but no quota is
6379 . If the maildir directory does not exist, it is created before any attempt
6380 to write a maildirsize file.
6382 . The quota value in the file is just a cache; if the quota is changed in
6383 the transport, the new value overrides.
6385 . A regular expression is available for excluding directories from the
6388 20. The autoreply transport checks the characters in options that define the
6389 message's headers; it allows continued headers, but it was checking with
6390 isspace() after an embedded newline instead of explicitly looking for a
6393 21. If all the "regular" hosts to which an address was routed had passed their
6394 expiry times, and had not reached their retry times, the address was
6395 bounced, even if fallback hosts were defined. Now Exim should go on to try
6398 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6399 equivalent code in the SMTP transport. Some hosts send humungous responses
6400 to HELO/EHLO, more than 1024 it seems.
6402 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6403 gives compiler warnings in some environments. I've now done it "properly",
6406 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6407 (because of problems with the built-in one) was declared to return uschar *
6408 instead of char *, causing compiler failure.
6410 25. Fixed a file descriptor leak when processing alias/forward files.
6412 26. Fixed a minor format string issue in dbfn.c.
6414 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6416 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6417 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6420 29. When a sender address is verified, it is cached, to save repeating the test
6421 when there is more than one recipient in a message. However, when the
6422 verification involves a callout, it is possible for different callout
6423 options to be set for different recipients. It is too complicated to keep
6424 track of this in the cache, so now Exim always runs a verification when a
6425 callout is required, relying on the callout cache for the optimization.
6426 The overhead is duplication of the address routing, but this should not be
6429 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6430 to be verified with callout=postmaster, and the main callout worked but the
6431 postmaster check failed, the verification correctly failed. However, if a
6432 subsequent RCPT command asked for sender verification *without* the
6433 postmaster check, incorrect caching caused this verification also to fail,
6436 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6437 it was not caching the DNS options (qualify_single, search_parents) that
6438 were used when the lookup failed. A subsequent lookup with different
6439 options therefore always gave the same answer, though there were cases
6440 where it should not have. (Example: a "domains = !$mx_any" option on a
6441 dnslookup router: the "domains" option is always processed without any
6442 widening, but the router might have qualify_single set.) Now Exim uses the
6443 cached value only when the same options are set.
6445 32. Added John Jetmore's "exipick" utility to the distribution.
6447 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6448 than a timeout (e.g. a certificate is required, and is not provided), an
6449 Exim server now closes the connection immediately. Previously it waited for
6450 the client to close - but if the client is SSL, it seems that they each
6451 wait for each other, leading to a delay before one of them times out.
6453 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6454 maintained 0.8.x compatibility because I don't think many are using it, and
6455 it is clearly obsolete.
6457 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6460 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6461 client certificate was expired. A simple patch fixes this, though I don't
6462 understand the full logic of why the verify callback is called multiple
6465 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6466 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6467 which causes problems with some clients (such as the Certicom SSL Plus
6468 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6469 disables the coutermeasure allowing Eudora to connect."
6471 38. Exim was not checking that a write() to a log file succeeded. This could
6472 lead to Bad Things if a log got too big, in particular if it hit a file
6473 size limit. Exim now panics and dies if it cannot write to a log file, just
6474 as it does if it cannot open a log file.
6476 39. Modified OS/Makefile-Linux so that it now contains
6478 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6480 The two -D definitions ensure that Exim is compiled with large file
6481 support, which makes it possible to handle log files that are bigger than
6484 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6485 instance) a domain was checked against a named list that involved a lookup,
6486 causing $domain_data to be set, then another domain was checked against the
6487 same list, then the first domain was re-checked, the value of $domain_data
6488 after the final check could be wrong. In particular, if the second check
6489 failed, it could be set empty. This bug probably also applied to
6492 41. The strip_trailing_dot option was not being applied to the address given
6493 with the -f command-line option.
6495 42. The code for reading a message's header from the spool was incrementing
6496 $received_count, but never initializing it. This meant that the value was
6497 incorrect (doubled) while delivering a message in the same process in which
6498 it was received. In the most common configuration of Exim, this never
6499 happens - a fresh exec is done - but it can happen when
6500 deliver_drop_privilege is set.
6502 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6503 it now includes up to 150 characters of the unexpected data in the log
6506 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6507 and building data strings. The size of both of these buffers was 10 000
6508 bytes - far larger than anybody would *ever* want, thought I. Needless to
6509 say, somebody hit the limit. I have increased the maximum line length to
6510 20 000 and the maximum data length of concatenated lines to 100 000. I have
6511 also fixed two bugs, because there was no checking on these buffers. Tsk,
6512 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6513 buffer is too small.
6515 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6516 lsearch lookups. Now it does.
6518 46. When parsing a route_list item in a manualroute router, a fixed-length
6519 buffer was used for the list of hosts. I made this 1024 bytes long,
6520 thinking that nobody would ever have a list of hosts that long. Wrong.
6521 Somebody had a whole pile of complicated expansion conditions, and the
6522 string was silently truncated, leading to an expansion error. It turns out
6523 that it is easier to change to an unlimited length (owing to other changes
6524 that have happened since this code was originally written) than to build
6525 structure for giving a limitation error. The length of the item that
6526 expands into the list of hosts is now unlimited.
6528 47. The lsearch lookup could not handle data where the length of text line was
6529 more than 4095 characters. Such lines were truncated, leading to shortened
6530 data being returned. It should now handle lines of any length.
6532 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6533 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6536 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6537 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6538 no longer happens in comment lines. A list of replacements is now placed
6539 at the head of all of the source files, except those whose only change is
6540 to replace PERL_COMMAND in the very first #! line.
6542 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6543 messages on the queue, with a bottom-up merge sort, using code contributed
6544 by Michael Haardt. This should make operations like -bp somewhat faster on
6545 large queues. It won't affect queue runners, except when queue_run_in_order
6548 51. Installed eximstats 1.31 in the distribution.
6550 52. Added support for SRV lookups to the dnslookup router.
6552 53. If an ACL referred to $message_body or $message_body_end, the value was not
6553 reset for any messages that followed in the same SMTP session.
6555 54. The store-handling optimization for building very long strings was not
6556 differentiating between the different store pools. I don't think this
6557 actually made any difference in practice, but I've tidied it.
6559 55. While running the routers to verify a sender address, $sender_address
6560 was still set to the sender address. This is wrong, because when routing to
6561 send a bounce to the sender, it would be empty. Therefore, I have changed
6562 it so that, while verifying a sender address, $sender_address is set to <>.
6563 (There is no change to what happens when verifying a recipient address.)
6565 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6566 target A or AAAA records (if not already returned) without resetting the
6567 qualify_single or search_parents options of the DNS resolver. These are
6568 inappropriate in this case because the targets of MX and SRV records must
6569 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6570 target that, when qualified, matched something in the local domain. These
6571 two options are now turned off when doing these lookups.
6573 57. It seems that at least some releases of Reiserfs (which does not have the
6574 concept of a fixed number of inodes) returns zero and not -1 for the
6575 number of available inodes. This interacted badly with check_spool_inodes,
6576 which assumed that -1 was the "no such thing" setting. What I have done is
6577 to check that the total number of inodes is greater than zero before doing
6578 the test of how many are available.
6580 58. When a "warn" ACL statement has a log_message modifier, the message is
6581 remembered, and not repeated. This is to avoid a lot of repetition when a
6582 message has many recipients that cause the same warning to be written.
6583 However, Exim was preserving the list of already written lines for an
6584 entire SMTP session, which doesn't seem right. The memory is now reset if a
6585 new message is started.
6587 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6588 debugging output unless log_rewrite was also set.
6590 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6591 of (int)(handle) when we know that handle contains (void *)(-1).
6593 61. The Exim daemon panic-logs an error return when it closes the incoming
6594 connection. However "connection reset by peer" seems to be common, and
6595 isn't really an error worthy of noting specially, so that particular error
6598 62. When Exim is trying to find all the local interfaces, it used to panic and
6599 die if the ioctl to get the interface flags failed. However, it seems that
6600 on at least one OS (Solaris 9) it is possible to have an interface that is
6601 included in the list of interfaces, but for which you get a failure error
6602 for this call. This happens when the interface is not "plumbed" into a
6603 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6604 failure of the "get flags" call assumes that the interface is down.
6606 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6607 makes life easier for people who are doing arithmetic on fields extracted
6608 from dates, where you often get leading zeros that should not be
6609 interpreted as octal.
6611 64. Added qualify_domain to the redirect router, to override the global
6614 65. If a pathologically long header line contained very many addresses (the
6615 report of this problem mentioned 10 000) and each of them was rewritten,
6616 Exim could use up a very large amount of memory. (It kept on making new
6617 copies of the header line as it rewrote, and never released the old ones.)
6618 At the expense of a bit more processing, the header rewriting function has
6619 been changed so that it no longer eats memory in this way.
6621 66. The generation of the Received: header has been moved from the time that a
6622 message starts to be received, to the time that it finishes. The timestamp
6623 in the Received: header should now be very close to that of the <= log
6624 line. There are two side-effects of this change:
6626 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6627 logged header lines no longer include the local Received: line, because
6628 it has not yet been created. The same applies to a copy of the message
6629 that is returned to a non-SMTP sender when a message is rejected.
6631 (b) When a filter file is tested using -bf, no additional Received: header
6632 is added to the test message. After some thought, I decided that this
6635 This change does not affect the value of $received_for. It is still set
6636 after address rewriting, but before local_scan() is called.
6638 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6640 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6641 gave an unhelpful panic error message, and a defer error. I have managed to
6642 change this behaviour so that it now rejects any supplied certificate,
6643 which seems right, as the list of acceptable certificates is empty.
6645 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6646 gave an unhelpful defer error. I have not managed to make this reject any
6647 supplied certificates, but the error message it gives is "no certificate
6648 supplied", which is not helpful.
6650 70. exigrep's output now also includes lines that are not associated with any
6651 message, but which match the given pattern. Implemented by a patch from
6652 Martin Sluka, which also tidied up the Perl a bit.
6654 71. Recipient callout verification, like sender verification, was using <> in
6655 the MAIL FROM command. This isn't really the right thing, since the actual
6656 sender may affect whether the remote host accepts the recipient or not. I
6657 have changed it to use the actual sender in the callout; this means that
6658 the cache record is now keyed on a recipient/sender pair, not just the
6659 recipient address. There doesn't seem to be a real danger of callout loops,
6660 since a callout by the remote host to check the sender would use <>.
6661 [SEE ABOVE: changed after hitting problems.]
6663 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6664 temporary errors. However, in the case of such a code being given after
6665 the end of a data transmission (i.e. after ".") Exim was failing to write
6666 a retry record for the message. (Yes, there was some broken host that was
6667 actually sending 8xx at this point.)
6669 73. An unknown lookup type in a host list could cause Exim to panic-die when
6670 the list was checked. (An example that provoked this was putting <; in the
6671 middle of a list instead of at the start.) If this happened during a DATA
6672 ACL check, a -D file could be left lying around. This kind of configuration
6673 error no longer causes Exim to die; instead it causes a defer error. The
6674 incident is still logged to the main and panic logs.
6676 74. Buglet left over from Exim 3 conversion. The message "too many messages
6677 in one connection" was written to the rejectlog but not the mainlog, except
6678 when address rewriting (yes!) was being logged.
6680 75. Added write_rejectlog option.
6682 76. When a system filter was run not as root (that is, when system_filter_user
6683 was set), the values of the $n variables were not being returned to the
6684 main process; thus, they were not subsequently available in the $sn
6687 77. Added +return_path_on_delivery log selector.
6689 78. A connection timeout was being treated differently from recipients deferred
6690 when testing hosts_max_try with a message that was older than the host's
6691 retry timeout. (The host should not be counted, thus allowing all hosts to
6692 be tried at least once before bouncing.) This may have been the cause of an
6693 occasionally reported bug whereby a message would remain on the queue
6694 longer than the retry timeout, but would be bounced if a delivery was
6695 forced. I say "may" because I never totally pinned down the problem;
6696 setting up timeout/retry tests is difficult. See also the next item.
6698 79. The ultimate address timeout was not being applied to errors that involved
6699 a combination of host plus message (for example, a timeout on a MAIL
6700 command). When an address resolved to a number of possible hosts, and they
6701 were not all tried for each delivery (e.g. because of hosts_max_try), a
6702 message could remain on the queue longer than the retry timeout.
6704 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6707 81. Fixed an obscure SMTP outgoing bug which required at least the following
6708 conditions: (a) there was another message waiting for the same server;
6709 (b) the server returned 5xx to all RCPT commands in the first message so
6710 that the message was not completed; (c) the server dropped the connection
6711 or gave a negative response to the RSET that Exim sends to abort the
6712 transaction. The observed case was a dropped connection after DATA that had
6713 been sent in pipelining mode. That is, the server had advertised PIPELINING
6714 but was not implementing it correctly. The effect of the bug was incorrect
6715 behaviour, such as trying another host, and this could lead to a crash.
6721 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6722 and daemon.c were passed as pointers to ints; they should have been
6723 pointers to socklen_t variables (which are typically unsigned ints).
6725 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6728 3. Fixed a really odd bug that affected only the testing scheme; patching a
6729 certain fixed string in the binary changed the value of another string that
6730 happened to be identical to the end of the original first string.
6732 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6733 name", it returns that address as the IP address. On some operating
6734 systems (e.g. Solaris), it also passes back the IP address string as the
6735 "host name". However, on others (e.g. Linux), it passes back an empty
6736 string. Exim wasn't checking for this, and was changing the host name to an
6737 empty string, assuming it had been canonicalized.
6739 5. Although rare, it is permitted to have more than one PTR record for a given
6740 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6741 all the names associated with an address, because they do in Solaris.
6742 However, it seems that they do not in Linux for data that comes from the
6743 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6744 I found this out when I moved to a new Linux workstation and tried to run
6745 the Exim test suite.
6747 To get round this problem I have changed the code so that it now does its
6748 own call to the DNS to look up PTR records when searching for a host name.
6749 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6750 addresses that are only in /etc/hosts are still found.
6752 This behaviour is, however, controlled by an option called host_lookup_
6753 order, which defaults to "bydns:byaddr". If people want to use the other
6754 order, or indeed, just use one or the other means of lookup, they can
6755 specify it in this variable.
6757 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6758 some operating systems, this comes back from gethostbyaddr() as an empty
6759 string, and this is what Exim used to test for. However, it seems that in
6760 other systems, "." is yielded. Exim now tests for this case too.
6762 7. The values of check_spool_space and check_log_space are now held internally
6763 as a number of kilobytes instead of an absolute number of bytes. If a
6764 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6765 kilobyte. This means that much larger values can be stored.
6767 8. Exim monitor: an attempt to get the action menu when not actually pointing
6768 at a message produces an empty menu entitled "No message selected". This
6769 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6770 no entries in it ("Shell widget menu has zero width and/or height"). So I
6771 have added a single, blank menu entry in this case.
6773 9. Added ${quote_local_part.
6775 10. MIME decoding is now applied to the contents of Subject: header lines when
6778 11. Now that a reference to $sender_host_address automatically causes a reverse
6779 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6780 host lookup before query-style lookups in lists that might use this
6781 variable. This has therefore been abolished, and the "net-" prefix is no
6782 longer necessary for query-style lookups.
6784 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6785 have been a typo for LFLAGS, so it has been changed.
6787 13. The install script calls Exim with "-C /dev/null" in order to find the
6788 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6789 to be output. However, since Exim outputs its version number before the
6790 error, it didn't break the script. It just looked ugly. I fixed this by
6791 always allowing "-C /dev/null" if the caller is root.
6793 14. Ignore overlarge ACL variable number when reading spool file - insurance
6794 against a later release with more variables having written the file.
6796 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6797 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6798 this, as well as the form without the "IPv6" on the front.
6800 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6801 OS/Makefile-Darwin file.
6803 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6804 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6806 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6807 variable. However, some OS use ints for inodes. I've added cast to long int
6808 to get rid of the compiler warning.
6810 19. I had forgotten to lock out "/../" in configuration file names when
6811 ALT_CONFIG_PREFIX was set.
6813 20. Routers used for verification do not need to specify transports. However,
6814 if such a router generated a host list, and callout was configured, Exim
6815 crashed, because it could not find a port number from the (non-existent)
6816 transport. It now assumes port 25 in this circumstance.
6818 21. Added the -t option to exigrep.
6820 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6821 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6822 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6823 them is set, LOOKUP_LSEARCH is forced.
6825 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6827 24. Added sender and host information to the "rejected by local_scan()" log
6828 line; previously there was no indication of these.
6830 25. Added .include_if_exists.
6832 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6833 Linux. It turns out that not all file systems support this. Apparently some
6834 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6835 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6836 not supported on the file descriptor, is now ignored when Exim is trying to
6837 sync a directory. This applies only to Linux.
6839 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6841 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6842 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6845 29. Added support for maildirsize files from supplied patch (modified a bit).
6847 30. The use of :fail: followed by an empty string could lead Exim to respond to
6848 sender verification failures with (e.g.):
6850 550 Verification failed for <xxx>
6851 550 Sender verify failed
6853 where the first response line was missing the '-' that indicates it is not
6854 the final line of the response.
6856 31. The loop for finding the name of the user that called Exim had a hardwired
6857 limit of 10; it now uses the value of finduser_retries, which is used for
6858 all other user lookups.
6860 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6863 33. Exim was neglecting to zero errno before one call of strtol() when
6864 expanding a string and expecting an integer value. On some systems this
6865 resulted in spurious "integer overflow" errors. Also, it was casting the
6866 result into an int without checking.
6868 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6869 did not work. The code looks as if it has *never* worked, though it appears
6870 to have been documented since at least release 1.62. I have made it work.
6872 35. The "timeout_DNS" error in retry rules, also documented since at least
6873 1.62, also never worked. As it isn't clear exactly what this means, and
6874 clearly it isn't a major issue, I have abolished the feature by treating it
6875 as "timeout", and writing a warning to the main and panic logs.
6877 36. The display of retry rules for -brt wasn't always showing the error code
6880 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6881 timeout_connect_A, timeout_connect_MX.
6883 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6884 to the empty sender.
6886 39. The daemon was not analysing the content of -oX till after it had closed
6887 stderr and disconnected from the controlling terminal. This meant that any
6888 syntax errors were only noted on the panic log, and the return code from
6889 the command was 0. By re-arranging the code a little, I've made the
6890 decoding happen first, so such errors now appear on stderr, and the return
6891 code is 1. However, the actual setting up of the sockets still happens in
6892 the disconnected process, so errors there are still only recorded on the
6895 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6896 connections (as happens on some IP stacks) was logged at start up time as
6897 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6898 it from "IPv6 and IPv4", which means that two separate sockets are being
6901 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6902 says whether AF_INET or AF_INET6 was passed as an argument.
6904 42. Exiwhat output was messed up when time zones were included in log
6907 43. Exiwhat now gives more information about the daemon's listening ports,
6908 and whether -tls-on-connect was used.
6910 44. The "port" option of the smtp transport is now expanded.
6912 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6913 silently ignored. Now an error message is written to the main and panic
6916 46. There's a new ACL modifier called "logwrite" which writes to a log file
6917 as soon as it is encountered.
6919 47. Added $local_user_uid and $local_user_gid at routing time.
6921 48. Exim crashed when trying to verify a sender address that was being
6924 49. Exim was recognizing only a space character after ".include". It now also
6925 recognizes a tab character.
6927 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6928 extracting the relevant information from the specification. The man page no
6929 longer contains scrambled data for the -d option, and I've added a section
6930 at the front about calling Exim under different names.
6932 51. Added "extra_headers" argument to the "mail" command in filter files.
6934 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6937 53. Installed eximstats 1.29.
6939 54. Added transport_filter_timeout as a generic transport option.
6941 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6942 Cc: header lines. This was required by RFC 822, but it not required by RFC
6945 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6946 incoming messages that did not have them. Now it does so only if the
6947 message originates locally, that is, if there is no associated remote host
6948 address. When Resent- header lines are present, this applies to the Resent-
6949 lines rather than the non-Resent- lines.
6951 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6952 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6954 58. Messages for configuration errors now include the name of the main
6955 configuration file - useful now that there may be more than one file in a
6956 list (.included file names were always shown).
6958 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6959 for those rare installations that do not start the daemon as root or run it
6960 setuid root. I've cut out the call to initgroups() if the daemon is not
6963 60. The Exim user and group can now be bound into the binary as text strings
6964 that are looked up at the start of Exim's processing.
6966 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6968 62. Added $mailstore_basename variable.
6970 63. Installed patch to sieve.c from Michael Haardt.
6972 64. When Exim failed to open the panic log after failing to open the main log,
6973 the original message it was trying to log was written to stderr and debug
6974 output, but if they were not available (the usual case in production), it
6975 was lost. Now it is written to syslog before the two lines that record the
6976 failures to open the logs.
6978 65. Users' Exim filters run in subprocesses under the user's uid. It is
6979 possible for a "deliver" command or an alias in a "personal" command to
6980 provoke an address rewrite. If logging of address rewriting is configured,
6981 this fails because the process is not running as root or exim. There may be
6982 a better way of dealing with this, but for the moment (because 4.30 needs
6983 to be released), I have disabled address rewrite logging when running a
6984 filter in a non-root, non-exim process.
6990 1. The buildconfig auxiliary program wasn't quoting the value set for
6991 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6992 not defined. This bug was masked in 4.22 by the effect that was fixed in
6995 2. Some messages that were rejected after a message id was allocated were
6996 shown as "incomplete" by exigrep. It no longer does this for messages that
6997 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6999 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7000 have allow_domain_literals set, the ID did not get logged in the <= line.
7001 Domain literals are now always recognized in Message-ID: header lines.
7003 4. The first argument for a ${extract expansion item is the key name or field
7004 number. Leading and trailing spaces in this item were not being ignored,
7005 causing some misleading effects.
7007 5. When deliver_drop_privilege was set, single queue runner processes started
7008 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7009 same command in the process it spins off) were not dropping privilege.
7011 6. When the daemon running as "exim" started a queue runner, it always
7012 re-executed Exim in the spun-off process. This is a waste of effort when
7013 deliver_drop_privilege is set. The new process now just calls the
7014 queue-runner function directly.
7020 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7023 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7024 ignored. Though the use of -oP was forcing the writing of a pid file, it
7025 was always written to the default place.
7027 3. If the message "no IP address found for host xxxx" is generated during
7028 incoming verification, it is now followed by identification of the incoming
7029 connection (so you can more easily find what provoked it).
7031 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7033 5. Added some features to "harden" Exim a bit more against certain attacks:
7035 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7036 be put in Local/Makefile. This is like the never_users runtime option,
7037 but it cannot be overridden. The default setting is "root".
7039 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7040 prefix string with which any file named in a -C command line option
7043 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7044 is retained for -C and -D only if the caller of Exim is root. Without
7045 it, the exim user may also use -C and -D and retain privilege.
7047 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7048 command line option is disabled.
7050 6. Macro names set by the -D option must start with an upper case letter, just
7051 like macro names defined in the configuration file.
7053 7. Added "dereference=" facility to LDAP.
7055 8. Two instances of the typo "uknown" in the source files are fixed.
7057 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7058 the Configure-Makefile script screwed up while processing it.
7060 10. Incorporated PCRE 4.4.
7062 11. The SMTP synchronization check was not operating right at the start of an
7063 SMTP session. For example, it could not catch a HELO sent before the client
7064 waited for the greeting. There is now a check for outstanding input at the
7065 point when the greeting is written. Because of the duplex, asynchronous
7066 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7067 way, but not yet received, when the check is performed.
7069 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7070 on TCP/IP sockets, because this apparently causes some broken clients to
7073 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7074 unchanged) from the Cygwin maintainer.
7076 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7077 was supported instead of testing for mbx. Effectively a typo.
7079 15. The spa authenticator server code was not checking that the input it
7080 received was valid base64.
7082 16. The debug output line for the "set" modifier in ACLs was not showing the
7083 name of the variable that was being set.
7085 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7087 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7088 lookup on its own. Something else was needed to trigger the lookup. For
7089 example, a match in host_lookup or the need for a host name in a host list.
7090 Now, if $sender_host_name is referenced and the host name has not yet been
7091 looked up, a lookup is performed. If the lookup fails, the variable remains
7092 empty, and $host_lookup_failed is set to "1".
7094 19. Added "eqi" as a case-independent comparison operator.
7096 20. The saslauthd authentication condition could segfault if neither service
7097 nor realm was specified.
7099 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7100 error message that was logged was misleading, and incoming SMTP
7101 connections were dropped. The message is now more accurate, and temporary
7102 errors are given to SMTP connections.
7104 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7105 (except RSET) with a 421 error, until QUIT is received. However, it was
7106 failing to send a response to QUIT.
7108 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7109 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7110 ACL is now run after the helo_try_verify_hosts code.
7112 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7113 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7114 case-independent, so other case variants are also recognized). Apparently
7115 some systems use these upper case variants.
7117 25. If more than two messages were waiting for the same host, and a transport
7118 filter was specified for the transport, Exim sent two messages over the
7119 same TCP/IP connection, and then failed with "socket operation on non-
7120 socket" when it tried to send the third.
7122 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7124 27. The extern definition of crypt16() in expand.c was not being excluded when
7125 the OS had its own crypt16() function.
7127 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7128 as a preferred synonym for return_size_limit, both as an option and as an
7131 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7133 30. Changed the default configuration ACL to relax the local part checking rule
7134 for addresses that are not in any local domains. For these addresses,
7135 slashes and pipe symbols are allowed within local parts, but the sequence
7136 /../ is explicitly forbidden.
7138 31. SPA server authentication was not clearing the challenge buffer before
7141 32. log_message in a "warn" ACL statement was writing to the reject log as
7142 well as to the main log, which contradicts the documentation and doesn't
7143 seem right (because no rejection is happening). So I have stopped it.
7145 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7146 However, I am unable to do any testing of this.
7148 34. Fixed an infelicity in the appendfile transport. When checking directories
7149 for a mailbox, to see if any needed to be created, it was accidentally
7150 using path names with one or more superfluous leading slashes; tracing
7151 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7153 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7154 contents are added to the log line that is written for every discarded
7155 recipient. (Previously a log_message setting was ignored.)
7157 36. The ${quote: operator now quotes the string if it is empty.
7159 37. The install script runs exim in order to find its version number. If for
7160 some reason other than non-existence or emptiness, which it checks, it
7161 could not run './exim', it was installing it with an empty version number,
7162 i.e. as "exim-". This error state is now caught, and the installation is
7165 38. An argument was missing from the function that creates an error message
7166 when Exim fails to connect to the socket for saslauthd authentication.
7167 This could cause Exim to crash, or give a corrupted message.
7169 39. Added isip, isip4, and isip6 to ${if conditions.
7171 40. The ACL variables $acl_xx are now saved with the message, and can be
7172 accessed later in routers, transports, and filters.
7174 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7175 strings in the file are not string-expanded.
7177 42. If a MAIL command specified a SIZE value that was too large to fit into an
7178 int variable, the check against message_size_limit failed. Such values are
7179 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7180 day this will have to be increased, but I don't think I want to be around
7181 when emails are that large.
7188 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7189 iconv() is not standard in FreeBSD.
7191 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7192 IPv6 enabled. The observed symptom was a segmentation fault on return from
7193 the function os_common_find_running_interfaces() in src/os.c.
7195 3. In the check_special_case() function in daemon.c I had used "errno" as an
7196 argument name, which causes warnings on some systems. This was basically a
7197 typo, since it was named "eno" in the comments!
7199 4. The code that waits for the clock to tick (at a resolution of some fraction
7200 of a second) so as to ensure message-id uniqueness was always waiting for
7201 at least one whole tick, when it could have waited for less. [This is
7202 almost certainly not relevant at current processor speeds, where it is
7203 unlikely to ever wait at all. But we try to future-proof.]
7205 5. The function that sleeps for a time interval that includes fractions of a
7206 second contained a race. It did not block SIGALRM between setting the
7207 timer, and suspending (a couple of lines later). If the interval was short
7208 and the sigsuspend() was delayed until after it had expired, the suspension
7209 never ended. On busy systems this could lead to processes getting stuck for
7212 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7213 process, before it forks any delivery processes. The open lookup caching
7214 mechanism meant that the open file or database connection was passed into
7215 the delivery process. The problem was that delivery processes always tidy
7216 up cached lookup data. This could cause a problem for the next delivery
7217 process started by the queue runner, because the external queue runner
7218 process does not know about the closure. So the next delivery process
7219 still has data in the lookup cache. In the case of a file lookup, there was
7220 no problem because closing a file descriptor in a subprocess doesn't affect
7221 the parent. However, if the lookup was caching a connection to a database,
7222 the connection was closed, and the second delivery process was likely to
7223 see errors such as "PGSQL: query failed: server closed the connection
7224 unexpectedly". The problem has been fixed by closing all cached lookups
7225 in a queue runner before running a delivery process.
7227 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7228 seem to have the "const" qualifier which it has on other OS. I've
7231 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7232 *of the same type* (client or server) with the same public name that an
7233 error should be diagnosed.
7235 9. When Exim looked up a host name for an IP address, but failed to find the
7236 original IP address when looking up the host name (a safety check), it
7237 output the message "<ip address> does not match any IP for NULL", which was
7238 confusing, to say the least. The bug was that the host name should have
7239 appeared instead of "NULL".
7241 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7242 user that is built into the binary, and the -C or -D options is used, root
7243 privilege is dropped before the configuration file is read. In addition,
7244 logging is switched to stderr instead of the normal log files. If the
7245 configuration then re-defines the Exim user, the unprivileged environment
7246 is probably not what is expected, so Exim logs a panic warning message (but
7249 However, if deliver_drop_privilege is set, the unprivileged state may well
7250 be exactly what is intended, so the warning has been cut out in that case,
7251 and Exim is allowed to try to write to its normal log files.
7257 1. smtp_return_error_details was not giving details for temporary sender
7258 or receiver verification errors.
7260 2. Diagnose a configuration error if two authenticators have the same public
7263 3. Exim used not to create the message log file for a message until the first
7264 delivery attempt. This could be confusing when incoming messages were held
7265 for policy or load reasons. The message log file is now created at the time
7266 the message is received, and an initial "Received" line is written to it.
7268 4. The automatically generated man page for command line options had a minor
7269 bug that caused no ill effects; however, a more serious problem was that
7270 the procedure for building the man page automatically didn't always
7271 operate. Consequently, release 4.20 contains an out-of-date version. This
7272 shouldn't happen again.
7274 5. When building Exim with embedded Perl support, the script that builds the
7275 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7276 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7278 6. The freeze_tell option was not being used for messages that were frozen on
7279 arrival, either by an ACL or by local_scan().
7281 7. Added the smtp_incomplete_transaction log selector.
7283 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7284 was accepting AUTH without a new EHLO.
7286 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7287 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7290 10. Logging of TCP/IP connections (when configured) now happens in the main
7291 daemon process instead of the child process, so that the TCP/IP connection
7292 count is more accurate (but it can never be perfect).
7294 11. The use of "drop" in a nested ACL was not being handled correctly in the
7295 outer ACL. Now, if condition failure induced by the nested "drop" causes
7296 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7297 or "require"), the connection is dropped.
7299 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7300 that yield "discard" can now be used with an "accept" or a "discard" verb,
7301 but an error is generated for any others (because I can't see a useful way
7302 to define what should happen).
7304 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7305 are now processed in the same way as lines in the Exim configuration file.
7306 In particular, continuation lines are supported.
7308 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7310 15. Added -ti meaning -t -i.
7312 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7313 domains, and warn by logging if others are found.
7315 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7316 returned from GIFCONF when Exim is trying to find the list of interfaces on
7317 a host. The code in os.c has been modified to copy each ifreq to an aligned
7318 structure in all cases.
7320 Also, in some cases, the returned ifreq's were being copied to a 'struct
7321 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7322 means the last couple of bytes of an IPv6 address could be chopped if the
7323 ifreq contained only a normal sockaddr (14 bytes storage).
7325 18. Named domain lists were not supported in the hosts_treat_as_local option.
7326 An entry such as +xxxx was not recognized, and was treated as a literal
7329 19. Ensure that header lines added by a DATA ACL are included in the reject log
7330 if the ACL subsequently rejects the message.
7332 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7333 MD5 (which is deprecated).
7335 21. When testing a filter file using -bf, Exim was writing a message when it
7336 took the sender from a "From " line in the message, but it was not doing so
7337 when it took $return_path from a Return-Path: header line. It now does.
7339 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7340 with a valid header line field name (a series of printing characters
7341 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7343 23. Changed "disc" in the source to "disk" to conform to the documentation and
7344 the book and for uniformity.
7346 24. Ignore Sendmail's -Ooption=value command line item.
7348 25. When execve() failed while trying to run a command in a pipe transport,
7349 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7350 could be confused with a return value of 69 from the command itself. This
7351 has been changed to 127, the value the shell returns if it is asked to run
7352 a non-existent command. The wording for the related log line suggests a
7353 non-existent command as the problem.
7355 26. If received_header_text expands to an empty string, do not add a Received:
7356 header line to the message. (Well, it adds a token one on the spool, but
7357 marks it "old" so that it doesn't get used or transmitted.)
7359 27. Installed eximstats 1.28 (addition of -nt option).
7361 28. There was no check for failure on the call to getsockname() in the daemon
7362 code. This can fail if there is a shortage of resources on the system, with
7363 ENOMEM, for example. A temporary error is now given on failure.
7365 29. Contrary to the C standard, it seems that in some environments, the
7366 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7367 program. Exim now does this explicitly; it affects the formatting of
7368 timestamps using strftime().
7370 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7371 complaints. I've now initialized all the variables, to avoid this.
7373 32. Header lines added by a system filter were not being "seen" during
7374 transport-time rewrites.
7376 33. The info_callback() function passed to OpenSSL is set up with type void
7377 (*)(SSL *, int, int), as described somewhere. However, when calling the
7378 function (actually a macro) that sets it up, the type void(*)() is
7379 expected. I've put in a cast to prevent warnings from picky compilers.
7381 34. If a DNS black list lookup found a CNAME record, but there were no A
7382 records associated with the domain it pointed at, Exim crashed.
7384 35. If a DNS black list lookup returned more than one A record, Exim ignored
7385 all but the first. It now scans all returned addresses if a particular IP
7386 value is being sought. In this situation, the contents of the
7387 $dnslist_value variable are a list of all the addresses, separated by a
7390 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7391 to accept a host name and all its aliases if the forward lookup for any of
7392 them yielded the IP address of the incoming connection. Now it accepts only
7393 those names whose forward lookup yields the correct IP address. Any other
7394 names are discarded. This closes a loophole whereby a rogue DNS
7395 administrator could create reverse DNS records to break through a
7396 wildcarded host restriction in an ACL.
7398 37. If a user filter or a system filter that ran in a subprocess used any of
7399 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7400 the wrong values were passed to the pipe command ($thisaddress had the
7401 value of $0, $0 had the value of $1, etc). This bug was introduced by
7402 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7404 38. Improved the line breaking for long SMTP error messages from ACLs.
7405 Previously, if there was no break point between 40 and 75 characters, Exim
7406 left the rest of the message alone. Two changes have been made: (a) I've
7407 reduced the minimum length to 35 characters; (b) if it can't find a break
7408 point between 35 and 75 characters, it looks ahead and uses the first one
7409 that it finds. This may give the occasional overlong line, but at least the
7410 remaining text gets split now.
7412 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7413 file descriptors might be low, and that setting 1000 would always raise it.
7414 It turns out that in some environments, the limit is already over 1000 and
7415 that lowering it causes trouble. So now Exim takes care not to decrease it.
7417 40. When delivering a message, the value of $return_path is set to $sender_
7418 address at the start of routing (routers may change the value). By an
7419 oversight, this default was not being set up when an address was tested by
7420 -bt or -bv, which affected the outcome if any router or filter referred to
7423 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7424 log only when "message" or "log_message" are set. However, if one of the
7425 conditions was an address verification, or a call to a nested ACL, the
7426 messages generated by the underlying test were being passed through. This
7427 no longer happens. The underlying message is available in $acl_verify_
7428 message for both "message" and "log_message" expansions, so it can be
7429 passed through if needed.
7431 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7432 new expansion $bh_ to give the encoded byte string without charset
7433 translation. Translation happens only if iconv() is available; HAVE_ICONV
7434 indicates this at build time. HEADERS_CHARSET gives the charset to
7435 translate to; headers_charset can change it in the configuration, and
7436 "headers charset" can change it in an individual filter file.
7438 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7439 that creates RFC 2047 encoded "words" labels them as that charset instead
7440 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7441 expansion operator; (ii) when Exim creates a From: line for a local
7442 message; (iii) when a header line is rewritten to include a "phrase" part.
7444 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7445 buggy, causing it to skip the first lines of messages whose message ID
7446 ended in 'D'. This would not have bitten before Exim release 4.14, because
7447 message IDs were unlikely to end in 'D' before then. The effect was to have
7448 incorrect size information for certain domains.
7450 45. #include "config.h" was missing at the start of the crypt16.c module. This
7451 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7454 46. If there was a timeout during a "random" callout check, Exim treated it as
7455 a failure of the random address, and carried on sending RSET and the real
7456 address. If the delay was just some slowness somewhere, the response to the
7457 original RCPT would be taken as a response to RSET and so on, causing
7458 mayhem of various kinds.
7460 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7461 when I implemented it. It didn't allow for the fact that some option values
7462 may legitimately be negative (e.g. size_addition), and it didn't even do
7463 the right test for positive values.
7465 48. Domain names in DNS records are case-independent. Exim always looks them up
7466 in lower case. Some resolvers return domain names in exactly the case they
7467 appear in the zone file, that is, they may contain uppercase letters. Not
7468 all resolvers do this - some return always lower case. Exim was treating a
7469 change of case by a resolver as a change of domain, similar to a widening
7470 of a domain abbreviation. This triggered its re-routing code and so it was
7471 trying to route what was effectively the same domain again. This normally
7472 caused routing to fail (because the router wouldn't handle the domain
7473 twice). Now Exim checks for this case specially, and just changes the
7474 casing of the domain that it ultimately uses when it transmits the message
7477 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7480 50. If a filter generated a file delivery with a non-absolute name (possible if
7481 no home directory exists for the router), the forbid_file option was not
7484 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7485 the existing equality matching.
7487 52. Exim was using ints instead of ino_t variables in some places where it was
7488 dealing with inode numbers.
7490 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7491 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7492 called TMPDIR, and if it finds it is different, it changes its value.
7494 54. The smtp_printf() function is now made available to local_scan() so
7495 additional output lines can be written before returning. There is also an
7496 smtp_fflush() function to enable the detection of a dropped connection.
7497 The variables smtp_input and smtp_batched_input are exported to
7500 55. Changed the default runtime configuration: the message "Unknown user"
7501 has been removed from the ACL, and instead placed on the localuser router,
7502 using the cannot_route_message feature. This means that any verification
7503 failures that generate their own messages won't get overridden. Similarly,
7504 the "Unrouteable address" message that was in the ACL for unverifiable
7505 relay addresses has also been removed.
7507 56. Added hosts_avoid_esmtp to the smtp transport.
7509 57. The exicyclog script was not checking for the esoteric option
7510 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7511 will work only if exicyclog is run under the appropriate euid.
7513 58. Following a discussion on the list, the rules by which Exim recognises line
7514 endings on incoming messages have been changed. The -dropcr and drop_cr
7515 options are now no-ops, retained only for backwards compatibility. The
7516 following line terminators are recognized: LF CRLF CR. However, special
7517 processing applies to CR:
7519 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7520 nor a local message in the state where . is a terminator.
7522 (ii) If a bare CR is encountered in a header line, an extra space is added
7523 after the line terminator so as not to end the header. The reasoning
7524 behind this is that bare CRs in header lines are most likely either
7525 to be mistakes, or people trying to play silly games.
7527 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7528 was being incorrectly given as 18 bytes larger than it should have been.
7529 This is a VOB (very old bug).
7531 60. This may never have affected anything current, but just in case it has:
7532 When the local host is found other than at the start of a list of hosts,
7533 the local host, those with the same MX, and any that follow, are discarded.
7534 When the list in question was part of a longer list of hosts, the following
7535 hosts (not currently being processed) were also being discarded. This no
7536 longer happens. I'm not sure if this situation could ever has previously
7539 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7542 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7543 header. This is something that is recommended in a new Internet Draft, and
7544 is something that is documented as being done by Sendmail. There are two
7545 possible values. For messages generated by the autoreply transport, Exim
7548 Auto-Submitted: auto-replied
7550 whereas for all other generated messages (e.g. bounces) it adds
7552 Auto-Submitted: auto-generated
7554 63. The "personal" condition in filters now includes a test for the
7555 Auto-Submitted: header. If it contains the string "auto-" the message it
7556 not considered personal.
7558 64. Added rcpt_include_affixes as a generic transport option.
7560 65. Added queue_only_override (default true).
7562 66. Added the syslog_duplication option.
7564 67. If what should have been the first header line of a message consisted of
7565 a space followed by a colon, Exim was mis-interpreting it as a header line.
7566 It isn't of course - it is syntactically invalid and should therefore be
7567 treated as the start of the message body. The misbehaviour could have
7568 caused a number of strange effects, including loss of data in subsequent
7569 header lines, and spool format errors.
7571 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7572 client host had authenticated. This control can now be exercised by an ACL
7573 for more flexibility.
7575 69. By default, callouts do not happen when testing with -bh. There is now a
7576 variant, -bhc, which does actually run the callout code, including
7577 consulting and updating the callout cache.
7579 70. Added support for saslauthd authentication, courtesy of Alexander
7582 71. If statvfs() failed on the spool or log directories while checking their
7583 size for availability, Exim confusingly gave the error "space shortage".
7584 Furthermore, in debugging mode it crashed with a floating point exception.
7585 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7586 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7587 serious problem, Exim now writes to the main and panic logs when this
7588 happens, with details of the failure. It then refuses to accept the
7589 incoming message, giving the message "spool directory problem" or "log
7590 directory problem" with a 421 code for SMTP messages.
7592 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7593 0, 1, and 2 exist, because some OS complain for execs without them (see
7594 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7595 descriptors. However, the code omitted to check that the open succeeded,
7596 causing mysterious errors if for some reason the permissions on /dev/null
7597 got screwed. Now Exim writes a message to the main and panic logs, and
7598 bombs out if it can't open /dev/null.
7600 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7601 interact so that it is all more flexible. It is supposed to remain
7602 backwards compatible. Also added extra_local_interfaces.
7604 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7605 to bomb out with an assertion failure - to the client this appears as a
7606 connection drop. This problem occurs in the part of the code that was taken
7607 from the Samba project. Fortunately, the assertion is in a very simple
7608 function, so I have fixed this by reproducing the function inline in the
7609 one place where it is called, and arranging for authentication to fail
7610 instead of killing the process with assert().
7612 75. The SPA client code was not working when the server requested OEM rather
7613 than Unicode encoding.
7615 76. Added code to make require_files with a specific uid setting more usable in
7616 the case where statting the file as root fails - usually a non-root-mounted
7617 NFS file system. When this happens and the failure is EACCES, Exim now
7618 forks a subprocess and does the per-uid checking as the relevant uid.
7620 77. Added process_log_path.
7622 78. If log_file_path was not explicitly set, a setting of check_log_space or
7623 check_log_inodes was ignored.
7625 79. If a space check for the spool or log partitions fails, the incident is now
7626 logged. Of course, in the latter case the data may get lost...
7628 80. Added the %p formatting code to string_format() so that it can be used to
7629 print addresses in debug_print(). Adjusted all the address printing in the
7630 debugging in store.c to use %p rather than %d.
7632 81. There was a concern that a line of code in smtp_in.c could overflow a
7633 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7634 initially expressed, the concern was not well-founded, because trailing
7635 spaces are removed early. However, if the trailing spaces were followed by
7636 a NULL, they did not get removed, so the overflow was possible. Two fixes
7639 (a) I re-wrote the offending code in a cleaner fashion.
7640 (b) If an incoming SMTP command contains a NULL character, it is rejected
7643 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7644 runs initgroups(), so that if the Exim user is in any additional groups,
7645 they will be used during message reception.
7651 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob