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 Bug 1329: Fix format of Maildir-format filenames to match other mail-
10 related applications. Previously an "H" was used where available info
11 says that "M" should be, so change to match.
13 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
14 as arguments, so an implementation trying to copy these into local
15 buffer was taking a taint-enformance trap. Fix by using dynamically
22 JH/01 Avoid costly startup code when not strictly needed. This reduces time
23 for some exim process initialisations. It does mean that the logging
24 of TLS configuration problems is only done for the daemon startup.
26 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
28 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
29 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
32 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
33 DKIM signing is being done. Previously a transport_filter would always
34 disable CHUNKING, falling back to traditional DATA.
36 JH/05 Regard command-line receipients as tainted.
38 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
40 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
41 PAM library frees one of the arguments given to it, despite the
42 documentation. Therefore a plain malloc must be used.
44 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
45 on-stack buffers were used, resulting in a taint trap when DSN information
46 copied from a received message was written into the buffer.
48 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
49 the ordering of its ARC headers. This caused a crash.
51 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
52 a new record was being constructed with information from the peer, a trap
55 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
56 installation would get error messages from DMARC verify, when it hit the
57 nonexistent file indicated by the default. Distros wanting DMARC enabled
58 should both provide the file and set the option.
59 Also enforce no DMARC verification for command-line sourced messages.
61 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
62 could, depending on the platform, hang at the STARTTLS response.
64 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
65 message on a connection. Previously if one message had ARC headers and
66 the following one did not, a crash could result when adding an
67 Authentication-Results: header.
69 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
70 Exim main code and Exim-related utities. The introduction of taint
71 tracking also did many adjustments to string handling. Since then, eximon
72 frequently terminated with an assert failure.
74 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
75 check for 452 responses. This slightly helps the inefficieny of doing
76 a large alias-expansion into a recipient-limited target. The max_rcpt
77 transport option still applies (and at the current default, will override
78 the new feature). The check is done for either cause of synch, and forces
79 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
80 connection. The new facility is not tunable at this time.
82 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
83 library live data was being used, so the results became garbage. Make
84 copies while it is still usable.
86 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
87 on delivery deferred (==) and failed (**) lines (if a delivery was
88 attemtped). Previously it was only on completion (=>) lines.
90 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
91 for the expansion of the server_scram_iter and server_scram_salt options.
93 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
94 are now specifically given a NO_DATA response without hitting the system
95 resolver. The library goes on to do the now-standard TXT lookup.
96 Use of dnsdb lookups is not affected.
98 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
99 only retrieve the errormessage once. Previously two calls to dlerror()
100 were used, and the second one (for mainlog/paniclog) retrieved null
103 JH/20 Taint checking: disallow use of tainted data for
104 - the appendfile transport file and directory options
105 - the pipe transport command
106 - the autoreply transport file, log and once options
107 - file names used by the redirect router (including filter files)
109 - paths used by single-key lookups
110 Previously this was permitted.
112 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
113 adjusted the size of a major service buffer; this failed because the
114 buffer was in use at the time. Change to a compile-time increase in the
115 buffer size, when this authenticator is compiled into exim.
117 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
118 previous fast-mode was untenable in the face of glibs using mmap to
119 support larger malloc requests.
121 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
122 New values supported, if defined on system where compiled:
123 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
124 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
126 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
127 running a limited number of proceses in parallel, a benefit is gained. The
128 amount varies with the platform hardware and load. The use of the option
129 queue_run_in_order means we cannot do this, as ordering becomes
132 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
133 had introduced a string-copy (for ensuring NUL-termination) which was not
134 appropriate for that case, which can include embedded NUL bytes in the
135 block of data. Investigation showed the copy to actually be needless, the
136 data being length-specified.
138 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
139 done during a receiving connection, and both used TLS, global info was
140 used rather than per-connection info for tracking the state of data
141 queued for transmission. This could result in a connection hang.
143 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
144 Previously, when delivering serveral messages down a single connection
145 only the first would provide a SIZE. This was due to the size information
146 not being properly tracked.
148 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
149 TAI (at 37 seconds currently), pretend to be in UTC for time-related
150 expansion and logging. Previously, spurious values such as a future
151 minute could be seen.
153 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
154 it could crash from a null-deref. This could also affect the
155 ${addresses: } operator and ${readsock } item.
157 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
158 message following a mime one, the variable was not reset.
160 JH/30 When an pipelined-connect fails at the first response, assume incorrect
161 cached capability (perhaps the peer reneged?) and immediately retry in
164 JH/31 Fix spurious detection of timeout while writing to transport filter.
166 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
167 an attempt to copy the string was made before checking it.
169 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
170 taint of the lookup key was maintained; we now regard the presence in the
171 filesystem as sufficient validation.
173 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
174 argument is supplied.
176 JH/35 The dsearch lookup now requires that the directory is an absolute path.
177 Previously this was not checked, and nonempty relative paths made an
178 access under Exim's current working directory.
180 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
181 Previously no event was raised.
183 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
184 parameter supplied by the sender MAIL FROM command. Previously it was
185 ignored, and only the check_spool_space option value for the required
188 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
189 the size of the signing public-key. Previously it was instead giving
190 the size of the signature hash.
192 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
193 the default. See the (new) dkim_verify_min_keysizes option.
195 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
196 and an ACL use a lookup for checking either the local_part or domain,
197 stale data could be accessed. Ensure that variable references are
198 dropped between messages.
200 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
201 by the client was not checked as pointing within response data before
202 being used. A malicious client could thus cause an out-of-bounds read and
203 possibly gain authentication. Fix by adding the check.
205 JH/42 Internationalisation: change the default for downconversion in the smtp
206 transport to be "if needed". Previously it was "as previously set" for
207 the message, which usually meant "if needed" for message-submission but
208 "no" for everything else. However, MTAs have been seen using SMTPUTF8
209 even when the envelope addresses did not need it, resulting in forwarding
210 failures to non-supporting MTAs. A downconvert in such cases will be
211 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
212 transport. The change does mean that addresses needing conversion will
213 be converted when previously a delivery failure would occur.
215 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
216 response was received it would be used unchecked in a fail-DSN, violating
217 standards on line-length limits. Truncate if needed.
219 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
220 doesn't work. (Additionally add a new main config option to configure the
227 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
228 side, to decode a TLS 1.3 packet capture.
230 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
231 Previously the default library behaviour applied, sending two, each in
234 JH/03 Debug output for ACL now gives the config file name and line number for
237 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
239 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
241 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
242 buffer overrun for (non-chunking) other transports.
244 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
245 TLS1.3, means that a server rejecting a client certificate is not visible
246 to the client until the first read of encrypted data (typically the
247 response to EHLO). Add detection for that case and treat it as a failed
248 TLS connection attempt, so that the normal retry-in-clear can work (if
249 suitably configured).
251 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
252 and/or domain. Found and fixed by Jason Betts.
254 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
255 configuration). If a CNAME target was not a wellformed name pattern, a
258 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
259 the OS reports them interleaved with other addresses.
261 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
262 used both for input and for a verify callout, both encrypted, SMTP
263 responses being sent by the server could be lost. This resulted in
264 dropped connections and sometimes bounces generated by a peer sending
267 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
268 string. Previously it was possible to cause undefined behaviour in a
269 library routine (usually a crash). Found by "zerons".
271 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
274 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
275 API was removed, so update to use the newer ones.
277 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
278 any timeout set, is taking a long time. Previously we would hang on to a
279 rotated logfile "forever" if the input was arriving with long gaps
280 (a previous attempt to fix addressed lack, for a long time, of initial
283 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
284 shared (NFS) environment. The length of the tempfile name is now
285 4 + 16 ("hdr.$message_exim_id") which might break on file
286 systems which restrict the file name length to lower values.
289 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
290 shared (NFS) environment.
292 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
293 did for all versions <4.90). Notably -M, -m, --invert, -I may be
296 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
297 on some platforms for bit 31.
299 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
300 to changes apparently associated with TLS1.3 handling some of the APIs
301 previously used were either nonfunctional or inappropriate. Strings
302 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
303 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
304 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
305 This affects log line X= elements, the $tls_{in,out}_cipher variables,
306 and the use of specific cipher names in the encrypted= ACL condition.
308 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
310 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
311 verification result was not updated unless hosts_require_ocsp applied.
313 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
314 queue_list_requires_admin set to false, non-admin users were denied the
317 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
318 directory-of-certs mode. Previously they were advertised despite the
321 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
322 A single TCP connection by a client will now hold a TLS connection open
323 for multiple message deliveries, by default. Previoud the default was to
326 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
327 default. If built with the facility, DANE will be used. The facility
328 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
330 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
331 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
332 must be defined and you must still, unless you define DISABLE_TLS, manage
333 the the include-dir and library-file requirements that go with that
334 choice. Non-TLS builds are still supported.
336 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
339 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
340 default. If the platform supports and has the facility enabled, it will
341 be requested on all coneections.
343 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
344 controlled by the build-time option SUPPORT_PIPE_CONNECT.
346 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
348 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
349 success-DSN messages. Previously the From: header was always the default
350 one for these; the option was ignored.
352 JH/28 Fix the timeout on smtp response to apply to the whole response.
353 Previously it was reset for every read, so a teergrubing peer sending
354 single bytes within the time limit could extend the connection for a
355 long time. Credit to Qualsys Security Advisory Team for the discovery.
357 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
358 delivery address, which leaked information of the results of local
359 forwarding. Change to the original envelope recipient address, per
362 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
363 requested. Previously not bounce was generated and a log entry of
364 error ignored was made.
366 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
368 JH/32 Introduce a general tainting mechanism for values read from the input
369 channel, and values derived from them. Refuse to expand any tainted
370 values, to catch one form of exploit.
372 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
373 was unused and the unexpanded text used for the test. Found and
374 fixed by Ruben Jenster.
376 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
377 an attempt to use a TLS library read routine dereffed a nul pointer,
380 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
381 them in our smtp response.
383 JH/36 Have the general router option retry_use_local_part default to true when
384 any of the restrictive preconditions are set (to anything). Previously it
385 was only for check_local user. The change removes one item of manual
386 configuration which is required for proper retries when a remote router
387 handles a subset of addresses for a domain.
389 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
390 link count into consideration.
392 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
393 caused the extension of big_buffer, the following lines were ignored.
395 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
396 accordance with RFC 2308. Previously there was no expiry, so a longlived
397 receive process (eg. due to ACL delays) versus a short SOA value could
400 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
402 JH/39 Promote DMARC support to mainline.
404 JH/40 Bug 2452: Add a References: header to DSNs.
406 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
407 parameters. The relevant library call is documented as "Deprecated: This
408 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
409 3.6.0, DH parameters are negotiated following RFC7919."
411 HS/06 Change the default of dnssec_request_domains to "*"
413 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
414 carried on and emitted a BDAT command, even when PIPELINING was not
417 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
418 buffer was used for the filename, resulting in a trap when tainted
419 arguments (eg. $domain) were used.
421 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
422 recommended to avoid a possible server-load attack. The feature can be
423 re-enabled via the openssl_options main cofiguration option.
425 JH/45 local_scan API: documented the current smtp_printf() call. This changed
426 for version 4.90 - adding a "more data" boolean to the arguments.
427 Bumped the ABI version number also, this having been missed previously;
428 release versions 4.90 to 4.92.3 inclusive were effectively broken in
429 respect of usage of smtp_printf() by either local_scan code or libraries
430 accessed via the ${dlfunc } expansion item. Both will need coding
431 adjustment for any calls to smtp_printf() to match the new function
432 signature; a FALSE value for the new argument is always safe.
434 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
435 the file-offset (which the Linux syscall does, and exim expects); this
436 resulted in an indefinite loop.
438 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
439 to do ARC verification. The Authentication-Results: header line added
440 by the configuration then had no ARC item.
446 JH/01 Remove code calling the customisable local_scan function, unless a new
447 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
449 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
450 non-signal-safe functions being used.
452 JH/03 Bug 2269: When presented with a received message having a stupidly large
453 number of DKIM-Signature headers, disable DKIM verification to avoid
454 a resource-consumption attack. The limit is set at twenty.
456 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
457 report of oldest_pass in ${authres } in consequence, and separate out
458 some descriptions of reasons for verification fail.
460 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
461 files in the spool were present and unlocked. A queue-runner could spot
462 them, resulting in a duplicate delivery. Fix that by doing the unlock
463 after the unlink. Investigation by Tim Stewart. Take the opportunity to
464 add more error-checking on spoolfile handling while that code is being
467 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
468 No known attacks, no CVE, this is defensive hardening.
470 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
471 a queue-runner could start a delivery while other operations were ongoing.
472 Cutthrough delivery was a common victim, resulting in duplicate delivery.
473 Found and investigated by Tim Stewart. Fix by using the open message data
474 file handle rather than opening another, and not locally closing it (which
475 releases a lock) for that case, while creating the temporary .eml format
476 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
478 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
479 $sender_verify_failure/$recipient_verify_failure to "random".
481 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
484 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
485 Previously this would segfault.
487 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
490 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
491 like zero, since the resolver should be doing this for us, But we need one
492 as a CNAME but no MX presence gets the CNAME returned; we need to check
493 that doesn't point to an MX to declare it "no MX returned" rather than
494 "error, loop". A new main option is added so the older capability of
495 following some limited number of chain links is maintained.
497 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
499 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
500 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
501 now-deprecated earlier definitions used only specified the range up to TLS
502 1.2 (in the older-version library docs).
504 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
506 JH/15 Rework TLS client-side context management. Stop using a global, and
507 explicitly pass a context around. This enables future use of TLS for
508 connections to service-daemons (eg. malware scanning) while a client smtp
509 connection is using TLS; with cutthrough connections this is quite likely.
511 JH/16 Fix ARC verification to do AS checks in reverse order.
513 JH/17 Support a "tls" option on the ${readsocket } expansion item.
515 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
516 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
517 Previously the "utf8" would be re-prepended for every additional message.
519 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
520 Previously thery were accepted, resulting in issues when attempting to
521 forward messages to a non-supporting MTA.
523 PP/02 Let -n work with printing macros too, not just options.
525 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
526 one parent address was copied, and bogus data was used at delivery-logging
527 time. Either a crash (after delivery) or bogus log data could result.
528 Discovery and analysis by Tim Stewart.
530 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
531 Previously if the string ended mid-character, we did not insert the
532 promised '?' replacement.
534 PP/04 Documentation: current string operators work on bytes, not codepoints.
536 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
537 should pack well giving a smaller memory footprint so better caching and
538 therefore performance. Group the declarations where this can't be done so
539 that the byte-sized flag variables are not interspersed among pointer
540 variables, giving a better chance of good packing by the compiler.
542 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
543 non-null, to avoid issues with sites running BATV. Previously reports were
544 sent with an empty envelope sender so looked like bounces.
546 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
547 The ignore_error flag wasn't being returned from the filter subprocess so
548 was not set for later routers. Investigation and fix by Matthias Kurz.
550 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
551 and a msg:complete for the whole, when a message is manually removed using
552 -Mrm. Developement by Matthias Kurz, hacked on by JH.
554 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
555 a "Gnu special" function, asprintf() in the DB utility binary builds; I
556 hope that is portable enough.
558 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
559 requiring a known-CA anchor certificate; make it now rely entirely on the
560 TLSA as an anchor. Checking the name on the leaf cert against the name
561 on the A-record for the host is still done for TA (but not for EE mode).
563 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
564 included in delivery lines for non-proxied connections, when compiled with
565 SUPPORT_SOCKS and running with proxy logging enabled.
567 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
568 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
569 move the existing event to fire before the normal logging of message
570 failure so that custom logging is bracketed by normal logging.
572 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
573 msg:fail:internal event. Developement by Matthias Kurz.
575 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
576 far too small for todays use of crypto signatures stored there. Go all
577 the way to the max DNS message size of 64kB, even though this might be
578 overmuch for IOT constrained device use.
580 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
581 copy a string over itself. The library routine is documented as not
582 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
584 JH/32 For main options check_spool_space and check_inode_space, where the
585 platform supports 64b integers, support more than the previous 2^31 kB
586 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
587 the previous G, M, k.
589 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
590 $authenticated_fail_id variable on authentication failure. Previously
593 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
594 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
595 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
596 GNUTLS_SEC_PARAM_MEDIUM.
598 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
599 side. Previously we would continue as if no SNI had been received.
601 JH/36 Harden the handling of string-lists. When a list consisted of a sole
602 "<" character, which should be a list-separator specification, we walked
603 off past the nul-terimation.
605 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
606 causes) even when the retry time is not yet met. Previously they were
607 not, meaning that when (say) an account was over-quota and temp-rejecting,
608 and multiple senders' messages were queued, only one sender would get
609 notified on each configured delay_warning cycle.
611 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
613 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
614 cipher-suites, an error can be left on the stack even for a succeeding
615 accept; this results in impossible error messages when a later operation
618 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
619 return error codes indicating retry. Under TLS1.3 this becomes required.
621 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
622 it only wrote the new authenticators, resulting in a lack of tracking of
623 peer changes of ESMTP extensions until the next cache flush.
625 JH/41 Fix the loop reading a message header line to check for integer overflow,
626 and more-often against header_maxsize. Previously a crafted message could
627 induce a crash of the recive process; now the message is cleanly rejected.
629 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
630 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
636 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
637 When redis_servers is set to a list of > 1 element, and the Redis servers
638 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
639 case of MOVED into a DEFER case instead, thus moving the query onto the
640 next server in the list. For a cluster of N elements, all N servers must
641 be defined in redis_servers.
643 GF/02 Catch and remove uninitialized value warning in exiqsumm
644 Check for existence of @ARGV before looking at $ARGV[0]
646 JH/01 Replace the store_release() internal interface with store_newblock(),
647 which internalises the check required to safely use the old one, plus
648 the allocate and data copy operations duplicated in both (!) of the
649 extant use locations.
651 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
652 modifier. This matches the restriction on the commandline.
654 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
655 Previously only the last row was returned.
657 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
658 we assumed that tags in the header were well-formed, and parsed the
659 element content after inspecting only the first char of the tag.
660 Assumptions at that stage could crash the receive process on malformed
663 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
664 While running the DKIM ACL we operate on the Permanent memory pool so that
665 variables created with "set" persist to the DATA ACL. Also (at any time)
666 DNS lookups that fail create cache records using the Permanent pool. But
667 expansions release any allocations made on the current pool - so a dnsdb
668 lookup expansion done in the DKIM ACL releases the memory used for the
669 DNS negative-cache, and bad things result. Solution is to switch to the
670 Main pool for expansions.
671 While we're in that code, add checks on the DNS cache during store_reset,
672 active in the testsuite.
673 Problem spotted, and debugging aided, by Wolfgang Breyha.
675 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
676 When none of the hosts presented to a transport match an already-open
677 connection, close it and proceed with the list. Previously we would
678 queue the message. Spotted by Lena with Yahoo, probably involving
681 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
682 Previously a spurious "250 OK id=" response was appended to the proper
685 JH/08 The "support for" informational output now, which built with Content
686 Scanning support, has a line for the malware scanner interfaces compiled
687 in. Interface can be individually included or not at build time.
689 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
690 by the template makefile "src/EDITME". The "STREAM" support for an older
691 ClamAV interface method is removed.
693 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
694 rows affected is given instead).
696 JH/11 The runtime Berkeley DB library version is now additionally output by
697 "exim -d -bV". Previously only the compile-time version was shown.
699 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
700 SMTP connection. Previously, when one had more receipients than the
701 first, an abortive onward connection was made. Move to full support for
702 multiple onward connections in sequence, handling cutthrough connection
703 for all multi-message initiating connections.
705 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
706 routers. Previously, a multi-recipient message would fail to match the
707 onward-connection opened for the first recipient, and cause its closure.
709 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
710 a timeout on read on a GnuTLS initiating connection, resulting in the
711 initiating connection being dropped. This mattered most when the callout
712 was marked defer_ok. Fix to keep the two timeout-detection methods
715 JH/15 Relax results from ACL control request to enable cutthrough, in
716 unsupported situations, from error to silently (except under debug)
717 ignoring. This covers use with PRDR, frozen messages, queue-only and
720 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
722 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
723 metadata, resulting in a crash in free().
725 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
726 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
727 Broken also in d185889f4, with init system revamp.
729 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
730 Previously we abruptly closed the connection after reading a malware-
731 found indication; now we go on to read the "scan ok" response line,
734 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
735 ACL. Previously, a crash would result.
737 JH/19 Speed up macro lookups during configuration file read, by skipping non-
738 macro text after a replacement (previously it was only once per line) and
739 by skipping builtin macros when searching for an uppercase lead character.
741 JH/20 DANE support moved from Experimental to mainline. The Makefile control
742 for the build is renamed.
744 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
745 was allocated for every new TLS startup, meaning one per message. Fix
746 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
748 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
749 reported the original. Fix to report (as far as possible) the ACL
750 result replacing the original.
752 JH/23 Fix memory leak during multi-message connections using STARTTLS under
753 OpenSSL. Certificate information is loaded for every new TLS startup,
754 and the resources needed to be freed.
756 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
758 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
761 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
762 DATA response info to the (existing) per-recipient response info for
763 the "C=" log element. It can have useful tracking info from the
764 destination system. Patch from Simon Arlott.
766 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
767 length value. Previously this would segfault.
769 HS/02 Support Avast multiline protoocol, this allows passing flags to
770 newer versions of the scanner.
772 JH/28 Ensure that variables possibly set during message acceptance are marked
773 dead before release of memory in the daemon loop. This stops complaints
774 about them when the debug_store option is enabled. Discovered specifically
775 for sender_rate_period, but applies to a whole set of variables.
776 Do the same for the queue-runner and queue-list loops, for variables set
777 from spool message files. Do the same for the SMTP per-message loop, for
778 certain variables indirectly set in ACL operations.
780 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
781 as a multi-recipient message from a mailinglist manager). The coding had
782 an arbitrary cutoff number of characters while checking for more input;
783 enforced by writing a NUL into the buffer. This corrupted long / fast
784 input. The problem was exposed more widely when more pipelineing of SMTP
785 responses was introduced, and one Exim system was feeding another.
786 The symptom is log complaints of SMTP syntax error (NUL chars) on the
787 receiving system, and refused recipients seen by the sending system
788 (propating to people being dropped from mailing lists).
789 Discovered and pinpointed by David Carter.
791 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
792 replaced by the ${authresults } expansion.
794 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
796 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
797 allows proper process termination in container environments.
799 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
800 Previously the "final dot" had a newline after it; ensure it is CR,LF.
802 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
803 and "err_perm", deprecated since 4.83 when the RFC-defined words
804 "temperror" and "permerror" were introduced.
806 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
807 transport-filters or DKIM-signing. The restriction was lost in the
808 consolidation of verify-callout and delivery SMTP handling.
809 Extend the restriction to also cover ARC-signing.
811 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
812 in defer=pass mode supply a 450 to the initiator. Previously the message
815 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
816 tls_require_ciphers is used as before.
818 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
819 "pass_unscanned". Only tmpfails from the scanner are written to
820 the paniclog, as they may require admin intervention (permission
821 denied, license issues). Other scanner errors (like decompression
822 bombs) do not cause a paniclog entry.
824 JH/36 Fix reinitialisation of DKIM logging variable between messages.
825 Previously it was possible to log spurious information in receive log
828 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
829 triggered odd behaviour from Outlook Express clients.
831 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
834 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
835 since the IETF WG has not yet settled on that versus the original
836 "bare" representation.
838 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
839 Previously the millisecond value corrupted the output.
840 Fix also for syslog_pid=no and log_selector +pid, for which the pid
841 corrupted the output.
847 JH/01 Rework error string handling in TLS interface so that the caller in
848 more cases is responsible for logging. This permits library-sourced
849 string to be attached to addresses during delivery, and collapses
850 pairs of long lines into single ones.
852 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
853 during configuration. Wildcards are allowed and expanded.
855 JH/02 Rework error string handling in DKIM to pass more info back to callers.
856 This permits better logging.
858 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
859 do not close it down and have the child transport start it up again on
860 the passed-on TCP connection. Instead, proxy the child (and any
861 subsequent ones) for TLS via a unix-domain socket channel. Logging is
862 affected: the continued delivery log lines do not have any DNSSEC, TLS
863 Certificate or OCSP information. TLS cipher information is still logged.
865 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
866 identical IP addresses on different listening ports. Will also affect
869 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
870 add noisy ifdef guards to special-case this sillyness.
871 Patch from Bernd Kuhls.
873 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
874 than 255 are no longer allowed.
876 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
877 Disable the session-cache too, which might reduce our load. Since we
878 currrectly use a new context for every connection, both as server and
879 client, there is no benefit for these.
880 GnuTLS appears to not support tickets server-side by default (we don't
881 call gnutls_session_ticket_enable_server()) but client side is enabled
882 by default on recent versions (3.1.3 +) unless the PFS priority string
885 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
886 <https://reproducible-builds.org/specs/source-date-epoch/>.
888 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
889 the check for any unsuccessful recipients did not notice the limit, and
890 erroneously found still-pending ones.
892 JH/08 Pipeline CHUNKING command and data together, on kernels that support
893 MSG_MORE. Only in-clear (not on TLS connections).
895 JH/09 Avoid using a temporary file during transport using dkim. Unless a
896 transport-filter is involved we can buffer the headers in memory for
897 creating the signature, and read the spool data file once for the
898 signature and again for transmission.
900 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
901 4.77 as the kernel support then wasn't solid, having issues in 64bit
902 mode. Now, it's been long enough. Add support for FreeBSD also.
904 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
905 case where the routing stage had gathered several addresses to send to
906 a host before calling the transport for the first, we previously failed
907 to close down TLS in the old transport process before passing the TCP
908 connection to the new process. The new one sent a STARTTLS command
909 which naturally failed, giving a failed delivery and bloating the retry
910 database. Investigation and fix prototype from Wolfgang Breyha.
912 JH/12 Fix check on SMTP command input synchronisation. Previously there were
913 false-negatives in the check that the sender had not preempted a response
914 or prompt from Exim (running as a server), due to that code's lack of
915 awareness of the SMTP input buffering.
917 PP/04 Add commandline_checks_require_admin option.
918 Exim drops privileges sanely, various checks such as -be aren't a
919 security problem, as long as you trust local users with access to their
920 own account. When invoked by services which pass untrusted data to
921 Exim, this might be an issue. Set this option in main configuration
922 AND make fixes to the calling application, such as using `--` to stop
925 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
926 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
927 responses to those, into a single TLS record each way (this usually means
928 a single packet). As a side issue, smtp_enforce_sync now works on TLS
931 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
932 affects you only if you're dancing at the edge of the param size limits.
933 If you are, and this message makes sense to you, then: raise the
934 configured limit or use OpenSSL 1.1. Nothing we can do for older
937 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
938 cmdline element to get the documented default one. Previously it was
941 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
942 are unsuable for non-message-specific reasons - eg. connection timeout,
943 banner-time rejection.
945 JH/16 Fix logging of delivery remote port, when specified by router, under
948 PP/06 Repair manualroute's ability to take options in any order, even if one
949 is the name of a transport.
952 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
954 JH/17 Change the list-building routines interface to use the expanding-string
955 triplet model, for better allocation and copying behaviour.
957 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
958 Previously it was constructed the first time a possibly-matching string
959 was met in the configuration file input during startup; now it is done
962 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
963 compatible one, to avoid the (poorly documented) possibility of a config
964 file in the working directory redirecting the DB files, possibly correpting
965 some existing file. CVE-2017-10140 assigned for BDB.
967 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
968 cache-hot. Previously, although the result was properly cached, the
969 initial verify call returned a defer.
971 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
972 the main verify for receipient in uncached-mode.
974 JH/22 Retire historical build files to an "unsupported" subdir. These are
975 defined as "ones for which we have no current evidence of testing".
977 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
978 if present. Previously it was ignored.
980 JH/24 Start using specified-initialisers in C structure init coding. This is
981 a C99 feature (it's 2017, so now considered safe).
983 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
984 if was a fixed-sized field and bitmask ops via macros; it is now more
987 PP/07 GitHub PR 56: Apply MariaDB build fix.
988 Patch provided by Jaroslav Škarvada.
990 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
991 during Coverity cleanups [4.87 JH/47]
992 Diagnosis and fix provided by Michael Fischer v. Mollard.
994 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
995 the right size to place the terminating semicolon on its own folded
996 line, the header hash was calculated to an incorrect value thanks to
997 the (relaxed) space the fold became.
999 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1000 and confused the parent.
1002 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1003 which could crash as a result. This could lead to undeliverable messages.
1005 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1008 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1009 matters on fast-turnover and PID-randomising systems, which were getting
1010 out-of-order delivery.
1012 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1013 a possibly-overlapping copy. The symptom was that "Remote host closed
1014 connection in response to HELO" was logged instead of the actual 4xx
1017 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1018 Previously only that bufferd was discarded, resulting in SYMTP command
1021 JH/32 DKIM: when a message has multiple signatures matching an identity given
1022 in dkim_verify_signers, run the dkim acl once for each. Previously only
1023 one run was done. Bug 2189.
1025 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1026 "panic the current process" to "deliberately defer". The panic log is
1027 still written with the problem list name; the mail and reject logs now
1028 get a temp-reject line for the message that was being handled, saying
1029 something like "domains check lookup or other defer". The SMTP 451
1030 message is still "Temporary local problem".
1032 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1033 A crafted sequence of BDAT commands could result in in-use memory beeing
1034 freed. CVE-2017-16943.
1036 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1037 from SMTP input. Previously it was always done; now only done for DATA
1038 and not BDAT commands. CVE-2017-16944.
1040 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1041 to the message (such as an overlong header line). Previously this was
1042 not done and we did not exit BDAT mode. Followon from the previous item
1043 though a different problem.
1049 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1050 than -2003 did; needs libidn2 in addition to libidn.
1052 JH/02 The path option on a pipe transport is now expanded before use.
1054 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1055 Patch provided by "Björn", documentation fix added too.
1057 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1058 missing a wire-to-host endian conversion.
1060 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1061 close after a BDAT command line could be taken as a following command,
1062 giving a synch failure. Fix by only checking for synch immediately
1063 before acknowledging the chunk.
1065 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1066 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1067 macro. Patches provided by Josh Soref.
1069 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1070 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1071 by IANA as a service extension. However, John Klensin suggests that we
1074 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1075 the dkim code may be unix-mode line endings rather than smtp wire-format
1076 CRLF, so prepend a CR to any bare LF.
1078 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1079 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1081 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1082 the first were themselves being wrongly included in the feed into dkim
1083 processing; with most chunk sizes in use this resulted in an incorrect
1084 body hash calculated value.
1086 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1087 DKIM signature block, for verification. Although advised against by
1088 standards it is specifically not ruled illegal.
1090 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1092 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1093 missing a body hash (the bh= tag).
1095 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1096 It seems that HAProxy sends the Proxy Protocol information in clear and
1097 only then does a TLS startup, so do the same.
1099 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1100 TCP connections (such as for Spamd) unless the daemon successfully set
1101 Fast Open mode on its listening sockets. This fixes breakage seen on
1102 too-old kernels or those not configured for Fast Open, at the cost of
1103 requiring both directions being enabled for TFO, and TFO never being used
1104 by non-daemon-related Exim processes.
1106 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1107 endings, at least on the first header line. Try to canonify any that get
1108 past that check, despite the cost.
1110 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1111 now limited to an arbitrary five deep, while parsing addresses with the
1112 strip_excess_angle_brackets option enabled.
1114 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1115 instead leave the unprompted TLS handshake in socket buffer for the
1116 TLS library to consume.
1118 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1120 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1122 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1123 region in one operation, for speed, and this leaves assigned pointers
1124 dangling. Add checks run only under the testsuite which checks all
1125 variables at a store-reset and panics on a dangling pointer; add code
1126 explicitly nulling out all the variables discovered. Fixes one known
1127 bug: a transport crash, where a dangling pointer for $sending_ip_address
1128 originally assigned in a verify callout, is re-used.
1130 PP/06 Drop '.' from @INC in various Perl scripts.
1132 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1134 PP/08 Reduce a number of compilation warnings under clang; building with
1135 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1136 should be warning-free.
1138 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1140 HS/01 Fix portability problems introduced by PP/08 for platforms where
1141 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1143 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1144 chunk. This allows us to accept broken chunked messages. We need a more
1145 general solution here.
1147 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1148 already-broken messages in the queue.
1150 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1152 JH/19 Fix reference counting bug in routing-generated-address tracking.
1158 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1159 supports it and a size is available (ie. the sending peer gave us one).
1161 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1162 years of being deprecated). The replacements are the ACLs
1163 acl_smtp_mime and acl_not_smtp_mime.
1165 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1166 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1167 or even an in-clear connection were permitted. Now, if the host lookup
1168 was dnssec and dane was requested then the host is only used if the
1169 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1170 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1171 if one fails this test.
1172 This means that a poorly-configured remote DNS will make it incommunicado;
1173 but it protects against a DNS-interception attack on it.
1175 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1176 non-noisy when a race steals the message being considered.
1178 JH/05 If main configuration option tls_certificate is unset, generate a
1179 self-signed certificate for inbound TLS connections.
1181 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1182 in rewrites and routers.
1184 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1185 and logged a warning sing 4.83; now they are a configuration file error.
1187 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1188 (lacking @domain). Apply the same qualification processing as RCPT.
1190 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1192 JH/10 Support ${sha256:} applied to a string (as well as the previous
1195 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1196 a cutthrough deliver is pending, as we always want to make a connection.
1197 This also avoids re-routing the message when later placing the cutthrough
1198 connection after a verify cache hit.
1199 Do not update it with the verify result either.
1201 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1202 when routing results in more than one destination address.
1204 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1205 signing (which inhibits the cutthrough capability). Previously only
1206 the presence of an option was tested; now an expansion evaluating as
1207 empty is permissible (obviously it should depend only on data available
1208 when the cutthrough connection is made).
1210 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1211 the relevant preceding SMTP command did not note the pipelining mode.
1213 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1214 Previously they were not counted.
1216 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1217 as one having no matching records. Previously we deferred the message
1218 that needed the lookup.
1220 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1221 distinguished as "(=".
1223 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1224 for missing MX records. Previously it only worked for missing A records.
1226 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1228 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1229 after the data-go-ahead and data-ack. Patch from Jason Betts.
1231 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1232 even for a "none" policy. Patch from Tony Meyer.
1234 JH/22 Fix continued use of a connection for further deliveries. If a port was
1235 specified by a router, it must also match for the delivery to be
1238 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1239 When one of the recipients of a message was unsuitable for the connection
1240 (has no matching addresses), we lost track of needing to mark it
1241 deferred. As a result mail would be lost.
1243 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1245 JH/25 Decoding ACL controls is now done using a binary search; the source code
1246 takes up less space and should be simpler to maintain. Merge the ACL
1247 condition decode tables also, with similar effect.
1249 JH/26 Fix problem with one_time used on a redirect router which returned the
1250 parent address unchanged. A retry would see the parent address marked as
1251 delivered, so not attempt the (identical) child. As a result mail would
1254 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1255 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1256 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1259 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1260 Main config options check_{log,spool}_{inodes,space} are now
1261 100 inodes, 10MB unless set otherwise in the configuration.
1263 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1264 Previously it only applied to the main-section connection policy
1267 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1269 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1270 by me. Added RFC7919 DH primes as an alternative.
1272 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1273 are not in the system include path.
1275 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1276 GnuTLS, when a session startup failed (eg because the client disconnected)
1277 Exim did stdio operations after fclose. This was exposed by a recent
1278 change which nulled out the file handle after the fclose.
1280 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1281 signed directly by the cert-signing cert, rather than an intermediate
1282 OCSP-signing cert. This is the model used by LetsEncrypt.
1284 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1286 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1287 an incoming connection.
1289 HS/02 Bug 1802: Do not half-close the connection after sending a request
1292 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1293 fallback to "prime256v1".
1295 JH/34 SECURITY: Use proper copy of DATA command in error message.
1296 Could leak key material. Remotely exploitable. CVE-2016-9963.
1302 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1303 and 3.4.4 - once the server is enabled to respond to an OCSP request
1304 it does even when not requested, resulting in a stapling non-aware
1305 client dropping the TLS connection.
1307 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1308 support variable-length bit vectors. No functional change.
1310 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1311 The I= interface field on outgoing lines is now after the H= remote
1312 host field, same as incoming lines. There is a separate
1313 outgoing_interface log selector which allows you to disable the
1316 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1317 If not running log_selector +smtp_connection the mainlog would be held
1318 open indefinitely after a "too many connections" event, including to a
1319 deleted file after a log rotate. Leave the per net connection logging
1320 leaving it open for efficiency as that will be quickly detected by the
1321 check on the next write.
1323 HS/01 Bug 1671: Fix post transport crash.
1324 Processing the wait-<transport> messages could crash the delivery
1325 process if the message IDs didn't exist for some reason. When
1326 using 'split_spool_directory=yes' the construction of the spool
1327 file name failed already, exposing the same netto behaviour.
1329 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1330 mime_regex ACL conditions.
1332 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1333 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1334 to HELO, local diagnostic string.
1336 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1337 log line to debug. Even when configured with a tls authenticator many
1338 client connections are expected to not authenticate in this way, so
1339 an authenticate fail is not an error.
1341 HS/02 Add the Exim version string to the process info. This way exiwhat
1342 gives some more detail about the running daemon.
1344 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1345 matter for fast-change records such as DNSBLs.
1347 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1348 retry record, even if constant. There may be multiple transports with
1349 different interface settings and the retry behaviour needs to be kept
1352 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1354 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1356 JH/10 Bug 840: fix log_defer_output option of pipe transport
1358 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1359 hosts_randomize. This matters a lot when combined with mua_wrapper.
1361 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1362 ${quote_pgsql:<string>} operator.
1364 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1366 JH/14 Change header file name for freeradius-client. Relevant if compiling
1367 with Radius support; from the Gentoo tree and checked under Fedora.
1369 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1371 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1372 option transport_filter, meaning no filtering.
1374 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1376 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1377 defaults to "*" (all hosts). The variable is now available when not built
1378 with TLS, default unset, mainly to enable keeping the testsuite sane.
1379 If a server certificate is not supplied (via tls_certificate) an error is
1380 logged, and clients will find TLS connections fail on startup. Presumably
1381 they will retry in-clear.
1382 Packagers of Exim are strongly encouraged to create a server certificate
1383 at installation time.
1385 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1386 with the $config_file variable.
1388 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1389 in transport context, after the attempt, and per-recipient. The latter type
1390 is per host attempted. The event data is the error message, and the errno
1391 information encodes the lookup type (A vs. MX) used for the (first) host,
1392 and the trailing two digits of the smtp 4xx response.
1394 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1395 to write to mainlog (or rejectlog, paniclog) in the window between file
1396 creation and permissions/ownership being changed. Particularly affects
1397 installations where exicyclog is run as root, rather than exim user;
1398 result is that the running daemon panics and dies.
1400 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1402 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1403 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1404 "pri" and "weight". Note that the previous implicit priority given by the
1405 list order is no longer honoured.
1407 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1408 for DKIM processing.
1410 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1411 by defining SUPPORT_SOCKS.
1413 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1414 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1415 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1416 variables are renamed to proxy_{local,external}_{address,port}.
1418 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1419 for a build by defining SUPPORT_I18N
1421 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1422 of the query string, and make ${quote_redis:} do that quoting.
1424 JH/29 Move Events support from Experimental to mainline, enabled by default
1425 and removable for a build by defining DISABLE_EVENT.
1427 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1429 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1430 cached by the daemon.
1432 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1433 by defining LOOKUP_REDIS. The libhiredis library is required.
1435 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1436 keys are given for lookup.
1438 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1439 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1440 only supported when built with TLS support. The PolarSSL SHA routines
1441 are still used when the TLS library is too old for convenient support.
1443 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1444 openssl_options), for security. OpenSSL forces this from version 1.1.0
1445 server-side so match that on older versions.
1447 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1448 allocation for $value could be released as the expansion processing
1449 concluded, but leaving the global pointer active for it.
1451 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1452 and to use the domains and local_parts ACL conditions.
1454 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1455 incorrectly not doubled on cutthrough transmission, hence seen as a
1456 body-termination at the receiving system - resulting in truncated mails.
1457 Commonly the sender saw a TCP-level error, and retransmitted the message
1458 via the normal store-and-forward channel. This could result in duplicates
1459 received - but deduplicating mailstores were liable to retain only the
1460 initial truncated version.
1462 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1464 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1466 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1467 we're in there, support oversigning also; bug 1309.
1469 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1471 HS/04 Add support for keep_environment and add_environment options.
1473 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1474 either intentional arithmetic overflow during PRNG, or testing config-
1477 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1478 delivery resulted in actual delivery. Cancel cutthrough before DATA
1481 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1482 rejects a recipient: pass the reject to the originator.
1484 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1485 Many were false-positives and ignorable, but it's worth fixing the
1488 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1489 for the new environment-manipulation done at startup. Move the routines
1490 from being local to tls.c to being global via the os.c file.
1492 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1493 an extract embedded as result-arg for a map, the first arg for extract
1494 is unavailable so we cannot tell if this is a numbered or keyed
1495 extraction. Accept either.
1501 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1504 JH/02 The smtp transport option "multi_domain" is now expanded.
1506 JH/03 The smtp transport now requests PRDR by default, if the server offers
1509 JH/04 Certificate name checking on server certificates, when exim is a client,
1510 is now done by default. The transport option tls_verify_cert_hostnames
1511 can be used to disable this per-host. The build option
1512 EXPERIMENTAL_CERTNAMES is withdrawn.
1514 JH/05 The value of the tls_verify_certificates smtp transport and main options
1515 default to the word "system" to access the system default CA bundle.
1516 For GnuTLS, only version 3.0.20 or later.
1518 JH/06 Verification of the server certificate for a TLS connection is now tried
1519 (but not required) by default. The verification status is now logged by
1520 default, for both outbound TLS and client-certificate supplying inbound
1523 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1526 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1527 Status Notification (bounce) messages are now MIME format per RFC 3464.
1528 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1529 under the control of the dsn_advertise_hosts option, and routers may
1530 have a dsn_lasthop option.
1532 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1533 default, modifiable by a malware= option. The list separator for
1534 the options can now be changed in the usual way. Bug 68.
1536 JH/10 The smtp_receive_timeout main option is now expanded before use.
1538 JH/11 The incoming_interface log option now also enables logging of the
1539 local interface on delivery outgoing connections.
1541 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1542 if the interface and destination host and port all match.
1544 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1547 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1548 Patch from Andrew Lewis.
1550 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1551 now supports optional time-restrictions, weighting, and priority
1552 modifiers per server. Patch originally by <rommer@active.by>.
1554 JH/16 The spamd_address main option now supports a mixed list of local
1555 and remote servers. Remote servers can be IPv6 addresses, and
1556 specify a port-range.
1558 JH/17 Bug 68: The spamd_address main option now supports an optional
1559 timeout value per server.
1561 JH/18 Bug 1581: Router and transport options headers_add/remove can
1562 now have the list separator specified.
1564 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1567 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1570 JH/21 Support for the A6 type of dns record is withdrawn.
1572 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1573 rather than the verbs used.
1575 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1576 from 255 to 1024 chars.
1578 JH/24 Verification callouts now attempt to use TLS by default.
1580 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1581 are generic router options now. The defaults didn't change.
1583 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1584 Original patch from Alexander Shikoff, worked over by JH.
1586 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1587 files if ZCAT_COMMAND is not executable.
1589 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1591 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1593 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1594 Normally benign, it bites when the pair was led to by a CNAME;
1595 modern usage is to not canonicalize the domain to a CNAME target
1596 (and we were inconsistent anyway for A-only vs AAAA+A).
1598 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1600 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1601 when evaluating $sender_host_dnssec.
1603 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1604 $sender_helo_dnssec variable.
1606 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1608 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1610 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1612 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1613 documented as working, but never had. Support all but $spam_report.
1615 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1616 added for tls authenticator.
1618 HS/03 Add perl_taintmode main config option
1624 TL/01 When running the test suite, the README says that variables such as
1625 no_msglog_check are global and can be placed anywhere in a specific
1626 test's script, however it was observed that placement needed to be near
1627 the beginning for it to behave that way. Changed the runtest perl
1628 script to read through the entire script once to detect and set these
1629 variables, reset to the beginning of the script, and then run through
1630 the script parsing/test process like normal.
1632 TL/02 The BSD's have an arc4random API. One of the functions to induce
1633 adding randomness was arc4random_stir(), but it has been removed in
1634 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1635 function when detected.
1637 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1638 cause callback expansion.
1640 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1641 syntax errors in an expansion can be treated as a string instead of
1642 logging or causing an error, due to the internal use of bool_lax
1643 instead of bool when processing it.
1645 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1646 server certificates when making smtp deliveries.
1648 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1650 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1652 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1654 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1655 Merged patch from Sebastian Wiedenroth.
1657 JH/05 Fix results-pipe from transport process. Several recipients, combined
1658 with certificate use, exposed issues where response data items split
1659 over buffer boundaries were not parsed properly. This eventually
1660 resulted in duplicates being sent. This issue only became common enough
1661 to notice due to the introduction of connection certificate information,
1662 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1664 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1665 size buffer was used, resulting in syntax errors when an expansion
1668 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1669 version 3.3.6 or later.
1671 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1672 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1673 both become "event_action", the variables become $event_name, $event_data
1674 and $event_defer_errno. There is a new variable $verify_mode, usable in
1675 routers, transports and related events. The tls:cert event is now also
1676 raised for inbound connections, if the main configuration event_action
1679 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1680 early OCSP support, but no stapling (appears to be less than 1.0.0).
1682 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1683 server certificate names available under the smtp transport option
1684 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1687 JH/10 Time-related extraction expansions from certificates now use the main
1688 option "timezone" setting for output formatting, and are consistent
1689 between OpenSSL and GnuTLS compilations. Bug 1541.
1691 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1692 encoded parameter in the incoming message. Bug 1558.
1694 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1695 include certificate info, eximon was claiming there were spoolfile
1698 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1700 JH/14 Log delivery-related information more consistently, using the sequence
1701 "H=<name> [<ip>]" wherever possible.
1703 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1704 are problematic for Debian distribution, omit them from the release
1707 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1709 JH/16 Fix string representation of time values on 64bit time_t architectures.
1712 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1713 output list separator was used.
1718 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1719 checkers that were complaining about end of non-void function with no
1722 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1723 This was a regression introduced in 4.83 by another bugfix.
1725 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1727 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1728 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1734 TF/01 Correctly close the server side of TLS when forking for delivery.
1736 When a message was received over SMTP with TLS, Exim failed to clear up
1737 the incoming connection properly after forking off the child process to
1738 deliver the message. In some situations the subsequent outgoing
1739 delivery connection happened to have the same fd number as the incoming
1740 connection previously had. Exim would try to use TLS and fail, logging
1741 a "Bad file descriptor" error.
1743 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1744 utilities have not been installed.
1746 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1747 temporary space as the ACL may create new global variables.
1749 TL/01 LDAP support uses per connection or global context settings, depending
1750 upon the detected version of the libraries at build time.
1752 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1753 to extract and use the src ip:port in logging and expansions as if it
1754 were a direct connection from the outside internet. PPv2 support was
1755 updated based on HAProxy spec change in May 2014.
1757 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1759 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1760 Properly escape header and check for NULL return.
1762 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1765 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1767 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1768 characters in header names, implemented as a verify condition.
1769 Contributed by Michael Fischer v. Mollard.
1771 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1772 results permerror and temperror. Previous values are deprecated but
1773 still accepted. In a future release, err_perm and err_temp will be
1774 completely removed, which will be a backward incompatibility if the
1775 ACL tests for either of these two old results. Patch contributed by
1776 user bes-internal on the mailing list.
1778 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1780 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1781 selectors, in both main and reject logs.
1783 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1786 JH/07 Add malware type "sock" for talking to simple daemon.
1788 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1790 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1791 routers/transports under cutthrough routing.
1793 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1794 numbers. Touch up "bool" conditional to keep the same definition.
1796 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1798 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1800 JH/12 Expand items in router/transport headers_add or headers_remove lists
1801 individually rather than the list as a whole. Bug 1452.
1803 Required for reasonable handling of multiple headers_ options when
1804 they may be empty; requires that headers_remove items with embedded
1805 colons must have them doubled (or the list-separator changed).
1807 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1808 view the policy declared in the DMARC record. Currently, $dmarc_status
1809 is a combined value of both the record presence and the result of the
1812 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1814 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1815 dnslookup router and the smtp transport (applying to the forward
1818 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1819 of ldap servers used for a specific lookup. Patch provided by Heiko
1822 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1823 New variable $lookup_dnssec_authenticated for observability.
1825 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1826 Patch submitted by Lars Timman.
1828 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1830 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1831 Requires trusted mode and valid format message id, aborts otherwise.
1832 Patch contributed by Heiko Schlichting.
1834 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1835 certextract with support for various fields. Bug 1358.
1837 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1838 is requested by default, modifiable by smtp transport option
1841 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1842 operate on certificate variables to give certificate fingerprints
1843 Also new ${sha256:cert_variable}.
1845 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1847 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1848 Christian Aistleitner.
1850 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1852 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1853 file. Patch from Wolfgang Breyha.
1855 JH/25 Expand the coverage of the delivery $host and $host_address to
1856 client authenticators run in verify callout. Bug 1476.
1858 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1859 align with daemon_smtp_ports. Bug 72.
1861 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1862 support and error reporting did not work properly.
1864 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1865 and is readable. Patch from Andrew Colin Kissa.
1867 TL/14 Enhance documentation of ${run expansion and how it parses the
1868 commandline after expansion, particularly in the case when an
1869 unquoted variable expansion results in an empty value.
1871 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1873 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1874 by renaming away from it. Addresses GH issue 15, reported by
1877 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1879 TL/15 SECURITY: prevent double expansion in math comparison functions
1880 (can expand unsanitized data). Not remotely exploitable.
1887 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1889 PP/02 Make -n do something, by making it not do something.
1890 When combined with -bP, the name of an option is not output.
1892 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1895 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1896 $sender_host_name and config options to manage this, and basic check
1899 PP/05 DSCP support for outbound connections and control modifier for inbound.
1901 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1902 (Only plugin which currently uses this is kerberos4, which nobody should
1903 be using, but we should make it available and other future plugins might
1904 conceivably use it, even though it would break NAT; stuff *should* be
1905 using channel bindings instead).
1907 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1908 name; added for Sendmail compatibility; requires admin caller.
1909 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1910 just ignore it); requires trusted caller.
1911 Also parse but ignore: -Ac -Am -X<logfile>
1914 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1916 TL/02 Add +smtp_confirmation as a default logging option.
1918 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1919 Patch by Magnus Holmgren from 2007-02-20.
1921 TL/04 Bugzilla 1281 - Spec typo.
1922 Bugzilla 1283 - Spec typo.
1923 Bugzilla 1290 - Spec grammar fixes.
1925 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1927 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1929 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1930 mailing list by by Dmitry Isaikin.
1932 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1934 JH/02 Support "G" suffix to numbers in ${if comparisons.
1936 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1938 NM/01 Bugzilla 1197 - Spec typo
1939 Bugzilla 1196 - Spec examples corrections
1941 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1943 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1944 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1947 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1948 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1950 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1951 "acl {{name}{arg}...}", and optional args on acl condition
1954 JH/05 Permit multiple router/transport headers_add/remove lines.
1956 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1958 JH/07 Avoid using a waiting database for a single-message-only transport.
1959 Performance patch from Paul Fisher. Bugzilla 1262.
1961 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1964 JH/09 Add $headers_added variable, with content from use of ACL modifier
1965 add_header (but not yet added to the message). Bugzilla 199.
1967 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1968 Pulled from Bugzilla 817 by Wolfgang Breyha.
1970 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1972 (nb: this is the same fix as in Exim 4.80.1)
1974 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1977 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1980 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1982 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1983 advertise SMTP AUTH mechanism to us, instead of a generic
1984 protocol violation error. Also, make Exim more robust to bad
1985 data from the Dovecot auth socket.
1987 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1989 When a queue runner is handling a message, Exim first routes the
1990 recipient addresses, during which it prunes them based on the retry
1991 hints database. After that it attempts to deliver the message to
1992 any remaining recipients. It then updates the hints database using
1995 So if a recipient address works intermittently, it can get repeatedly
1996 deferred at routing time. The retry hints record remains fresh so the
1997 address never reaches the final cutoff time.
1999 This is a fairly common occurrence when a user is bumping up against
2000 their storage quota. Exim had some logic in its local delivery code
2001 to deal with this. However it did not apply to per-recipient defers
2002 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2004 This change adds a proper retry rule check during routing so that the
2005 final cutoff time is checked against the message's age. We only do
2006 this check if there is an address retry record and there is not a
2007 domain retry record; this implies that previous attempts to handle
2008 the address had the retry_use_local_parts option turned on. We use
2009 this as an approximation for the destination being like a local
2010 delivery, as in LMTP.
2012 I suspect this new check makes the old local delivery cutoff check
2013 redundant, but I have not verified this so I left the code in place.
2015 TF/02 Correct gecos expansion when From: is a prefix of the username.
2017 Test 0254 submits a message to Exim with the header
2021 When I ran the test suite under the user fanf2, Exim expanded
2022 the header to contain my full name, whereas it should have added
2023 a Resent-Sender: header. It erroneously treats any prefix of the
2024 username as equal to the username.
2026 This change corrects that bug.
2028 GF/01 DCC debug and logging tidyup
2029 Error conditions log to paniclog rather than rejectlog.
2030 Debug lines prefixed by "DCC: " to remove any ambiguity.
2032 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2034 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2035 Bug spotted by Jeremy Harris; was flawed since initial commit.
2036 Would have resulted in OCSP responses post-SNI triggering an Exim
2037 NULL dereference and crash.
2039 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2041 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2042 Bug detection, analysis and fix by Samuel Thibault.
2043 Bugzilla 1331, Debian bug #698092.
2045 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2047 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2048 Server implementation by Todd Lyons, client by JH.
2049 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2050 config variable "prdr_enable" controls whether the server
2051 advertises the facility. If the client requests PRDR a new
2052 acl_data_smtp_prdr ACL is called once for each recipient, after
2053 the body content is received and before the acl_smtp_data ACL.
2054 The client is controlled by both of: a hosts_try_prdr option
2055 on the smtp transport, and the server advertisement.
2056 Default client logging of deliveries and rejections involving
2057 PRDR are flagged with the string "PRDR".
2059 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2060 fclose(). Diagnosis by Todd Lyons.
2062 PP/17 Update configure.default to handle IPv6 localhost better.
2063 Patch by Alain Williams (plus minor tweaks).
2066 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2067 This is now consistent with GnuTLS, and is now documented: the
2068 previous undocumented portable approach to treating the option as
2069 unset was to force an expansion failure. That still works, and
2070 an empty string is now equivalent.
2072 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2073 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2074 not performing validation itself.
2076 PP/20 Added force_command boolean option to pipe transport.
2077 Patch from Nick Koston, of cPanel Inc.
2079 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2082 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2084 PP/21 Fix eximon continuous updating with timestamped log-files.
2085 Broken in a format-string cleanup in 4.80, missed when I repaired the
2086 other false fix of the same issue.
2087 Report and fix from Heiko Schlichting.
2090 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2091 Report from Prashanth Katuri.
2093 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2094 It's SecureTransport, so affects any MacOS clients which use the
2095 system-integrated TLS libraries, including email clients.
2097 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2098 using a MIME ACL for non-SMTP local injection.
2099 Report and assistance in diagnosis by Warren Baker.
2101 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2103 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2105 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2106 last id that failed so it may be referenced in subsequent ACL's.
2108 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2111 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2112 ldap library initialization, allowing self-signed CA's to be
2113 used. Also properly sets require_cert option later in code by
2114 using NULL (global ldap config) instead of ldap handle (per
2115 session). Bug diagnosis and testing by alxgomz.
2117 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2118 the src/util/ subdirectory.
2120 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2121 renamed to Transport Post Delivery Action by Jeremy Harris, as
2124 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2125 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2126 redis_servers = needs to be configured which will be used by the redis
2127 lookup. Patch from Warren Baker, of The Packet Hub.
2129 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2131 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2132 hostname or reverse DNS when processing a host list. Used suggestions
2133 from multiple comments on this bug.
2135 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2137 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2138 Missed a few lines, added it to make the runtest require no keyboard
2141 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2142 contains upper case chars. Make router use caseful_local_part.
2144 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2145 support when GnuTLS has been built with p11-kit.
2151 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2153 This, or similar/improved, will also be change PP/11 of 4.82.
2159 PP/01 Handle short writes when writing local log-files.
2160 In practice, only affects FreeBSD (8 onwards).
2161 Bugzilla 1053, with thanks to Dmitry Isaikin.
2163 NM/01 Bugzilla 949 - Documentation tweak
2165 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2168 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2170 PP/02 Implemented gsasl authenticator.
2172 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2174 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2175 `pkg-config foo` for cflags/libs.
2177 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2178 with rest of GSASL and with heimdal_gssapi.
2180 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2181 `pkg-config foo` for cflags/libs for the TLS implementation.
2183 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2184 properties get this fed in as external SSF. A number of robustness
2185 and debugging improvements to the cyrus_sasl authenticator.
2187 PP/08 cyrus_sasl server now expands the server_realm option.
2189 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2190 Patch by Jeremy Harris.
2192 PP/10 Added dbmjz lookup type.
2194 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2196 PP/12 MAIL args handles TAB as well as SP, for better interop with
2197 non-compliant senders.
2198 Analysis and variant patch by Todd Lyons.
2200 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2201 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2202 Patch from Dirk Mueller <dmueller@suse.com>
2204 PP/13 tls_peerdn now print-escaped for spool files.
2205 Observed some $tls_peerdn in wild which contained \n, which resulted
2206 in spool file corruption.
2208 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2209 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2210 or write after TLS renegotiation, which otherwise led to messages
2213 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2214 as a tracking header (ie: a signed header comes before the signature).
2215 Patch from Wolfgang Breyha.
2217 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2218 comma-sep list; embedded commas doubled.
2220 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2222 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2224 Report and patch from Dmitry Banschikov.
2226 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2227 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2228 failures. We appear to now support TLS1.1+ with Exim.
2230 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2231 lets Exim select keys and certificates based upon TLS SNI from client.
2232 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2233 before an outbound SMTP session. New log_selector, +tls_sni.
2235 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2236 NULL dereference. Report and patch from Alun Jones.
2238 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2239 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2240 Not seeing resolver debug output on NetBSD, but suspect this is a
2241 resolver implementation change.
2243 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2244 Left warnings. Added "eximon gdb" invocation mode.
2246 PP/21 Defaulting "accept_8bitmime" to true, not false.
2248 PP/22 Added -bw for inetd wait mode support.
2250 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2251 locate the relevant includes and libraries. Made this the default.
2253 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2254 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2256 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2257 This may cause build issues on older platforms.
2259 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2260 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2261 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2262 Added SNI support via GnuTLS too.
2263 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2265 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2267 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2268 Applied second patch from Janne, implementing suggestion to default
2269 multiple-strings-in-record handling to match SPF spec.
2271 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2273 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2274 read-only, out of scope).
2275 Patch from Wolfgang Breyha, report from Stuart Northfield.
2277 PP/29 Fix three issues highlighted by clang analyser static analysis.
2278 Only crash-plausible issue would require the Cambridge-specific
2279 iplookup router and a misconfiguration.
2280 Report from Marcin Mirosław.
2282 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2284 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2285 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2286 As part of this, removing so much warning spew let me fix some minor
2287 real issues in debug logging.
2289 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2290 assignment on my part. Fixed.
2292 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2293 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2294 Janne Snabb (who went above and beyond: thank you).
2296 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2297 string otherwise requires a connection and a bunch more work and it's
2298 relatively easy to get wrong. Should also expose TLS library linkage
2301 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2302 64-bit ${eval} (JH/03).
2304 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2305 GNU libc to support some of the 64-bit stuff, should not lead to
2306 conflicts. Defined before os.h is pulled in, so if a given platform
2307 needs to override this, it can.
2309 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2310 protection layer was required, which is not implemented.
2311 Bugzilla 1254, patch from Wolfgang Breyha.
2313 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2314 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2315 tls_dhparam take prime identifiers. Also unbreak combination of
2316 OpenSSL+DH_params+TLSSNI.
2318 PP/39 Disable SSLv2 by default in OpenSSL support.
2324 PP/01 Solaris build fix for Oracle's LDAP libraries.
2325 Bugzilla 1109, patch from Stephen Usher.
2327 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2329 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2332 TF/02 Fix a couple more cases where we did not log the error message
2333 when unlink() failed. See also change 4.74-TF/03.
2335 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2336 lock up or crash if it happened to be inside a call to libc when it
2337 got a SIGUSR1 from exiwhat.
2339 The SIGUSR1 handler appends the current process status to the process
2340 log which is later printed by exiwhat. It used to use the general
2341 purpose logging code to do this, but several functions it calls are
2342 not safe for signals.
2344 The new output code in the SIGUSR1 handler is specific to the process
2345 log, and simple enough that it's easy to inspect for signal safety.
2346 Removing some special cases also simplifies the general logging code.
2347 Removing the spurious timestamps from the process log simplifies
2350 TF/04 Improved ratelimit ACL condition.
2352 The /noupdate option has been deprecated in favour of /readonly which
2353 has clearer semantics. The /leaky, /strict, and /readonly update modes
2354 are mutually exclusive. The update mode is no longer included in the
2355 database key; it just determines when the database is updated. (This
2356 means that when you upgrade Exim will forget old rate measurements.)
2358 Exim now checks that the per_* options are used with an update mode that
2359 makes sense for the current ACL. For example, when Exim is processing a
2360 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2361 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2362 must specify per_mail/readonly. If you omit the update mode it defaults to
2363 /leaky where that makes sense (as before) or /readonly where required.
2365 The /noupdate option is now undocumented but still supported for
2366 backwards compatibility. It is equivalent to /readonly except that in
2367 ACLs where /readonly is required you may specify /leaky/noupdate or
2368 /strict/noupdate which are treated the same as /readonly.
2370 A useful new feature is the /count= option. This is a generalization
2371 of the per_byte option, so that you can measure the throughput of other
2372 aggregate values. For example, the per_byte option is now equivalent
2373 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2375 The per_rcpt option has been generalized using the /count= mechanism
2376 (though it's more complicated than the per_byte equivalence). When it is
2377 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2378 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2379 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2380 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2381 Note that using per_rcpt with a non-readonly update mode in more than
2382 one ACL will cause the recipients to be double-counted. (The per_mail
2383 and per_byte options don't have this problem.)
2385 The handling of very low rates has changed slightly. If the computed rate
2386 is less than the event's count (usually one) then this event is the first
2387 after a long gap. In this case the rate is set to the same as this event's
2388 count, so that the first message of a spam run is counted properly.
2390 The major new feature is a mechanism for counting the rate of unique
2391 events. The new per_addr option counts the number of different
2392 recipients that someone has sent messages to in the last time period. It
2393 behaves like per_rcpt if all the recipient addresses are different, but
2394 duplicate recipient addresses do not increase the measured rate. Like
2395 the /count= option this is a general mechanism, so the per_addr option
2396 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2397 example, measure the rate that a client uses different sender addresses
2398 with the options per_mail/unique=$sender_address. There are further
2399 details in the main documentation.
2401 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2403 TF/06 Removed a few PCRE remnants.
2405 TF/07 Automatically extract Exim's version number from tags in the git
2406 repository when doing development or release builds.
2408 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2409 Bugzilla 879. Patch from Paul Fisher.
2411 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2412 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2415 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2417 PP/05 Variable $av_failed, true if the AV scanner deferred.
2418 Bugzilla 1078. Patch from John Horne.
2420 PP/06 Stop make process more reliably on build failure.
2421 Bugzilla 1087. Patch from Heiko Schlittermann.
2423 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2424 Bugzilla 1089. Patch from Heiko Schlittermann.
2426 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2427 Bugzilla 1131. Patch from Holger Weiß.
2429 PP/09 Handle IPv6 addresses with SPF.
2430 Bugzilla 860. Patch from Wolfgang Breyha.
2432 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2434 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2437 PP/11 match_* no longer expand right-hand-side by default.
2438 New compile-time build option, EXPAND_LISTMATCH_RHS.
2439 New expansion conditions, "inlist", "inlisti".
2441 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2443 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2445 PP/14 fix log_write() format string regression from TF/03.
2446 Bugzilla 1152. Patch from Dmitry Isaikin.
2452 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2454 PP/02 Harmonised TLS library version reporting; only show if debugging.
2455 Layout now matches that introduced for other libraries in 4.74 PP/03.
2457 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2459 PP/04 New "dns_use_edns0" global option.
2461 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2464 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2465 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2467 TK/01 Updated PolarSSL code to 0.14.2.
2468 Bugzilla 1097. Patch from Andreas Metzler.
2470 PP/07 Catch divide-by-zero in ${eval:...}.
2471 Fixes bugzilla 1102.
2473 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2476 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2477 format-string attack -- SECURITY: remote arbitrary code execution.
2479 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2480 time unintentionally subject to list matching rules, letting the header
2481 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2482 arbitrary string expansion). This allowed for information disclosure.
2484 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2485 INT_MIN/-1 -- value coerced to INT_MAX.
2491 NM/01 Workaround for PCRE version dependency in version reporting
2494 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2495 This fixes portability to compilers other than gcc, notably
2496 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2498 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2499 makefiles for portability to HP-UX and POSIX correctness.
2501 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2502 Also via indented variable definition in the Makefile.
2503 (Debugging by Oliver Heesakkers).
2505 PP/02 Restore caching of spamd results with expanded spamd_address.
2506 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2508 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2509 Improves build reliability. Fix from: Frank Elsner
2511 NM/02 Fix wide character breakage in the rfc2047 coding
2512 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2514 NM/03 Allow underscore in dnslist lookups
2515 Fixes bug 1026. Patch from Graeme Fowler
2517 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2518 Code patches from Adam Ciarcinski of NetBSD.
2520 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2523 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2524 is logged, to avoid truncation. Patch from John Horne.
2526 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2527 Patch from Jakob Hirsch.
2529 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2530 SQL string expansion failure details.
2531 Patch from Andrey Oktyabrski.
2533 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2534 Patch from Simon Arlott.
2536 PP/09 New lookups functionality failed to compile on old gcc which rejects
2537 extern declarations in function scope.
2538 Patch from Oliver Fleischmann
2540 PP/10 Use sig_atomic_t for flags set from signal handlers.
2541 Check getgroups() return and improve debugging.
2542 Fixed developed for diagnosis in bug 927 (which turned out to be
2545 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2546 Patch from Mark Zealey.
2548 PP/12 Bugzilla 1056: Improved spamd server selection.
2549 Patch from Mark Zealey.
2551 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2552 Based on patch from Heiko Schlittermann.
2554 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2555 Patch from Uwe Doering, sign-off by Michael Haardt.
2557 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2558 variable declaration deep within a block. Bug and patch from
2561 PP/15 lookups-Makefile IRIX compatibility coercion.
2563 PP/16 Make DISABLE_DKIM build knob functional.
2565 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2566 Patch by Simon Arlott
2568 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2569 variable argument macros. Our copy now differs from upstream.
2575 TF/01 Failure to get a lock on a hints database can have serious
2576 consequences so log it to the panic log.
2578 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2579 controlled using the smtp_confirmation log selector.
2581 TF/03 Include the error message when we fail to unlink a spool file.
2583 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2584 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2585 for maintaining out-of-tree patches for some time.
2587 PP/01 Bugzilla 139: Documentation and portability issues.
2588 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2589 Handle per-OS dynamic-module compilation flags.
2591 PP/02 Let /dev/null have normal permissions.
2592 The 4.73 fixes were a little too stringent and complained about the
2593 permissions on /dev/null. Exempt it from some checks.
2594 Reported by Andreas M. Kirchwitz.
2596 PP/03 Report version information for many libraries, including
2597 Exim version information for dynamically loaded libraries. Created
2598 version.h, now support a version extension string for distributors
2599 who patch heavily. Dynamic module ABI change.
2601 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2602 privilege escalation vulnerability whereby the Exim run-time user
2603 can cause root to append content of the attacker's choosing to
2606 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2609 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2610 If dropping privileges for untrusted macros, we disabled normal logging
2611 on the basis that it would fail; for the Exim run-time user, this is not
2612 the case, and it resulted in successful deliveries going unlogged.
2613 Fixed. Reported by Andreas Metzler.
2619 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2620 only prepend for the Resent-* case. Fixes regression introduced in
2621 Exim 4.70 by NM/22 for Bugzilla 607.
2623 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2624 increasing numbers of administrators be bitten by this.
2626 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2628 PP/03 Bugzilla 994: added openssl_options main configuration option.
2630 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2632 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2634 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2636 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2637 without a peer certificate, leading to a segfault because of an
2638 assumption that peers always have certificates. Be a little more
2639 paranoid. Problem reported by Martin Tscholak.
2641 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2642 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2643 NB: ClamAV planning to remove STREAM in "middle of 2010".
2644 CL also introduces -bmalware, various -d+acl logging additions and
2645 more caution in buffer sizes.
2647 PP/09 Implemented reverse_ip expansion operator.
2649 PP/10 Bugzilla 937: provide a "debug" ACL control.
2651 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2653 PP/12 Bugzilla 973: Implement --version.
2655 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2657 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2659 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2661 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2662 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2663 ignore trailing whitespace.
2665 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2667 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2670 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2671 Notification from Dr Andrew Aitchison.
2673 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2674 ExtendedDetectionInfo response format.
2675 Notification from John Horne.
2677 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2680 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2681 XSL and documented dependency on system catalogs, with examples of how
2684 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2687 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2688 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2689 configuration file which is writeable by the Exim user or group.
2691 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2692 of configuration files to cover files specified with the -C option if
2693 they are going to be used with root privileges, not just the default
2696 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2697 option (effectively making it always true).
2699 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2700 files to be used while preserving root privileges.
2702 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2703 that rogue child processes cannot use them.
2705 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2706 run-time user, instead of root.
2708 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2709 Exim run-time user without dropping privileges.
2711 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2712 result string, instead of calling string_vformat() twice with the same
2715 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2716 for other users. Others should always drop root privileges if they use
2717 -C on the command line, even for a whitelisted configure file.
2719 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2721 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2727 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2728 $data_path, and $header_path variables; fixed documentation bugs and
2731 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2732 exipick to access non-standard spools, including the "frozen" queue
2735 NM/01 Bugzilla 965: Support mysql stored procedures.
2736 Patch from Alain Williams
2738 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2740 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2741 Patch from Andreas Metzler
2743 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2744 Patch from Kirill Miazine
2746 NM/05 Bugzilla 671: Added umask to procmail example.
2748 JJ/03 installed exipick 20100323.0, fixing doc bug
2750 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2751 directory. Notification and patch from Dan Rosenberg.
2753 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2755 TK/02 Improve log output when DKIM signing operation fails.
2757 MH/01 Treat the transport option dkim_domain as a colon separated
2758 list, not as a single string, and sign the message with each element,
2759 omitting multiple occurences of the same signer.
2761 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2762 Bugzilla 985, 986. Patch by Simon Arlott
2764 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2765 Patch by Simon Arlott
2767 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2768 MBX locking. Notification from Dan Rosenberg.
2774 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2776 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2778 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2780 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2782 NM/04 Bugzilla 915: Flag broken perl installation during build.
2788 TK/01 Added patch by Johannes Berg that expands the main option
2789 "spamd_address" if it starts with a dollar sign.
2791 TK/02 Write list of recipients to X-Envelope-Sender header when building
2792 the mbox-format spool file for content scanning (suggested by Jakob
2795 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2796 (http://www.dcc-servers.net/) support via dccifd. Activated by
2797 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2799 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2800 by Mark Daniel Reidel <mr@df.eu>.
2802 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2803 When building exim an external PCRE library is now needed -
2804 PCRE is a system library on the majority of modern systems.
2805 See entry on PCRE_LIBS in EDITME file.
2807 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2808 conversation. Added nologin parameter to request.
2809 Patch contributed by Kirill Miazine.
2811 TF/01 Do not log submission mode rewrites if they do not change the address.
2813 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2815 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2816 log files in place. Contributed by Roberto Lima.
2818 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2820 TF/03 Bugzilla 615: When checking the local_parts router precondition
2821 after a local_part_suffix or local_part_prefix option, Exim now
2822 does not use the address's named list lookup cache, since this
2823 contains cached lookups for the whole local part.
2825 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2826 Robert Millan. Documentation is in experimental-spec.txt.
2828 TF/04 Bugzilla 668: Fix parallel build (make -j).
2830 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2832 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2833 Patch provided by Jan Srzednicki.
2835 TF/05 Leading white space used to be stripped from $spam_report which
2836 wrecked the formatting. Now it is preserved.
2838 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2839 that they are available at delivery time.
2841 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2843 TF/08 TLS error reporting now respects the incoming_interface and
2844 incoming_port log selectors.
2846 TF/09 Produce a more useful error message if an SMTP transport's hosts
2847 setting expands to an empty string.
2849 NM/06 Bugzilla 744: EXPN did not work under TLS.
2850 Patch provided by Phil Pennock.
2852 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2853 Patch provided by Richard Godbee.
2855 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2856 acl_smtp_notquit, added index entry.
2858 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2859 Patch provided by Eugene Bujak.
2861 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2862 accept(). Patch provided by Maxim Dounin.
2864 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2865 Patch provided by Phil Pennock.
2867 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2869 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2870 Patch provided by Brad "anomie" Jorsch.
2872 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2873 Patch provided by Dean Brooks.
2875 TK/05 Add native DKIM support (does not depend on external libraries).
2877 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2878 Patch provided by Graeme Fowler.
2880 NM/16 Bugzilla 851: Documentation example syntax fix.
2882 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2884 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2887 NM/19 Bugzilla 745: TLS version reporting.
2888 Patch provided by Phil Pennock.
2890 NM/20 Bugzilla 167: bool: condition support.
2891 Patch provided by Phil Pennock.
2893 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2894 clients. Patch provided by Phil Pennock.
2896 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2897 Patch provided by Brad "anomie" Jorsch.
2899 NM/23 Bugzilla 687: Fix misparses in eximstats.
2900 Patch provided by Heiko Schlittermann.
2902 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2903 Patch provided by Heiko Schlittermann.
2905 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2906 plus update to original patch.
2908 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2910 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2911 Patch provided by David Brownlee.
2913 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2915 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2917 NM/30 Bugzilla 888: TLS documentation bugfixes.
2919 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2921 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2922 Unlike the original bugzilla I have changed all shell scripts in src tree.
2924 NM/33 Bugzilla 898: Transport filter timeout fix.
2925 Patch by Todd Rinaldo.
2927 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2928 Patch by Serge Demonchaux.
2930 NM/35 Bugzilla 39: Base64 decode bug fixes.
2931 Patch by Jakob Hirsch.
2933 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2935 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2937 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2939 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2945 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2946 ALT-N's libdkim that I have put here:
2947 http://duncanthrax.net/exim-experimental/
2949 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2950 were called 'true' and it seems that C99 defines that as a reserved
2951 keyword to be used with 'bool' variable types. That means you could
2952 not include C99-style headers which use bools without triggering
2953 build errors in sieve.c.
2955 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2956 as mailq or other aliases. Changed the --help handling significantly
2957 to do whats expected. exim_usage() emits usage/help information.
2959 SC/01 Added the -bylocaldomain option to eximstats.
2961 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2963 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2965 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2971 PH/01 Another patch from the Sieve maintainer.
2973 PH/02 When an IPv6 address is converted to a string for single-key lookup
2974 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2975 dots are used instead of colons so that keys in lsearch files need not
2976 contain colons. This was done some time before quoting was made available
2977 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2978 (notated using the quote facility) so as to distinguish them from IPv4
2979 keys. This meant that lookups for IP addresses in host lists did not work
2980 for iplsearch lookups.
2982 This has been fixed by arranging for IPv6 addresses to be expressed with
2983 colons if the lookup type is iplsearch. This is not incompatible, because
2984 previously such lookups could never work.
2986 The situation is now rather anomalous, since one *can* have colons in
2987 ordinary lsearch keys. However, making the change in all cases is
2988 incompatible and would probably break a number of configurations.
2990 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2993 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2994 conversion specification without a maximum field width, thereby enabling
2995 a rogue spamd server to cause a buffer overflow. While nobody in their
2996 right mind would setup Exim to query an untrusted spamd server, an
2997 attacker that gains access to a server running spamd could potentially
2998 exploit this vulnerability to run arbitrary code as the Exim user.
3000 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3001 $primary_hostname instead of what libspf2 thinks the hosts name is.
3003 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3004 a directory entry by the name of the lookup key. Previously, if a
3005 symlink pointed to a non-existing file or a file in a directory that
3006 Exim lacked permissions to read, a lookup for a key matching that
3007 symlink would fail. Now it is enough that a matching directory entry
3008 exists, symlink or not. (Bugzilla 503.)
3010 PH/03 The body_linecount and body_zerocount variables are now exported in the
3013 PH/04 Added the $dnslist_matched variable.
3015 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3016 This means they are set thereafter only if the connection becomes
3019 PH/06 Added the client_condition to authenticators so that some can be skipped
3020 by clients under certain conditions.
3022 PH/07 The error message for a badly-placed control=no_multiline_responses left
3023 "_responses" off the end of the name.
3025 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3027 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3028 (without spaces) instead of just copying the configuration text.
3030 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3032 PH/11 Added $max_received_linelength.
3034 PH/12 Added +ignore_defer and +include_defer to host lists.
3036 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3037 way in which PCRE > 7.0 is built.
3039 PH/14 Implemented queue_only_load_latch.
3041 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3042 MAIL command. The effect was to mangle the value on 64-bit systems.
3044 PH/16 Another patch from the Sieve maintainer.
3046 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3048 PH/18 If a system quota error occurred while trying to create the file for
3049 a maildir delivery, the message "Mailbox is full" was not appended to the
3050 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3051 only to a quota excession during the actual writing of the file.
3053 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3054 characters?) which causes problems in log lines. The DN values are now
3055 passed through string_printing() before being added to log lines.
3057 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3058 and InterBase are left for another time.)
3060 PH/21 Added message_body_newlines option.
3062 PH/22 Guard against possible overflow in moan_check_errorcopy().
3064 PH/23 POSIX allows open() to be a macro; guard against that.
3066 PH/24 If the recipient of an error message contained an @ in the local part
3067 (suitably quoted, of course), incorrect values were put in $domain and
3068 $local_part during the evaluation of errors_copy.
3074 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3075 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3078 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3079 issue a MAIL command.
3081 PH/02 In an ACL statement such as
3083 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3085 if a client was not listed at all, or was listed with a value other than
3086 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3087 the condition was not true (as it should be), so access was not denied.
3088 The bug was that the ! inversion was incorrectly passed on to the second
3089 item. This has been fixed.
3091 PH/03 Added additional dnslists conditions == and =& which are different from
3092 = and & when the dns lookup returns more than one IP address.
3094 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3095 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3097 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3098 FSYNC, which compiles an option called disable_fsync that allows for
3099 bypassing fsync(). The documentation is heavily laced with warnings.
3101 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3103 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3104 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3105 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3106 including adding "make clean"; (3) Added -fPIC when compiling the test
3107 dynamically loaded module, to get rid of a warning.
3109 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3110 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3111 = 0s. The bug is otherwise harmless.
3113 PH/07 There was a bug in the dovecot authenticator such that the value of
3114 $auth1 could be overwritten, and so not correctly preserved, after a
3115 successful authentication. This usually meant that the value preserved by
3116 the server_setid option was incorrect.
3118 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3120 PH/09 Installed PCRE release 7.0.
3122 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3123 run for batched SMTP input. It is now run at the start of every message
3124 in the batch. While fixing this I discovered that the process information
3125 (output by running exiwhat) was not always getting set for -bs and -bS
3126 input. This is fixed, and it now also says "batched" for BSMTP.
3128 PH/11 Added control=no_pipelining.
3130 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3131 patch, slightly modified), and move the expansion of helo_data till after
3132 the connection is made in the smtp transport (so it can use these
3135 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3137 PH/14 Added log_selector = +pid.
3139 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3141 PH/16 Add ${if forany and ${if forall.
3143 PH/17 Added dsn_from option to vary the From: line in DSNs.
3145 PH/18 Flush SMTP output before performing a callout, unless control =
3146 no_callout_flush is set.
3148 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3149 was true (the default) a successful delivery failed to delete the retry
3150 item, thus causing premature timeout of the address. The bug is now
3153 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3155 PH/21 Long custom messages for fakedefer and fakereject are now split up
3156 into multiline responses in the same way that messages for "deny" and
3157 other ACL rejections are.
3159 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3160 with slight modification.
3162 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3163 draft, changing the syntax and factoring some duplicate code".
3165 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3166 for deliveries of the second and subsequent messages over the same SMTP
3169 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3170 ${reduce, with only minor "tidies".
3172 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3174 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3175 expansion side effects.
3177 PH/27 When a message times out after an over-quota error from an Exim-imposed
3178 quota, the bounce message says "mailbox is full". This message was not
3179 being given when it was a system quota that was exceeded. It now should
3182 MH/03 Made $recipients available in local_scan(). local_scan() already has
3183 better access to the recipient list through recipients_list[], but
3184 $recipients can be useful in postmaster-provided expansion strings.
3186 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3187 in the case of a MAIL command with additional options following the
3188 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3189 were accidentally chopped off.
3191 PH/29 SMTP synchronization checks are implemented when a command is read -
3192 there is a check that no more input is waiting when there shouldn't be
3193 any. However, for some commands, a delay in an ACL can mean that it is
3194 some time before the response is written. In this time, more input might
3195 arrive, invalidly. So now there are extra checks after an ACL has run for
3196 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3197 pipelining has not been advertised.
3199 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3201 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3202 returned in the SMTP response when smtp_return_error_details was false.
3203 This has been fixed.
3205 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3206 instead of the C I/O that was originally supplied, because problems were
3207 reported on Solaris.
3209 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3210 Exim which did not show up earlier: it was assuming that a call to
3211 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3212 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3213 was a macro that became an assignment, so it seemed to work. This has
3214 changed to a proper function call with a void return, hence the compile
3215 error. Exim's code has been fixed.
3217 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3220 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3222 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3224 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3225 because it turns out to be too big or there is a timeout) there is no
3226 "Completed" line in the log. When some messages of this type were
3227 selected by exigrep, they were listed as "not completed". Others were
3228 picked up by some special patterns. I have improved the selection
3229 criteria to be more general.
3231 PH/38 The host_find_failed option in the manualroute router can now be set
3232 to "ignore", to completely ignore a host whose IP address cannot be
3233 found. If all hosts are ignored, the behaviour is controlled by the new
3234 host_all_ignored option.
3236 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3237 homing or because of multiple MX records with /mx) generated more than
3238 one IP address, and the following item turned out to be the local host,
3239 all the secondary addresses of the first item were incorrectly removed
3240 from the list, along with the local host and any following hosts (which
3241 is what is supposed to happen).
3243 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3244 whoever called Exim into the -H file. In the case of the daemon it was
3245 behaving confusingly. When first started, it used values for whoever
3246 started the daemon, but after a SIGHUP it used the Exim user (because it
3247 calls itself on a restart). I have changed the code so that it now always
3250 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3251 message are rejected with the same error (e.g. no authentication or bad
3252 sender address), and a DATA command is nevertheless sent (as can happen
3253 with PIPELINING or a stupid MUA), the error message that was given to the
3254 RCPT commands is included in the rejection of the DATA command. This is
3255 intended to be helpful for MUAs that show only the final error to their
3258 PH/42 Another patch from the Sieve maintainer.
3260 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3261 Eximstats - Fixed some broken HTML links and added missing column headers
3263 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3264 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3266 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3268 PH/43 Yet another patch from the Sieve maintainer.
3270 PH/44 I found a way to check for a TCP/IP connection going away before sending
3271 the response to the final '.' that terminates a message, but only in the
3272 case where the client has not sent further data following the '.'
3273 (unfortunately, this is allowed). However, in many cases there won't be
3274 any further data because there won't be any more messages to send. A call
3275 to select() can be used: if it shows that the input is "ready", there is
3276 either input waiting, or the socket has been closed. An attempt to read
3277 the next input character can distinguish the two cases. Previously, Exim
3278 would have sent an OK response which the client would never have see.
3279 This could lead to message repetition. This fix should cure that, at
3280 least in a lot of common cases.
3282 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3283 advertised in response to EHLO.
3289 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3290 fixed by 4.65/MH/01 (is this a record?) are fixed:
3292 (i) An empty string was always treated as zero by the numeric comparison
3293 operators. This behaviour has been restored.
3295 (ii) It is documented that the numeric comparison operators always treat
3296 their arguments as decimal numbers. This was broken in that numbers
3297 starting with 0 were being interpreted as octal.
3299 While fixing these problems I realized that there was another issue that
3300 hadn't been noticed. Values of message_size_limit (both the global option
3301 and the transport option) were treated as octal if they started with 0.
3302 The documentation was vague. These values are now always treated as
3303 decimal, and I will make that clear in the documentation.
3309 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3310 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3313 MH/01 Don't check that the operands of numeric comparison operators are
3314 integers when their expansion is in "skipping" mode (fixes bug
3315 introduced by 4.64-PH/07).
3317 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3318 child addresses, Exim now panics and dies. Previously, because the count
3319 is held in a short int, deliveries were likely to be lost. As such a
3320 large number of recipients for a single message is ridiculous
3321 (performance will be very, very poor), I have chosen to impose a limit
3322 rather than extend the field.
3328 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3329 leftover -K file (the existence of which was triggered by #402).
3330 While we were at it, introduced process PID as part of the -K
3331 filename. This should rule out race conditions when creating
3334 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3335 processing considerably. Previous code took too long for large mails,
3336 triggering a timeout which in turn triggers #401.
3338 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3339 in the DK code in transports.c. sendfile() is not really portable,
3340 hence the _LINUX specificness.
3342 TF/01 In the add_headers option to the mail command in an Exim filter,
3343 there was a bug that Exim would claim a syntax error in any
3344 header after the first one which had an odd number of characters
3347 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3348 callout verification, Exim cached a "reject" for the entire domain. This
3349 is correct for most verifications, but it is not correct for a recipient
3350 verification with use_sender or use_postmaster set, because in that case
3351 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3352 case of MAIL FROM:<> rejection from other early rejections (e.g.
3353 rejection of HELO). When verifying a recipient using a non-null MAIL
3354 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3355 Whatever the result of the callout, the value of the domain cache is
3356 left unchanged (for any other kind of callout, getting as far as trying
3357 RCPT means that the domain itself is ok).
3359 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3362 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3363 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3364 session, the daemon ignores ECONNECT errors and logs others; it now
3365 ignores EPIPE as well.
3367 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3368 (quoted-printable decoding).
3370 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3371 later the small subsequent patch to fix an introduced bug.
3373 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3375 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3377 PH/08 An error is now given if message_size_limit is specified negative.
3379 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3380 to be given (somewhat) arbitrary names.
3382 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3385 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3386 miscellaneous code fixes
3388 PH/10 Added the log_reject_target ACL modifier to specify where to log
3391 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3392 hostname. This is wrong, because it relates to the incoming message (and
3393 probably the interface on which it is arriving) and not to the outgoing
3394 callout (which could be using a different interface). This has been
3395 changed to use the value of the helo_data option from the smtp transport
3396 instead - this is what is used when a message is actually being sent. If
3397 there is no remote transport (possible with a router that sets up host
3398 addresses), $smtp_active_hostname is used.
3400 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3401 tweaks were necessary in order to get it to work (see also 21 below):
3402 (a) The code assumed that strncpy() returns a negative number on buffer
3403 overflow, which isn't the case. Replaced with Exim's string_format()
3405 (b) There were several signed/unsigned issues. I just did the minimum
3406 hacking in of casts. There is scope for a larger refactoring.
3407 (c) The code used strcasecmp() which is not a standard C function.
3408 Replaced with Exim's strcmpic() function.
3409 (d) The code set only $1; it now sets $auth1 as well.
3410 (e) A simple test gave the error "authentication client didn't specify
3411 service in request". It would seem that Dovecot has changed its
3412 interface. Fortunately there's a specification; I followed it and
3413 changed what the client sends and it appears to be working now.
3415 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3418 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3419 address A is aliased to B and C, where B exists and C does not. Without
3420 -v the output is "A verified" because verification stops after a
3421 successful redirection if more than one address is generated. However,
3422 with -v the child addresses are also verified. Exim was outputting "A
3423 failed to verify" and then showing the successful verification for C,
3424 with its parentage. It now outputs "B failed to verify", showing B's
3425 parentage before showing the successful verification of C.
3427 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3428 look up a TXT record in a specific list after matching in a combined
3431 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3432 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3433 they consult the DNS. I had assumed they would set it the way they
3434 wanted; and indeed my experiments on Linux seem to show that in some
3435 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3436 To be on the safe side, however, I have now made the interface to
3437 host_find_byname() similar to host_find_bydns(), with an argument
3438 containing the DNS resolver options. The host_find_byname() function now
3439 sets these options at its start, just as host_find_bydns() does. The smtp
3440 transport options dns_qualify_single and dns_search_parents are passed to
3441 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3442 of host_find_byname() use the default settings of RES_DEFNAMES
3443 (qualify_single) but not RES_DNSRCH (search_parents).
3445 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3446 spool_read_header() do less string testing, by means of a preliminary
3447 switch on the second character of optional "-foo" lines. (This is
3448 overdue, caused by the large number of possibilities that now exist.
3449 Originally there were few.) While I was there, I also converted the
3450 str(n)cmp tests so they don't re-test the leading "-" and the first
3451 character, in the hope this might squeeze out yet more improvement.
3453 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3454 flag allowing group syntax was set by the header_syntax check but not
3455 turned off, possible causing trouble later; (2) The flag was not being
3456 set at all for the header_verify test, causing "group"-style headers to
3457 be rejected. I have now set it in this case, and also caused header_
3458 verify to ignore an empty address taken from a group. While doing this, I
3459 came across some other cases where the code for allowing group syntax
3460 while scanning a header line wasn't quite right (mostly, not resetting
3461 the flag correctly in the right place). These bugs could have caused
3462 trouble for malformed header lines. I hope it is now all correct.
3464 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3465 with the "reply" argument non-NULL. The code, however (which originally
3466 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3467 but it didn't always do it. This confused somebody who was copying the
3468 code for some other use. I have removed all the tests.
3470 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3471 feature that was used to support insecure browsers during the U.S. crypto
3472 embargo. It requires special client support, and Exim is probably the
3473 only MTA that supported it -- and would never use it because real RSA is
3474 always available. This code has been removed, because it had the bad
3475 effect of slowing Exim down by computing (never used) parameters for the
3476 RSA_EXPORT functionality.
3478 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3479 authenticator to fail if there's a tab character in the incoming data
3480 (there should never be unless someone is messing about, as it's supposed
3481 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3482 if the connection is using TLS or if the remote IP is the same as the
3483 local IP, and the "valid-client-cert option" if a client certificate has
3486 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3487 authenticators. This can be used for authorization after authentication
3488 succeeds. (In the case of plaintext, it servers for both authentication
3491 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3492 if any retry times were supplied.
3494 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3495 connection, where there is no client IP address to check. In this
3496 situation, the verify now always succeeds.
3498 PH/25 Applied John Jetmore's -Mset patch.
3500 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3502 PH/27 In a string expansion for a processed (not raw) header when multiple
3503 headers of the same name were present, leading whitespace was being
3504 removed from all of them, but trailing whitespace was being removed only
3505 from the last one. Now trailing whitespace is removed from each header
3506 before concatenation. Completely empty headers in a concatenation (as
3507 before) are ignored.
3509 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3510 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3512 PH/29 [Removed. This was a change that I later backed out, and forgot to
3513 correct the ChangeLog entry (that I had efficiently created) before
3514 committing the later change.]
3516 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3517 address errors (4xx response to RCPT) over the same connection as other
3518 messages routed to the same hosts. Such deliveries are always "forced",
3519 so retry times are not inspected. This resulted in far too many retries
3520 for the affected addresses. The effect occurred only when there were more
3521 hosts than the hosts_max_try setting in the smtp transport when it had
3522 the 4xx errors. Those hosts that it had tried were not added to the list
3523 of hosts for which the message was waiting, so if all were tried, there
3524 was no problem. Two fixes have been applied:
3526 (i) If there are any address or message errors in an SMTP delivery, none
3527 of the hosts (tried or untried) are now added to the list of hosts
3528 for which the message is waiting, so the message should not be a
3529 candidate for sending over the same connection that was used for a
3530 successful delivery of some other message. This seems entirely
3531 reasonable: after all the message is NOT "waiting for some host".
3532 This is so "obvious" that I'm not sure why it wasn't done
3533 previously. Hope I haven't missed anything, but it can't do any
3534 harm, as the worst effect is to miss an optimization.
3536 (ii) If, despite (i), such a delivery is accidentally attempted, the
3537 routing retry time is respected, so at least it doesn't keep
3538 hammering the server.
3540 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3541 in ${readsocket because some servers need this prod.
3543 PH/32 Added some extra debug output when updating a wait-xxx database.
3545 PH/33 The hint "could be header name not terminated by colon", which has been
3546 given for certain expansion errors for a long time, was not being given
3547 for the ${if def:h_colon_omitted{... case.
3549 PH/34 The spec says: "With one important exception, whenever a domain list is
3550 being scanned, $domain contains the subject domain." There was at least
3551 one case where this was not true.
3553 PH/35 The error "getsockname() failed: connection reset by peer" was being
3554 written to the panic log as well as the main log, but it isn't really
3555 panic-worthy as it just means the connection died rather early on. I have
3556 removed the panic log writing for the ECONNRESET error when getsockname()
3559 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3560 runs only) independently of the message's sender address. This meant
3561 that, if the 4xx error was in fact related to the sender, a different
3562 message to the same recipient with a different sender could confuse
3563 things. In particular, this can happen when sending to a greylisting
3564 server, but other circumstances could also provoke similar problems.
3565 I have changed the default so that the retry time for these errors is now
3566 based a combination of the sender and recipient addresses. This change
3567 can be overridden by setting address_retry_include_sender=false in the
3570 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3571 remote server are returned as part of bounce messages. This was not
3572 happening for LMTP over a pipe (the lmtp transport), but now it is the
3573 same for both kinds of LMTP.
3575 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3576 in header lines that were in fact CNAMEs. This is no longer the case.
3578 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3579 and queue runs started by the daemon processed all messages. This has
3580 been fixed so that -R and -S can now usefully be given with -q<time>.
3582 PH/40 Import PCRE release 6.7 (fixes some bugs).
3584 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3586 PH/42 Give an error if -q is specified more than once.
3588 PH/43 Renamed the variables $interface_address and $interface_port as
3589 $received_ip_address and $received_port, to make it clear that these
3590 values apply to message reception, and not to the outgoing interface when
3591 a message is delivered. (The old names remain recognized, of course.)
3593 PH/44 There was no timeout on the connect() call when using a Unix domain
3594 socket in the ${readsocket expansion. There now is.
3596 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3597 be meaningful with "accept".
3599 SC/01 Eximstats V1.43
3600 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3602 SC/02 Eximstats V1.44
3603 Use a glob alias rather than an array ref in the generated
3604 parser. This improves both readability and performance.
3606 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3607 Collect SpamAssassin and rejection statistics.
3608 Don't display local sender or destination tables unless
3609 there is data to show.
3610 Added average volumes into the top table text output.
3612 SC/04 Eximstats V1.46
3613 Collect data on the number of addresses (recipients)
3614 as well as the number of messages.
3616 SC/05 Eximstats V1.47
3617 Added 'Message too big' to the list of mail rejection
3618 reasons (thanks to Marco Gaiarin).
3620 SC/06 Eximstats V1.48
3621 Mainlog lines which have GMT offsets and are too short to
3622 have a flag are now skipped.
3624 SC/07 Eximstats V1.49 (Alain Williams)
3625 Added the -emptyok flag.
3627 SC/08 Eximstats V1.50
3628 Fixes for obtaining the IP address from reject messages.
3630 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3631 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3632 whitespace changes from 4.64-PH/27
3634 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3637 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3638 are found, allow negative numbers in numeric criteria)
3640 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3642 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3643 to match changes made in 4.64-PH/43
3645 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3647 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3648 "domain" is actually an IP address, and force a failure. This locks out
3649 those revolvers/nameservers that support "A-for-A" lookups, in
3650 contravention of the specifications.
3652 PH/48 When a host name was looked up from an IP address, and the subsequent
3653 forward lookup of the name timed out, the host name was left in
3654 $sender_host_name, contrary to the specification.
3656 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3657 restricted to single-key lookups, Exim was not diagnosing an error if
3658 * or *@ was used with a query-style lookup.
3660 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3662 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3663 long ago, but noone interested enough thought of it. Let's just say that
3664 the "1.1" means that there are some new functions that weren't there at
3665 some point in the past.
3667 PH/51 Error processing for expansion failure of helo_data from an smtp
3668 transport during callout processing was broken.
3670 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3671 tested/used via the -bh/-bhc/-bs options.
3673 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3674 bug, fixed in subsequent PCRE releases).
3676 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3677 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3679 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3685 SC/01 Use a glob alias rather than an array ref in eximstats generated
3686 parser. This improves both readability and performance.
3688 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3689 Don't display local sender or destination tables in eximstats unless
3690 there is data to show.
3691 Added average volumes into the eximstats top table text output.
3693 SC/03 Collect data on the number of addresses (recipients) as well
3694 as the number of messages in eximstats.
3696 TF/01 Correct an error in the documentation for the redirect router. Exim
3697 does (usually) call initgroups() when daemonizing.
3699 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3700 with consistent privilege compared to when running as a daemon.
3702 TF/03 Note in the spec that $authenticated_id is not set for local
3703 submissions from trusted users.
3705 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3706 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3708 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3709 by adding some example configuration directives to the default
3710 configuration file. A little bit of work is required to uncomment the
3711 directives and define how usernames and passwords are checked, but
3712 there is now a framework to start from.
3714 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3715 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3716 without this. I don't know how relevant this is to other LDAP libraries.
3718 PH/02 Add the verb name to the "unknown ACL verb" error.
3720 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3722 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3724 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3725 directory not expanded when it should be if an expanded home directory
3726 was set for the address (which is overridden by the transport).
3728 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3731 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3732 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3733 because it is too late at that time, and has no effect.
3735 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3736 security issue with \' (bugzilla #107). I could not use the
3737 PQescapeStringConn() function, because it needs a PGconn value as one of
3740 PH/08 When testing addresses using -bt, indicate those final addresses that
3741 are duplicates that would not cause an additional delivery. At least one
3742 person was confused, thinking that -bt output corresponded to deliveries.
3743 (Suppressing duplicates isn't a good idea as you lose the information
3744 about possibly different redirections that led to the duplicates.)
3746 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3747 systems where poll() doesn't work, in particular OS X.
3749 PH/10 Added more information to debugging output for retry time not reached.
3751 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3752 operations in malware.c.
3754 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3757 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3758 syslog_duplication set false, log lines that would normally be written
3759 both the the main log and to the reject log were not written to syslog at
3762 PH/14 In the default configuration, change the use of "message" in ACL warn
3763 statements to "add_header".
3765 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3766 not followed by a command (e.g. "seen endif").
3768 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3769 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3772 PH/17 Added extra conditions to the default value of delay_warning_condition
3776 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3777 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3778 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3781 The Auto-Submitted: and various List- headers are standardised, whereas I
3782 don't think Precedence: ever was.
3784 PH/18 Refactored debugging code in route_finduser() to show more information,
3785 in particular, the error code if getpwnam() issues one.
3787 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3788 This is apparently needed in addition to the PH/07 change above to avoid
3789 any possible encoding problems.
3791 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3792 but not after initializing Perl.
3794 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3795 output them only if debugging. By default they are written stderr,
3796 apparently, which is not desirable.
3798 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3801 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3804 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3806 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3807 authenticated or an ident call has been made. Suppress the default
3808 values for $authenticated_id and $authenticated_sender (but permit -oMai
3809 and -oMas) when testing with -bh.
3811 PH/24 Re-jigged the order of the tests in the default configuration so that the
3812 tests for valid domains and recipients precede the DNS black list and CSA
3813 tests, on the grounds that those ones are more expensive.
3815 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3816 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3817 This bug exists in every version of Exim that I still have, right back to
3820 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3821 However, an attempt to turn on case-sensitivity in a regex key by
3822 including (?-i) didn't work because the subject string was already
3823 lowercased, and the effects were non-intuitive. It turns out that a
3824 one-line patch can be used to allow (?-i) to work as expected.
3830 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3831 other effects) broke the use of negated acl sub-conditions.
3833 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3836 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3837 "Deny" causes Exim to reject the incoming connection with a 554 error.
3838 Unfortunately, if there is a major crisis, such as a disk failure,
3839 tcp-wrappers gives "deny", whereas what one would like would be some
3840 kind of temporary error. A kludge has been added to help with this.
3841 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3842 554 error is used if errno is still zero or contains ENOENT (which occurs
3843 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3846 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3848 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3849 errors. Otherwise a message that provokes a temporary error (when other
3850 messages do not) can cause a whole host to time out.
3852 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3853 addresses were routed directly to files or pipes from a redirect router.
3854 File deliveries just didn't batch; pipe deliveries might have suffered
3857 PH/06 A failure to get a lock for a hints database would erroneously always say
3858 "Failed to get write lock", even when it was really a read lock.
3860 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3861 of 0600. This has been changed to use the value of the lockfile_mode
3862 option (which defaults to 0600).
3864 PH/08 Applied small patch from the Sieve maintainer.
3866 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3867 folder from quota calculations, a direct delivery into this folder messed
3868 up the contents of the maildirsize file. This was because the regex was
3869 used only to exclude .Trash (or whatever) when the size of the mailbox
3870 was calculated. There was no check that a delivery was happening into an
3871 excluded directory. This bug has been fixed by ignoring all quota
3872 processing for deliveries into excluded directories.
3874 PH/10 Added the maildirfolder_create_regex option to appendfile.
3880 PH/01 The code for finding all the local interface addresses on a FreeBSD
3881 system running IPv6 was broken. This may well have applied to all BSD
3882 systems, as well as to others that have similar system calls. The broken
3883 code found IPv4 interfaces correctly, but gave incorrect values for the
3884 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3885 that it would not match correctly against @[] and not recognize the IPv6
3888 PH/02 The ipliteral router was not recognizing addresses of the form user@
3889 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3891 PH/03 Added disable_ipv6.
3893 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3894 decoded form, because it is most often used to construct To: headers
3895 lines in autoreplies, and the decoded form may well be syntactically
3896 invalid. However, $reply_address has leading white space removed, and all
3897 newlines turned into spaces so that the autoreply transport does not
3900 PH/05 If group was specified without a user on a router, and no group or user
3901 was specified on a transport, the group from the router was ignored.
3903 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3904 for visible compile-time settings that can be used to change these
3905 numbers, for those that want even more. Backwards compatibility with old
3906 spool files has been maintained. However, going back to a previous Exim
3907 release will lost any variables that are in spool files.
3909 PH/07 Two small changes when running in the test harness: increase delay when
3910 passing a TCP/IP connection to a new process, in case the original
3911 process has to generate a bounce, and remove special handling of
3912 127.0.0.2 (sic), which is no longer necessary.
3914 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3915 be the same on different OS.
3917 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3920 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3921 whether --show-vars was specified or not
3923 JJ/02 exipick: Added support for new ACL variable spool format introduced
3926 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3927 syntactically invalid From: or Reply-to: line, and a filter used this to
3928 generate an autoreply, and therefore failed to obtain an address for the
3929 autoreply, Exim could try to deliver to a non-existent relative file
3930 name, causing unrelated and misleading errors. What now happens is that
3931 it logs this as a hard delivery error, but does not attempt to create a
3934 PH/11 The exinext utility has a -C option for testing purposes, but although
3935 the given file was scanned by exinext itself; it wasn't being passed on
3936 when Exim was called.
3938 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3939 an end-of-file indication when reading a command response.
3941 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3942 compiled. In many other places in Exim, IPv6 addresses are always
3943 recognized, so I have changed this. It also means that IPv4 domain
3944 literals of the form [IPV4:n.n.n.n] are now always recognized.
3946 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3947 used if the router is not running as root, for example, when verifying at
3948 ACL time, or when using -bh. The debugging output from this situation was
3949 non-existent - all you got was a failure to exec. I have made two
3952 (a) Failures to set uid/gid, the current directory, or a process leader
3953 in a subprocess such as that created by queryprogram now generate
3954 suitable debugging output when -d is set.
3956 (b) The queryprogram router detects when it is not running as root,
3957 outputs suitable debugging information if -d is set, and then runs
3958 the subprocess without attempting to change uid/gid.
3960 PH/15 Minor change to Makefile for building test_host (undocumented testing
3963 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3964 additional section of a DNS packet that returns MX or SRV records.
3965 Instead, it always explicitly searches for A/AAAA records. This avoids
3966 major problems that occur when a DNS server includes only records of one
3967 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3968 fixed another bug: if SRV records were looked up and the corresponding
3969 address records were *not* found in the additional section, the port
3970 values from the SRV records were lost.
3972 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3973 using the correct key (the original address) when searching the retry
3974 rules in order to find which one to use for generating the retry hint.
3976 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3977 adding the default one. Similarly, if it contains a Reply-To: header, the
3978 errors_reply_to option, if set, is not used.
3980 PH/19 When calculating a retry time, Exim used to measure the "time since
3981 failure" by looking at the "first failed" field in the retry record. Now
3982 it does not use this if it is later than than the arrival time of the
3983 message. Instead it uses the arrival time. This makes for better
3984 behaviour in cases where some deliveries succeed, thus re-setting the
3985 "first failed" field. An example is a quota failure for a huge message
3986 when small messages continue to be delivered. Without this change, the
3987 "time since failure" will always be short, possible causing more frequent
3988 delivery attempts for the huge message than are intended.
3989 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3991 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3992 $1, $2, $3) because the numerical variables can be reset during some
3993 expansion items (e.g. "match"), thereby losing the authentication data.
3995 PH/21 Make -bV show the size of off_t variables so that the test suite can
3996 decide whether to run tests for quotas > 2G.
3998 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3999 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4000 filecount value is greater than 2G or if a quota value is greater than 2G
4001 on a system where the size of off_t is not greater than 4, a panic error
4004 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4005 never match. The debug and -bh output now contains an explicit error
4006 message indicating a malformed IPv4 address or mask.
4008 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4009 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4010 PH/23 above applies.
4012 PH/25 Do not write to syslog when running in the test harness. The only
4013 occasion when this arises is a failure to open the main or panic logs
4014 (for which there is an explicit test).
4016 PH/26 Added the /no_tell option to "control=freeze".
4018 PH/27 If a host name lookup failed very early in a connection, for example, if
4019 the IP address matched host_lookup and the reverse lookup yielded a name
4020 that did not have a forward lookup, an error message of the form "no IP
4021 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4022 could be logged. Now it outputs the IP address instead of "NULL".
4024 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4025 allows the sender and the authenticated sender to be set when
4026 submitting a message from within Exim. Since child_open_exim() is
4027 documented for local_scan(), the new function should be too.
4029 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4030 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4031 results in an empty string is now treated as unset.
4033 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4035 PH/31 Added sender_verify_fail logging option.
4037 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4038 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4039 tidied the source and removed it altogether.
4041 PH/33 When a queue run was abandoned because the load average was too high, a
4042 log line was always written; now it is written only if the queue_run log
4043 selector is set. In addition, the log line for abandonment now contains
4044 information about the queue run such as the pid. This is always present
4045 in "start" and "stop" lines but was omitted from the "abandon" line.
4047 PH/34 Omit spaces between a header name and the colon in the error message that
4048 is given when verify = headers_syntax fails (if there are lots of them,
4049 the message gets confusing).
4051 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4052 names, as there are now some PTR records that contain slashes. This check
4053 is only to protect against broken name servers that fall over on strange
4054 characters, so the fact that it applies to all lookups doesn't matter.
4056 PH/36 Now that the new test suite is complete, we can remove some of the
4057 special code in Exim that was needed for the old test suite. For example,
4058 sorting DNS records because real resolvers return them in an arbitrary
4059 order. The new test suite's fake resolver always returns records in the
4062 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4063 bounces) except when queue_only is set, to avoid logging races between
4064 the different processes.
4066 PH/38 Panic-die if .include specifies a non-absolute path.
4068 PH/39 A tweak to the "H" retry rule from its user.
4070 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4071 a label. They prevented compilation on older perls.
4073 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4074 a warning to be raised on newish perls.
4076 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4077 on queue. Changes to match documented behaviour of showing count of
4078 messages matching specified criteria.
4080 PH/40 Changed the default ident timeout from 30s to 5s.
4082 PH/41 Added support for the use of login_cap features, on those BSD systems
4083 that have them, for controlling the resources used by pipe deliveries.
4085 PH/42 The content-scanning code uses fopen() to create files in which to put
4086 message data. Previously it was not paying any attention to the mode of
4087 the files. Exim runs with umask(0) because the rest of the code creates
4088 files with open(), and sets the required mode explicitly. Thus, these
4089 files were ending up world-writeable. This was not a big issue, because,
4090 being within the spool directory, they were not world-accessible. I have
4091 created a function called modefopen, which takes an additional mode
4092 argument. It sets umask(777), creates the file, chmods it to the required
4093 mode, then resets the umask. All the relevant calls to fopen() in the
4094 content scanning code have been changed to use this function.
4096 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4097 to 24 hours. This avoids potential overflow problems when processing G
4098 and H retry rules. I suspect nobody ever tinkers with this value.
4100 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4102 PH/45 When the plaintext authenticator is running as a client, the server's
4103 challenges are checked to ensure they are valid base64 strings. By
4104 default, the authentication attempt is cancelled if an invalid string is
4105 received. Setting client_ignore_invalid_base64 true ignores these errors.
4106 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4107 they are received. Thus, the responses can be made to depend on the
4108 challenges. If an invalid string is ignored, an empty string is placed in
4111 PH/46 Messages that are created by the autoreply transport now contains a
4112 References: header, in accordance with RFCs 2822 and 3834.
4114 PH/47 Added authenticated_sender_force to the smtp transport.
4116 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4118 PH/49 Installed latest patch from the Sieve maintainer.
4120 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4121 was also set, the appendfile transport was unnecessarily scanning a
4122 directory of message files (e.g. for maildir delivery) to find the count
4123 of files (along with the size), even though it did not need this
4124 information. It now does the scan only if it needs to find either the
4125 size of the count of files.
4127 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4129 PH/52 Two bugs concerned with error handling when the smtp transport is
4132 (i) Exim was not creating retry information for temporary errors given
4133 for individual recipients after the DATA command when the smtp transport
4134 was used in LMTP mode. This meant that they could be retried too
4135 frequently, and not timed out correctly.
4137 (ii) Exim was setting the flag that allows error details to be returned
4138 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4139 recipients that were returned after the DATA command.
4141 PH/53 This is related to PH/52, but is more general: for any failing address,
4142 when detailed error information was permitted to be returned to the
4143 sender, but the error was temporary, then after the final timeout, only
4144 "retry timeout exceeded" was returned. Now it returns the full error as
4145 well as "retry timeout exceeded".
4147 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4148 do this, and (what is worse) MTAs that accept it.
4150 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4151 will now be deprecated.
4153 PH/56 New os.c-cygwin from the Cygwin maintainer.
4155 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4156 formats (previously only available in exim formats via -bpr, -bpru,
4157 and -bpra. Now also available in native and exiqgrep formats)
4159 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4160 with very large, slow to parse queues
4162 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4164 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4166 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4167 responses to authentication challenges, though it was showing the
4168 challenges; (ii) I've removed the CR characters from the debug output for
4171 PH/58 Allow for the insertion of a newline as well as a space when a string
4172 is turned into more than one encoded-word during RFC 2047 encoding. The
4173 Sieve code now uses this.
4175 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4176 data_4xx, lost_connection, tls_required.
4178 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4179 message was being sent as an SMTP response.
4181 PH/61 Add -l and -k options to exicyclog.
4183 PH/62 When verifying, if an address was redirected to one new address, so that
4184 verification continued, and the new address failed or deferred after
4185 having set something in $address_data, the value of $address_data was not
4186 passed back to the ACL. This was different to the case when no
4187 redirection occurred. The value is now passed back in both cases.
4189 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4190 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4191 use login_cap.h, so on its own it isn't the distinguishing feature. The
4192 new name refers directly to the setclassresources() function.
4194 PH/65 Added configuration files for NetBSD3.
4196 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4198 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4199 is preferred over IPv4.
4201 PH/68 The bounce_return_message and bounce_return_body options were not being
4202 honoured for bounces generated during the reception of non-SMTP messages.
4203 In particular, this applied to messages rejected by the ACL. This bug has
4204 been fixed. However, if bounce_return_message is true and bounce_return_
4205 body is false, the headers that are returned for a non-SMTP message
4206 include only those that have been read before the error was detected.
4207 (In the case of an ACL rejection, they have all been read.)
4209 PH/69 The HTML version of the specification is now built in a directory called
4210 spec_html instead of spec.html, because the latter looks like a path with
4211 a MIME-type, and this confuses some software.
4213 PH/70 Catch two compiler warnings in sieve.c.
4215 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4216 function verify_get_ident() calls ip_connect() to connect a socket, but
4217 if the "connect()" function timed out, ip_connect() used to close the
4218 socket. However, verify_get_ident() also closes the socket later, and in
4219 between Exim writes to the log, which may get opened at this point. When
4220 the socket was closed in ip_connect(), the log could get the same file
4221 descriptor number as the socket. This naturally causes chaos. The fix is
4222 not to close the socket in ip_connect(); the socket should be closed by
4223 the function that creates it. There was only one place in the code where
4224 this was missing, in the iplookup router, which I don't think anybody now
4225 uses, but I've fixed it anyway.
4227 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4228 well as to direct DNS lookups. Otherwise the handling of names in host
4229 lists is inconsistent and therefore confusing.
4235 PH/01 Two changes to the default runtime configuration:
4237 (1) Move the checks for relay_from_hosts and authenticated clients from
4238 after to before the (commented out) DNS black list checks.
4240 (2) Add control=submission to the relay_from_hosts and authenticated
4241 clients checks, on the grounds that messages accepted by these
4242 statements are most likely to be submissions.
4244 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4246 (1) Generate an error if the third argument for the ${prvs expansion is
4249 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4252 (3) Reset the variables that are obtained from the first argument of
4253 ${prvscheck and used in the second argument before leaving the code,
4254 because their memory is reclaimed, so using them afterwards may do
4257 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4258 one (it's much easier than Tom thought :-).
4260 (5) Because of (4), we can now allow for the use of $prvscheck_result
4261 inside the third argument.
4263 PH/03 For some reason, the default setting of PATH when running a command from
4264 a pipe transport was just "/usr/bin". I have changed it to
4267 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4268 anything to be listed in the output from -bV.
4270 PH/05 When a filter generated an autoreply, the entire To: header line was
4271 quoted in the delivery log line, like this:
4273 => >A.N.Other <ano@some.domain> <original@ddress> ...
4275 This has been changed so that it extracts the operative address. There
4276 may be more than one such address. If so, they are comma-separated, like
4279 => >ano@some.domain,ona@other.domain <original@ddress> ...
4281 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4282 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4283 not being looked up in the rDNS to get a host name, Exim was showing the
4284 IP address twice in Received: lines, even though the IP addresses were
4285 identical. For example:
4287 Received: from [1.2.3.4] (helo=[1.2.3.4])
4289 However, if the real host name was known, it was omitting the HELO data
4290 if it matched the actual IP address. This has been tidied up so that it
4291 doesn't show the same IP address twice.
4293 PH/07 When both +timestamp and +memory debugging was on, the value given by
4294 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4295 called by the debug printing, thereby overwriting the timestamp buffer.
4296 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4298 PH/08 When the original message was included in an autoreply transport, it
4299 always said "this is a copy of the message, including all the headers",
4300 even if body_only or headers_only was set. It now gives an appropriate
4303 PH/09 Applied a patch from the Sieve maintainer which:
4305 o fixes some comments
4306 o adds the (disabled) notify extension core
4307 o adds some debug output for the result of if/elsif tests
4308 o points to the current vacation draft in the documentation
4309 and documents the missing references header update
4313 o fixes a bug in processing the envelope test (when testing
4314 multiple envelope elements, the last element determined the
4317 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4318 Electronic Mail") by including:
4320 Auto-submitted: auto-generated
4322 in the messages that it generates (bounce messages and others, such as
4323 warnings). In the case of bounce messages for non-SMTP messages, there was
4324 also a typo: it was using "Auto_submitted" (underscore instead of
4325 hyphen). Since every message generated by Exim is necessarily in response
4326 to another message, thes have all been changed to:
4328 Auto-Submitted: auto-replied
4330 in accordance with these statements in the RFC:
4332 The auto-replied keyword:
4334 - SHOULD be used on messages sent in direct response to another
4335 message by an automatic process,
4337 - MUST NOT be used on manually-generated messages,
4339 - MAY be used on Delivery Status Notifications (DSNs) and Message
4340 Disposition Notifications (MDNs),
4342 - MUST NOT be used on messages generated by automatic or periodic
4343 processes, except for messages which are automatic responses to
4346 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4347 to the default Received: header definition.
4349 PH/12 Added log selector acl_warn_skipped (default on).
4351 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4352 variables because (a) they are in the wrong storage pool and (b) even if
4353 they were copied, it wouldn't work properly because of the caching.
4355 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4356 checking when decoding. Apparently there are clients that generate
4357 overlong encoded strings. Why am I not surprised?
4359 PH/15 If the first argument of "${if match_address" was not empty, but did not
4360 contain an "@" character, Exim crashed. Now it writes a panic log message
4361 and treats the condition as false.
4363 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4365 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4366 extension "envelope-auth". The code is finished and in agreement with
4367 other implementations, but there is no documentation so far and in fact,
4368 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4369 not changing the active code.
4371 Print executed "if" and "elsif" statements when debugging is used. This
4372 helps a great deal to understand what a filter does.
4374 Document more things not specified clearly in RFC3028. I had all this
4375 sorted out, when out of a sudden new issues came to my mind. Oops."
4377 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4380 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4381 canonical form (as documented). However, after a host name lookup from
4382 the IP address, check_host() was doing a simple string comparison with
4383 addresses acquired from the DNS when checking that the found name did
4384 have the original IP as one of its addresses. Since any found IPv6
4385 addresses are likely to be in abbreviated form, the comparison could
4386 fail. Luckily, there already exists a function for doing the comparison
4387 by converting both addresses to binary, so now that is used instead of
4388 the text comparison.
4390 PH/20 There was another similar case to PH/19, when a complete host name was
4391 given in a host list; looking up its IP address could give an abbreviated
4392 form, whereas the current host's name might or might not be abbreviated.
4393 The same fix has been applied.
4399 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4400 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4403 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4404 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4406 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4408 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4409 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4410 Also, for "Auto-Submitted", treat anything other than "no" as
4411 non-personal, in accordance with RFC 3834. (Previously it treated
4412 anything starting "auto-" as non-personal.)
4414 TF/01 The control=submission/name=... option had a problem with syntax
4415 errors if the name included a slash character. The /name= option
4416 now slurps the rest of the string, so it can include any characters
4417 but it must come last in the list of options (after /sender_retain
4420 PH/05 Some modifications to the interface to the fake nameserver for the new
4428 TK/01 Added the "success_on_redirect" address verification option. See
4429 NewStuff for rationale and an example.
4431 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4433 PH/02 Patch to exigrep to allow it to work on syslog lines.
4435 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4436 fread() to skip over the body file's header line, because in Cygwin the
4437 header line is locked and is inaccessible.
4439 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4440 co-exist for some time) to make it clear that it is the Exim ID that is
4441 referenced, not the Message-ID: header line.
4443 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4444 string_format() function, because snprintf() does not exist on all
4447 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4448 ${stat: expansion item.
4450 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4451 protocol synchronization error", to keep the pedants happy.
4453 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4454 well as for IRIX systems, when gcc is being used. See the host.c source
4457 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4459 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4462 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4463 transport and to the smtp transport in LMTP mode.
4465 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4467 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4468 run a filter in a subprocess. This could lead to confusion in subsequent
4469 lookups in the parent process. There should also be a search_tidyup() at
4470 the end of the subprocess.
4472 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4473 only if the host matched helo_try_verify_hosts, which caused the
4474 verification to occur when the EHLO/HELO command was issued. The ACL just
4475 tested the remembered result. Now, if a previous verification attempt has
4476 not happened, "verify = helo" does it there and then.
4478 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4480 TK/03 Fix log output including CR from clamd.
4482 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4483 exist provoked a memory error which could cause a segfault.
4485 PH/15 Installed PCRE 6.2
4487 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4489 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4490 of the problem. Specifically, suggested +O2 rather than +O1 for the
4493 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4495 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4496 record, the port was not correctly logged when the outgoing_port log
4497 selector was set (it logged the transort's default port).
4499 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4500 fallback_hosts, and "hosts" in the smtp transport.
4502 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4503 host errors such as "Connection refused".
4505 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4506 authentication with radiusclient 0.4.9:
4508 - Error returned from rc_read_config was caught wrongly
4509 - Username/password not passed on to radius server due to wrong length.
4511 The presumption is that some radiusclient API changes for 4.51/PH/17
4512 were not taken care of correctly. The code is still untested by me (my
4513 Linux distribution still has 0.3.2 of radiusclient), but it was
4514 contributed by a Radius user.
4516 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4517 expanding the "port" option of the smtp transport.
4519 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4520 while reading a MIME header. Thanks to Tom Hughes for a patch.
4522 PH/24 Include config.h inside local_scan.h so that configuration settings are
4525 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4526 that in an ACL for RCPT (for example), you can examine exactly what was
4529 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4530 commands, but it was not correctly comparing the address with the actual
4531 client host address. Thus, it would show the EHLO address in Received:
4532 header lines when this was not necessary.
4534 PH/27 Added the % operator to ${eval:}.
4536 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4537 it should be ignoring failures (because with -C, for example, it has lost
4538 privilege). It wasn't ignoring creation failures other than "already
4541 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4544 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4545 because an input line was too long, either on its own, or by virtue of
4546 too many continuations, the temporary file was not being removed, and the
4547 return code was incorrect.
4549 PH/31 Missing "BOOL" in function definition in filtertest.c.
4551 PH/32 Applied Sieve patches from the maintainer.
4553 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4555 PH/33 Added "verify = not_blind".
4557 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4558 Local/Makefile (with some defaults set). These are used in built scripts
4559 such as exicyclog, but they have never been used in the exim_install
4560 script (though there are many overriding facilities there). I have
4561 arranged that the exim_install script now takes note of these two
4564 PH/35 Installed configuration files for Dragonfly.
4566 PH/36 When a locally submitted message by a trusted user did not contain a
4567 From: header, and the sender address was obtained from -f or from an SMTP
4568 MAIL command, and the trusted user did not use -F to supply a sender
4569 name, $originator_name was incorrectly used when constructing a From:
4570 header. Furthermore, $originator_name was used for submission mode
4571 messages from external hosts without From: headers in a similar way,
4572 which is clearly wrong.
4574 PH/37 Added control=suppress_local_fixups.
4576 PH/38 When log_selector = +received_sender was set, and the addition of the
4577 sender made the log line's construction buffer exactly full, or one byte
4578 less than full, an overflow happened when the terminating "\n" was
4581 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4582 when the result of a list match is failure because a DNS lookup failed.
4584 PH/40 RM_COMMAND is now used in the building process.
4586 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4587 the "build-* directories that it finds.
4589 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4590 address in a domain literal was a prefix of an interface address.
4592 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4593 when verifying a sender address, unless rewrite_headers is false.
4595 PH/44 Wrote a long comment about why errors_to addresses are verified as
4596 recipients, not senders.
4598 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4599 the ratelimit ACL was added.
4601 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4603 PH/46 Added extra information about PostgreSQL errors to the error string.
4605 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4606 avoiding the need to install special zones in a real server. This is
4607 backwards compatible; if it can't find the fake resolver, it drops back.
4608 Thus, both old and new test suites can be run.
4610 TF/02 Added util/ratelimit.pl
4612 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4613 clock is set back in time.
4615 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4616 Candler <B.Candler@pobox.com>.
4618 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4619 OK for addresses that are the result of redirections.
4621 PH/48 A number of further additions for the benefit of the new test suite,
4622 including a fake gethostbyname() that interfaces to the fake DNS resolver
4625 TF/06 The fix for widen_domains has also been applied to qualify_single and
4626 search_parents which are the other dnslookup options that can cause
4629 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4632 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4634 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4635 service" if it encountered an MX record with an empty target hostname.
4636 The message is now "an MX or SRV record indicated no SMTP service".
4638 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4639 if the DNS of the sending site is misconfigured. This is quite a
4640 common situation. This change restores the behaviour of treating a
4641 helo verification defer as a failure.
4643 PH/51 If self=fail was set on a router, the bounce message did not include the
4644 actual error message.
4650 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4652 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4653 command itself ended in error, the underlying message about the transport
4654 filter timeout was being overwritten with the pipe command error. Now the
4655 underlying error message should be appended to the second error message.
4657 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4659 PH/02 Reduce the amount of output that "make" produces by default. Full output
4660 can still be requested.
4662 PH/03 The warning log line about a condition test deferring for a "warn" verb
4663 was being output only once per connection, rather than after each
4664 occurrence (because it was using the same function as for successful
4665 "warn" verbs). This seems wrong, so I have changed it.
4667 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4668 it should not have, which might have caused a crash in the right
4669 circumstances, but probably never did.
4671 PH/04 Installed a modified version of Tony Finch's patch to make submission
4672 mode fix the return path as well as the Sender: header line, and to
4673 add a /name= option so that you can make the user's friendly name appear
4676 TF/03 Added the control = fakedefer ACL modifier.
4678 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4679 Mark Lowes for thorough testing.
4681 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4683 TK/03 Merged latest SRS patch from Miles Wilton.
4685 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4686 with the definition in sysexits.h (which is #included earlier).
4687 Fortunately, Exim does not actually use EX_OK. The code used to try to
4688 preserve the sysexits.h value, by assuming that macro definitions were
4689 scanned for macro replacements. I have been disabused of this notion,
4690 so now the code just undefines EX_OK before #including unistd.h.
4692 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4693 in the smtp transport. When a block could not be written in a single
4694 write() function, the timeout was being re-applied to each part-write.
4695 This seems wrong - if the receiver was accepting one byte at a time it
4696 would take for ever. The timeout is now adjusted when this happens. It
4697 doesn't have to be particularly precise.
4699 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4700 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4702 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4703 without a domain if the check to <postmaster@domain> fails.
4705 SC/01 Eximstats: added -xls and the ability to specify output files
4706 (patch written by Frank Heydlauf).
4708 SC/02 Eximstats: use FileHandles for outputting results.
4710 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4712 SC/04 Eximstats: fixed display of large numbers with -nvr option
4714 SC/05 Eximstats: fixed merging of reports with empty tables.
4716 SC/06 Eximstats: added the -include_original_destination flag
4718 SC/07 Eximstats: removed tabs and trailing whitespace.
4720 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4722 TK/06 MBOX spool code: Add real "From " MBOX separator line
4723 so the .eml file is really in mbox format (even though
4724 most programs do not really care). Patch from Alex Miller.
4726 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4727 The latter is generated from $received_to and is only set if the
4728 message has one envelope recipient. SA can use these headers,
4729 obviously out-of-the-box. Patch from Alex Miller.
4731 PH/08 The ${def test on a variable was returning false if the variable's
4732 value was "0", contrary to what the specification has always said!
4733 The result should be true unless the variable is empty.
4735 PH/09 The syntax error of a character other than { following "${if
4736 def:variable_name" (after optional whitespace) was not being diagnosed.
4737 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4738 accidental colon was present, for example, could give incorrect results.
4740 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4741 result is used (not including appendfile, where other changes are about
4744 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4745 This involved changing a lot of size variables from int to off_t. It
4746 should work with maildirs and everything.
4748 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4749 spamd dying while we are connected to it.
4751 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4754 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4755 function for BDB 4.3.
4757 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4759 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4760 the sources for PCRE have been re-arranged and more files are now
4763 PH/15 The code I had for printing potentially long long variables in PH/11
4764 above was not the best (it lost precision). The length of off_t variables
4765 is now inspected at build time, and an appropriate printing format (%ld
4766 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4767 to be "long long int" or "long int". This is needed for the internal
4768 formatting function string_vformat().
4770 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4771 the configuration file to be ":syslog", then the script "guesses" where
4772 the logs files are, rather than using the compiled in default. In our
4773 case the guess is not the same as the compiled default, so the script
4774 suddenly stopped working when I started to use syslog. The patch checks
4775 to see if log_file_path is "". If so, it attempts to read it from exim
4776 with no configuration file to get the compiled in version, before it
4777 falls back to the previous guessing code."
4779 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4780 implementing BATV in an Exim configuration. See NewStuff for the gory
4783 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4784 Makefile that are specific to HP-UX.
4786 PH/18 If the "use_postmaster" option was set for a recipient callout together
4787 with the "random" option, the postmaster address was used as the MAIL
4788 FROM address for the random test, but not for the subsequent recipient
4789 test. It is now used for both.
4791 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4792 patch removes a few documentation additions to RFC 3028, because the
4793 latest draft now contains them. It adds the new en;ascii-case comparator
4794 and a new error check for 8bit text in MIME parts. Comparator and
4795 require names are now matched exactly. I enabled the subaddress
4796 extension, but it is not well tested yet (read: it works for me)."
4798 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4799 rework some of the code of TK/09 above to avoid the hardwired use of
4800 "%lld" and "long long". Replaced the call to snprintf() with a call to
4803 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4804 records point to non-existent hosts", "retry timeout exceeded", and
4805 "retry time not reached for any host after a long failure period".
4807 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4808 experimental DomainKeys support:
4810 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4811 (2) On an error such as an illegally used "control", the wrong name for
4812 the control was given.
4814 These problems did NOT occur unless DomainKeys support was compiled.
4816 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4818 PH/24 Added ${if match_ip condition.
4820 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4821 databases so that it will be absolutely obvious if a crash occurs in the
4822 DB library. This is a regular occurrence (often caused by mis-matched
4825 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4826 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4827 picked up on a user's system that detects such things. There doesn't seem
4828 to be a gcc warning option for this - only an attribute that has to be
4829 put on the function's prototype. It seems that in Fedora Core 4 they have
4830 set this on a number of new functions. No doubt there will be more in due
4833 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4834 specify a transport. However, if an address that was verified by such a
4835 router was the subject of a callout, Exim crashed because it tried to
4836 read the rcpt_include_affixes from the non-existent transport. Now it
4837 just assumes that the setting of that option is false. This bug was
4838 introduced by 4.51/PH/31.
4840 PH/28 Changed -d+all to exclude +memory, because that information is very
4841 rarely of interest, but it makes the output a lot bigger. People tend to
4842 do -d+all out of habit.
4844 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4845 code in os-type was giving problems when libc.so lives in lib64, like on
4848 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4849 aren't the modern standard, and it seems that some systems' include files
4850 don't always have them. Exim was already checking for some of the newer
4851 ones like T_AAAA, and defining it itself. I've added checks for all the
4852 record types that Exim uses.
4854 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4855 not automatically generating a new one, as it is supposed to. This
4856 prevented TLS from working. If the file did exist, but contained invalid
4857 data, a new version was generated, as expected. It was only the case of a
4858 non-existent file that was broken.
4860 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4861 with a change in libdomainkeys > 0.64.
4863 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4864 from DNS. If the selector record carries the flag, it now has
4865 precedence over the domain-wide flag.
4867 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4869 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4870 the use of a port name that isn't defined in /etc/services) occurred, the
4871 message was deferred as in a normal delivery, and thus remained on the
4872 spool, instead of being failed because of the mua_wrapper setting. This
4873 is now fixed, and I tidied up some of the mua_wrapper messages at the
4876 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4877 the messages in a hash of arrays rather than using individual hashes.
4878 This is a bit cleaner and results in dramatic memory savings, albeit
4879 at a slight CPU cost.
4881 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4882 as requested by Marc Sherman.
4884 SC/10 Eximstats: added histograms for user specified patterns as requested
4887 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4889 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4890 fopen() in the content-scanning modules that did not already have it.
4896 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4897 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4899 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4901 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4903 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4904 to test Sieve filters that use "vacation".
4906 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4907 that changes the way the GnuTLS parameters are stored in the cache file.
4908 The new format can be generated externally. For backward compatibility,
4909 if the data in the cache doesn't make sense, Exim assumes it has read an
4910 old-format file, and it generates new data and writes a new file. This
4911 means that you can't go back to an older release without removing the
4914 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4915 work if there are any delivery delays because "one_time" forces the
4916 parent to be marked "delivered", so its unseen clone is never tried
4917 again. For this reason, Exim now forbids the simultaneous setting of
4920 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4921 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4922 entry if you want to know the details. The fix, however, neglected to
4923 consider the case when local delivery batching is involved. The test for
4924 "previously delivered" was not happening when checking to see if an
4925 address could be batched with a previous (undelivered) one; under
4926 certain circumstances this could lead to multiple deliveries to the same
4929 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4930 in its include files, and this causes problems building Exim.
4932 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4933 header_syntax) but Exim was just ignoring anything given after a slash.
4934 In particular, this caused confusion with an attempt to use "verify =
4935 reverse_host_lookup/defer_ok". An error is now given when options are
4936 supplied for verify items that do not have them. (Maybe reverse_host_
4937 lookup should have a defer_ok option, but that's a different point.)
4939 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4940 defined by RFC 821) to 2048, because there were problems with some AUTH
4941 commands, and RFC 1869 says the size should be increased for extended
4942 SMTP commands that take arguments.
4944 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4947 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4948 "unknown" error; now it says that the functionality isn't in the binary.
4950 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4951 an address' error message when a string expansion fails (syntax or
4952 whatever). Otherwise the password may appear in the log. Following change
4953 PH/42 below, there is no longer a chance of it appearing in a bounce
4956 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4958 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4959 of its addresses was ever tried. (Bugzilla bug #2.)
4961 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4962 the result incorrectly in the debug output. (It correctly added a newline
4963 to what was transported.)
4965 TF/01 Added $received_time.
4967 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4968 commented out examples of how to interface to a virus scanner and to
4969 SpamAssassin. Also added commented examples of av_scanner and
4970 spamd_address settings.
4972 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4973 and controls are allowed in which ACLs. There were a couple of minor
4974 errors. Some of the entries in the conditions table (which is a table of
4975 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4976 negation of where the condition IS allowed.
4978 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4980 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4981 header file does not have a version number, so I've had to invent a new
4982 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4983 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4984 radiusclient), but it was contributed by a Radius user.
4986 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4987 files or directories, for OpenSSL.
4989 PH/19 When an Exim process that is running as root has to create an Exim log
4990 file, it does so in a subprocess that runs as exim:exim so as to get the
4991 ownership right at creation (otherwise, other Exim processes might see
4992 the file with the wrong ownership). There was no test for failure of this
4993 fork() call, which would lead to the process getting stuck as it waited
4994 for a non-existent subprocess. Forks do occasionally fail when resources
4995 run out. I reviewed all the other calls to fork(); they all seem to check
4998 PH/20 When checking for unexpected SMTP input at connect time (before writing
4999 the banner), Exim was not dealing correctly with a non-positive return
5000 from the read() function. If the client had disconnected by this time,
5001 the result was a log entry for a synchronization error with an empty
5002 string after "input=" when read() returned zero. If read() returned -1
5003 (an event I could not check), uninitialized data bytes were printed.
5004 There were reports of junk text (parts of files, etc) appearing after
5007 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5009 PH/22 Added support for macro redefinition, and (re)definition in between
5010 driver and ACL definitions.
5012 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5013 forgetting to use the resulting value; it was using the unexpanded value.
5015 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5016 hadn't been configured. The fix is from Juergen Kreileder, who
5017 understands it better than I do:
5019 "Here's what I see happening with three configured cyrus_sasl
5020 authenticators configured (plain, login, cram-md5):
5022 On startup auth_cyrus_sasl_init() gets called for each of these.
5023 This means three calls to sasl_listmech() without a specified mech_list.
5024 => SASL tests which mechs of all available mechs actually work
5025 => three warnings about OTP not working
5026 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5028 With the patch, sasl_listmech() also gets called three times. But now
5029 SASL's mech_list option is set to the server_mech specified in the the
5030 authenticator. Or in other words, the answer from sasl_listmech()
5031 gets limited to just the mech you're testing for (which is different
5033 => the return list contains just 'plain' or 'login', 'cram-md5' or
5034 nothing depending on the value of ob->server_mech.
5036 I've just tested the patch: Authentication still works fine,
5037 unavailable mechs specified in the exim configuration are still
5038 caught, and the auth.log warnings about OTP are gone."
5040 PH/25 When debugging is enabled, the contents of the command line are added
5041 to the debugging output, even when log_selector=+arguments is not
5044 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5045 answer is "GNU", and only if the return is "GNU/something" is the answer
5048 PH/27 $acl_verify_message is now set immediately after the failure of a
5049 verification in an ACL, and so is available in subsequent modifiers. In
5050 particular, the message can be preserved by coding like this:
5052 warn !verify = sender
5053 set acl_m0 = $acl_verify_message
5055 Previously, $acl_verify_message was set only while expanding "message"
5056 and "log_message" when a very denied access.
5058 PH/28 Modified OS/os.c-Linux with
5060 -#ifndef OS_LOAD_AVERAGE
5061 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5063 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5064 nomenclature these days.)
5066 PH/29 Installed patch from the Sieve maintainer that adds the options
5067 sieve_useraddress and sieve_subaddress to the redirect router.
5069 PH/30 In these circumstances:
5070 . Two addresses routed to the same list of hosts;
5071 . First host does not offer TLS;
5072 . First host accepts first address;
5073 . First host gives temporary error to second address;
5074 . Second host offers TLS and a TLS session is established;
5075 . Second host accepts second address.
5076 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5077 and peerdn, if requested) that were in fact used only for the second
5080 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5081 attention to any local part prefix or suffix that was matched by the
5082 router that accepted the address. It now behaves in the same way as it
5083 does for delivery: the affixes are removed from the local part unless
5084 rcpt_include_affixes is set on the transport.
5086 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5087 timeout during the DATA phase of an incoming message.
5089 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5090 applied a patch sanctioned by the Sieve maintainer.
5092 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5093 the uid or gid is negative. A case of a negative gid caused this to be
5094 noticed. The fix allows for either to be negative.
5096 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5097 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5100 PH/36 The change PH/12 above was broken. Fixed it.
5102 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5103 the grounds that routing the same address twice would always produce the
5104 same answer. This might have been true once, but it is certainly no
5105 longer true now. Routing a child address may depend on the previous
5106 routing that produced that child. Some complicated redirection strategies
5107 went wrong when messages had multiple recipients, and made Exim's
5108 behaviour dependent on the order in which the addresses were given.
5110 I have moved the duplicate checking until after the routing is complete.
5111 Exim scans the addresses that are assigned to local and remote
5112 transports, and removes any duplicates. This means that more work will be
5113 done, as duplicates will always all be routed, but duplicates are
5114 presumably rare, so I don't expect this is of any significance.
5116 For deliveries to pipes, files, and autoreplies, the duplicate checking
5117 still happens during the routing process, since they are not going to be
5120 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5121 It corrects a timeout issue with spamd. This is Ian's comment: "The
5122 background is that sometimes spamd either never reads data from a
5123 connection it has accepted, or it never writes response data. The exiscan
5124 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5125 blindly assumes that writes won't block so it may never time out."
5127 PH/39 Allow G after quota size as well as K and M.
5129 PH/40 The value set for $authenticated_id in an authenticator may not contain
5130 binary zeroes or newlines because the value is written to log lines and
5131 to spool files. There was no check on this. Now the value is run through
5132 the string_printing() function so that such characters are converted to
5133 printable escape sequences.
5135 PH/41 $message_linecount is a new variable that contains the total number of
5136 lines in the message. Compare $body_linecount, which is the count for the
5139 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5140 bounce and delay warning messages, except in certain special cases, which
5143 (a) An SMTP error message from a remote host;
5144 (b) A message specified in a :fail: redirection;
5145 (c) A message specified in a "fail" command in a system filter;
5146 (d) A message specified in a FAIL return from the queryprogram router;
5147 (e) A message specified by the cannot_route_message router option.
5149 In these cases only, Exim does include the error details in bounce and
5150 warning messages. There are also a few cases where bland messages such
5151 as "unrouteable address" or "local delivery error" are given.
5153 PH/43 $value is now also set for the "else" part of a ${run expansion.
5155 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5156 being worked on, but at least Exim now implements the latest version to
5159 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5160 process to complete was treated as a delivery failure, a timeout while
5161 writing the message to the pipe was logged, but erroneously treated as a
5162 successful delivery. Such timeouts include transport filter timeouts. For
5163 consistency with the overall process timeout, these timeouts are now
5164 treated as errors, giving rise to delivery failures by default. However,
5165 there is now a new Boolean option for the pipe transport called
5166 timeout_defer, which, if set TRUE, converts the failures into defers for
5167 both kinds of timeout. A transport filter timeout is now identified in
5170 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5171 systems where "make" and "gmake" are different, calling "gmake" at top
5172 level broke things. I've arranged for the value of $(MAKE) to be passed
5173 from the Makefile to this script so that it can call the same version of
5177 A note about Exim versions 4.44 and 4.50
5178 ----------------------------------------
5180 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5181 changes of various kinds. As a consequence, a big documentation update was
5182 needed. This delayed the release for rather longer than seemed good, especially
5183 in the light of a couple of (minor) security issues. Therefore, the changes
5184 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5185 release. So 4.44 and 4.50 are in effect two different branches that both start
5188 I have left the 4.50 change log unchanged; it contains all the changes since
5189 4.43. The change log for 4.44 is below; many of its items are identical to
5190 those for 4.50. This seems to be the most sensible way to preserve the
5191 historical information.
5197 1. Minor wording change to the doc/README.SIEVE file.
5199 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5200 computation of the current number of files was incorrect.
5202 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5203 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5206 4. Give more explanation in the error message when the command for a transport
5207 filter fails to execute.
5209 5. There are several places where Exim runs a non-Exim command in a
5210 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5211 was being done only for the command run by the queryprogram router. It is
5212 now done for all such subprocesses. The other cases are: ${run, transport
5213 filters, and the commands run by the lmtp and pipe transports.
5215 6. Added CONFIGURE_GROUP build-time option.
5217 7. Some older OS have a limit of 256 on the maximum number of file
5218 descriptors. Exim was using setrlimit() to set 1000 as a large value
5219 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5220 systems. I've change it so that if it can't get 1000, it tries for 256.
5222 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5223 was an oversight, and furthermore, ever since the addition of extra
5224 controls (e.g. 4.43/32), the checks on when to allow different forms of
5225 "control" were broken. There should now be diagnostics for all cases when a
5226 control that does not make sense is encountered.
5228 9. Added the /retain_sender option to "control=submission".
5230 10. $recipients is now available in the predata ACL (oversight).
5232 11. Tidy the search cache before the fork to do a delivery from a message
5233 received from the command line. Otherwise the child will trigger a lookup
5234 failure and thereby defer the delivery if it tries to use (for example) a
5235 cached ldap connection that the parent has called unbind on.
5237 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5238 of $address_data from the recipient verification was clobbered by the
5239 sender verification.
5241 13. The value of address_data from a sender verification is now available in
5242 $sender_address_data in subsequent conditions in the ACL statement.
5244 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5246 15. Added a new option "connect=<time>" to callout options, to set a different
5249 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5250 was its contents. (It was OK if the option was not defined at all.)
5252 17. A "Completed" log line is now written for messages that are removed from
5253 the spool by the -Mrm option.
5255 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5256 information about exactly what failed.
5258 19. Added -dd to debug only the daemon process.
5260 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5261 handles timeouts, both on the server side and network timeouts. Renamed the
5262 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5264 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5265 It is now set to "smtps".
5267 22. $host_address is now set to the target address during the checking of
5268 ignore_target_hosts.
5270 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5271 being passed; this would have caused $sender_host_name to have been used if
5272 matching the list had actually called for a host name (not very likely,
5273 since this list is usually IP addresses). A host name is now passed as
5276 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5277 code that specifies a non-restarting handler (typically sigaction() in
5278 modern systems) in an attempt to fix a rare and obscure crash bug.
5280 25. Narrowed the window for a race in the daemon that could cause it to ignore
5281 SIGCHLD signals. This is not a major problem, because they are used only to
5282 wake it up if nothing else does.
5284 26. A malformed maildirsize file could cause Exim to calculate negative values
5285 for the mailbox size or file count. Odd effects could occur as a result.
5286 The maildirsize information is now recalculated if the size or filecount
5289 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5290 support for a long time. Removed HAVE_SYS_VFS_H.
5292 28. Installed the latest version of exipick from John Jetmore.
5294 29. In an address list, if the pattern was not a regular expression, an empty
5295 subject address (from a bounce message) matched only if the pattern was an
5296 empty string. Non-empty patterns were not even tested. This was the wrong
5297 because it is perfectly reasonable to use an empty address as part of a
5298 database query. An empty address is now tested by patterns that are
5299 lookups. However, all the other forms of pattern expect the subject to
5300 contain a local part and a domain, and therefore, for them, an empty
5301 address still always fails if the pattern is not itself empty.
5303 30. Exim went into a mad DNS loop when attempting to do a callout where the
5304 host was specified on an smtp transport, and looking it up yielded more
5305 than one IP address.
5307 31. Re-factored the code for checking spool and log partition space into a
5308 function that finds that data and another that does the check. The former
5309 is then used to implement four new variables: $spool_space, $log_space,
5310 $spool_inodes, and $log_inodes.
5312 32. The RFC2047 encoding function was originally intended for short strings
5313 such as real names; it was not keeping to the 75-character limit for
5314 encoded words that the RFC imposes. It now respects the limit, and
5315 generates multiple encoded words if necessary. To be on the safe side, I
5316 have increased the buffer size for the ${rfc2047: expansion operator from
5319 33. It is now permitted to omit both strings after an "if" condition; if the
5320 condition is true, the result is "true". As before, when the second string
5321 is omitted, a false condition yields an empty string. This makes it less
5322 cumbersome to write custom ACL and router conditions.
5324 34. Failure to deliver a bounce message always caused it to be frozen, even if
5325 there was an errors_to setting on the router. The errors_to setting is now
5328 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5329 canonical form (fully expanded) before being placed in
5330 $sender_host_address.
5332 36. The table in the code that translates DNS record types into text (T_A to
5333 "A" for instance) was missing entries for NS and CNAME. It is just possible
5334 that this could have caused confusion if both these types were looked up
5335 for the same domain, because the text type is used as part of Exim's
5336 per-process caching. But the chance of anyone hitting this buglet seems
5339 37. The dnsdb lookup has been extended in a number of ways.
5341 (1) There is a new type, "zns", which walks up the domain tree until it
5342 finds some nameserver records. It should be used with care.
5344 (2) There is a new type, "mxh", which is like "mx" except that it returns
5345 just the host names, not the priorities.
5347 (3) It is now possible to give a list of domains (or IP addresses) to be
5348 looked up. The behaviour when one of the lookups defers can be
5349 controlled by a keyword.
5351 (4) It is now possible to specify the separator character for use when
5352 multiple records are returned.
5354 38. The dnslists ACL condition has been extended: it is now possible to supply
5355 a list of IP addresses and/or domains to be looked up in a particular DNS
5358 39. Added log_selector=+queue_time_overall.
5360 40. When running the queue in the test harness, wait just a tad after forking a
5361 delivery process, to get repeatability of debugging output.
5363 41. Include certificate and key file names in error message when GnuTLS fails
5364 to set them up, because the GnuTLS error message doesn't include the name
5365 of the failing file when there is a problem reading it.
5367 42. Allow both -bf and -bF in the same test run.
5369 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5371 44. The "Exiscan patch" is now merged into the mainline Exim source.
5373 45. Sometimes the final signoff response after QUIT could fail to get
5374 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5375 < 0 before doing a fflush(). This bug looks as though it goes back to the
5376 introduction of TLS in release 3.20, but "sometimes" must have been rare
5377 because the tests only now provoked it.
5379 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5380 (this can affect the format of dates).
5382 47. exim_tidydb, when checking for the continued existence of a message for
5383 which it has found a message-specific retry record, was not finding
5384 messages that were in split spool directories. Consequently, it was
5385 deleting retry records that should have stayed in existence.
5387 48. Steve fixed some bugs in eximstats.
5389 49. The SPA authentication driver was not abandoning authentication and moving
5390 on to the next authenticator when an expansion was forced to fail,
5391 contradicting the general specification for all authenticators. Instead it
5392 was generating a temporary error. It now behaves as specified.
5394 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5395 (the order specifies the preference for clients). The order is now AES256,
5396 AES128, 3DES, ARCFOUR128.
5398 51. Small patch to Sieve code - explicitly set From: when generating an
5401 52. Exim crashed if a remote delivery caused a very long error message to be
5402 recorded - for instance if somebody sent an entire SpamAssassin report back
5403 as a large number of 550 error lines. This bug was coincidentally fixed by
5404 increasing the size of one of Exim's internal buffers (big_buffer) that
5405 happened as part of the Exiscan merge. However, to be on the safe side, I
5406 have made the code more robust (and fixed the comments that describe what
5409 53. Now that there can be additional text after "Completed" in log lines (if
5410 the queue_time_overall log selector is set), a one-byte patch to exigrep
5411 was needed to allow it to recognize "Completed" as not the last thing in
5414 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5415 patch that reportedly fixes this has been added. I am not expert enough to
5416 create a test for it. This is what the patch creator wrote:
5418 "I found a little strange behaviour of ldap code when working with
5419 Windows 2003 AD Domain, where users was placed in more than one
5420 Organization Units. When I tried to give exim partial DN, the exit code
5421 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5422 But simultaneously result of request was absolutely normal ldap result,
5423 so I produce this patch..."
5425 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5426 REFERENCE, so I have modified the code to exclude the patch when that macro
5429 55. Some experimental protocols are using DNS PTR records for new purposes. The
5430 keys for these records are domain names, not reversed IP addresses. The
5431 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5432 leaves it alone. Component reversal etc. now happens only for IP addresses.
5435 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5437 57. Double the size of the debug message buffer (to 2048) so that more of very
5438 long debug lines gets shown.
5440 58. The exicyclog utility now does better if the number of log files to keep
5441 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5443 59. Two changes related to the smtp_active_hostname option:
5445 (1) $smtp_active_hostname is now available as a variable.
5446 (2) The default for smtp_banner uses $smtp_active_hostname instead
5447 of $primary_hostname.
5449 60. The host_aton() function is supposed to be passed a string that is known
5450 to be a valid IP address. However, in the case of IPv6 addresses, it was
5451 not checking this. This is a hostage to fortune. Exim now panics and dies
5452 if the condition is not met. A case was found where this could be provoked
5453 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5454 components; fortuitously, this particular loophole had already been fixed
5455 by change 4.50/55 above.
5457 If there are any other similar loopholes, the new check in host_aton()
5458 itself should stop them being exploited. The report I received stated that
5459 data on the command line could provoke the exploit when Exim was running as
5460 exim, but did not say which command line option was involved. All I could
5461 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5462 running as the user.
5465 61. There was a buffer overflow vulnerability in the SPA authentication code
5466 (which came originally from the Samba project). I have added a test to the
5467 spa_base64_to_bits() function which I hope fixes it.
5470 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5471 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5473 63. The daemon start-up calls getloadavg() while still root for those OS that
5474 need the first call to be done as root, but it missed one case: when
5475 deliver_queue_load_max is set with deliver_drop_privilege. This is
5476 necessary for the benefit of the queue runner, because there is no re-exec
5477 when deliver_drop_privilege is set.
5479 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5480 This has been fixed.
5482 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5483 was in use, was not putting the data itself into the right store pool;
5484 consequently, it could be overwritten for a subsequent message in the same
5485 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5488 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5490 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5491 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5492 the function were treating the return as a boolean value, which happened to
5493 work because 0=false and not-0=true, but is not correct code.
5495 68. The host_aton() function was not handling scoped IPv6 addresses (those
5496 with, for example, "%eth0" on the end) correctly.
5498 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5499 negated items (that is, ~something) in unsigned ints. Some compilers
5500 apparently mutter when there is no cast.
5502 70. If an address verification called from an ACL failed, and did not produce a
5503 user-specific message (i.e. there was only a "system" message), nothing was
5504 put in $acl_verify_message. In this situation, it now puts the system
5507 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5508 session; change 4.31/43 added the unwanted input to the log line - except
5509 that it did not do this in the start of session case. It now does.
5511 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5512 This is wrong and can cause the other end to generate a synchronization
5513 error if it is another Exim or anything else that does the synchronization
5514 check. A QUIT command is no longer sent after a timeout.
5516 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5517 during host lookups.
5519 74. The defer_ok option of callout verification was not working if it was used
5520 when verifying addresses in header lines, that is, for this case:
5522 verify = header_sender/callout=defer_ok
5524 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5525 those file descriptors could be used for SMTP connections. If anything
5526 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5527 could be sent to the SMTP client, causing chaos. The daemon now opens
5528 stdin, stdout, and stderr to /dev/null when it puts itself into the
5531 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5532 log by default. The user can override this with suitable Perl magic.
5534 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5535 the log message when discard triggers, was not working for the DATA ACL or
5536 for the non-SMTP ACL.
5538 78. Error message wording change in sieve.c.
5540 79. If smtp_accept_max_per_host was set, the number of connections could be
5541 restricted to fewer than expected, because the daemon was trying to set up
5542 a new connection before checking whether the processes handling previous
5543 connections had finished. The check for completed processes is now done
5544 earlier. On busy systems, this bug wouldn't be noticed because something
5545 else would have woken the daemon, and it would have reaped the completed
5548 80. If a message was submitted locally by a user whose login name contained one
5549 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5550 It caused a spool format error. I have fixed the spool reading code. A
5551 related problem was that the "from" clause in the Received: line became
5552 illegal because of the space(s). It is now covered by ${quote_local_part.
5554 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5557 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5559 83. Patch from Sieve maintainer to fix unterminated string problem in
5560 "vacation" handling.
5562 84. Some minor changes to the Linux configuration files to help with other
5563 OS variants using glibc.
5565 85. One more patch for Sieve to update vacation handling to latest spec.
5568 ----------------------------------------------------
5569 See the note above about the 4.44 and 4.50 releases.
5570 ----------------------------------------------------
5576 1. Change 4.43/35 introduced a bug that caused file counts to be
5577 incorrectly computed when quota_filecount was set in an appendfile
5580 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5581 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5584 3. Give more explanation in the error message when the command for a transport
5585 filter fails to execute.
5587 4. There are several places where Exim runs a non-Exim command in a
5588 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5589 was being done only for the command run by the queryprogram router. It is
5590 now done for all such subprocesses. The other cases are: ${run, transport
5591 filters, and the commands run by the lmtp and pipe transports.
5593 5. Some older OS have a limit of 256 on the maximum number of file
5594 descriptors. Exim was using setrlimit() to set 1000 as a large value
5595 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5596 systems. I've change it so that if it can't get 1000, it tries for 256.
5598 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5599 was an oversight, and furthermore, ever since the addition of extra
5600 controls (e.g. 4.43/32), the checks on when to allow different forms of
5601 "control" were broken. There should now be diagnostics for all cases when a
5602 control that does not make sense is encountered.
5604 7. $recipients is now available in the predata ACL (oversight).
5606 8. Tidy the search cache before the fork to do a delivery from a message
5607 received from the command line. Otherwise the child will trigger a lookup
5608 failure and thereby defer the delivery if it tries to use (for example) a
5609 cached ldap connection that the parent has called unbind on.
5611 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5612 of $address_data from the recipient verification was clobbered by the
5613 sender verification.
5615 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5616 was its contents. (It was OK if the option was not defined at all.)
5618 11. A "Completed" log line is now written for messages that are removed from
5619 the spool by the -Mrm option.
5621 12. $host_address is now set to the target address during the checking of
5622 ignore_target_hosts.
5624 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5625 being passed; this would have caused $sender_host_name to have been used if
5626 matching the list had actually called for a host name (not very likely,
5627 since this list is usually IP addresses). A host name is now passed as
5630 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5631 code that specifies a non-restarting handler (typically sigaction() in
5632 modern systems) in an attempt to fix a rare and obscure crash bug.
5634 15. Narrowed the window for a race in the daemon that could cause it to ignore
5635 SIGCHLD signals. This is not a major problem, because they are used only to
5636 wake it up if nothing else does.
5638 16. A malformed maildirsize file could cause Exim to calculate negative values
5639 for the mailbox size or file count. Odd effects could occur as a result.
5640 The maildirsize information is now recalculated if the size or filecount
5643 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5644 support for a long time. Removed HAVE_SYS_VFS_H.
5646 18. Updated exipick to current release from John Jetmore.
5648 19. Allow an empty sender to be matched against a lookup in an address list.
5649 Previously the only cases considered were a regular expression, or an
5652 20. Exim went into a mad DNS lookup loop when doing a callout where the
5653 host was specified on the transport, if the DNS lookup yielded more than
5656 21. The RFC2047 encoding function was originally intended for short strings
5657 such as real names; it was not keeping to the 75-character limit for
5658 encoded words that the RFC imposes. It now respects the limit, and
5659 generates multiple encoded words if necessary. To be on the safe side, I
5660 have increased the buffer size for the ${rfc2047: expansion operator from
5663 22. Failure to deliver a bounce message always caused it to be frozen, even if
5664 there was an errors_to setting on the router. The errors_to setting is now
5667 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5668 canonical form (fully expanded) before being placed in
5669 $sender_host_address.
5671 24. Updated eximstats to version 1.33
5673 25. Include certificate and key file names in error message when GnuTLS fails
5674 to set them up, because the GnuTLS error message doesn't include the name
5675 of the failing file when there is a problem reading it.
5677 26. Expand error message when OpenSSL has problems setting up cert/key files.
5680 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5681 (this can affect the format of dates).
5683 28. exim_tidydb, when checking for the continued existence of a message for
5684 which it has found a message-specific retry record, was not finding
5685 messages that were in split spool directories. Consequently, it was
5686 deleting retry records that should have stayed in existence.
5688 29. eximstats updated to version 1.35
5689 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5690 1.35 - bugfix such that pie charts by volume are generated correctly
5692 30. The SPA authentication driver was not abandoning authentication and moving
5693 on to the next authenticator when an expansion was forced to fail,
5694 contradicting the general specification for all authenticators. Instead it
5695 was generating a temporary error. It now behaves as specified.
5697 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5698 (the order specifies the preference for clients). The order is now AES256,
5699 AES128, 3DES, ARCFOUR128.
5701 31. Small patch to Sieve code - explicitly set From: when generating an
5704 32. Exim crashed if a remote delivery caused a very long error message to be
5705 recorded - for instance if somebody sent an entire SpamAssassin report back
5706 as a large number of 550 error lines. This bug was coincidentally fixed by
5707 increasing the size of one of Exim's internal buffers (big_buffer) that
5708 happened as part of the Exiscan merge. However, to be on the safe side, I
5709 have made the code more robust (and fixed the comments that describe what
5712 33. Some experimental protocols are using DNS PTR records for new purposes. The
5713 keys for these records are domain names, not reversed IP addresses. The
5714 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5715 leaves it alone. Component reversal etc. now happens only for IP addresses.
5718 34. The host_aton() function is supposed to be passed a string that is known
5719 to be a valid IP address. However, in the case of IPv6 addresses, it was
5720 not checking this. This is a hostage to fortune. Exim now panics and dies
5721 if the condition is not met. A case was found where this could be provoked
5722 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5723 components; fortuitously, this particular loophole had already been fixed
5724 by change 4.50/55 or 4.44/33 above.
5726 If there are any other similar loopholes, the new check in host_aton()
5727 itself should stop them being exploited. The report I received stated that
5728 data on the command line could provoke the exploit when Exim was running as
5729 exim, but did not say which command line option was involved. All I could
5730 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5731 running as the user.
5734 35. There was a buffer overflow vulnerability in the SPA authentication code
5735 (which came originally from the Samba project). I have added a test to the
5736 spa_base64_to_bits() function which I hope fixes it.
5739 36. The daemon start-up calls getloadavg() while still root for those OS that
5740 need the first call to be done as root, but it missed one case: when
5741 deliver_queue_load_max is set with deliver_drop_privilege. This is
5742 necessary for the benefit of the queue runner, because there is no re-exec
5743 when deliver_drop_privilege is set.
5745 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5746 was in use, was not putting the data itself into the right store pool;
5747 consequently, it could be overwritten for a subsequent message in the same
5748 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5751 38. Sometimes the final signoff response after QUIT could fail to get
5752 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5753 < 0 before doing a fflush(). This bug looks as though it goes back to the
5754 introduction of TLS in release 3.20, but "sometimes" must have been rare
5755 because the tests only now provoked it.
5761 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5762 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5763 Now it returns an integer. A number of calls were still expecting a T/F
5764 return. Fortuitously, in all cases, the tests worked in OK situations,
5765 which is the norm. However, things would have gone wrong on any write
5766 failures on the smtp file descriptor. This function is used when sending
5767 messages over SMTP and also when doing verify callouts.
5769 2. When Exim is called to do synchronous delivery of a locally submitted
5770 message (the -odf or -odi options), it no longer closes stderr before doing
5773 3. Implemented the mua_wrapper option.
5775 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5777 5. Implemented the functions header_remove(), header_testname(),
5778 header_add_at_position(), and receive_remove_recipient(), and exported them
5781 6. If an ACL "warn" statement specified the addition of headers, Exim already
5782 inserted X-ACL-Warn: at the start if there was no header name. However, it
5783 was not making this test for the second and subsequent header lines if
5784 there were newlines in the string. This meant that an invalid header could
5785 be inserted if Exim was badly configured.
5787 7. Allow an ACL "warn" statement to add header lines at the start or after all
5788 the Received: headers, as well as at the end.
5790 8. Added the rcpt_4xx retry error code.
5792 9. Added postmaster_mailfrom=xxx to callout verification option.
5794 10. Added mailfrom=xxxx to the callout verification option, for verify=
5797 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5798 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5800 12. Inserted some casts to stop certain compilers warning when using pointer
5801 differences as field lengths or precisions in printf-type calls (mostly
5802 affecting debugging statements).
5804 13. Added optional readline() support for -be (dynamically loaded).
5806 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5807 same clock tick as a message's arrival, so that its received time was the
5808 same as the "first fail" time on the retry record, and that message
5809 remained on the queue past the ultimate address timeout, every queue runner
5810 would try a delivery (because it was past the ultimate address timeout) but
5811 after another failure, the ultimate address timeout, which should have then
5812 bounced the address, did not kick in. This was a "< instead of <=" error;
5813 in most cases the first failure would have been in the next clock tick
5814 after the received time, and all would be well.
5816 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5817 being recognized when the domain list was tested by the match_domain
5818 condition in an expansion string.
5820 16. Added the ${str2b64: operator.
5822 17. Exim was always calling setrlimit() to set a large limit for the number of
5823 processes, without checking whether the existing limit was already
5824 adequate. (It did check for the limit on file descriptors.) Furthermore,
5825 errors from getrlimit() and setrlimit() were being ignored. Now they are
5826 logged to the main and panic logs, but Exim does carry on, to try to do its
5827 job under whatever limits there are.
5829 18. Imported PCRE 5.0.
5831 19. Trivial typo in log message " temporarily refused connection" (the leading
5834 20. If the log selector return_path_on_delivery was set and an address was
5835 redirected to /dev/null, the delivery process crashed because it assumed
5836 that a return path would always be set for a "successful" delivery. In this
5837 case, the whole delivery is bypassed as an optimization, and therefore no
5840 21. Internal re-arrangement: the function for sending a challenge and reading
5841 a response while authentication was assuming a zero-terminated challenge
5842 string. It's now changed to take a pointer and a length, to allow for
5843 binary data in such strings.
5845 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5847 23. Exim was not respecting finduser_retries when seeking the login of the
5848 uid under which it was called; it was always trying 10 times. (The default
5849 setting of finduser_retries is zero.) Also, it was sleeping after the final
5850 failure, which is pointless.
5852 24. Implemented tls_on_connect_ports.
5854 25. Implemented acl_smtp_predata.
5856 26. If the domain in control=submission is set empty, Exim assumes that the
5857 authenticated id is a complete email address when it generates From: or
5858 Sender: header lines.
5860 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5861 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5862 chown and chgrp in /bin and hostname in /usr/bin.
5864 28. Exim was keeping the "process log" file open after each use, just as it
5865 does for the main log. This opens the possibility of it remaining open for
5866 long periods when the USR1 signal hits a daemon. Occasional processlog
5867 errors were reported, that could have been caused by this. Anyway, it seems
5868 much more sensible not to leave this file open at all, so that is what now
5871 29. The long-running daemon process does not normally write to the log once it
5872 has entered its main loop, and it closes the log before doing so. This is
5873 so that log files can straightforwardly be renamed and moved. However,
5874 there are a couple of unusual error situations where the daemon does write
5875 log entries, and I had neglected to close the log afterwards.
5877 30. The text of an SMTP error response that was received during a remote
5878 delivery was being truncated at 512 bytes. This is too short for some of
5879 the long messages that one sometimes sees. I've increased the limit to
5882 31. It is now possible to make retry rules that apply only when a message has a
5883 specific sender, in particular, an empty sender.
5885 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5886 it possible to be selective about when SMTP synchronization is enforced.
5888 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5890 32. Implemented hosts_connection_nolog.
5892 33. Added an ACL for QUIT.
5894 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5897 35. Added mailbox_size and mailbox_filecount to appendfile.
5899 36. Added control = no_multiline_responses to ACLs.
5901 37. There was a bug in the logic of the code that waits for the clock to tick
5902 in the case where the clock went backwards by a substantial amount such
5903 that the microsecond fraction of "now" was more than the microsecond
5904 fraction of "then" (but the whole seconds number was less).
5906 38. Added support for the libradius Radius client library this is found on
5907 FreeBSD (previously only the radiusclient library was supported).
5913 1. When certain lookups returned multiple values in the form name=value, the
5914 quoting of the values was not always being done properly. Specifically:
5915 (a) If the value started with a double quote, but contained no whitespace,
5917 (b) If the value contained whitespace other than a space character (i.e.
5918 tabs or newlines or carriage returns) it was not quoted.
5919 This fix has been applied to the mysql and pgsql lookups by writing a
5920 separate quoting function and calling it from the lookup code. The fix
5921 should probably also be applied to nisplus, ibase and oracle lookups, but
5922 since I cannot test any of those, I have not disturbed their existing code.
5924 2. A hit in the callout cache for a specific address caused a log line with no
5925 reason for rejecting RCPT. Now it says "Previous (cached) callout
5926 verification failure".
5928 3. There was an off-by-one bug in the queryprogram router. An over-long
5929 return line was truncated at 256 instead of 255 characters, thereby
5930 overflowing its buffer with the terminating zero. As well as fixing this, I
5931 have increased the buffer size to 1024 (and made a note to document this).
5933 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5934 when Exim is waiting for an SMTP response from a remote server, Exim
5935 restarts its select() call on the socket, thereby resetting its timeout.
5936 This is not a problem when such interrupts are rare. Somebody set up a cron
5937 job to run exiwhat every 2 minutes, which is less than the normal select()
5938 timeout (5 or 10 minutes). This meant that the select() timeout never
5939 kicked in because it was always reset. I have fixed this by comparing the
5940 time when an interrupt arrives with the time at the start of the first call
5941 to select(). If more time than the timeout has elapsed, the interrupt is
5942 treated as a timeout.
5944 5. Some internal re-factoring in preparation for the addition of Sieve
5945 extensions (by MH). In particular, the "personal" test is moved to a
5946 separate function, and given an option for scanning Cc: and Bcc: (which is
5947 not set for Exim filters).
5949 6. When Exim created an email address using the login of the caller as the
5950 local part (e.g. when creating a From: or Sender: header line), it was not
5951 quoting the local part when it contained special characters such as @.
5953 7. Installed new OpenBSD configuration files.
5955 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5956 try to make them clearer.
5958 9. Callout options, other than the timeout value, were being ignored when
5959 verifying sender addresses in header lines. For example, when using
5961 verify = header_sender/callout=no_cache
5963 the cache was (incorrectly) being used.
5965 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5966 only the Cygwin environment.
5968 11. When return_path_on_delivery was set as a log selector, if different remote
5969 addresses in the same message used different return paths and parallel
5970 remote delivery occurred, the wrong values would sometimes be logged.
5971 (Whenever a remote delivery process finished, the return path value from
5972 the most recently started remote delivery process was logged.)
5974 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5975 lines when AUTH and/or TLS are in use. This is the "received protocol"
5976 field. Exim used to use "asmtp" for authenticated SMTP, without any
5977 indication (in the protocol name) for TLS use. Now it follows the RFC and
5978 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5979 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5980 names appear in log lines as well as in Received: header lines.
5982 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5983 extensions, and comparison tests, and to fix some bugs.
5985 14. Changes to the "personal" filter test:
5987 (1) The test was buggy in that it was just doing the equivalent of
5988 "contains" tests on header lines. For example, if a user's address was
5989 anne@some.where, the "personal" test would incorrectly be true for
5991 To: susanne@some.where
5993 This test is now done by extracting each address from the header in turn,
5994 and checking the entire address. Other tests that are part of "personal"
5995 are now done using regular expressions (for example, to check local parts
5996 of addresses in From: header lines).
5998 (2) The list of non-personal local parts in From: addresses has been
5999 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6000 taken from the Sieve specification recommendations.
6002 (3) If the message contains any header line starting with "List-" it is
6003 treated as non-personal.
6005 (4) The test for "circular" in the Subject: header line has been removed
6006 because it now seems ill-conceived.
6008 15. Minor typos in src/EDITME comments corrected.
6010 16. Installed latest exipick from John Jetmore.
6012 17. If headers_add on a router specified a text string that was too long for
6013 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6014 of string_sprintf() is now avoided.
6016 18. $message_body_size was not set (it was always zero) when running the DATA
6017 ACL and the local_scan() function.
6019 19. For the "mail" command in an Exim filter, no default was being set for
6020 the once_repeat time, causing a random time value to be used if "once" was
6021 specified. (If the value happened to be <= 0, no repeat happened.) The
6022 default is now 0s, meaning "never repeat". The "vacation" command was OK
6023 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6024 (I found it when inspecting the code).
6026 20. There is now an overall timeout for performing a callout verification. It
6027 defaults to 4 times the callout timeout, which applies to individual SMTP
6028 commands during the callout. The overall timeout applies when there is more
6029 than one host that can be tried. The timeout is checked before trying the
6030 next host. This prevents very long delays if there are a large number of
6031 hosts and all are timing out (e.g. when the network connections are timing
6032 out). The value of the overall timeout can be changed by specifying an
6033 additional sub-option for "callout", called "maxwait". For example:
6035 verify = sender/callout=5s,maxwait=20s
6037 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6038 to the end before writing, but this should make it even safer).
6040 22. Exim was forgetting that it had advertised PIPELINING for the second and
6041 subsequent messages on an SMTP connection. It was also not resetting its
6042 memory on STARTTLS and an internal HELO.
6044 23. When Exim logs an SMTP synchronization error within a session, it now
6045 records whether PIPELINING has been advertised or not.
6047 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6048 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6049 rather than long int.
6051 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6053 26. Added the never_mail option to autoreply.
6059 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6060 crash if the getsockname() call failed; this can happen if a connection is
6061 closed very soon after it is established. The problem was simply in the
6062 order in which certain operations were done, causing Exim to try to write
6063 to the SMTP stream before it had set up the file descriptor. The bug has
6064 been fixed by making things happen in the correct order.
6070 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6071 before the connection was closed, thus losing the rejection response.
6073 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6074 some early Solaris releases, but causes trouble in current releases where
6075 socklen_t is defined.
6077 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6080 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6083 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6084 that releasing the top of it at the end releases what was used for sub-
6085 expansions (unless the block got too big). However, discard this block if
6086 the first thing is a variable or header, so that we can use its block when
6087 it is dynamic (useful for very large $message_headers, for example).
6089 6. Lookups now cache *every* query, not just the most recent. A new, separate
6090 store pool is used for this. It can be recovered when all lookup caches are
6091 flushed. Lookups now release memory at the end of their result strings.
6092 This has involved some general refactoring of the lookup sources.
6094 7. Some code has been added to the store_xxx() functions to reduce the amount
6095 of flapping under certain conditions.
6097 8. log_incoming_interface used to affect only the <= reception log lines. Now
6098 it causes the local interface and port to be added to several more SMTP log
6099 lines, for example "SMTP connection from", and rejection lines.
6101 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6103 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6105 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6106 could be overwritten at the end of the current message (or the start of a
6107 new message if it was set in a HELO ACL). The value is now preserved for
6108 the duration of the SMTP connection.
6110 12. If a transport had a headers_rewrite setting, and a matching header line
6111 contained an unqualified address, that address was qualified, even if it
6112 did not match any rewriting rules. The underlying bug was that the values
6113 of the flags that permit the existence of unqualified sender and recipient
6114 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6115 non-local messages, and by -bnq for local messages) were not being
6116 preserved with the message after it was received.
6118 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6119 "next input=" as part of the text comprising the host identity instead of
6120 the correct text. The code was using the same buffer for two different
6121 strings. However, depending on which order the printing function evaluated
6122 its arguments, the bug did not always show up. Under Linux, for example, my
6123 test suite worked just fine.
6125 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6126 doesn't work with some older versions of Perl. It has been changed to "my",
6127 which in any case is probably the better facility to use.
6129 15. A really picky compiler found some instances of statements for creating
6130 error messages that either had too many or two few arguments for the format
6133 16. The size of the buffer for calls to the DNS resolver has been increased
6134 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6135 for addresses that have a lot of PTR records. This alleviates a problem; it
6136 does not fully solve it.
6138 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6139 buffer now truncates the list and logs the incident, which is the same
6140 action as happens when Exim is looking up a host name and its aliases.
6141 Previously in this situation something unpredictable would happen;
6142 sometimes it was "internal error: store_reset failed".
6144 18. If a server dropped the connection unexpectedly when an Exim client was
6145 using GnuTLS and trying to read a response, the client delivery process
6146 crashed while trying to generate an error log message.
6148 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6149 string, for example:
6151 warn message = H1: something\nH2: something
6153 the text was added as a single header line from Exim's point of view
6154 though it ended up OK in the delivered message. However, searching for the
6155 second and subsequent header lines using $h_h2: did not work. This has been
6156 fixed. Similarly, if a system filter added multiple headers in this way,
6157 the routers could not see them.
6159 20. Expanded the error message when iplsearch is called with an invalid key to
6160 suggest using net-iplsearch in a host list.
6162 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6163 ACLs are no longer actually imposed (and a message to that effect is
6166 22. If a "gecos" field in a passwd entry contained escaped characters, in
6167 particular, if it contained a \" sequence, Exim got it wrong when building
6168 a From: or a Sender: header from that name. A second bug also caused
6169 incorrect handling when an unquoted " was present following a character
6170 that needed quoting.
6172 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6173 was not being matched caselessly.
6175 24. Arranged for all hyphens in the exim.8 source to be escaped with
6178 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6179 properly. Recipient callout cache records were still being keyed to include
6180 the sender, even when use_sender was set false. This led to far more
6181 callouts that were necessary. The sender is no longer included in the key
6182 when use_sender is false.
6184 26. Added "control = submission" modifier to ACLs.
6186 27. Added the ${base62d: operator to decode base 62 numbers.
6188 28. dnsdb lookups can now access SRV records.
6190 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6191 the configuration file.
6193 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6194 (-v) mode. This makes the output for a verbose queue run more intelligible.
6196 31. Added a use_postmaster feature to recipient callouts.
6198 32. Added the $body_zerocount variable, containing the number of binary zero
6199 bytes in the message body.
6201 33. The time of last modification of the "new" subdirectory is now used as the
6202 "mailbox time last read" when there is a quota error for a maildir
6205 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6207 35. Added +ignore_unknown as a special item in host lists.
6209 36. Code for decoding IPv6 addresses in host lists is now included, even if
6210 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6211 address was recognized as an IP address, but was then not correctly decoded
6212 into binary, causing unexpected and incorrect effects when compared with
6219 1. Very minor rewording of debugging text in manualroute to say "list of
6220 hosts" instead of "hostlist".
6222 2. If verify=header_syntax was set, and a header line with an unqualified
6223 address (no domain) and a large number of spaces between the end of the
6224 name and the colon was received, the reception process suffered a buffer
6225 overflow, and (when I tested it) crashed. This was caused by some obsolete
6226 code that should have been removed. The fix is to remove it!
6228 3. When running in the test harness, delay a bit after writing a bounce
6229 message to get a bit more predictability in the log output.
6231 4. Added a call to search_tidyup() just before forking a reception process. In
6232 theory, someone could use a lookup in the expansion of smtp_accept_max_
6233 per_host which, without the tidyup, could leave open a database connection.
6235 5. Added the variables $recipient_data and $sender_data which get set from a
6236 lookup success in an ACL "recipients" or "senders" condition, or a router
6237 "senders" option, similar to $domain_data and $local_part_data.
6239 6. Moved the writing of debug_print from before to after the "senders" test
6242 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6243 problems for message scanning, either using a data ACL, or using
6244 local_scan() because the Received: header was not generated till after they
6245 were called (in order to set the time as the time of reception completion).
6246 I have revised the way this works. The header is now generated after the
6247 body is received, but before the ACL or local_scan() are called. After they
6248 are run, the timestamp in the header is updated.
6254 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6255 before starting a queue runner without re-exec. This happened only when
6256 deliver_drop_privilege was set or when the Exim user was set to root. The
6257 effect of the bug was that timeouts during subsequent deliveries caused
6258 crashes instead of being properly handled. The handler is now left at its
6259 default (and expected) setting.
6261 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6262 message, again when deliver_drop_privilege is set or Exim is run as root.
6263 The bug described in (1) was not present in this case, but the tidying up
6264 of the other signals was missing. I have made the two cases consistent.
6266 3. The ignore_target_hosts setting on a manualroute router was being ignored
6267 for hosts that were looked up using the /MX notation.
6269 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6272 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6273 operated on the sender address. After changing the $sender_address to <>
6274 for the sender address verify, Exim was re-instated it as the original
6275 (before rewriting) address, but remembering that it had rewritten it, so it
6276 wasn't rewriting it again. This bug also had the effect of breaking the
6277 sender address verification caching when the sender address was rewritten.
6279 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6280 This has been changed so that if the ip literal address matches
6281 ignore_target_hosts, the router declines.
6283 7. Added expansion conditions match_domain, match_address, and match_local_
6284 part (NOT match_host).
6286 8. The placeholder for the Received: header didn't have a length field set.
6288 9. Added code to Exim itself and to exim_lock to test for a specific race
6289 condition that could lead to file corruption when using MBX delivery. The
6290 issue is with the lockfile that is created in /tmp. If this file is removed
6291 after a process has opened it but before that process has acquired a lock,
6292 there is the potential for a second process to recreate the file and also
6293 acquire a lock. This could lead to two Exim processes writing to the file
6294 at the same time. The added code performs the same test as UW imapd; it
6295 checks after acquiring the lock that its file descriptor still refers to
6296 the same named file.
6298 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6299 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6302 11. Added the smtp_active_hostname option. If used, this will typically be made
6303 to depend on the incoming interface address. Because $interface_address is
6304 not set up until the daemon has forked a reception process, error responses
6305 that can happen earlier (such as "too many connections") no longer contain
6308 12. If an expansion in a condition on a "warn" statement fails because a lookup
6309 defers, the "warn" statement is abandoned, and the next ACL statement is
6310 processed. Previously this caused the whole ACL to be aborted.
6312 13. Added the iplsearch lookup type.
6314 14. Added ident_timeout as a log selector.
6316 15. Added tls_certificate_verified as a log selector.
6318 16. Added a global option tls_require_ciphers (compare the smtp transport
6319 option of the same name). This controls incoming TLS connections.
6321 17. I finally figured out how to make tls_require_ciphers do a similar thing
6322 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6323 before starting the TLS session.
6325 18. Tabs are now shown as \t in -bP output.
6327 19. If the log selector return_path_on_delivery was set, Exim crashed when
6328 bouncing a message because it had too many Received: header lines.
6330 20. If two routers both had headers_remove settings, and the first one included
6331 a superfluous trailing colon, the final name in the first list and the
6332 first name in the second list were incorrectly joined into one item (with a
6333 colon in the middle).
6339 1. Added -C and -D options to the exinext utility, mainly to make it easier
6340 to include in the automated testing, but these could be helpful when
6341 multiple configurations are in use.
6343 2. The exinext utility was not formatting the output nicely when there was
6344 an alternate port involved in the retry record key, nor when there was a
6345 message id as well (for retries that were specific to a specific message
6346 and a specific host). It was also confused by IPv6 addresses, because of
6347 the additional colons they contain. I have fixed the IPv4 problem, and
6348 patched it up to do a reasonable job for IPv6.
6350 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6351 delivery, the log line now contains "pipelined" if PIPELINING was used.
6353 4. An SMTP transport process used to panic and die if the bind() call to set
6354 an explicit outgoing interface failed. This has been changed; it is now
6355 treated in the same way as a connect() failure.
6357 5. A reference to $sender_host_name in the part of a conditional expansion
6358 that was being skipped was still causing a DNS lookup. This no longer
6361 6. The def: expansion condition was not recognizing references to header lines
6362 that used bh_ and bheader_.
6364 7. Added the _cache feature to named lists.
6366 8. The code for checking quota_filecount in the appendfile transport was
6367 allowing one more file than it should have been.
6369 9. For compatibility with Sendmail, the command line option
6377 and sets the incoming protocol and host name (for trusted callers). The
6378 host name and its colon can be omitted when only the protocol is to be set.
6379 Note the Exim already has two private options, -pd and -ps, that refer to
6380 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6381 "s", but I don't think that's a major issue.
6383 10. A number of refactoring changes to the code, none of which should affect
6386 (a) The number of logging options was getting close to filling up the
6387 32-bit word that was used as a bit map. I have split them into two classes:
6388 those that are passed in the argument to log_write(), and those that are
6389 only ever tested independently outside of that function. These are now in
6390 separate 32-bit words, so there is plenty of room for expansion again.
6391 There is no change in the user interface or the logging behaviour.
6393 (b) When building, for example, log lines, the code previously used a
6394 macro that called string_cat() twice, in order to add two strings. This is
6395 not really sufficiently general. Furthermore, there was one instance where
6396 it was actually wrong because one of the argument was used twice, and in
6397 one call a function was used. (As it happened, calling the function twice
6398 did not affect the overall behaviour.) The macro has been replaced by a
6399 function that can join an arbitrary number of extra strings onto a growing
6402 (c) The code for expansion conditions now uses a table and a binary chop
6403 instead of a serial search (which was left over from when there were very
6404 few conditions). Also, it now recognizes conditions like "pam" even when
6405 the relevant support is not compiled in: a suitably worded error message is
6406 given if an attempt is made to use such a condition.
6408 11. Added ${time_interval:xxxxx}.
6410 12. A bug was causing one of the ddress fields not to be passed back correctly
6411 from remote delivery subprocesses. The field in question was not being
6412 subsequently used, so this caused to problems in practice.
6414 13. Added new log selectors queue_time and deliver_time.
6416 14. Might have fixed a bug in maildirsizefile handling that threw up
6417 "unexpected character" debug warnings, and recalculated the data
6418 unnecessarily. In any case, I expanded the warning message to give more
6421 15. Added the message "Restricted characters in address" to the statements in
6422 the default ACL that block characters like @ and % in local parts.
6424 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6425 Three changes have been made:
6427 (a) There was a serious bug; a negative response to MAIL caused the whole
6428 recipient domain to be cached as invalid, thereby blocking all messages
6429 to all local parts at the same domain, from all senders. This bug has
6430 been fixed. The domain is no longer cached after a negative response to
6431 MAIL if the sender used is not empty.
6433 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6436 (c) A new callout option, "use_sender" has been added for people who want
6437 the modified behaviour.
6443 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6446 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6447 indeed breaks things for older releases.
6449 3. Added additional logging to the case where there is a problem reading data
6450 from a filter that is running in a subprocess using a pipe, in order to
6451 try to track down a specific problem.
6453 4. Testing facility fudge: when running in the test harness and attempting
6454 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6455 getting "No route to host". Convert this to a timeout.
6457 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6460 6. Some OS don't have socklen_t but use size_t instead. This affects the
6461 fifth argument of getsockopt() amongst other things. This is now
6462 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6463 can be set for individual OS. I have set it for SunOS5, OSF1, and
6464 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6465 some earlier ones do not.
6467 7. Change 4.30/15 was not doing the test caselessly.
6469 8. The standard form for an IPv6 address literal was being rejected by address
6470 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6471 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6472 this, as well as the form without the "IPv6" on the front (but only when
6473 address literals are enabled, of course).
6475 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6477 10. Exim crashed if a message with an empty sender address specified by -f
6478 encountered a router with an errors_to setting. This could be provoked only
6479 by a command such as
6483 where an empty string was supplied; "<>" did not hit this bug.
6485 11. Installed PCRE release 4.5.
6487 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6488 remained set. It is now erased.
6490 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6491 times from message ids (which are base 36 rather than the normal 62).
6493 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6494 were being counted as actual protocol errors, and logged if the log
6495 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6496 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6497 and DATA following a set of rejected RCPTs do not count as protocol errors.
6498 In other words, Exim assumes they were pipelined, though this may not
6499 actually be the case. Of course, in all cases the client gets an
6500 appropriate error code.
6502 15. If a lookup fails in an ACL condition, a message about the failure may
6503 be available; it is used if testing the ACL cannot continue, because most
6504 such messages specify what the cause of the deferral is. However, some
6505 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6506 that caused an old message to be retained and used if a later statement
6507 caused a defer, replacing the real cause of the deferral.
6509 16. If an IP address had so many PTR records that the DNS lookup buffer
6510 was not large enough to hold them, Exim could crash while trying to process
6511 the truncated data. It now detects and logs this case.
6513 17. Further to 4.21/58, another change has been made: if (and only if) the
6514 first line of a message (the first header line) ends with CRLF, a bare LF
6515 in a subsequent header line has a space inserted after it, so as not to
6516 terminate the header.
6518 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6519 unnecessarily, used atoi() instead of strtol(), and didn't check the
6520 termination when getting file sizes from file names by regex.
6522 19. Completely re-implemented the support for maildirsize files, in the light
6523 of a number of problems with the previous contributed implementation
6524 (4.30/29). In particular:
6526 . If the quota is zero, the maildirsize file is maintained, but no quota is
6529 . If the maildir directory does not exist, it is created before any attempt
6530 to write a maildirsize file.
6532 . The quota value in the file is just a cache; if the quota is changed in
6533 the transport, the new value overrides.
6535 . A regular expression is available for excluding directories from the
6538 20. The autoreply transport checks the characters in options that define the
6539 message's headers; it allows continued headers, but it was checking with
6540 isspace() after an embedded newline instead of explicitly looking for a
6543 21. If all the "regular" hosts to which an address was routed had passed their
6544 expiry times, and had not reached their retry times, the address was
6545 bounced, even if fallback hosts were defined. Now Exim should go on to try
6548 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6549 equivalent code in the SMTP transport. Some hosts send humungous responses
6550 to HELO/EHLO, more than 1024 it seems.
6552 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6553 gives compiler warnings in some environments. I've now done it "properly",
6556 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6557 (because of problems with the built-in one) was declared to return uschar *
6558 instead of char *, causing compiler failure.
6560 25. Fixed a file descriptor leak when processing alias/forward files.
6562 26. Fixed a minor format string issue in dbfn.c.
6564 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6566 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6567 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6570 29. When a sender address is verified, it is cached, to save repeating the test
6571 when there is more than one recipient in a message. However, when the
6572 verification involves a callout, it is possible for different callout
6573 options to be set for different recipients. It is too complicated to keep
6574 track of this in the cache, so now Exim always runs a verification when a
6575 callout is required, relying on the callout cache for the optimization.
6576 The overhead is duplication of the address routing, but this should not be
6579 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6580 to be verified with callout=postmaster, and the main callout worked but the
6581 postmaster check failed, the verification correctly failed. However, if a
6582 subsequent RCPT command asked for sender verification *without* the
6583 postmaster check, incorrect caching caused this verification also to fail,
6586 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6587 it was not caching the DNS options (qualify_single, search_parents) that
6588 were used when the lookup failed. A subsequent lookup with different
6589 options therefore always gave the same answer, though there were cases
6590 where it should not have. (Example: a "domains = !$mx_any" option on a
6591 dnslookup router: the "domains" option is always processed without any
6592 widening, but the router might have qualify_single set.) Now Exim uses the
6593 cached value only when the same options are set.
6595 32. Added John Jetmore's "exipick" utility to the distribution.
6597 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6598 than a timeout (e.g. a certificate is required, and is not provided), an
6599 Exim server now closes the connection immediately. Previously it waited for
6600 the client to close - but if the client is SSL, it seems that they each
6601 wait for each other, leading to a delay before one of them times out.
6603 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6604 maintained 0.8.x compatibility because I don't think many are using it, and
6605 it is clearly obsolete.
6607 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6610 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6611 client certificate was expired. A simple patch fixes this, though I don't
6612 understand the full logic of why the verify callback is called multiple
6615 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6616 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6617 which causes problems with some clients (such as the Certicom SSL Plus
6618 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6619 disables the coutermeasure allowing Eudora to connect."
6621 38. Exim was not checking that a write() to a log file succeeded. This could
6622 lead to Bad Things if a log got too big, in particular if it hit a file
6623 size limit. Exim now panics and dies if it cannot write to a log file, just
6624 as it does if it cannot open a log file.
6626 39. Modified OS/Makefile-Linux so that it now contains
6628 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6630 The two -D definitions ensure that Exim is compiled with large file
6631 support, which makes it possible to handle log files that are bigger than
6634 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6635 instance) a domain was checked against a named list that involved a lookup,
6636 causing $domain_data to be set, then another domain was checked against the
6637 same list, then the first domain was re-checked, the value of $domain_data
6638 after the final check could be wrong. In particular, if the second check
6639 failed, it could be set empty. This bug probably also applied to
6642 41. The strip_trailing_dot option was not being applied to the address given
6643 with the -f command-line option.
6645 42. The code for reading a message's header from the spool was incrementing
6646 $received_count, but never initializing it. This meant that the value was
6647 incorrect (doubled) while delivering a message in the same process in which
6648 it was received. In the most common configuration of Exim, this never
6649 happens - a fresh exec is done - but it can happen when
6650 deliver_drop_privilege is set.
6652 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6653 it now includes up to 150 characters of the unexpected data in the log
6656 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6657 and building data strings. The size of both of these buffers was 10 000
6658 bytes - far larger than anybody would *ever* want, thought I. Needless to
6659 say, somebody hit the limit. I have increased the maximum line length to
6660 20 000 and the maximum data length of concatenated lines to 100 000. I have
6661 also fixed two bugs, because there was no checking on these buffers. Tsk,
6662 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6663 buffer is too small.
6665 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6666 lsearch lookups. Now it does.
6668 46. When parsing a route_list item in a manualroute router, a fixed-length
6669 buffer was used for the list of hosts. I made this 1024 bytes long,
6670 thinking that nobody would ever have a list of hosts that long. Wrong.
6671 Somebody had a whole pile of complicated expansion conditions, and the
6672 string was silently truncated, leading to an expansion error. It turns out
6673 that it is easier to change to an unlimited length (owing to other changes
6674 that have happened since this code was originally written) than to build
6675 structure for giving a limitation error. The length of the item that
6676 expands into the list of hosts is now unlimited.
6678 47. The lsearch lookup could not handle data where the length of text line was
6679 more than 4095 characters. Such lines were truncated, leading to shortened
6680 data being returned. It should now handle lines of any length.
6682 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6683 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6686 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6687 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6688 no longer happens in comment lines. A list of replacements is now placed
6689 at the head of all of the source files, except those whose only change is
6690 to replace PERL_COMMAND in the very first #! line.
6692 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6693 messages on the queue, with a bottom-up merge sort, using code contributed
6694 by Michael Haardt. This should make operations like -bp somewhat faster on
6695 large queues. It won't affect queue runners, except when queue_run_in_order
6698 51. Installed eximstats 1.31 in the distribution.
6700 52. Added support for SRV lookups to the dnslookup router.
6702 53. If an ACL referred to $message_body or $message_body_end, the value was not
6703 reset for any messages that followed in the same SMTP session.
6705 54. The store-handling optimization for building very long strings was not
6706 differentiating between the different store pools. I don't think this
6707 actually made any difference in practice, but I've tidied it.
6709 55. While running the routers to verify a sender address, $sender_address
6710 was still set to the sender address. This is wrong, because when routing to
6711 send a bounce to the sender, it would be empty. Therefore, I have changed
6712 it so that, while verifying a sender address, $sender_address is set to <>.
6713 (There is no change to what happens when verifying a recipient address.)
6715 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6716 target A or AAAA records (if not already returned) without resetting the
6717 qualify_single or search_parents options of the DNS resolver. These are
6718 inappropriate in this case because the targets of MX and SRV records must
6719 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6720 target that, when qualified, matched something in the local domain. These
6721 two options are now turned off when doing these lookups.
6723 57. It seems that at least some releases of Reiserfs (which does not have the
6724 concept of a fixed number of inodes) returns zero and not -1 for the
6725 number of available inodes. This interacted badly with check_spool_inodes,
6726 which assumed that -1 was the "no such thing" setting. What I have done is
6727 to check that the total number of inodes is greater than zero before doing
6728 the test of how many are available.
6730 58. When a "warn" ACL statement has a log_message modifier, the message is
6731 remembered, and not repeated. This is to avoid a lot of repetition when a
6732 message has many recipients that cause the same warning to be written.
6733 However, Exim was preserving the list of already written lines for an
6734 entire SMTP session, which doesn't seem right. The memory is now reset if a
6735 new message is started.
6737 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6738 debugging output unless log_rewrite was also set.
6740 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6741 of (int)(handle) when we know that handle contains (void *)(-1).
6743 61. The Exim daemon panic-logs an error return when it closes the incoming
6744 connection. However "connection reset by peer" seems to be common, and
6745 isn't really an error worthy of noting specially, so that particular error
6748 62. When Exim is trying to find all the local interfaces, it used to panic and
6749 die if the ioctl to get the interface flags failed. However, it seems that
6750 on at least one OS (Solaris 9) it is possible to have an interface that is
6751 included in the list of interfaces, but for which you get a failure error
6752 for this call. This happens when the interface is not "plumbed" into a
6753 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6754 failure of the "get flags" call assumes that the interface is down.
6756 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6757 makes life easier for people who are doing arithmetic on fields extracted
6758 from dates, where you often get leading zeros that should not be
6759 interpreted as octal.
6761 64. Added qualify_domain to the redirect router, to override the global
6764 65. If a pathologically long header line contained very many addresses (the
6765 report of this problem mentioned 10 000) and each of them was rewritten,
6766 Exim could use up a very large amount of memory. (It kept on making new
6767 copies of the header line as it rewrote, and never released the old ones.)
6768 At the expense of a bit more processing, the header rewriting function has
6769 been changed so that it no longer eats memory in this way.
6771 66. The generation of the Received: header has been moved from the time that a
6772 message starts to be received, to the time that it finishes. The timestamp
6773 in the Received: header should now be very close to that of the <= log
6774 line. There are two side-effects of this change:
6776 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6777 logged header lines no longer include the local Received: line, because
6778 it has not yet been created. The same applies to a copy of the message
6779 that is returned to a non-SMTP sender when a message is rejected.
6781 (b) When a filter file is tested using -bf, no additional Received: header
6782 is added to the test message. After some thought, I decided that this
6785 This change does not affect the value of $received_for. It is still set
6786 after address rewriting, but before local_scan() is called.
6788 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6790 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6791 gave an unhelpful panic error message, and a defer error. I have managed to
6792 change this behaviour so that it now rejects any supplied certificate,
6793 which seems right, as the list of acceptable certificates is empty.
6795 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6796 gave an unhelpful defer error. I have not managed to make this reject any
6797 supplied certificates, but the error message it gives is "no certificate
6798 supplied", which is not helpful.
6800 70. exigrep's output now also includes lines that are not associated with any
6801 message, but which match the given pattern. Implemented by a patch from
6802 Martin Sluka, which also tidied up the Perl a bit.
6804 71. Recipient callout verification, like sender verification, was using <> in
6805 the MAIL FROM command. This isn't really the right thing, since the actual
6806 sender may affect whether the remote host accepts the recipient or not. I
6807 have changed it to use the actual sender in the callout; this means that
6808 the cache record is now keyed on a recipient/sender pair, not just the
6809 recipient address. There doesn't seem to be a real danger of callout loops,
6810 since a callout by the remote host to check the sender would use <>.
6811 [SEE ABOVE: changed after hitting problems.]
6813 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6814 temporary errors. However, in the case of such a code being given after
6815 the end of a data transmission (i.e. after ".") Exim was failing to write
6816 a retry record for the message. (Yes, there was some broken host that was
6817 actually sending 8xx at this point.)
6819 73. An unknown lookup type in a host list could cause Exim to panic-die when
6820 the list was checked. (An example that provoked this was putting <; in the
6821 middle of a list instead of at the start.) If this happened during a DATA
6822 ACL check, a -D file could be left lying around. This kind of configuration
6823 error no longer causes Exim to die; instead it causes a defer error. The
6824 incident is still logged to the main and panic logs.
6826 74. Buglet left over from Exim 3 conversion. The message "too many messages
6827 in one connection" was written to the rejectlog but not the mainlog, except
6828 when address rewriting (yes!) was being logged.
6830 75. Added write_rejectlog option.
6832 76. When a system filter was run not as root (that is, when system_filter_user
6833 was set), the values of the $n variables were not being returned to the
6834 main process; thus, they were not subsequently available in the $sn
6837 77. Added +return_path_on_delivery log selector.
6839 78. A connection timeout was being treated differently from recipients deferred
6840 when testing hosts_max_try with a message that was older than the host's
6841 retry timeout. (The host should not be counted, thus allowing all hosts to
6842 be tried at least once before bouncing.) This may have been the cause of an
6843 occasionally reported bug whereby a message would remain on the queue
6844 longer than the retry timeout, but would be bounced if a delivery was
6845 forced. I say "may" because I never totally pinned down the problem;
6846 setting up timeout/retry tests is difficult. See also the next item.
6848 79. The ultimate address timeout was not being applied to errors that involved
6849 a combination of host plus message (for example, a timeout on a MAIL
6850 command). When an address resolved to a number of possible hosts, and they
6851 were not all tried for each delivery (e.g. because of hosts_max_try), a
6852 message could remain on the queue longer than the retry timeout.
6854 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6857 81. Fixed an obscure SMTP outgoing bug which required at least the following
6858 conditions: (a) there was another message waiting for the same server;
6859 (b) the server returned 5xx to all RCPT commands in the first message so
6860 that the message was not completed; (c) the server dropped the connection
6861 or gave a negative response to the RSET that Exim sends to abort the
6862 transaction. The observed case was a dropped connection after DATA that had
6863 been sent in pipelining mode. That is, the server had advertised PIPELINING
6864 but was not implementing it correctly. The effect of the bug was incorrect
6865 behaviour, such as trying another host, and this could lead to a crash.
6871 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6872 and daemon.c were passed as pointers to ints; they should have been
6873 pointers to socklen_t variables (which are typically unsigned ints).
6875 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6878 3. Fixed a really odd bug that affected only the testing scheme; patching a
6879 certain fixed string in the binary changed the value of another string that
6880 happened to be identical to the end of the original first string.
6882 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6883 name", it returns that address as the IP address. On some operating
6884 systems (e.g. Solaris), it also passes back the IP address string as the
6885 "host name". However, on others (e.g. Linux), it passes back an empty
6886 string. Exim wasn't checking for this, and was changing the host name to an
6887 empty string, assuming it had been canonicalized.
6889 5. Although rare, it is permitted to have more than one PTR record for a given
6890 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6891 all the names associated with an address, because they do in Solaris.
6892 However, it seems that they do not in Linux for data that comes from the
6893 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6894 I found this out when I moved to a new Linux workstation and tried to run
6895 the Exim test suite.
6897 To get round this problem I have changed the code so that it now does its
6898 own call to the DNS to look up PTR records when searching for a host name.
6899 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6900 addresses that are only in /etc/hosts are still found.
6902 This behaviour is, however, controlled by an option called host_lookup_
6903 order, which defaults to "bydns:byaddr". If people want to use the other
6904 order, or indeed, just use one or the other means of lookup, they can
6905 specify it in this variable.
6907 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6908 some operating systems, this comes back from gethostbyaddr() as an empty
6909 string, and this is what Exim used to test for. However, it seems that in
6910 other systems, "." is yielded. Exim now tests for this case too.
6912 7. The values of check_spool_space and check_log_space are now held internally
6913 as a number of kilobytes instead of an absolute number of bytes. If a
6914 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6915 kilobyte. This means that much larger values can be stored.
6917 8. Exim monitor: an attempt to get the action menu when not actually pointing
6918 at a message produces an empty menu entitled "No message selected". This
6919 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6920 no entries in it ("Shell widget menu has zero width and/or height"). So I
6921 have added a single, blank menu entry in this case.
6923 9. Added ${quote_local_part.
6925 10. MIME decoding is now applied to the contents of Subject: header lines when
6928 11. Now that a reference to $sender_host_address automatically causes a reverse
6929 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6930 host lookup before query-style lookups in lists that might use this
6931 variable. This has therefore been abolished, and the "net-" prefix is no
6932 longer necessary for query-style lookups.
6934 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6935 have been a typo for LFLAGS, so it has been changed.
6937 13. The install script calls Exim with "-C /dev/null" in order to find the
6938 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6939 to be output. However, since Exim outputs its version number before the
6940 error, it didn't break the script. It just looked ugly. I fixed this by
6941 always allowing "-C /dev/null" if the caller is root.
6943 14. Ignore overlarge ACL variable number when reading spool file - insurance
6944 against a later release with more variables having written the file.
6946 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6947 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6948 this, as well as the form without the "IPv6" on the front.
6950 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6951 OS/Makefile-Darwin file.
6953 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6954 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6956 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6957 variable. However, some OS use ints for inodes. I've added cast to long int
6958 to get rid of the compiler warning.
6960 19. I had forgotten to lock out "/../" in configuration file names when
6961 ALT_CONFIG_PREFIX was set.
6963 20. Routers used for verification do not need to specify transports. However,
6964 if such a router generated a host list, and callout was configured, Exim
6965 crashed, because it could not find a port number from the (non-existent)
6966 transport. It now assumes port 25 in this circumstance.
6968 21. Added the -t option to exigrep.
6970 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6971 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6972 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6973 them is set, LOOKUP_LSEARCH is forced.
6975 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6977 24. Added sender and host information to the "rejected by local_scan()" log
6978 line; previously there was no indication of these.
6980 25. Added .include_if_exists.
6982 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6983 Linux. It turns out that not all file systems support this. Apparently some
6984 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6985 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6986 not supported on the file descriptor, is now ignored when Exim is trying to
6987 sync a directory. This applies only to Linux.
6989 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6991 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6992 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6995 29. Added support for maildirsize files from supplied patch (modified a bit).
6997 30. The use of :fail: followed by an empty string could lead Exim to respond to
6998 sender verification failures with (e.g.):
7000 550 Verification failed for <xxx>
7001 550 Sender verify failed
7003 where the first response line was missing the '-' that indicates it is not
7004 the final line of the response.
7006 31. The loop for finding the name of the user that called Exim had a hardwired
7007 limit of 10; it now uses the value of finduser_retries, which is used for
7008 all other user lookups.
7010 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7013 33. Exim was neglecting to zero errno before one call of strtol() when
7014 expanding a string and expecting an integer value. On some systems this
7015 resulted in spurious "integer overflow" errors. Also, it was casting the
7016 result into an int without checking.
7018 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7019 did not work. The code looks as if it has *never* worked, though it appears
7020 to have been documented since at least release 1.62. I have made it work.
7022 35. The "timeout_DNS" error in retry rules, also documented since at least
7023 1.62, also never worked. As it isn't clear exactly what this means, and
7024 clearly it isn't a major issue, I have abolished the feature by treating it
7025 as "timeout", and writing a warning to the main and panic logs.
7027 36. The display of retry rules for -brt wasn't always showing the error code
7030 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7031 timeout_connect_A, timeout_connect_MX.
7033 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7034 to the empty sender.
7036 39. The daemon was not analysing the content of -oX till after it had closed
7037 stderr and disconnected from the controlling terminal. This meant that any
7038 syntax errors were only noted on the panic log, and the return code from
7039 the command was 0. By re-arranging the code a little, I've made the
7040 decoding happen first, so such errors now appear on stderr, and the return
7041 code is 1. However, the actual setting up of the sockets still happens in
7042 the disconnected process, so errors there are still only recorded on the
7045 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7046 connections (as happens on some IP stacks) was logged at start up time as
7047 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7048 it from "IPv6 and IPv4", which means that two separate sockets are being
7051 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7052 says whether AF_INET or AF_INET6 was passed as an argument.
7054 42. Exiwhat output was messed up when time zones were included in log
7057 43. Exiwhat now gives more information about the daemon's listening ports,
7058 and whether -tls-on-connect was used.
7060 44. The "port" option of the smtp transport is now expanded.
7062 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7063 silently ignored. Now an error message is written to the main and panic
7066 46. There's a new ACL modifier called "logwrite" which writes to a log file
7067 as soon as it is encountered.
7069 47. Added $local_user_uid and $local_user_gid at routing time.
7071 48. Exim crashed when trying to verify a sender address that was being
7074 49. Exim was recognizing only a space character after ".include". It now also
7075 recognizes a tab character.
7077 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7078 extracting the relevant information from the specification. The man page no
7079 longer contains scrambled data for the -d option, and I've added a section
7080 at the front about calling Exim under different names.
7082 51. Added "extra_headers" argument to the "mail" command in filter files.
7084 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7087 53. Installed eximstats 1.29.
7089 54. Added transport_filter_timeout as a generic transport option.
7091 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7092 Cc: header lines. This was required by RFC 822, but it not required by RFC
7095 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7096 incoming messages that did not have them. Now it does so only if the
7097 message originates locally, that is, if there is no associated remote host
7098 address. When Resent- header lines are present, this applies to the Resent-
7099 lines rather than the non-Resent- lines.
7101 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7102 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7104 58. Messages for configuration errors now include the name of the main
7105 configuration file - useful now that there may be more than one file in a
7106 list (.included file names were always shown).
7108 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7109 for those rare installations that do not start the daemon as root or run it
7110 setuid root. I've cut out the call to initgroups() if the daemon is not
7113 60. The Exim user and group can now be bound into the binary as text strings
7114 that are looked up at the start of Exim's processing.
7116 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7118 62. Added $mailstore_basename variable.
7120 63. Installed patch to sieve.c from Michael Haardt.
7122 64. When Exim failed to open the panic log after failing to open the main log,
7123 the original message it was trying to log was written to stderr and debug
7124 output, but if they were not available (the usual case in production), it
7125 was lost. Now it is written to syslog before the two lines that record the
7126 failures to open the logs.
7128 65. Users' Exim filters run in subprocesses under the user's uid. It is
7129 possible for a "deliver" command or an alias in a "personal" command to
7130 provoke an address rewrite. If logging of address rewriting is configured,
7131 this fails because the process is not running as root or exim. There may be
7132 a better way of dealing with this, but for the moment (because 4.30 needs
7133 to be released), I have disabled address rewrite logging when running a
7134 filter in a non-root, non-exim process.
7140 1. The buildconfig auxiliary program wasn't quoting the value set for
7141 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7142 not defined. This bug was masked in 4.22 by the effect that was fixed in
7145 2. Some messages that were rejected after a message id was allocated were
7146 shown as "incomplete" by exigrep. It no longer does this for messages that
7147 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7149 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7150 have allow_domain_literals set, the ID did not get logged in the <= line.
7151 Domain literals are now always recognized in Message-ID: header lines.
7153 4. The first argument for a ${extract expansion item is the key name or field
7154 number. Leading and trailing spaces in this item were not being ignored,
7155 causing some misleading effects.
7157 5. When deliver_drop_privilege was set, single queue runner processes started
7158 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7159 same command in the process it spins off) were not dropping privilege.
7161 6. When the daemon running as "exim" started a queue runner, it always
7162 re-executed Exim in the spun-off process. This is a waste of effort when
7163 deliver_drop_privilege is set. The new process now just calls the
7164 queue-runner function directly.
7170 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7173 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7174 ignored. Though the use of -oP was forcing the writing of a pid file, it
7175 was always written to the default place.
7177 3. If the message "no IP address found for host xxxx" is generated during
7178 incoming verification, it is now followed by identification of the incoming
7179 connection (so you can more easily find what provoked it).
7181 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7183 5. Added some features to "harden" Exim a bit more against certain attacks:
7185 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7186 be put in Local/Makefile. This is like the never_users runtime option,
7187 but it cannot be overridden. The default setting is "root".
7189 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7190 prefix string with which any file named in a -C command line option
7193 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7194 is retained for -C and -D only if the caller of Exim is root. Without
7195 it, the exim user may also use -C and -D and retain privilege.
7197 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7198 command line option is disabled.
7200 6. Macro names set by the -D option must start with an upper case letter, just
7201 like macro names defined in the configuration file.
7203 7. Added "dereference=" facility to LDAP.
7205 8. Two instances of the typo "uknown" in the source files are fixed.
7207 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7208 the Configure-Makefile script screwed up while processing it.
7210 10. Incorporated PCRE 4.4.
7212 11. The SMTP synchronization check was not operating right at the start of an
7213 SMTP session. For example, it could not catch a HELO sent before the client
7214 waited for the greeting. There is now a check for outstanding input at the
7215 point when the greeting is written. Because of the duplex, asynchronous
7216 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7217 way, but not yet received, when the check is performed.
7219 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7220 on TCP/IP sockets, because this apparently causes some broken clients to
7223 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7224 unchanged) from the Cygwin maintainer.
7226 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7227 was supported instead of testing for mbx. Effectively a typo.
7229 15. The spa authenticator server code was not checking that the input it
7230 received was valid base64.
7232 16. The debug output line for the "set" modifier in ACLs was not showing the
7233 name of the variable that was being set.
7235 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7237 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7238 lookup on its own. Something else was needed to trigger the lookup. For
7239 example, a match in host_lookup or the need for a host name in a host list.
7240 Now, if $sender_host_name is referenced and the host name has not yet been
7241 looked up, a lookup is performed. If the lookup fails, the variable remains
7242 empty, and $host_lookup_failed is set to "1".
7244 19. Added "eqi" as a case-independent comparison operator.
7246 20. The saslauthd authentication condition could segfault if neither service
7247 nor realm was specified.
7249 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7250 error message that was logged was misleading, and incoming SMTP
7251 connections were dropped. The message is now more accurate, and temporary
7252 errors are given to SMTP connections.
7254 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7255 (except RSET) with a 421 error, until QUIT is received. However, it was
7256 failing to send a response to QUIT.
7258 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7259 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7260 ACL is now run after the helo_try_verify_hosts code.
7262 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7263 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7264 case-independent, so other case variants are also recognized). Apparently
7265 some systems use these upper case variants.
7267 25. If more than two messages were waiting for the same host, and a transport
7268 filter was specified for the transport, Exim sent two messages over the
7269 same TCP/IP connection, and then failed with "socket operation on non-
7270 socket" when it tried to send the third.
7272 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7274 27. The extern definition of crypt16() in expand.c was not being excluded when
7275 the OS had its own crypt16() function.
7277 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7278 as a preferred synonym for return_size_limit, both as an option and as an
7281 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7283 30. Changed the default configuration ACL to relax the local part checking rule
7284 for addresses that are not in any local domains. For these addresses,
7285 slashes and pipe symbols are allowed within local parts, but the sequence
7286 /../ is explicitly forbidden.
7288 31. SPA server authentication was not clearing the challenge buffer before
7291 32. log_message in a "warn" ACL statement was writing to the reject log as
7292 well as to the main log, which contradicts the documentation and doesn't
7293 seem right (because no rejection is happening). So I have stopped it.
7295 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7296 However, I am unable to do any testing of this.
7298 34. Fixed an infelicity in the appendfile transport. When checking directories
7299 for a mailbox, to see if any needed to be created, it was accidentally
7300 using path names with one or more superfluous leading slashes; tracing
7301 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7303 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7304 contents are added to the log line that is written for every discarded
7305 recipient. (Previously a log_message setting was ignored.)
7307 36. The ${quote: operator now quotes the string if it is empty.
7309 37. The install script runs exim in order to find its version number. If for
7310 some reason other than non-existence or emptiness, which it checks, it
7311 could not run './exim', it was installing it with an empty version number,
7312 i.e. as "exim-". This error state is now caught, and the installation is
7315 38. An argument was missing from the function that creates an error message
7316 when Exim fails to connect to the socket for saslauthd authentication.
7317 This could cause Exim to crash, or give a corrupted message.
7319 39. Added isip, isip4, and isip6 to ${if conditions.
7321 40. The ACL variables $acl_xx are now saved with the message, and can be
7322 accessed later in routers, transports, and filters.
7324 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7325 strings in the file are not string-expanded.
7327 42. If a MAIL command specified a SIZE value that was too large to fit into an
7328 int variable, the check against message_size_limit failed. Such values are
7329 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7330 day this will have to be increased, but I don't think I want to be around
7331 when emails are that large.
7338 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7339 iconv() is not standard in FreeBSD.
7341 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7342 IPv6 enabled. The observed symptom was a segmentation fault on return from
7343 the function os_common_find_running_interfaces() in src/os.c.
7345 3. In the check_special_case() function in daemon.c I had used "errno" as an
7346 argument name, which causes warnings on some systems. This was basically a
7347 typo, since it was named "eno" in the comments!
7349 4. The code that waits for the clock to tick (at a resolution of some fraction
7350 of a second) so as to ensure message-id uniqueness was always waiting for
7351 at least one whole tick, when it could have waited for less. [This is
7352 almost certainly not relevant at current processor speeds, where it is
7353 unlikely to ever wait at all. But we try to future-proof.]
7355 5. The function that sleeps for a time interval that includes fractions of a
7356 second contained a race. It did not block SIGALRM between setting the
7357 timer, and suspending (a couple of lines later). If the interval was short
7358 and the sigsuspend() was delayed until after it had expired, the suspension
7359 never ended. On busy systems this could lead to processes getting stuck for
7362 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7363 process, before it forks any delivery processes. The open lookup caching
7364 mechanism meant that the open file or database connection was passed into
7365 the delivery process. The problem was that delivery processes always tidy
7366 up cached lookup data. This could cause a problem for the next delivery
7367 process started by the queue runner, because the external queue runner
7368 process does not know about the closure. So the next delivery process
7369 still has data in the lookup cache. In the case of a file lookup, there was
7370 no problem because closing a file descriptor in a subprocess doesn't affect
7371 the parent. However, if the lookup was caching a connection to a database,
7372 the connection was closed, and the second delivery process was likely to
7373 see errors such as "PGSQL: query failed: server closed the connection
7374 unexpectedly". The problem has been fixed by closing all cached lookups
7375 in a queue runner before running a delivery process.
7377 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7378 seem to have the "const" qualifier which it has on other OS. I've
7381 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7382 *of the same type* (client or server) with the same public name that an
7383 error should be diagnosed.
7385 9. When Exim looked up a host name for an IP address, but failed to find the
7386 original IP address when looking up the host name (a safety check), it
7387 output the message "<ip address> does not match any IP for NULL", which was
7388 confusing, to say the least. The bug was that the host name should have
7389 appeared instead of "NULL".
7391 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7392 user that is built into the binary, and the -C or -D options is used, root
7393 privilege is dropped before the configuration file is read. In addition,
7394 logging is switched to stderr instead of the normal log files. If the
7395 configuration then re-defines the Exim user, the unprivileged environment
7396 is probably not what is expected, so Exim logs a panic warning message (but
7399 However, if deliver_drop_privilege is set, the unprivileged state may well
7400 be exactly what is intended, so the warning has been cut out in that case,
7401 and Exim is allowed to try to write to its normal log files.
7407 1. smtp_return_error_details was not giving details for temporary sender
7408 or receiver verification errors.
7410 2. Diagnose a configuration error if two authenticators have the same public
7413 3. Exim used not to create the message log file for a message until the first
7414 delivery attempt. This could be confusing when incoming messages were held
7415 for policy or load reasons. The message log file is now created at the time
7416 the message is received, and an initial "Received" line is written to it.
7418 4. The automatically generated man page for command line options had a minor
7419 bug that caused no ill effects; however, a more serious problem was that
7420 the procedure for building the man page automatically didn't always
7421 operate. Consequently, release 4.20 contains an out-of-date version. This
7422 shouldn't happen again.
7424 5. When building Exim with embedded Perl support, the script that builds the
7425 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7426 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7428 6. The freeze_tell option was not being used for messages that were frozen on
7429 arrival, either by an ACL or by local_scan().
7431 7. Added the smtp_incomplete_transaction log selector.
7433 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7434 was accepting AUTH without a new EHLO.
7436 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7437 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7440 10. Logging of TCP/IP connections (when configured) now happens in the main
7441 daemon process instead of the child process, so that the TCP/IP connection
7442 count is more accurate (but it can never be perfect).
7444 11. The use of "drop" in a nested ACL was not being handled correctly in the
7445 outer ACL. Now, if condition failure induced by the nested "drop" causes
7446 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7447 or "require"), the connection is dropped.
7449 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7450 that yield "discard" can now be used with an "accept" or a "discard" verb,
7451 but an error is generated for any others (because I can't see a useful way
7452 to define what should happen).
7454 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7455 are now processed in the same way as lines in the Exim configuration file.
7456 In particular, continuation lines are supported.
7458 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7460 15. Added -ti meaning -t -i.
7462 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7463 domains, and warn by logging if others are found.
7465 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7466 returned from GIFCONF when Exim is trying to find the list of interfaces on
7467 a host. The code in os.c has been modified to copy each ifreq to an aligned
7468 structure in all cases.
7470 Also, in some cases, the returned ifreq's were being copied to a 'struct
7471 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7472 means the last couple of bytes of an IPv6 address could be chopped if the
7473 ifreq contained only a normal sockaddr (14 bytes storage).
7475 18. Named domain lists were not supported in the hosts_treat_as_local option.
7476 An entry such as +xxxx was not recognized, and was treated as a literal
7479 19. Ensure that header lines added by a DATA ACL are included in the reject log
7480 if the ACL subsequently rejects the message.
7482 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7483 MD5 (which is deprecated).
7485 21. When testing a filter file using -bf, Exim was writing a message when it
7486 took the sender from a "From " line in the message, but it was not doing so
7487 when it took $return_path from a Return-Path: header line. It now does.
7489 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7490 with a valid header line field name (a series of printing characters
7491 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7493 23. Changed "disc" in the source to "disk" to conform to the documentation and
7494 the book and for uniformity.
7496 24. Ignore Sendmail's -Ooption=value command line item.
7498 25. When execve() failed while trying to run a command in a pipe transport,
7499 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7500 could be confused with a return value of 69 from the command itself. This
7501 has been changed to 127, the value the shell returns if it is asked to run
7502 a non-existent command. The wording for the related log line suggests a
7503 non-existent command as the problem.
7505 26. If received_header_text expands to an empty string, do not add a Received:
7506 header line to the message. (Well, it adds a token one on the spool, but
7507 marks it "old" so that it doesn't get used or transmitted.)
7509 27. Installed eximstats 1.28 (addition of -nt option).
7511 28. There was no check for failure on the call to getsockname() in the daemon
7512 code. This can fail if there is a shortage of resources on the system, with
7513 ENOMEM, for example. A temporary error is now given on failure.
7515 29. Contrary to the C standard, it seems that in some environments, the
7516 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7517 program. Exim now does this explicitly; it affects the formatting of
7518 timestamps using strftime().
7520 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7521 complaints. I've now initialized all the variables, to avoid this.
7523 32. Header lines added by a system filter were not being "seen" during
7524 transport-time rewrites.
7526 33. The info_callback() function passed to OpenSSL is set up with type void
7527 (*)(SSL *, int, int), as described somewhere. However, when calling the
7528 function (actually a macro) that sets it up, the type void(*)() is
7529 expected. I've put in a cast to prevent warnings from picky compilers.
7531 34. If a DNS black list lookup found a CNAME record, but there were no A
7532 records associated with the domain it pointed at, Exim crashed.
7534 35. If a DNS black list lookup returned more than one A record, Exim ignored
7535 all but the first. It now scans all returned addresses if a particular IP
7536 value is being sought. In this situation, the contents of the
7537 $dnslist_value variable are a list of all the addresses, separated by a
7540 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7541 to accept a host name and all its aliases if the forward lookup for any of
7542 them yielded the IP address of the incoming connection. Now it accepts only
7543 those names whose forward lookup yields the correct IP address. Any other
7544 names are discarded. This closes a loophole whereby a rogue DNS
7545 administrator could create reverse DNS records to break through a
7546 wildcarded host restriction in an ACL.
7548 37. If a user filter or a system filter that ran in a subprocess used any of
7549 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7550 the wrong values were passed to the pipe command ($thisaddress had the
7551 value of $0, $0 had the value of $1, etc). This bug was introduced by
7552 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7554 38. Improved the line breaking for long SMTP error messages from ACLs.
7555 Previously, if there was no break point between 40 and 75 characters, Exim
7556 left the rest of the message alone. Two changes have been made: (a) I've
7557 reduced the minimum length to 35 characters; (b) if it can't find a break
7558 point between 35 and 75 characters, it looks ahead and uses the first one
7559 that it finds. This may give the occasional overlong line, but at least the
7560 remaining text gets split now.
7562 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7563 file descriptors might be low, and that setting 1000 would always raise it.
7564 It turns out that in some environments, the limit is already over 1000 and
7565 that lowering it causes trouble. So now Exim takes care not to decrease it.
7567 40. When delivering a message, the value of $return_path is set to $sender_
7568 address at the start of routing (routers may change the value). By an
7569 oversight, this default was not being set up when an address was tested by
7570 -bt or -bv, which affected the outcome if any router or filter referred to
7573 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7574 log only when "message" or "log_message" are set. However, if one of the
7575 conditions was an address verification, or a call to a nested ACL, the
7576 messages generated by the underlying test were being passed through. This
7577 no longer happens. The underlying message is available in $acl_verify_
7578 message for both "message" and "log_message" expansions, so it can be
7579 passed through if needed.
7581 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7582 new expansion $bh_ to give the encoded byte string without charset
7583 translation. Translation happens only if iconv() is available; HAVE_ICONV
7584 indicates this at build time. HEADERS_CHARSET gives the charset to
7585 translate to; headers_charset can change it in the configuration, and
7586 "headers charset" can change it in an individual filter file.
7588 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7589 that creates RFC 2047 encoded "words" labels them as that charset instead
7590 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7591 expansion operator; (ii) when Exim creates a From: line for a local
7592 message; (iii) when a header line is rewritten to include a "phrase" part.
7594 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7595 buggy, causing it to skip the first lines of messages whose message ID
7596 ended in 'D'. This would not have bitten before Exim release 4.14, because
7597 message IDs were unlikely to end in 'D' before then. The effect was to have
7598 incorrect size information for certain domains.
7600 45. #include "config.h" was missing at the start of the crypt16.c module. This
7601 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7604 46. If there was a timeout during a "random" callout check, Exim treated it as
7605 a failure of the random address, and carried on sending RSET and the real
7606 address. If the delay was just some slowness somewhere, the response to the
7607 original RCPT would be taken as a response to RSET and so on, causing
7608 mayhem of various kinds.
7610 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7611 when I implemented it. It didn't allow for the fact that some option values
7612 may legitimately be negative (e.g. size_addition), and it didn't even do
7613 the right test for positive values.
7615 48. Domain names in DNS records are case-independent. Exim always looks them up
7616 in lower case. Some resolvers return domain names in exactly the case they
7617 appear in the zone file, that is, they may contain uppercase letters. Not
7618 all resolvers do this - some return always lower case. Exim was treating a
7619 change of case by a resolver as a change of domain, similar to a widening
7620 of a domain abbreviation. This triggered its re-routing code and so it was
7621 trying to route what was effectively the same domain again. This normally
7622 caused routing to fail (because the router wouldn't handle the domain
7623 twice). Now Exim checks for this case specially, and just changes the
7624 casing of the domain that it ultimately uses when it transmits the message
7627 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7630 50. If a filter generated a file delivery with a non-absolute name (possible if
7631 no home directory exists for the router), the forbid_file option was not
7634 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7635 the existing equality matching.
7637 52. Exim was using ints instead of ino_t variables in some places where it was
7638 dealing with inode numbers.
7640 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7641 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7642 called TMPDIR, and if it finds it is different, it changes its value.
7644 54. The smtp_printf() function is now made available to local_scan() so
7645 additional output lines can be written before returning. There is also an
7646 smtp_fflush() function to enable the detection of a dropped connection.
7647 The variables smtp_input and smtp_batched_input are exported to
7650 55. Changed the default runtime configuration: the message "Unknown user"
7651 has been removed from the ACL, and instead placed on the localuser router,
7652 using the cannot_route_message feature. This means that any verification
7653 failures that generate their own messages won't get overridden. Similarly,
7654 the "Unrouteable address" message that was in the ACL for unverifiable
7655 relay addresses has also been removed.
7657 56. Added hosts_avoid_esmtp to the smtp transport.
7659 57. The exicyclog script was not checking for the esoteric option
7660 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7661 will work only if exicyclog is run under the appropriate euid.
7663 58. Following a discussion on the list, the rules by which Exim recognises line
7664 endings on incoming messages have been changed. The -dropcr and drop_cr
7665 options are now no-ops, retained only for backwards compatibility. The
7666 following line terminators are recognized: LF CRLF CR. However, special
7667 processing applies to CR:
7669 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7670 nor a local message in the state where . is a terminator.
7672 (ii) If a bare CR is encountered in a header line, an extra space is added
7673 after the line terminator so as not to end the header. The reasoning
7674 behind this is that bare CRs in header lines are most likely either
7675 to be mistakes, or people trying to play silly games.
7677 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7678 was being incorrectly given as 18 bytes larger than it should have been.
7679 This is a VOB (very old bug).
7681 60. This may never have affected anything current, but just in case it has:
7682 When the local host is found other than at the start of a list of hosts,
7683 the local host, those with the same MX, and any that follow, are discarded.
7684 When the list in question was part of a longer list of hosts, the following
7685 hosts (not currently being processed) were also being discarded. This no
7686 longer happens. I'm not sure if this situation could ever has previously
7689 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7692 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7693 header. This is something that is recommended in a new Internet Draft, and
7694 is something that is documented as being done by Sendmail. There are two
7695 possible values. For messages generated by the autoreply transport, Exim
7698 Auto-Submitted: auto-replied
7700 whereas for all other generated messages (e.g. bounces) it adds
7702 Auto-Submitted: auto-generated
7704 63. The "personal" condition in filters now includes a test for the
7705 Auto-Submitted: header. If it contains the string "auto-" the message it
7706 not considered personal.
7708 64. Added rcpt_include_affixes as a generic transport option.
7710 65. Added queue_only_override (default true).
7712 66. Added the syslog_duplication option.
7714 67. If what should have been the first header line of a message consisted of
7715 a space followed by a colon, Exim was mis-interpreting it as a header line.
7716 It isn't of course - it is syntactically invalid and should therefore be
7717 treated as the start of the message body. The misbehaviour could have
7718 caused a number of strange effects, including loss of data in subsequent
7719 header lines, and spool format errors.
7721 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7722 client host had authenticated. This control can now be exercised by an ACL
7723 for more flexibility.
7725 69. By default, callouts do not happen when testing with -bh. There is now a
7726 variant, -bhc, which does actually run the callout code, including
7727 consulting and updating the callout cache.
7729 70. Added support for saslauthd authentication, courtesy of Alexander
7732 71. If statvfs() failed on the spool or log directories while checking their
7733 size for availability, Exim confusingly gave the error "space shortage".
7734 Furthermore, in debugging mode it crashed with a floating point exception.
7735 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7736 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7737 serious problem, Exim now writes to the main and panic logs when this
7738 happens, with details of the failure. It then refuses to accept the
7739 incoming message, giving the message "spool directory problem" or "log
7740 directory problem" with a 421 code for SMTP messages.
7742 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7743 0, 1, and 2 exist, because some OS complain for execs without them (see
7744 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7745 descriptors. However, the code omitted to check that the open succeeded,
7746 causing mysterious errors if for some reason the permissions on /dev/null
7747 got screwed. Now Exim writes a message to the main and panic logs, and
7748 bombs out if it can't open /dev/null.
7750 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7751 interact so that it is all more flexible. It is supposed to remain
7752 backwards compatible. Also added extra_local_interfaces.
7754 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7755 to bomb out with an assertion failure - to the client this appears as a
7756 connection drop. This problem occurs in the part of the code that was taken
7757 from the Samba project. Fortunately, the assertion is in a very simple
7758 function, so I have fixed this by reproducing the function inline in the
7759 one place where it is called, and arranging for authentication to fail
7760 instead of killing the process with assert().
7762 75. The SPA client code was not working when the server requested OEM rather
7763 than Unicode encoding.
7765 76. Added code to make require_files with a specific uid setting more usable in
7766 the case where statting the file as root fails - usually a non-root-mounted
7767 NFS file system. When this happens and the failure is EACCES, Exim now
7768 forks a subprocess and does the per-uid checking as the relevant uid.
7770 77. Added process_log_path.
7772 78. If log_file_path was not explicitly set, a setting of check_log_space or
7773 check_log_inodes was ignored.
7775 79. If a space check for the spool or log partitions fails, the incident is now
7776 logged. Of course, in the latter case the data may get lost...
7778 80. Added the %p formatting code to string_format() so that it can be used to
7779 print addresses in debug_print(). Adjusted all the address printing in the
7780 debugging in store.c to use %p rather than %d.
7782 81. There was a concern that a line of code in smtp_in.c could overflow a
7783 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7784 initially expressed, the concern was not well-founded, because trailing
7785 spaces are removed early. However, if the trailing spaces were followed by
7786 a NULL, they did not get removed, so the overflow was possible. Two fixes
7789 (a) I re-wrote the offending code in a cleaner fashion.
7790 (b) If an incoming SMTP command contains a NULL character, it is rejected
7793 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7794 runs initgroups(), so that if the Exim user is in any additional groups,
7795 they will be used during message reception.
7801 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob