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 a local
15 buffer was taking a taint-enforcement trap. Fix by using dynamically
16 created buffers. Similar fix for radius expansion condition.
18 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
19 reasonable, eg. to count headers. Fix by using dynamically created
20 buffers rather than a local. Do similar fixes for ACL actions "dcc",
21 "log_reject_target", "malware" and "spam"; the arguments are expanded
22 so could be handling tainted values.
24 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
25 broken the (no-op) support for this sendmail command. Restore it
26 to doing nothing, silently, and returning good status.
28 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
29 record path was given (or the default used) without a leading directory
30 path, an error occurred on trying to open it. Use the transport's working
33 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
34 transport. Previously it was the name on the DNS A-record; use instead
35 the head of the CNAME chain leading there (if there is one). This seems
36 to align better with RFC 6125.
38 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
39 smtp_accept_max_per_host allocated resources which were not released
40 when the limit was exceeded. This eventually crashed the daemon. Fix
41 by adding a relase action in that path.
43 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
44 expanded; previously using tainted values was rejected. Fix by using
45 dynamically-created buffers.
47 JH/09 Relax restrictions on ACL verify condition needing access to message
48 headers. Previously they were only permitted in data and non-smtp ACLs;
49 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
50 not_blind, header_sender and header_names_ascii verification.
52 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
53 Previously a macro used one argument twice; when called with the
54 argument as an expression having side-effects, incorrect operation
55 resulted. Use an inlineable function.
57 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
58 held open for a verify callout. Previously this wan not accounted for
59 and a corrupt onward SMTP conversation resulted.
61 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
62 Previously they were embedded naively in the constructed address; when
63 needed, strip the quoting and quote the entire local_part.
64 Also make the inbound_srs expansion condition handle quoting.
66 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
67 excluded, not matching the documentation.
69 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
70 was given for the sqlite_dbfile a trap resulted.
72 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
73 "name" argument resulted in a trap. There is no reason to disallow such;
74 this was a coding error.
76 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
77 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
78 spent suspended, ignoring the Posix definition. Previously we assumed
79 it did and a constant offset from real time could be used as a correction.
80 Change to using the same clock source for the start-of-message and the
81 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
82 exists, just to get a clock slightly more aligned to reality.
84 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
85 RFC says it is optional some validators care. The missing char was not
86 intended but triggered by a line-wrap alignement. Discovery and fix by
87 Guillaume Outters, hacked on by JH.
89 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
90 name being quoted was tainted a trap would be taken. Fix by using
91 dynamicaly created buffers. The routine could have been called by a
92 rewrite with the "h" flag, by using the "-F" command-line option, or
93 by using a "name=" option on a control=submission ACL modifier.
95 JH/19 SPF: change the Authentication-Results expansion component to give
96 smtp.helo when the sender domain is empty. Previously it gave
99 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
100 not in 127.0.0.0/8 to help in spotting list domains taken over by a
101 domain-parking registrar.
103 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
104 Previously when a whitespace character was specified it was not inserted
105 after removing the newline.
107 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
108 the domain part of the recipient address. This overrides any tls_sni
109 option set, which was previously used.
111 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
114 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
115 is_tainted() had an off-by-one error in the overenthusiastic direction.
116 Find and fix by Gavan. Although NetBSD is not a supported platform for
117 4.94 this bug could affect other platforms.
119 PP/01 Fix default prime selection to be consistent.
120 One path used ike23 still, instead of exim.dev.20160529.3; now both
121 execution flows will use the same DH primes (currently
122 exim.dev.20160529.3).
124 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
125 option in smtp transport, to match the documentation. Previously
126 verification was not being done in some cases where it should have been.
128 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
129 than one server was defined and depending on the platform memory layout
130 details, an internal consistency trap could be hit while walking the list
133 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
134 files. The value is used by the authresults expansion item. Previously
135 if this was used in a router or transport, a crash could result.
137 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
138 have a sufficient incidence of EINTR returns from select that an
139 interaction with other operations done by the main daemon loop exposed
140 a bug in the error-handling. This was benign apart from the log
143 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
144 for consistency with delivered "=>" and failed "**" lines. While we're
145 there, handle PRX and TFO.
147 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
148 applied. This resulted, if any header-line rewrite rules were configured,
149 in a panic-log trigerrable by sending a message with a long address in
150 a header. Fix by increaing the arbitrary limit to larger than a single
151 (dewrapped) 5322 header line maximum size.
153 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
154 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
155 RFC 6648 which deprecates X- options in protocols as a general practice.
156 Changeover between the implementations is handled by the mechanisms
159 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
160 a remote which uses callout/hold. Previously the local was queued.
162 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
165 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
166 log line. The asterisk is now consistently appended to the remote IP
167 (and port, if given), and will also be provided on defer and fail log
168 lines. Previously it could be placed on the local IP if that was being
169 logged, and was only provided on delivery lines.
171 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
173 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
174 authenticator client_send option. Previously the next char, after a pair
175 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
176 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
177 documentation. There is still no way to get a leading ^ immediately
178 after a NUL (ie. for the password of a PLAIN method authenticator.
180 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
181 files. For bad sizes read, delete the record and whine to paniclog.
183 JH/38 When logging an AUTH failure, as server, do not include sensitive
184 information. Previously, the credentials would be included if given
185 as part of the AUTH command line and an ACL denied authentidcation.
192 JH/01 Avoid costly startup code when not strictly needed. This reduces time
193 for some exim process initialisations. It does mean that the logging
194 of TLS configuration problems is only done for the daemon startup.
196 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
198 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
199 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
202 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
203 DKIM signing is being done. Previously a transport_filter would always
204 disable CHUNKING, falling back to traditional DATA.
206 JH/05 Regard command-line recipients as tainted.
208 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
210 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
211 PAM library frees one of the arguments given to it, despite the
212 documentation. Therefore a plain malloc must be used.
214 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
215 on-stack buffers were used, resulting in a taint trap when DSN information
216 copied from a received message was written into the buffer.
218 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
219 the ordering of its ARC headers. This caused a crash.
221 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
222 a new record was being constructed with information from the peer, a trap
225 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
226 installation would get error messages from DMARC verify, when it hit the
227 nonexistent file indicated by the default. Distros wanting DMARC enabled
228 should both provide the file and set the option.
229 Also enforce no DMARC verification for command-line sourced messages.
231 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
232 could, depending on the platform, hang at the STARTTLS response.
234 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
235 message on a connection. Previously if one message had ARC headers and
236 the following one did not, a crash could result when adding an
237 Authentication-Results: header.
239 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
240 Exim main code and Exim-related utities. The introduction of taint
241 tracking also did many adjustments to string handling. Since then, eximon
242 frequently terminated with an assert failure.
244 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
245 check for 452 responses. This slightly helps the inefficieny of doing
246 a large alias-expansion into a recipient-limited target. The max_rcpt
247 transport option still applies (and at the current default, will override
248 the new feature). The check is done for either cause of synch, and forces
249 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
250 connection. The new facility is not tunable at this time.
252 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
253 library live data was being used, so the results became garbage. Make
254 copies while it is still usable.
256 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
257 on delivery deferred (==) and failed (**) lines (if a delivery was
258 attemtped). Previously it was only on completion (=>) lines.
260 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
261 for the expansion of the server_scram_iter and server_scram_salt options.
263 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
264 are now specifically given a NO_DATA response without hitting the system
265 resolver. The library goes on to do the now-standard TXT lookup.
266 Use of dnsdb lookups is not affected.
268 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
269 only retrieve the errormessage once. Previously two calls to dlerror()
270 were used, and the second one (for mainlog/paniclog) retrieved null
273 JH/20 Taint checking: disallow use of tainted data for
274 - the appendfile transport file and directory options
275 - the pipe transport command
276 - the autoreply transport file, log and once options
277 - file names used by the redirect router (including filter files)
279 - paths used by single-key lookups
280 Previously this was permitted.
282 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
283 adjusted the size of a major service buffer; this failed because the
284 buffer was in use at the time. Change to a compile-time increase in the
285 buffer size, when this authenticator is compiled into exim.
287 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
288 previous fast-mode was untenable in the face of glibs using mmap to
289 support larger malloc requests.
291 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
292 New values supported, if defined on system where compiled:
293 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
294 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
296 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
297 running a limited number of proceses in parallel, a benefit is gained. The
298 amount varies with the platform hardware and load. The use of the option
299 queue_run_in_order means we cannot do this, as ordering becomes
302 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
303 had introduced a string-copy (for ensuring NUL-termination) which was not
304 appropriate for that case, which can include embedded NUL bytes in the
305 block of data. Investigation showed the copy to actually be needless, the
306 data being length-specified.
308 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
309 done during a receiving connection, and both used TLS, global info was
310 used rather than per-connection info for tracking the state of data
311 queued for transmission. This could result in a connection hang.
313 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
314 Previously, when delivering serveral messages down a single connection
315 only the first would provide a SIZE. This was due to the size information
316 not being properly tracked.
318 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
319 TAI (at 37 seconds currently), pretend to be in UTC for time-related
320 expansion and logging. Previously, spurious values such as a future
321 minute could be seen.
323 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
324 it could crash from a null-deref. This could also affect the
325 ${addresses: } operator and ${readsock } item.
327 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
328 message following a mime one, the variable was not reset.
330 JH/30 When an pipelined-connect fails at the first response, assume incorrect
331 cached capability (perhaps the peer reneged?) and immediately retry in
334 JH/31 Fix spurious detection of timeout while writing to transport filter.
336 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
337 an attempt to copy the string was made before checking it.
339 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
340 taint of the lookup key was maintained; we now regard the presence in the
341 filesystem as sufficient validation.
343 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
344 argument is supplied.
346 JH/35 The dsearch lookup now requires that the directory is an absolute path.
347 Previously this was not checked, and nonempty relative paths made an
348 access under Exim's current working directory.
350 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
351 Previously no event was raised.
353 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
354 parameter supplied by the sender MAIL FROM command. Previously it was
355 ignored, and only the check_spool_space option value for the required
358 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
359 the size of the signing public-key. Previously it was instead giving
360 the size of the signature hash.
362 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
363 the default. See the (new) dkim_verify_min_keysizes option.
365 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
366 and an ACL use a lookup for checking either the local_part or domain,
367 stale data could be accessed. Ensure that variable references are
368 dropped between messages.
370 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
371 by the client was not checked as pointing within response data before
372 being used. A malicious client could thus cause an out-of-bounds read and
373 possibly gain authentication. Fix by adding the check.
375 JH/42 Internationalisation: change the default for downconversion in the smtp
376 transport to be "if needed". Previously it was "as previously set" for
377 the message, which usually meant "if needed" for message-submission but
378 "no" for everything else. However, MTAs have been seen using SMTPUTF8
379 even when the envelope addresses did not need it, resulting in forwarding
380 failures to non-supporting MTAs. A downconvert in such cases will be
381 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
382 transport. The change does mean that addresses needing conversion will
383 be converted when previously a delivery failure would occur.
385 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
386 response was received it would be used unchecked in a fail-DSN, violating
387 standards on line-length limits. Truncate if needed.
389 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
390 doesn't work. (Additionally add a new main config option to configure the
397 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
398 side, to decode a TLS 1.3 packet capture.
400 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
401 Previously the default library behaviour applied, sending two, each in
404 JH/03 Debug output for ACL now gives the config file name and line number for
407 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
409 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
411 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
412 buffer overrun for (non-chunking) other transports.
414 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
415 TLS1.3, means that a server rejecting a client certificate is not visible
416 to the client until the first read of encrypted data (typically the
417 response to EHLO). Add detection for that case and treat it as a failed
418 TLS connection attempt, so that the normal retry-in-clear can work (if
419 suitably configured).
421 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
422 and/or domain. Found and fixed by Jason Betts.
424 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
425 configuration). If a CNAME target was not a wellformed name pattern, a
428 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
429 the OS reports them interleaved with other addresses.
431 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
432 used both for input and for a verify callout, both encrypted, SMTP
433 responses being sent by the server could be lost. This resulted in
434 dropped connections and sometimes bounces generated by a peer sending
437 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
438 string. Previously it was possible to cause undefined behaviour in a
439 library routine (usually a crash). Found by "zerons".
441 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
444 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
445 API was removed, so update to use the newer ones.
447 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
448 any timeout set, is taking a long time. Previously we would hang on to a
449 rotated logfile "forever" if the input was arriving with long gaps
450 (a previous attempt to fix addressed lack, for a long time, of initial
453 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
454 shared (NFS) environment. The length of the tempfile name is now
455 4 + 16 ("hdr.$message_exim_id") which might break on file
456 systems which restrict the file name length to lower values.
459 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
460 shared (NFS) environment.
462 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
463 did for all versions <4.90). Notably -M, -m, --invert, -I may be
466 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
467 on some platforms for bit 31.
469 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
470 to changes apparently associated with TLS1.3 handling some of the APIs
471 previously used were either nonfunctional or inappropriate. Strings
472 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
473 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
474 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
475 This affects log line X= elements, the $tls_{in,out}_cipher variables,
476 and the use of specific cipher names in the encrypted= ACL condition.
478 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
480 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
481 verification result was not updated unless hosts_require_ocsp applied.
483 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
484 queue_list_requires_admin set to false, non-admin users were denied the
487 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
488 directory-of-certs mode. Previously they were advertised despite the
491 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
492 A single TCP connection by a client will now hold a TLS connection open
493 for multiple message deliveries, by default. Previously the default was to
496 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
497 default. If built with the facility, DANE will be used. The facility
498 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
500 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
501 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
502 must be defined and you must still, unless you define DISABLE_TLS, manage
503 the the include-dir and library-file requirements that go with that
504 choice. Non-TLS builds are still supported.
506 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
509 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
510 default. If the platform supports and has the facility enabled, it will
511 be requested on all coneections.
513 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
514 controlled by the build-time option SUPPORT_PIPE_CONNECT.
516 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
518 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
519 success-DSN messages. Previously the From: header was always the default
520 one for these; the option was ignored.
522 JH/28 Fix the timeout on smtp response to apply to the whole response.
523 Previously it was reset for every read, so a teergrubing peer sending
524 single bytes within the time limit could extend the connection for a
525 long time. Credit to Qualsys Security Advisory Team for the discovery.
527 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
528 delivery address, which leaked information of the results of local
529 forwarding. Change to the original envelope recipient address, per
532 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
533 requested. Previously not bounce was generated and a log entry of
534 error ignored was made.
536 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
538 JH/32 Introduce a general tainting mechanism for values read from the input
539 channel, and values derived from them. Refuse to expand any tainted
540 values, to catch one form of exploit.
542 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
543 was unused and the unexpanded text used for the test. Found and
544 fixed by Ruben Jenster.
546 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
547 an attempt to use a TLS library read routine dereffed a nul pointer,
550 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
551 them in our smtp response.
553 JH/36 Have the general router option retry_use_local_part default to true when
554 any of the restrictive preconditions are set (to anything). Previously it
555 was only for check_local user. The change removes one item of manual
556 configuration which is required for proper retries when a remote router
557 handles a subset of addresses for a domain.
559 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
560 link count into consideration.
562 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
563 caused the extension of big_buffer, the following lines were ignored.
565 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
566 accordance with RFC 2308. Previously there was no expiry, so a longlived
567 receive process (eg. due to ACL delays) versus a short SOA value could
570 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
572 JH/39 Promote DMARC support to mainline.
574 JH/40 Bug 2452: Add a References: header to DSNs.
576 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
577 parameters. The relevant library call is documented as "Deprecated: This
578 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
579 3.6.0, DH parameters are negotiated following RFC7919."
581 HS/06 Change the default of dnssec_request_domains to "*"
583 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
584 carried on and emitted a BDAT command, even when PIPELINING was not
587 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
588 buffer was used for the filename, resulting in a trap when tainted
589 arguments (eg. $domain) were used.
591 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
592 recommended to avoid a possible server-load attack. The feature can be
593 re-enabled via the openssl_options main cofiguration option.
595 JH/45 local_scan API: documented the current smtp_printf() call. This changed
596 for version 4.90 - adding a "more data" boolean to the arguments.
597 Bumped the ABI version number also, this having been missed previously;
598 release versions 4.90 to 4.92.3 inclusive were effectively broken in
599 respect of usage of smtp_printf() by either local_scan code or libraries
600 accessed via the ${dlfunc } expansion item. Both will need coding
601 adjustment for any calls to smtp_printf() to match the new function
602 signature; a FALSE value for the new argument is always safe.
604 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
605 the file-offset (which the Linux syscall does, and exim expects); this
606 resulted in an indefinite loop.
608 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
609 to do ARC verification. The Authentication-Results: header line added
610 by the configuration then had no ARC item.
616 JH/01 Remove code calling the customisable local_scan function, unless a new
617 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
619 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
620 non-signal-safe functions being used.
622 JH/03 Bug 2269: When presented with a received message having a stupidly large
623 number of DKIM-Signature headers, disable DKIM verification to avoid
624 a resource-consumption attack. The limit is set at twenty.
626 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
627 report of oldest_pass in ${authres } in consequence, and separate out
628 some descriptions of reasons for verification fail.
630 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
631 files in the spool were present and unlocked. A queue-runner could spot
632 them, resulting in a duplicate delivery. Fix that by doing the unlock
633 after the unlink. Investigation by Tim Stewart. Take the opportunity to
634 add more error-checking on spoolfile handling while that code is being
637 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
638 No known attacks, no CVE, this is defensive hardening.
640 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
641 a queue-runner could start a delivery while other operations were ongoing.
642 Cutthrough delivery was a common victim, resulting in duplicate delivery.
643 Found and investigated by Tim Stewart. Fix by using the open message data
644 file handle rather than opening another, and not locally closing it (which
645 releases a lock) for that case, while creating the temporary .eml format
646 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
648 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
649 $sender_verify_failure/$recipient_verify_failure to "random".
651 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
654 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
655 Previously this would segfault.
657 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
660 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
661 like zero, since the resolver should be doing this for us, But we need one
662 as a CNAME but no MX presence gets the CNAME returned; we need to check
663 that doesn't point to an MX to declare it "no MX returned" rather than
664 "error, loop". A new main option is added so the older capability of
665 following some limited number of chain links is maintained.
667 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
669 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
670 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
671 now-deprecated earlier definitions used only specified the range up to TLS
672 1.2 (in the older-version library docs).
674 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
676 JH/15 Rework TLS client-side context management. Stop using a global, and
677 explicitly pass a context around. This enables future use of TLS for
678 connections to service-daemons (eg. malware scanning) while a client smtp
679 connection is using TLS; with cutthrough connections this is quite likely.
681 JH/16 Fix ARC verification to do AS checks in reverse order.
683 JH/17 Support a "tls" option on the ${readsocket } expansion item.
685 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
686 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
687 Previously the "utf8" would be re-prepended for every additional message.
689 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
690 Previously thery were accepted, resulting in issues when attempting to
691 forward messages to a non-supporting MTA.
693 PP/02 Let -n work with printing macros too, not just options.
695 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
696 one parent address was copied, and bogus data was used at delivery-logging
697 time. Either a crash (after delivery) or bogus log data could result.
698 Discovery and analysis by Tim Stewart.
700 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
701 Previously if the string ended mid-character, we did not insert the
702 promised '?' replacement.
704 PP/04 Documentation: current string operators work on bytes, not codepoints.
706 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
707 should pack well giving a smaller memory footprint so better caching and
708 therefore performance. Group the declarations where this can't be done so
709 that the byte-sized flag variables are not interspersed among pointer
710 variables, giving a better chance of good packing by the compiler.
712 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
713 non-null, to avoid issues with sites running BATV. Previously reports were
714 sent with an empty envelope sender so looked like bounces.
716 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
717 The ignore_error flag wasn't being returned from the filter subprocess so
718 was not set for later routers. Investigation and fix by Matthias Kurz.
720 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
721 and a msg:complete for the whole, when a message is manually removed using
722 -Mrm. Developement by Matthias Kurz, hacked on by JH.
724 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
725 a "Gnu special" function, asprintf() in the DB utility binary builds; I
726 hope that is portable enough.
728 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
729 requiring a known-CA anchor certificate; make it now rely entirely on the
730 TLSA as an anchor. Checking the name on the leaf cert against the name
731 on the A-record for the host is still done for TA (but not for EE mode).
733 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
734 included in delivery lines for non-proxied connections, when compiled with
735 SUPPORT_SOCKS and running with proxy logging enabled.
737 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
738 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
739 move the existing event to fire before the normal logging of message
740 failure so that custom logging is bracketed by normal logging.
742 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
743 msg:fail:internal event. Developement by Matthias Kurz.
745 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
746 far too small for todays use of crypto signatures stored there. Go all
747 the way to the max DNS message size of 64kB, even though this might be
748 overmuch for IOT constrained device use.
750 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
751 copy a string over itself. The library routine is documented as not
752 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
754 JH/32 For main options check_spool_space and check_inode_space, where the
755 platform supports 64b integers, support more than the previous 2^31 kB
756 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
757 the previous G, M, k.
759 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
760 $authenticated_fail_id variable on authentication failure. Previously
763 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
764 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
765 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
766 GNUTLS_SEC_PARAM_MEDIUM.
768 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
769 side. Previously we would continue as if no SNI had been received.
771 JH/36 Harden the handling of string-lists. When a list consisted of a sole
772 "<" character, which should be a list-separator specification, we walked
773 off past the nul-terimation.
775 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
776 causes) even when the retry time is not yet met. Previously they were
777 not, meaning that when (say) an account was over-quota and temp-rejecting,
778 and multiple senders' messages were queued, only one sender would get
779 notified on each configured delay_warning cycle.
781 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
783 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
784 cipher-suites, an error can be left on the stack even for a succeeding
785 accept; this results in impossible error messages when a later operation
788 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
789 return error codes indicating retry. Under TLS1.3 this becomes required.
791 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
792 it only wrote the new authenticators, resulting in a lack of tracking of
793 peer changes of ESMTP extensions until the next cache flush.
795 JH/41 Fix the loop reading a message header line to check for integer overflow,
796 and more-often against header_maxsize. Previously a crafted message could
797 induce a crash of the recive process; now the message is cleanly rejected.
799 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
800 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
806 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
807 When redis_servers is set to a list of > 1 element, and the Redis servers
808 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
809 case of MOVED into a DEFER case instead, thus moving the query onto the
810 next server in the list. For a cluster of N elements, all N servers must
811 be defined in redis_servers.
813 GF/02 Catch and remove uninitialized value warning in exiqsumm
814 Check for existence of @ARGV before looking at $ARGV[0]
816 JH/01 Replace the store_release() internal interface with store_newblock(),
817 which internalises the check required to safely use the old one, plus
818 the allocate and data copy operations duplicated in both (!) of the
819 extant use locations.
821 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
822 modifier. This matches the restriction on the commandline.
824 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
825 Previously only the last row was returned.
827 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
828 we assumed that tags in the header were well-formed, and parsed the
829 element content after inspecting only the first char of the tag.
830 Assumptions at that stage could crash the receive process on malformed
833 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
834 While running the DKIM ACL we operate on the Permanent memory pool so that
835 variables created with "set" persist to the DATA ACL. Also (at any time)
836 DNS lookups that fail create cache records using the Permanent pool. But
837 expansions release any allocations made on the current pool - so a dnsdb
838 lookup expansion done in the DKIM ACL releases the memory used for the
839 DNS negative-cache, and bad things result. Solution is to switch to the
840 Main pool for expansions.
841 While we're in that code, add checks on the DNS cache during store_reset,
842 active in the testsuite.
843 Problem spotted, and debugging aided, by Wolfgang Breyha.
845 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
846 When none of the hosts presented to a transport match an already-open
847 connection, close it and proceed with the list. Previously we would
848 queue the message. Spotted by Lena with Yahoo, probably involving
851 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
852 Previously a spurious "250 OK id=" response was appended to the proper
855 JH/08 The "support for" informational output now, which built with Content
856 Scanning support, has a line for the malware scanner interfaces compiled
857 in. Interface can be individually included or not at build time.
859 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
860 by the template makefile "src/EDITME". The "STREAM" support for an older
861 ClamAV interface method is removed.
863 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
864 rows affected is given instead).
866 JH/11 The runtime Berkeley DB library version is now additionally output by
867 "exim -d -bV". Previously only the compile-time version was shown.
869 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
870 SMTP connection. Previously, when one had more recipients than the
871 first, an abortive onward connection was made. Move to full support for
872 multiple onward connections in sequence, handling cutthrough connection
873 for all multi-message initiating connections.
875 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
876 routers. Previously, a multi-recipient message would fail to match the
877 onward-connection opened for the first recipient, and cause its closure.
879 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
880 a timeout on read on a GnuTLS initiating connection, resulting in the
881 initiating connection being dropped. This mattered most when the callout
882 was marked defer_ok. Fix to keep the two timeout-detection methods
885 JH/15 Relax results from ACL control request to enable cutthrough, in
886 unsupported situations, from error to silently (except under debug)
887 ignoring. This covers use with PRDR, frozen messages, queue-only and
890 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
892 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
893 metadata, resulting in a crash in free().
895 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
896 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
897 Broken also in d185889f4, with init system revamp.
899 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
900 Previously we abruptly closed the connection after reading a malware-
901 found indication; now we go on to read the "scan ok" response line,
904 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
905 ACL. Previously, a crash would result.
907 JH/19 Speed up macro lookups during configuration file read, by skipping non-
908 macro text after a replacement (previously it was only once per line) and
909 by skipping builtin macros when searching for an uppercase lead character.
911 JH/20 DANE support moved from Experimental to mainline. The Makefile control
912 for the build is renamed.
914 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
915 was allocated for every new TLS startup, meaning one per message. Fix
916 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
918 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
919 reported the original. Fix to report (as far as possible) the ACL
920 result replacing the original.
922 JH/23 Fix memory leak during multi-message connections using STARTTLS under
923 OpenSSL. Certificate information is loaded for every new TLS startup,
924 and the resources needed to be freed.
926 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
928 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
931 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
932 DATA response info to the (existing) per-recipient response info for
933 the "C=" log element. It can have useful tracking info from the
934 destination system. Patch from Simon Arlott.
936 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
937 length value. Previously this would segfault.
939 HS/02 Support Avast multiline protoocol, this allows passing flags to
940 newer versions of the scanner.
942 JH/28 Ensure that variables possibly set during message acceptance are marked
943 dead before release of memory in the daemon loop. This stops complaints
944 about them when the debug_store option is enabled. Discovered specifically
945 for sender_rate_period, but applies to a whole set of variables.
946 Do the same for the queue-runner and queue-list loops, for variables set
947 from spool message files. Do the same for the SMTP per-message loop, for
948 certain variables indirectly set in ACL operations.
950 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
951 as a multi-recipient message from a mailinglist manager). The coding had
952 an arbitrary cutoff number of characters while checking for more input;
953 enforced by writing a NUL into the buffer. This corrupted long / fast
954 input. The problem was exposed more widely when more pipelineing of SMTP
955 responses was introduced, and one Exim system was feeding another.
956 The symptom is log complaints of SMTP syntax error (NUL chars) on the
957 receiving system, and refused recipients seen by the sending system
958 (propating to people being dropped from mailing lists).
959 Discovered and pinpointed by David Carter.
961 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
962 replaced by the ${authresults } expansion.
964 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
966 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
967 allows proper process termination in container environments.
969 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
970 Previously the "final dot" had a newline after it; ensure it is CR,LF.
972 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
973 and "err_perm", deprecated since 4.83 when the RFC-defined words
974 "temperror" and "permerror" were introduced.
976 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
977 transport-filters or DKIM-signing. The restriction was lost in the
978 consolidation of verify-callout and delivery SMTP handling.
979 Extend the restriction to also cover ARC-signing.
981 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
982 in defer=pass mode supply a 450 to the initiator. Previously the message
985 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
986 tls_require_ciphers is used as before.
988 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
989 "pass_unscanned". Only tmpfails from the scanner are written to
990 the paniclog, as they may require admin intervention (permission
991 denied, license issues). Other scanner errors (like decompression
992 bombs) do not cause a paniclog entry.
994 JH/36 Fix reinitialisation of DKIM logging variable between messages.
995 Previously it was possible to log spurious information in receive log
998 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
999 triggered odd behaviour from Outlook Express clients.
1001 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1004 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1005 since the IETF WG has not yet settled on that versus the original
1006 "bare" representation.
1008 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1009 Previously the millisecond value corrupted the output.
1010 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1011 corrupted the output.
1017 JH/01 Rework error string handling in TLS interface so that the caller in
1018 more cases is responsible for logging. This permits library-sourced
1019 string to be attached to addresses during delivery, and collapses
1020 pairs of long lines into single ones.
1022 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1023 during configuration. Wildcards are allowed and expanded.
1025 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1026 This permits better logging.
1028 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1029 do not close it down and have the child transport start it up again on
1030 the passed-on TCP connection. Instead, proxy the child (and any
1031 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1032 affected: the continued delivery log lines do not have any DNSSEC, TLS
1033 Certificate or OCSP information. TLS cipher information is still logged.
1035 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1036 identical IP addresses on different listening ports. Will also affect
1039 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1040 add noisy ifdef guards to special-case this sillyness.
1041 Patch from Bernd Kuhls.
1043 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1044 than 255 are no longer allowed.
1046 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1047 Disable the session-cache too, which might reduce our load. Since we
1048 currrectly use a new context for every connection, both as server and
1049 client, there is no benefit for these.
1050 GnuTLS appears to not support tickets server-side by default (we don't
1051 call gnutls_session_ticket_enable_server()) but client side is enabled
1052 by default on recent versions (3.1.3 +) unless the PFS priority string
1055 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1056 <https://reproducible-builds.org/specs/source-date-epoch/>.
1058 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1059 the check for any unsuccessful recipients did not notice the limit, and
1060 erroneously found still-pending ones.
1062 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1063 MSG_MORE. Only in-clear (not on TLS connections).
1065 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1066 transport-filter is involved we can buffer the headers in memory for
1067 creating the signature, and read the spool data file once for the
1068 signature and again for transmission.
1070 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1071 4.77 as the kernel support then wasn't solid, having issues in 64bit
1072 mode. Now, it's been long enough. Add support for FreeBSD also.
1074 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1075 case where the routing stage had gathered several addresses to send to
1076 a host before calling the transport for the first, we previously failed
1077 to close down TLS in the old transport process before passing the TCP
1078 connection to the new process. The new one sent a STARTTLS command
1079 which naturally failed, giving a failed delivery and bloating the retry
1080 database. Investigation and fix prototype from Wolfgang Breyha.
1082 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1083 false-negatives in the check that the sender had not preempted a response
1084 or prompt from Exim (running as a server), due to that code's lack of
1085 awareness of the SMTP input buffering.
1087 PP/04 Add commandline_checks_require_admin option.
1088 Exim drops privileges sanely, various checks such as -be aren't a
1089 security problem, as long as you trust local users with access to their
1090 own account. When invoked by services which pass untrusted data to
1091 Exim, this might be an issue. Set this option in main configuration
1092 AND make fixes to the calling application, such as using `--` to stop
1095 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1096 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1097 responses to those, into a single TLS record each way (this usually means
1098 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1101 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1102 affects you only if you're dancing at the edge of the param size limits.
1103 If you are, and this message makes sense to you, then: raise the
1104 configured limit or use OpenSSL 1.1. Nothing we can do for older
1107 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1108 cmdline element to get the documented default one. Previously it was
1111 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1112 are unsuable for non-message-specific reasons - eg. connection timeout,
1113 banner-time rejection.
1115 JH/16 Fix logging of delivery remote port, when specified by router, under
1118 PP/06 Repair manualroute's ability to take options in any order, even if one
1119 is the name of a transport.
1122 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1124 JH/17 Change the list-building routines interface to use the expanding-string
1125 triplet model, for better allocation and copying behaviour.
1127 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1128 Previously it was constructed the first time a possibly-matching string
1129 was met in the configuration file input during startup; now it is done
1132 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1133 compatible one, to avoid the (poorly documented) possibility of a config
1134 file in the working directory redirecting the DB files, possibly correpting
1135 some existing file. CVE-2017-10140 assigned for BDB.
1137 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1138 cache-hot. Previously, although the result was properly cached, the
1139 initial verify call returned a defer.
1141 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1142 the main verify for receipient in uncached-mode.
1144 JH/22 Retire historical build files to an "unsupported" subdir. These are
1145 defined as "ones for which we have no current evidence of testing".
1147 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1148 if present. Previously it was ignored.
1150 JH/24 Start using specified-initialisers in C structure init coding. This is
1151 a C99 feature (it's 2017, so now considered safe).
1153 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1154 if was a fixed-sized field and bitmask ops via macros; it is now more
1157 PP/07 GitHub PR 56: Apply MariaDB build fix.
1158 Patch provided by Jaroslav Škarvada.
1160 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1161 during Coverity cleanups [4.87 JH/47]
1162 Diagnosis and fix provided by Michael Fischer v. Mollard.
1164 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1165 the right size to place the terminating semicolon on its own folded
1166 line, the header hash was calculated to an incorrect value thanks to
1167 the (relaxed) space the fold became.
1169 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1170 and confused the parent.
1172 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1173 which could crash as a result. This could lead to undeliverable messages.
1175 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1178 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1179 matters on fast-turnover and PID-randomising systems, which were getting
1180 out-of-order delivery.
1182 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1183 a possibly-overlapping copy. The symptom was that "Remote host closed
1184 connection in response to HELO" was logged instead of the actual 4xx
1187 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1188 Previously only that bufferd was discarded, resulting in SYMTP command
1191 JH/32 DKIM: when a message has multiple signatures matching an identity given
1192 in dkim_verify_signers, run the dkim acl once for each. Previously only
1193 one run was done. Bug 2189.
1195 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1196 "panic the current process" to "deliberately defer". The panic log is
1197 still written with the problem list name; the mail and reject logs now
1198 get a temp-reject line for the message that was being handled, saying
1199 something like "domains check lookup or other defer". The SMTP 451
1200 message is still "Temporary local problem".
1202 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1203 A crafted sequence of BDAT commands could result in in-use memory beeing
1204 freed. CVE-2017-16943.
1206 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1207 from SMTP input. Previously it was always done; now only done for DATA
1208 and not BDAT commands. CVE-2017-16944.
1210 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1211 to the message (such as an overlong header line). Previously this was
1212 not done and we did not exit BDAT mode. Followon from the previous item
1213 though a different problem.
1219 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1220 than -2003 did; needs libidn2 in addition to libidn.
1222 JH/02 The path option on a pipe transport is now expanded before use.
1224 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1225 Patch provided by "Björn", documentation fix added too.
1227 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1228 missing a wire-to-host endian conversion.
1230 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1231 close after a BDAT command line could be taken as a following command,
1232 giving a synch failure. Fix by only checking for synch immediately
1233 before acknowledging the chunk.
1235 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1236 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1237 macro. Patches provided by Josh Soref.
1239 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1240 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1241 by IANA as a service extension. However, John Klensin suggests that we
1244 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1245 the dkim code may be unix-mode line endings rather than smtp wire-format
1246 CRLF, so prepend a CR to any bare LF.
1248 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1249 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1251 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1252 the first were themselves being wrongly included in the feed into dkim
1253 processing; with most chunk sizes in use this resulted in an incorrect
1254 body hash calculated value.
1256 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1257 DKIM signature block, for verification. Although advised against by
1258 standards it is specifically not ruled illegal.
1260 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1262 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1263 missing a body hash (the bh= tag).
1265 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1266 It seems that HAProxy sends the Proxy Protocol information in clear and
1267 only then does a TLS startup, so do the same.
1269 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1270 TCP connections (such as for Spamd) unless the daemon successfully set
1271 Fast Open mode on its listening sockets. This fixes breakage seen on
1272 too-old kernels or those not configured for Fast Open, at the cost of
1273 requiring both directions being enabled for TFO, and TFO never being used
1274 by non-daemon-related Exim processes.
1276 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1277 endings, at least on the first header line. Try to canonify any that get
1278 past that check, despite the cost.
1280 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1281 now limited to an arbitrary five deep, while parsing addresses with the
1282 strip_excess_angle_brackets option enabled.
1284 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1285 instead leave the unprompted TLS handshake in socket buffer for the
1286 TLS library to consume.
1288 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1290 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1292 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1293 region in one operation, for speed, and this leaves assigned pointers
1294 dangling. Add checks run only under the testsuite which checks all
1295 variables at a store-reset and panics on a dangling pointer; add code
1296 explicitly nulling out all the variables discovered. Fixes one known
1297 bug: a transport crash, where a dangling pointer for $sending_ip_address
1298 originally assigned in a verify callout, is re-used.
1300 PP/06 Drop '.' from @INC in various Perl scripts.
1302 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1304 PP/08 Reduce a number of compilation warnings under clang; building with
1305 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1306 should be warning-free.
1308 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1310 HS/01 Fix portability problems introduced by PP/08 for platforms where
1311 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1313 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1314 chunk. This allows us to accept broken chunked messages. We need a more
1315 general solution here.
1317 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1318 already-broken messages in the queue.
1320 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1322 JH/19 Fix reference counting bug in routing-generated-address tracking.
1328 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1329 supports it and a size is available (ie. the sending peer gave us one).
1331 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1332 years of being deprecated). The replacements are the ACLs
1333 acl_smtp_mime and acl_not_smtp_mime.
1335 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1336 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1337 or even an in-clear connection were permitted. Now, if the host lookup
1338 was dnssec and dane was requested then the host is only used if the
1339 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1340 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1341 if one fails this test.
1342 This means that a poorly-configured remote DNS will make it incommunicado;
1343 but it protects against a DNS-interception attack on it.
1345 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1346 non-noisy when a race steals the message being considered.
1348 JH/05 If main configuration option tls_certificate is unset, generate a
1349 self-signed certificate for inbound TLS connections.
1351 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1352 in rewrites and routers.
1354 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1355 and logged a warning sing 4.83; now they are a configuration file error.
1357 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1358 (lacking @domain). Apply the same qualification processing as RCPT.
1360 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1362 JH/10 Support ${sha256:} applied to a string (as well as the previous
1365 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1366 a cutthrough deliver is pending, as we always want to make a connection.
1367 This also avoids re-routing the message when later placing the cutthrough
1368 connection after a verify cache hit.
1369 Do not update it with the verify result either.
1371 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1372 when routing results in more than one destination address.
1374 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1375 signing (which inhibits the cutthrough capability). Previously only
1376 the presence of an option was tested; now an expansion evaluating as
1377 empty is permissible (obviously it should depend only on data available
1378 when the cutthrough connection is made).
1380 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1381 the relevant preceding SMTP command did not note the pipelining mode.
1383 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1384 Previously they were not counted.
1386 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1387 as one having no matching records. Previously we deferred the message
1388 that needed the lookup.
1390 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1391 distinguished as "(=".
1393 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1394 for missing MX records. Previously it only worked for missing A records.
1396 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1398 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1399 after the data-go-ahead and data-ack. Patch from Jason Betts.
1401 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1402 even for a "none" policy. Patch from Tony Meyer.
1404 JH/22 Fix continued use of a connection for further deliveries. If a port was
1405 specified by a router, it must also match for the delivery to be
1408 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1409 When one of the recipients of a message was unsuitable for the connection
1410 (has no matching addresses), we lost track of needing to mark it
1411 deferred. As a result mail would be lost.
1413 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1415 JH/25 Decoding ACL controls is now done using a binary search; the source code
1416 takes up less space and should be simpler to maintain. Merge the ACL
1417 condition decode tables also, with similar effect.
1419 JH/26 Fix problem with one_time used on a redirect router which returned the
1420 parent address unchanged. A retry would see the parent address marked as
1421 delivered, so not attempt the (identical) child. As a result mail would
1424 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1425 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1426 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1429 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1430 Main config options check_{log,spool}_{inodes,space} are now
1431 100 inodes, 10MB unless set otherwise in the configuration.
1433 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1434 Previously it only applied to the main-section connection policy
1437 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1439 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1440 by me. Added RFC7919 DH primes as an alternative.
1442 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1443 are not in the system include path.
1445 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1446 GnuTLS, when a session startup failed (eg because the client disconnected)
1447 Exim did stdio operations after fclose. This was exposed by a recent
1448 change which nulled out the file handle after the fclose.
1450 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1451 signed directly by the cert-signing cert, rather than an intermediate
1452 OCSP-signing cert. This is the model used by LetsEncrypt.
1454 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1456 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1457 an incoming connection.
1459 HS/02 Bug 1802: Do not half-close the connection after sending a request
1462 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1463 fallback to "prime256v1".
1465 JH/34 SECURITY: Use proper copy of DATA command in error message.
1466 Could leak key material. Remotely exploitable. CVE-2016-9963.
1472 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1473 and 3.4.4 - once the server is enabled to respond to an OCSP request
1474 it does even when not requested, resulting in a stapling non-aware
1475 client dropping the TLS connection.
1477 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1478 support variable-length bit vectors. No functional change.
1480 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1481 The I= interface field on outgoing lines is now after the H= remote
1482 host field, same as incoming lines. There is a separate
1483 outgoing_interface log selector which allows you to disable the
1486 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1487 If not running log_selector +smtp_connection the mainlog would be held
1488 open indefinitely after a "too many connections" event, including to a
1489 deleted file after a log rotate. Leave the per net connection logging
1490 leaving it open for efficiency as that will be quickly detected by the
1491 check on the next write.
1493 HS/01 Bug 1671: Fix post transport crash.
1494 Processing the wait-<transport> messages could crash the delivery
1495 process if the message IDs didn't exist for some reason. When
1496 using 'split_spool_directory=yes' the construction of the spool
1497 file name failed already, exposing the same netto behaviour.
1499 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1500 mime_regex ACL conditions.
1502 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1503 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1504 to HELO, local diagnostic string.
1506 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1507 log line to debug. Even when configured with a tls authenticator many
1508 client connections are expected to not authenticate in this way, so
1509 an authenticate fail is not an error.
1511 HS/02 Add the Exim version string to the process info. This way exiwhat
1512 gives some more detail about the running daemon.
1514 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1515 matter for fast-change records such as DNSBLs.
1517 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1518 retry record, even if constant. There may be multiple transports with
1519 different interface settings and the retry behaviour needs to be kept
1522 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1524 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1526 JH/10 Bug 840: fix log_defer_output option of pipe transport
1528 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1529 hosts_randomize. This matters a lot when combined with mua_wrapper.
1531 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1532 ${quote_pgsql:<string>} operator.
1534 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1536 JH/14 Change header file name for freeradius-client. Relevant if compiling
1537 with Radius support; from the Gentoo tree and checked under Fedora.
1539 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1541 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1542 option transport_filter, meaning no filtering.
1544 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1546 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1547 defaults to "*" (all hosts). The variable is now available when not built
1548 with TLS, default unset, mainly to enable keeping the testsuite sane.
1549 If a server certificate is not supplied (via tls_certificate) an error is
1550 logged, and clients will find TLS connections fail on startup. Presumably
1551 they will retry in-clear.
1552 Packagers of Exim are strongly encouraged to create a server certificate
1553 at installation time.
1555 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1556 with the $config_file variable.
1558 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1559 in transport context, after the attempt, and per-recipient. The latter type
1560 is per host attempted. The event data is the error message, and the errno
1561 information encodes the lookup type (A vs. MX) used for the (first) host,
1562 and the trailing two digits of the smtp 4xx response.
1564 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1565 to write to mainlog (or rejectlog, paniclog) in the window between file
1566 creation and permissions/ownership being changed. Particularly affects
1567 installations where exicyclog is run as root, rather than exim user;
1568 result is that the running daemon panics and dies.
1570 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1572 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1573 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1574 "pri" and "weight". Note that the previous implicit priority given by the
1575 list order is no longer honoured.
1577 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1578 for DKIM processing.
1580 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1581 by defining SUPPORT_SOCKS.
1583 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1584 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1585 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1586 variables are renamed to proxy_{local,external}_{address,port}.
1588 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1589 for a build by defining SUPPORT_I18N
1591 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1592 of the query string, and make ${quote_redis:} do that quoting.
1594 JH/29 Move Events support from Experimental to mainline, enabled by default
1595 and removable for a build by defining DISABLE_EVENT.
1597 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1599 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1600 cached by the daemon.
1602 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1603 by defining LOOKUP_REDIS. The libhiredis library is required.
1605 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1606 keys are given for lookup.
1608 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1609 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1610 only supported when built with TLS support. The PolarSSL SHA routines
1611 are still used when the TLS library is too old for convenient support.
1613 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1614 openssl_options), for security. OpenSSL forces this from version 1.1.0
1615 server-side so match that on older versions.
1617 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1618 allocation for $value could be released as the expansion processing
1619 concluded, but leaving the global pointer active for it.
1621 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1622 and to use the domains and local_parts ACL conditions.
1624 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1625 incorrectly not doubled on cutthrough transmission, hence seen as a
1626 body-termination at the receiving system - resulting in truncated mails.
1627 Commonly the sender saw a TCP-level error, and retransmitted the message
1628 via the normal store-and-forward channel. This could result in duplicates
1629 received - but deduplicating mailstores were liable to retain only the
1630 initial truncated version.
1632 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1634 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1636 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1637 we're in there, support oversigning also; bug 1309.
1639 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1641 HS/04 Add support for keep_environment and add_environment options.
1643 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1644 either intentional arithmetic overflow during PRNG, or testing config-
1647 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1648 delivery resulted in actual delivery. Cancel cutthrough before DATA
1651 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1652 rejects a recipient: pass the reject to the originator.
1654 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1655 Many were false-positives and ignorable, but it's worth fixing the
1658 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1659 for the new environment-manipulation done at startup. Move the routines
1660 from being local to tls.c to being global via the os.c file.
1662 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1663 an extract embedded as result-arg for a map, the first arg for extract
1664 is unavailable so we cannot tell if this is a numbered or keyed
1665 extraction. Accept either.
1671 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1674 JH/02 The smtp transport option "multi_domain" is now expanded.
1676 JH/03 The smtp transport now requests PRDR by default, if the server offers
1679 JH/04 Certificate name checking on server certificates, when exim is a client,
1680 is now done by default. The transport option tls_verify_cert_hostnames
1681 can be used to disable this per-host. The build option
1682 EXPERIMENTAL_CERTNAMES is withdrawn.
1684 JH/05 The value of the tls_verify_certificates smtp transport and main options
1685 default to the word "system" to access the system default CA bundle.
1686 For GnuTLS, only version 3.0.20 or later.
1688 JH/06 Verification of the server certificate for a TLS connection is now tried
1689 (but not required) by default. The verification status is now logged by
1690 default, for both outbound TLS and client-certificate supplying inbound
1693 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1696 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1697 Status Notification (bounce) messages are now MIME format per RFC 3464.
1698 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1699 under the control of the dsn_advertise_hosts option, and routers may
1700 have a dsn_lasthop option.
1702 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1703 default, modifiable by a malware= option. The list separator for
1704 the options can now be changed in the usual way. Bug 68.
1706 JH/10 The smtp_receive_timeout main option is now expanded before use.
1708 JH/11 The incoming_interface log option now also enables logging of the
1709 local interface on delivery outgoing connections.
1711 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1712 if the interface and destination host and port all match.
1714 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1717 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1718 Patch from Andrew Lewis.
1720 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1721 now supports optional time-restrictions, weighting, and priority
1722 modifiers per server. Patch originally by <rommer@active.by>.
1724 JH/16 The spamd_address main option now supports a mixed list of local
1725 and remote servers. Remote servers can be IPv6 addresses, and
1726 specify a port-range.
1728 JH/17 Bug 68: The spamd_address main option now supports an optional
1729 timeout value per server.
1731 JH/18 Bug 1581: Router and transport options headers_add/remove can
1732 now have the list separator specified.
1734 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1737 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1740 JH/21 Support for the A6 type of dns record is withdrawn.
1742 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1743 rather than the verbs used.
1745 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1746 from 255 to 1024 chars.
1748 JH/24 Verification callouts now attempt to use TLS by default.
1750 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1751 are generic router options now. The defaults didn't change.
1753 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1754 Original patch from Alexander Shikoff, worked over by JH.
1756 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1757 files if ZCAT_COMMAND is not executable.
1759 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1761 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1763 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1764 Normally benign, it bites when the pair was led to by a CNAME;
1765 modern usage is to not canonicalize the domain to a CNAME target
1766 (and we were inconsistent anyway for A-only vs AAAA+A).
1768 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1770 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1771 when evaluating $sender_host_dnssec.
1773 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1774 $sender_helo_dnssec variable.
1776 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1778 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1780 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1782 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1783 documented as working, but never had. Support all but $spam_report.
1785 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1786 added for tls authenticator.
1788 HS/03 Add perl_taintmode main config option
1794 TL/01 When running the test suite, the README says that variables such as
1795 no_msglog_check are global and can be placed anywhere in a specific
1796 test's script, however it was observed that placement needed to be near
1797 the beginning for it to behave that way. Changed the runtest perl
1798 script to read through the entire script once to detect and set these
1799 variables, reset to the beginning of the script, and then run through
1800 the script parsing/test process like normal.
1802 TL/02 The BSD's have an arc4random API. One of the functions to induce
1803 adding randomness was arc4random_stir(), but it has been removed in
1804 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1805 function when detected.
1807 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1808 cause callback expansion.
1810 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1811 syntax errors in an expansion can be treated as a string instead of
1812 logging or causing an error, due to the internal use of bool_lax
1813 instead of bool when processing it.
1815 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1816 server certificates when making smtp deliveries.
1818 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1820 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1822 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1824 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1825 Merged patch from Sebastian Wiedenroth.
1827 JH/05 Fix results-pipe from transport process. Several recipients, combined
1828 with certificate use, exposed issues where response data items split
1829 over buffer boundaries were not parsed properly. This eventually
1830 resulted in duplicates being sent. This issue only became common enough
1831 to notice due to the introduction of connection certificate information,
1832 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1834 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1835 size buffer was used, resulting in syntax errors when an expansion
1838 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1839 version 3.3.6 or later.
1841 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1842 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1843 both become "event_action", the variables become $event_name, $event_data
1844 and $event_defer_errno. There is a new variable $verify_mode, usable in
1845 routers, transports and related events. The tls:cert event is now also
1846 raised for inbound connections, if the main configuration event_action
1849 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1850 early OCSP support, but no stapling (appears to be less than 1.0.0).
1852 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1853 server certificate names available under the smtp transport option
1854 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1857 JH/10 Time-related extraction expansions from certificates now use the main
1858 option "timezone" setting for output formatting, and are consistent
1859 between OpenSSL and GnuTLS compilations. Bug 1541.
1861 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1862 encoded parameter in the incoming message. Bug 1558.
1864 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1865 include certificate info, eximon was claiming there were spoolfile
1868 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1870 JH/14 Log delivery-related information more consistently, using the sequence
1871 "H=<name> [<ip>]" wherever possible.
1873 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1874 are problematic for Debian distribution, omit them from the release
1877 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1879 JH/16 Fix string representation of time values on 64bit time_t architectures.
1882 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1883 output list separator was used.
1888 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1889 checkers that were complaining about end of non-void function with no
1892 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1893 This was a regression introduced in 4.83 by another bugfix.
1895 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1897 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1898 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1904 TF/01 Correctly close the server side of TLS when forking for delivery.
1906 When a message was received over SMTP with TLS, Exim failed to clear up
1907 the incoming connection properly after forking off the child process to
1908 deliver the message. In some situations the subsequent outgoing
1909 delivery connection happened to have the same fd number as the incoming
1910 connection previously had. Exim would try to use TLS and fail, logging
1911 a "Bad file descriptor" error.
1913 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1914 utilities have not been installed.
1916 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1917 temporary space as the ACL may create new global variables.
1919 TL/01 LDAP support uses per connection or global context settings, depending
1920 upon the detected version of the libraries at build time.
1922 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1923 to extract and use the src ip:port in logging and expansions as if it
1924 were a direct connection from the outside internet. PPv2 support was
1925 updated based on HAProxy spec change in May 2014.
1927 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1929 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1930 Properly escape header and check for NULL return.
1932 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1935 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1937 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1938 characters in header names, implemented as a verify condition.
1939 Contributed by Michael Fischer v. Mollard.
1941 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1942 results permerror and temperror. Previous values are deprecated but
1943 still accepted. In a future release, err_perm and err_temp will be
1944 completely removed, which will be a backward incompatibility if the
1945 ACL tests for either of these two old results. Patch contributed by
1946 user bes-internal on the mailing list.
1948 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1950 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1951 selectors, in both main and reject logs.
1953 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1956 JH/07 Add malware type "sock" for talking to simple daemon.
1958 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1960 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1961 routers/transports under cutthrough routing.
1963 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1964 numbers. Touch up "bool" conditional to keep the same definition.
1966 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1968 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1970 JH/12 Expand items in router/transport headers_add or headers_remove lists
1971 individually rather than the list as a whole. Bug 1452.
1973 Required for reasonable handling of multiple headers_ options when
1974 they may be empty; requires that headers_remove items with embedded
1975 colons must have them doubled (or the list-separator changed).
1977 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1978 view the policy declared in the DMARC record. Currently, $dmarc_status
1979 is a combined value of both the record presence and the result of the
1982 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1984 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1985 dnslookup router and the smtp transport (applying to the forward
1988 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1989 of ldap servers used for a specific lookup. Patch provided by Heiko
1992 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1993 New variable $lookup_dnssec_authenticated for observability.
1995 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1996 Patch submitted by Lars Timman.
1998 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2000 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2001 Requires trusted mode and valid format message id, aborts otherwise.
2002 Patch contributed by Heiko Schlichting.
2004 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2005 certextract with support for various fields. Bug 1358.
2007 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2008 is requested by default, modifiable by smtp transport option
2011 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2012 operate on certificate variables to give certificate fingerprints
2013 Also new ${sha256:cert_variable}.
2015 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2017 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2018 Christian Aistleitner.
2020 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2022 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2023 file. Patch from Wolfgang Breyha.
2025 JH/25 Expand the coverage of the delivery $host and $host_address to
2026 client authenticators run in verify callout. Bug 1476.
2028 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2029 align with daemon_smtp_ports. Bug 72.
2031 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2032 support and error reporting did not work properly.
2034 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2035 and is readable. Patch from Andrew Colin Kissa.
2037 TL/14 Enhance documentation of ${run expansion and how it parses the
2038 commandline after expansion, particularly in the case when an
2039 unquoted variable expansion results in an empty value.
2041 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2043 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2044 by renaming away from it. Addresses GH issue 15, reported by
2047 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2049 TL/15 SECURITY: prevent double expansion in math comparison functions
2050 (can expand unsanitized data). Not remotely exploitable.
2057 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2059 PP/02 Make -n do something, by making it not do something.
2060 When combined with -bP, the name of an option is not output.
2062 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2065 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2066 $sender_host_name and config options to manage this, and basic check
2069 PP/05 DSCP support for outbound connections and control modifier for inbound.
2071 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2072 (Only plugin which currently uses this is kerberos4, which nobody should
2073 be using, but we should make it available and other future plugins might
2074 conceivably use it, even though it would break NAT; stuff *should* be
2075 using channel bindings instead).
2077 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2078 name; added for Sendmail compatibility; requires admin caller.
2079 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2080 just ignore it); requires trusted caller.
2081 Also parse but ignore: -Ac -Am -X<logfile>
2084 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2086 TL/02 Add +smtp_confirmation as a default logging option.
2088 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2089 Patch by Magnus Holmgren from 2007-02-20.
2091 TL/04 Bugzilla 1281 - Spec typo.
2092 Bugzilla 1283 - Spec typo.
2093 Bugzilla 1290 - Spec grammar fixes.
2095 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2097 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2099 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2100 mailing list by by Dmitry Isaikin.
2102 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2104 JH/02 Support "G" suffix to numbers in ${if comparisons.
2106 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2108 NM/01 Bugzilla 1197 - Spec typo
2109 Bugzilla 1196 - Spec examples corrections
2111 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2113 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2114 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2117 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2118 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2120 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2121 "acl {{name}{arg}...}", and optional args on acl condition
2124 JH/05 Permit multiple router/transport headers_add/remove lines.
2126 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2128 JH/07 Avoid using a waiting database for a single-message-only transport.
2129 Performance patch from Paul Fisher. Bugzilla 1262.
2131 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2134 JH/09 Add $headers_added variable, with content from use of ACL modifier
2135 add_header (but not yet added to the message). Bugzilla 199.
2137 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2138 Pulled from Bugzilla 817 by Wolfgang Breyha.
2140 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2142 (nb: this is the same fix as in Exim 4.80.1)
2144 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2147 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2150 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2152 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2153 advertise SMTP AUTH mechanism to us, instead of a generic
2154 protocol violation error. Also, make Exim more robust to bad
2155 data from the Dovecot auth socket.
2157 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2159 When a queue runner is handling a message, Exim first routes the
2160 recipient addresses, during which it prunes them based on the retry
2161 hints database. After that it attempts to deliver the message to
2162 any remaining recipients. It then updates the hints database using
2165 So if a recipient address works intermittently, it can get repeatedly
2166 deferred at routing time. The retry hints record remains fresh so the
2167 address never reaches the final cutoff time.
2169 This is a fairly common occurrence when a user is bumping up against
2170 their storage quota. Exim had some logic in its local delivery code
2171 to deal with this. However it did not apply to per-recipient defers
2172 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2174 This change adds a proper retry rule check during routing so that the
2175 final cutoff time is checked against the message's age. We only do
2176 this check if there is an address retry record and there is not a
2177 domain retry record; this implies that previous attempts to handle
2178 the address had the retry_use_local_parts option turned on. We use
2179 this as an approximation for the destination being like a local
2180 delivery, as in LMTP.
2182 I suspect this new check makes the old local delivery cutoff check
2183 redundant, but I have not verified this so I left the code in place.
2185 TF/02 Correct gecos expansion when From: is a prefix of the username.
2187 Test 0254 submits a message to Exim with the header
2191 When I ran the test suite under the user fanf2, Exim expanded
2192 the header to contain my full name, whereas it should have added
2193 a Resent-Sender: header. It erroneously treats any prefix of the
2194 username as equal to the username.
2196 This change corrects that bug.
2198 GF/01 DCC debug and logging tidyup
2199 Error conditions log to paniclog rather than rejectlog.
2200 Debug lines prefixed by "DCC: " to remove any ambiguity.
2202 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2204 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2205 Bug spotted by Jeremy Harris; was flawed since initial commit.
2206 Would have resulted in OCSP responses post-SNI triggering an Exim
2207 NULL dereference and crash.
2209 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2211 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2212 Bug detection, analysis and fix by Samuel Thibault.
2213 Bugzilla 1331, Debian bug #698092.
2215 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2217 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2218 Server implementation by Todd Lyons, client by JH.
2219 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2220 config variable "prdr_enable" controls whether the server
2221 advertises the facility. If the client requests PRDR a new
2222 acl_data_smtp_prdr ACL is called once for each recipient, after
2223 the body content is received and before the acl_smtp_data ACL.
2224 The client is controlled by both of: a hosts_try_prdr option
2225 on the smtp transport, and the server advertisement.
2226 Default client logging of deliveries and rejections involving
2227 PRDR are flagged with the string "PRDR".
2229 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2230 fclose(). Diagnosis by Todd Lyons.
2232 PP/17 Update configure.default to handle IPv6 localhost better.
2233 Patch by Alain Williams (plus minor tweaks).
2236 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2237 This is now consistent with GnuTLS, and is now documented: the
2238 previous undocumented portable approach to treating the option as
2239 unset was to force an expansion failure. That still works, and
2240 an empty string is now equivalent.
2242 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2243 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2244 not performing validation itself.
2246 PP/20 Added force_command boolean option to pipe transport.
2247 Patch from Nick Koston, of cPanel Inc.
2249 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2252 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2254 PP/21 Fix eximon continuous updating with timestamped log-files.
2255 Broken in a format-string cleanup in 4.80, missed when I repaired the
2256 other false fix of the same issue.
2257 Report and fix from Heiko Schlichting.
2260 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2261 Report from Prashanth Katuri.
2263 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2264 It's SecureTransport, so affects any MacOS clients which use the
2265 system-integrated TLS libraries, including email clients.
2267 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2268 using a MIME ACL for non-SMTP local injection.
2269 Report and assistance in diagnosis by Warren Baker.
2271 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2273 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2275 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2276 last id that failed so it may be referenced in subsequent ACL's.
2278 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2281 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2282 ldap library initialization, allowing self-signed CA's to be
2283 used. Also properly sets require_cert option later in code by
2284 using NULL (global ldap config) instead of ldap handle (per
2285 session). Bug diagnosis and testing by alxgomz.
2287 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2288 the src/util/ subdirectory.
2290 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2291 renamed to Transport Post Delivery Action by Jeremy Harris, as
2294 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2295 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2296 redis_servers = needs to be configured which will be used by the redis
2297 lookup. Patch from Warren Baker, of The Packet Hub.
2299 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2301 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2302 hostname or reverse DNS when processing a host list. Used suggestions
2303 from multiple comments on this bug.
2305 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2307 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2308 Missed a few lines, added it to make the runtest require no keyboard
2311 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2312 contains upper case chars. Make router use caseful_local_part.
2314 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2315 support when GnuTLS has been built with p11-kit.
2321 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2323 This, or similar/improved, will also be change PP/11 of 4.82.
2329 PP/01 Handle short writes when writing local log-files.
2330 In practice, only affects FreeBSD (8 onwards).
2331 Bugzilla 1053, with thanks to Dmitry Isaikin.
2333 NM/01 Bugzilla 949 - Documentation tweak
2335 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2338 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2340 PP/02 Implemented gsasl authenticator.
2342 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2344 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2345 `pkg-config foo` for cflags/libs.
2347 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2348 with rest of GSASL and with heimdal_gssapi.
2350 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2351 `pkg-config foo` for cflags/libs for the TLS implementation.
2353 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2354 properties get this fed in as external SSF. A number of robustness
2355 and debugging improvements to the cyrus_sasl authenticator.
2357 PP/08 cyrus_sasl server now expands the server_realm option.
2359 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2360 Patch by Jeremy Harris.
2362 PP/10 Added dbmjz lookup type.
2364 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2366 PP/12 MAIL args handles TAB as well as SP, for better interop with
2367 non-compliant senders.
2368 Analysis and variant patch by Todd Lyons.
2370 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2371 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2372 Patch from Dirk Mueller <dmueller@suse.com>
2374 PP/13 tls_peerdn now print-escaped for spool files.
2375 Observed some $tls_peerdn in wild which contained \n, which resulted
2376 in spool file corruption.
2378 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2379 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2380 or write after TLS renegotiation, which otherwise led to messages
2383 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2384 as a tracking header (ie: a signed header comes before the signature).
2385 Patch from Wolfgang Breyha.
2387 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2388 comma-sep list; embedded commas doubled.
2390 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2392 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2394 Report and patch from Dmitry Banschikov.
2396 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2397 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2398 failures. We appear to now support TLS1.1+ with Exim.
2400 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2401 lets Exim select keys and certificates based upon TLS SNI from client.
2402 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2403 before an outbound SMTP session. New log_selector, +tls_sni.
2405 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2406 NULL dereference. Report and patch from Alun Jones.
2408 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2409 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2410 Not seeing resolver debug output on NetBSD, but suspect this is a
2411 resolver implementation change.
2413 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2414 Left warnings. Added "eximon gdb" invocation mode.
2416 PP/21 Defaulting "accept_8bitmime" to true, not false.
2418 PP/22 Added -bw for inetd wait mode support.
2420 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2421 locate the relevant includes and libraries. Made this the default.
2423 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2424 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2426 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2427 This may cause build issues on older platforms.
2429 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2430 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2431 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2432 Added SNI support via GnuTLS too.
2433 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2435 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2437 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2438 Applied second patch from Janne, implementing suggestion to default
2439 multiple-strings-in-record handling to match SPF spec.
2441 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2443 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2444 read-only, out of scope).
2445 Patch from Wolfgang Breyha, report from Stuart Northfield.
2447 PP/29 Fix three issues highlighted by clang analyser static analysis.
2448 Only crash-plausible issue would require the Cambridge-specific
2449 iplookup router and a misconfiguration.
2450 Report from Marcin Mirosław.
2452 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2454 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2455 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2456 As part of this, removing so much warning spew let me fix some minor
2457 real issues in debug logging.
2459 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2460 assignment on my part. Fixed.
2462 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2463 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2464 Janne Snabb (who went above and beyond: thank you).
2466 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2467 string otherwise requires a connection and a bunch more work and it's
2468 relatively easy to get wrong. Should also expose TLS library linkage
2471 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2472 64-bit ${eval} (JH/03).
2474 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2475 GNU libc to support some of the 64-bit stuff, should not lead to
2476 conflicts. Defined before os.h is pulled in, so if a given platform
2477 needs to override this, it can.
2479 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2480 protection layer was required, which is not implemented.
2481 Bugzilla 1254, patch from Wolfgang Breyha.
2483 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2484 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2485 tls_dhparam take prime identifiers. Also unbreak combination of
2486 OpenSSL+DH_params+TLSSNI.
2488 PP/39 Disable SSLv2 by default in OpenSSL support.
2494 PP/01 Solaris build fix for Oracle's LDAP libraries.
2495 Bugzilla 1109, patch from Stephen Usher.
2497 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2499 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2502 TF/02 Fix a couple more cases where we did not log the error message
2503 when unlink() failed. See also change 4.74-TF/03.
2505 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2506 lock up or crash if it happened to be inside a call to libc when it
2507 got a SIGUSR1 from exiwhat.
2509 The SIGUSR1 handler appends the current process status to the process
2510 log which is later printed by exiwhat. It used to use the general
2511 purpose logging code to do this, but several functions it calls are
2512 not safe for signals.
2514 The new output code in the SIGUSR1 handler is specific to the process
2515 log, and simple enough that it's easy to inspect for signal safety.
2516 Removing some special cases also simplifies the general logging code.
2517 Removing the spurious timestamps from the process log simplifies
2520 TF/04 Improved ratelimit ACL condition.
2522 The /noupdate option has been deprecated in favour of /readonly which
2523 has clearer semantics. The /leaky, /strict, and /readonly update modes
2524 are mutually exclusive. The update mode is no longer included in the
2525 database key; it just determines when the database is updated. (This
2526 means that when you upgrade Exim will forget old rate measurements.)
2528 Exim now checks that the per_* options are used with an update mode that
2529 makes sense for the current ACL. For example, when Exim is processing a
2530 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2531 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2532 must specify per_mail/readonly. If you omit the update mode it defaults to
2533 /leaky where that makes sense (as before) or /readonly where required.
2535 The /noupdate option is now undocumented but still supported for
2536 backwards compatibility. It is equivalent to /readonly except that in
2537 ACLs where /readonly is required you may specify /leaky/noupdate or
2538 /strict/noupdate which are treated the same as /readonly.
2540 A useful new feature is the /count= option. This is a generalization
2541 of the per_byte option, so that you can measure the throughput of other
2542 aggregate values. For example, the per_byte option is now equivalent
2543 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2545 The per_rcpt option has been generalized using the /count= mechanism
2546 (though it's more complicated than the per_byte equivalence). When it is
2547 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2548 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2549 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2550 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2551 Note that using per_rcpt with a non-readonly update mode in more than
2552 one ACL will cause the recipients to be double-counted. (The per_mail
2553 and per_byte options don't have this problem.)
2555 The handling of very low rates has changed slightly. If the computed rate
2556 is less than the event's count (usually one) then this event is the first
2557 after a long gap. In this case the rate is set to the same as this event's
2558 count, so that the first message of a spam run is counted properly.
2560 The major new feature is a mechanism for counting the rate of unique
2561 events. The new per_addr option counts the number of different
2562 recipients that someone has sent messages to in the last time period. It
2563 behaves like per_rcpt if all the recipient addresses are different, but
2564 duplicate recipient addresses do not increase the measured rate. Like
2565 the /count= option this is a general mechanism, so the per_addr option
2566 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2567 example, measure the rate that a client uses different sender addresses
2568 with the options per_mail/unique=$sender_address. There are further
2569 details in the main documentation.
2571 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2573 TF/06 Removed a few PCRE remnants.
2575 TF/07 Automatically extract Exim's version number from tags in the git
2576 repository when doing development or release builds.
2578 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2579 Bugzilla 879. Patch from Paul Fisher.
2581 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2582 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2585 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2587 PP/05 Variable $av_failed, true if the AV scanner deferred.
2588 Bugzilla 1078. Patch from John Horne.
2590 PP/06 Stop make process more reliably on build failure.
2591 Bugzilla 1087. Patch from Heiko Schlittermann.
2593 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2594 Bugzilla 1089. Patch from Heiko Schlittermann.
2596 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2597 Bugzilla 1131. Patch from Holger Weiß.
2599 PP/09 Handle IPv6 addresses with SPF.
2600 Bugzilla 860. Patch from Wolfgang Breyha.
2602 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2604 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2607 PP/11 match_* no longer expand right-hand-side by default.
2608 New compile-time build option, EXPAND_LISTMATCH_RHS.
2609 New expansion conditions, "inlist", "inlisti".
2611 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2613 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2615 PP/14 fix log_write() format string regression from TF/03.
2616 Bugzilla 1152. Patch from Dmitry Isaikin.
2622 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2624 PP/02 Harmonised TLS library version reporting; only show if debugging.
2625 Layout now matches that introduced for other libraries in 4.74 PP/03.
2627 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2629 PP/04 New "dns_use_edns0" global option.
2631 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2634 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2635 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2637 TK/01 Updated PolarSSL code to 0.14.2.
2638 Bugzilla 1097. Patch from Andreas Metzler.
2640 PP/07 Catch divide-by-zero in ${eval:...}.
2641 Fixes bugzilla 1102.
2643 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2646 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2647 format-string attack -- SECURITY: remote arbitrary code execution.
2649 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2650 time unintentionally subject to list matching rules, letting the header
2651 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2652 arbitrary string expansion). This allowed for information disclosure.
2654 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2655 INT_MIN/-1 -- value coerced to INT_MAX.
2661 NM/01 Workaround for PCRE version dependency in version reporting
2664 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2665 This fixes portability to compilers other than gcc, notably
2666 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2668 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2669 makefiles for portability to HP-UX and POSIX correctness.
2671 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2672 Also via indented variable definition in the Makefile.
2673 (Debugging by Oliver Heesakkers).
2675 PP/02 Restore caching of spamd results with expanded spamd_address.
2676 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2678 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2679 Improves build reliability. Fix from: Frank Elsner
2681 NM/02 Fix wide character breakage in the rfc2047 coding
2682 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2684 NM/03 Allow underscore in dnslist lookups
2685 Fixes bug 1026. Patch from Graeme Fowler
2687 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2688 Code patches from Adam Ciarcinski of NetBSD.
2690 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2693 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2694 is logged, to avoid truncation. Patch from John Horne.
2696 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2697 Patch from Jakob Hirsch.
2699 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2700 SQL string expansion failure details.
2701 Patch from Andrey Oktyabrski.
2703 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2704 Patch from Simon Arlott.
2706 PP/09 New lookups functionality failed to compile on old gcc which rejects
2707 extern declarations in function scope.
2708 Patch from Oliver Fleischmann
2710 PP/10 Use sig_atomic_t for flags set from signal handlers.
2711 Check getgroups() return and improve debugging.
2712 Fixed developed for diagnosis in bug 927 (which turned out to be
2715 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2716 Patch from Mark Zealey.
2718 PP/12 Bugzilla 1056: Improved spamd server selection.
2719 Patch from Mark Zealey.
2721 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2722 Based on patch from Heiko Schlittermann.
2724 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2725 Patch from Uwe Doering, sign-off by Michael Haardt.
2727 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2728 variable declaration deep within a block. Bug and patch from
2731 PP/15 lookups-Makefile IRIX compatibility coercion.
2733 PP/16 Make DISABLE_DKIM build knob functional.
2735 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2736 Patch by Simon Arlott
2738 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2739 variable argument macros. Our copy now differs from upstream.
2745 TF/01 Failure to get a lock on a hints database can have serious
2746 consequences so log it to the panic log.
2748 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2749 controlled using the smtp_confirmation log selector.
2751 TF/03 Include the error message when we fail to unlink a spool file.
2753 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2754 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2755 for maintaining out-of-tree patches for some time.
2757 PP/01 Bugzilla 139: Documentation and portability issues.
2758 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2759 Handle per-OS dynamic-module compilation flags.
2761 PP/02 Let /dev/null have normal permissions.
2762 The 4.73 fixes were a little too stringent and complained about the
2763 permissions on /dev/null. Exempt it from some checks.
2764 Reported by Andreas M. Kirchwitz.
2766 PP/03 Report version information for many libraries, including
2767 Exim version information for dynamically loaded libraries. Created
2768 version.h, now support a version extension string for distributors
2769 who patch heavily. Dynamic module ABI change.
2771 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2772 privilege escalation vulnerability whereby the Exim run-time user
2773 can cause root to append content of the attacker's choosing to
2776 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2779 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2780 If dropping privileges for untrusted macros, we disabled normal logging
2781 on the basis that it would fail; for the Exim run-time user, this is not
2782 the case, and it resulted in successful deliveries going unlogged.
2783 Fixed. Reported by Andreas Metzler.
2789 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2790 only prepend for the Resent-* case. Fixes regression introduced in
2791 Exim 4.70 by NM/22 for Bugzilla 607.
2793 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2794 increasing numbers of administrators be bitten by this.
2796 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2798 PP/03 Bugzilla 994: added openssl_options main configuration option.
2800 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2802 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2804 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2806 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2807 without a peer certificate, leading to a segfault because of an
2808 assumption that peers always have certificates. Be a little more
2809 paranoid. Problem reported by Martin Tscholak.
2811 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2812 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2813 NB: ClamAV planning to remove STREAM in "middle of 2010".
2814 CL also introduces -bmalware, various -d+acl logging additions and
2815 more caution in buffer sizes.
2817 PP/09 Implemented reverse_ip expansion operator.
2819 PP/10 Bugzilla 937: provide a "debug" ACL control.
2821 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2823 PP/12 Bugzilla 973: Implement --version.
2825 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2827 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2829 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2831 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2832 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2833 ignore trailing whitespace.
2835 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2837 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2840 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2841 Notification from Dr Andrew Aitchison.
2843 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2844 ExtendedDetectionInfo response format.
2845 Notification from John Horne.
2847 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2850 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2851 XSL and documented dependency on system catalogs, with examples of how
2854 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2857 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2858 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2859 configuration file which is writeable by the Exim user or group.
2861 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2862 of configuration files to cover files specified with the -C option if
2863 they are going to be used with root privileges, not just the default
2866 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2867 option (effectively making it always true).
2869 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2870 files to be used while preserving root privileges.
2872 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2873 that rogue child processes cannot use them.
2875 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2876 run-time user, instead of root.
2878 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2879 Exim run-time user without dropping privileges.
2881 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2882 result string, instead of calling string_vformat() twice with the same
2885 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2886 for other users. Others should always drop root privileges if they use
2887 -C on the command line, even for a whitelisted configure file.
2889 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2891 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2897 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2898 $data_path, and $header_path variables; fixed documentation bugs and
2901 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2902 exipick to access non-standard spools, including the "frozen" queue
2905 NM/01 Bugzilla 965: Support mysql stored procedures.
2906 Patch from Alain Williams
2908 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2910 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2911 Patch from Andreas Metzler
2913 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2914 Patch from Kirill Miazine
2916 NM/05 Bugzilla 671: Added umask to procmail example.
2918 JJ/03 installed exipick 20100323.0, fixing doc bug
2920 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2921 directory. Notification and patch from Dan Rosenberg.
2923 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2925 TK/02 Improve log output when DKIM signing operation fails.
2927 MH/01 Treat the transport option dkim_domain as a colon separated
2928 list, not as a single string, and sign the message with each element,
2929 omitting multiple occurences of the same signer.
2931 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2932 Bugzilla 985, 986. Patch by Simon Arlott
2934 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2935 Patch by Simon Arlott
2937 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2938 MBX locking. Notification from Dan Rosenberg.
2944 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2946 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2948 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2950 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2952 NM/04 Bugzilla 915: Flag broken perl installation during build.
2958 TK/01 Added patch by Johannes Berg that expands the main option
2959 "spamd_address" if it starts with a dollar sign.
2961 TK/02 Write list of recipients to X-Envelope-Sender header when building
2962 the mbox-format spool file for content scanning (suggested by Jakob
2965 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2966 (http://www.dcc-servers.net/) support via dccifd. Activated by
2967 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2969 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2970 by Mark Daniel Reidel <mr@df.eu>.
2972 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2973 When building exim an external PCRE library is now needed -
2974 PCRE is a system library on the majority of modern systems.
2975 See entry on PCRE_LIBS in EDITME file.
2977 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2978 conversation. Added nologin parameter to request.
2979 Patch contributed by Kirill Miazine.
2981 TF/01 Do not log submission mode rewrites if they do not change the address.
2983 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2985 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2986 log files in place. Contributed by Roberto Lima.
2988 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2990 TF/03 Bugzilla 615: When checking the local_parts router precondition
2991 after a local_part_suffix or local_part_prefix option, Exim now
2992 does not use the address's named list lookup cache, since this
2993 contains cached lookups for the whole local part.
2995 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2996 Robert Millan. Documentation is in experimental-spec.txt.
2998 TF/04 Bugzilla 668: Fix parallel build (make -j).
3000 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3002 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3003 Patch provided by Jan Srzednicki.
3005 TF/05 Leading white space used to be stripped from $spam_report which
3006 wrecked the formatting. Now it is preserved.
3008 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3009 that they are available at delivery time.
3011 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3013 TF/08 TLS error reporting now respects the incoming_interface and
3014 incoming_port log selectors.
3016 TF/09 Produce a more useful error message if an SMTP transport's hosts
3017 setting expands to an empty string.
3019 NM/06 Bugzilla 744: EXPN did not work under TLS.
3020 Patch provided by Phil Pennock.
3022 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3023 Patch provided by Richard Godbee.
3025 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3026 acl_smtp_notquit, added index entry.
3028 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3029 Patch provided by Eugene Bujak.
3031 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3032 accept(). Patch provided by Maxim Dounin.
3034 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3035 Patch provided by Phil Pennock.
3037 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3039 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3040 Patch provided by Brad "anomie" Jorsch.
3042 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3043 Patch provided by Dean Brooks.
3045 TK/05 Add native DKIM support (does not depend on external libraries).
3047 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3048 Patch provided by Graeme Fowler.
3050 NM/16 Bugzilla 851: Documentation example syntax fix.
3052 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3054 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3057 NM/19 Bugzilla 745: TLS version reporting.
3058 Patch provided by Phil Pennock.
3060 NM/20 Bugzilla 167: bool: condition support.
3061 Patch provided by Phil Pennock.
3063 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3064 clients. Patch provided by Phil Pennock.
3066 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3067 Patch provided by Brad "anomie" Jorsch.
3069 NM/23 Bugzilla 687: Fix misparses in eximstats.
3070 Patch provided by Heiko Schlittermann.
3072 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3073 Patch provided by Heiko Schlittermann.
3075 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3076 plus update to original patch.
3078 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3080 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3081 Patch provided by David Brownlee.
3083 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3085 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3087 NM/30 Bugzilla 888: TLS documentation bugfixes.
3089 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3091 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3092 Unlike the original bugzilla I have changed all shell scripts in src tree.
3094 NM/33 Bugzilla 898: Transport filter timeout fix.
3095 Patch by Todd Rinaldo.
3097 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3098 Patch by Serge Demonchaux.
3100 NM/35 Bugzilla 39: Base64 decode bug fixes.
3101 Patch by Jakob Hirsch.
3103 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3105 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3107 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3109 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3115 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3116 ALT-N's libdkim that I have put here:
3117 http://duncanthrax.net/exim-experimental/
3119 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3120 were called 'true' and it seems that C99 defines that as a reserved
3121 keyword to be used with 'bool' variable types. That means you could
3122 not include C99-style headers which use bools without triggering
3123 build errors in sieve.c.
3125 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3126 as mailq or other aliases. Changed the --help handling significantly
3127 to do whats expected. exim_usage() emits usage/help information.
3129 SC/01 Added the -bylocaldomain option to eximstats.
3131 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3133 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3135 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3141 PH/01 Another patch from the Sieve maintainer.
3143 PH/02 When an IPv6 address is converted to a string for single-key lookup
3144 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3145 dots are used instead of colons so that keys in lsearch files need not
3146 contain colons. This was done some time before quoting was made available
3147 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3148 (notated using the quote facility) so as to distinguish them from IPv4
3149 keys. This meant that lookups for IP addresses in host lists did not work
3150 for iplsearch lookups.
3152 This has been fixed by arranging for IPv6 addresses to be expressed with
3153 colons if the lookup type is iplsearch. This is not incompatible, because
3154 previously such lookups could never work.
3156 The situation is now rather anomalous, since one *can* have colons in
3157 ordinary lsearch keys. However, making the change in all cases is
3158 incompatible and would probably break a number of configurations.
3160 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3163 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3164 conversion specification without a maximum field width, thereby enabling
3165 a rogue spamd server to cause a buffer overflow. While nobody in their
3166 right mind would setup Exim to query an untrusted spamd server, an
3167 attacker that gains access to a server running spamd could potentially
3168 exploit this vulnerability to run arbitrary code as the Exim user.
3170 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3171 $primary_hostname instead of what libspf2 thinks the hosts name is.
3173 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3174 a directory entry by the name of the lookup key. Previously, if a
3175 symlink pointed to a non-existing file or a file in a directory that
3176 Exim lacked permissions to read, a lookup for a key matching that
3177 symlink would fail. Now it is enough that a matching directory entry
3178 exists, symlink or not. (Bugzilla 503.)
3180 PH/03 The body_linecount and body_zerocount variables are now exported in the
3183 PH/04 Added the $dnslist_matched variable.
3185 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3186 This means they are set thereafter only if the connection becomes
3189 PH/06 Added the client_condition to authenticators so that some can be skipped
3190 by clients under certain conditions.
3192 PH/07 The error message for a badly-placed control=no_multiline_responses left
3193 "_responses" off the end of the name.
3195 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3197 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3198 (without spaces) instead of just copying the configuration text.
3200 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3202 PH/11 Added $max_received_linelength.
3204 PH/12 Added +ignore_defer and +include_defer to host lists.
3206 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3207 way in which PCRE > 7.0 is built.
3209 PH/14 Implemented queue_only_load_latch.
3211 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3212 MAIL command. The effect was to mangle the value on 64-bit systems.
3214 PH/16 Another patch from the Sieve maintainer.
3216 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3218 PH/18 If a system quota error occurred while trying to create the file for
3219 a maildir delivery, the message "Mailbox is full" was not appended to the
3220 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3221 only to a quota excession during the actual writing of the file.
3223 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3224 characters?) which causes problems in log lines. The DN values are now
3225 passed through string_printing() before being added to log lines.
3227 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3228 and InterBase are left for another time.)
3230 PH/21 Added message_body_newlines option.
3232 PH/22 Guard against possible overflow in moan_check_errorcopy().
3234 PH/23 POSIX allows open() to be a macro; guard against that.
3236 PH/24 If the recipient of an error message contained an @ in the local part
3237 (suitably quoted, of course), incorrect values were put in $domain and
3238 $local_part during the evaluation of errors_copy.
3244 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3245 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3248 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3249 issue a MAIL command.
3251 PH/02 In an ACL statement such as
3253 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3255 if a client was not listed at all, or was listed with a value other than
3256 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3257 the condition was not true (as it should be), so access was not denied.
3258 The bug was that the ! inversion was incorrectly passed on to the second
3259 item. This has been fixed.
3261 PH/03 Added additional dnslists conditions == and =& which are different from
3262 = and & when the dns lookup returns more than one IP address.
3264 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3265 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3267 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3268 FSYNC, which compiles an option called disable_fsync that allows for
3269 bypassing fsync(). The documentation is heavily laced with warnings.
3271 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3273 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3274 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3275 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3276 including adding "make clean"; (3) Added -fPIC when compiling the test
3277 dynamically loaded module, to get rid of a warning.
3279 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3280 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3281 = 0s. The bug is otherwise harmless.
3283 PH/07 There was a bug in the dovecot authenticator such that the value of
3284 $auth1 could be overwritten, and so not correctly preserved, after a
3285 successful authentication. This usually meant that the value preserved by
3286 the server_setid option was incorrect.
3288 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3290 PH/09 Installed PCRE release 7.0.
3292 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3293 run for batched SMTP input. It is now run at the start of every message
3294 in the batch. While fixing this I discovered that the process information
3295 (output by running exiwhat) was not always getting set for -bs and -bS
3296 input. This is fixed, and it now also says "batched" for BSMTP.
3298 PH/11 Added control=no_pipelining.
3300 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3301 patch, slightly modified), and move the expansion of helo_data till after
3302 the connection is made in the smtp transport (so it can use these
3305 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3307 PH/14 Added log_selector = +pid.
3309 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3311 PH/16 Add ${if forany and ${if forall.
3313 PH/17 Added dsn_from option to vary the From: line in DSNs.
3315 PH/18 Flush SMTP output before performing a callout, unless control =
3316 no_callout_flush is set.
3318 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3319 was true (the default) a successful delivery failed to delete the retry
3320 item, thus causing premature timeout of the address. The bug is now
3323 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3325 PH/21 Long custom messages for fakedefer and fakereject are now split up
3326 into multiline responses in the same way that messages for "deny" and
3327 other ACL rejections are.
3329 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3330 with slight modification.
3332 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3333 draft, changing the syntax and factoring some duplicate code".
3335 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3336 for deliveries of the second and subsequent messages over the same SMTP
3339 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3340 ${reduce, with only minor "tidies".
3342 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3344 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3345 expansion side effects.
3347 PH/27 When a message times out after an over-quota error from an Exim-imposed
3348 quota, the bounce message says "mailbox is full". This message was not
3349 being given when it was a system quota that was exceeded. It now should
3352 MH/03 Made $recipients available in local_scan(). local_scan() already has
3353 better access to the recipient list through recipients_list[], but
3354 $recipients can be useful in postmaster-provided expansion strings.
3356 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3357 in the case of a MAIL command with additional options following the
3358 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3359 were accidentally chopped off.
3361 PH/29 SMTP synchronization checks are implemented when a command is read -
3362 there is a check that no more input is waiting when there shouldn't be
3363 any. However, for some commands, a delay in an ACL can mean that it is
3364 some time before the response is written. In this time, more input might
3365 arrive, invalidly. So now there are extra checks after an ACL has run for
3366 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3367 pipelining has not been advertised.
3369 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3371 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3372 returned in the SMTP response when smtp_return_error_details was false.
3373 This has been fixed.
3375 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3376 instead of the C I/O that was originally supplied, because problems were
3377 reported on Solaris.
3379 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3380 Exim which did not show up earlier: it was assuming that a call to
3381 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3382 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3383 was a macro that became an assignment, so it seemed to work. This has
3384 changed to a proper function call with a void return, hence the compile
3385 error. Exim's code has been fixed.
3387 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3390 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3392 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3394 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3395 because it turns out to be too big or there is a timeout) there is no
3396 "Completed" line in the log. When some messages of this type were
3397 selected by exigrep, they were listed as "not completed". Others were
3398 picked up by some special patterns. I have improved the selection
3399 criteria to be more general.
3401 PH/38 The host_find_failed option in the manualroute router can now be set
3402 to "ignore", to completely ignore a host whose IP address cannot be
3403 found. If all hosts are ignored, the behaviour is controlled by the new
3404 host_all_ignored option.
3406 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3407 homing or because of multiple MX records with /mx) generated more than
3408 one IP address, and the following item turned out to be the local host,
3409 all the secondary addresses of the first item were incorrectly removed
3410 from the list, along with the local host and any following hosts (which
3411 is what is supposed to happen).
3413 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3414 whoever called Exim into the -H file. In the case of the daemon it was
3415 behaving confusingly. When first started, it used values for whoever
3416 started the daemon, but after a SIGHUP it used the Exim user (because it
3417 calls itself on a restart). I have changed the code so that it now always
3420 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3421 message are rejected with the same error (e.g. no authentication or bad
3422 sender address), and a DATA command is nevertheless sent (as can happen
3423 with PIPELINING or a stupid MUA), the error message that was given to the
3424 RCPT commands is included in the rejection of the DATA command. This is
3425 intended to be helpful for MUAs that show only the final error to their
3428 PH/42 Another patch from the Sieve maintainer.
3430 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3431 Eximstats - Fixed some broken HTML links and added missing column headers
3433 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3434 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3436 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3438 PH/43 Yet another patch from the Sieve maintainer.
3440 PH/44 I found a way to check for a TCP/IP connection going away before sending
3441 the response to the final '.' that terminates a message, but only in the
3442 case where the client has not sent further data following the '.'
3443 (unfortunately, this is allowed). However, in many cases there won't be
3444 any further data because there won't be any more messages to send. A call
3445 to select() can be used: if it shows that the input is "ready", there is
3446 either input waiting, or the socket has been closed. An attempt to read
3447 the next input character can distinguish the two cases. Previously, Exim
3448 would have sent an OK response which the client would never have see.
3449 This could lead to message repetition. This fix should cure that, at
3450 least in a lot of common cases.
3452 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3453 advertised in response to EHLO.
3459 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3460 fixed by 4.65/MH/01 (is this a record?) are fixed:
3462 (i) An empty string was always treated as zero by the numeric comparison
3463 operators. This behaviour has been restored.
3465 (ii) It is documented that the numeric comparison operators always treat
3466 their arguments as decimal numbers. This was broken in that numbers
3467 starting with 0 were being interpreted as octal.
3469 While fixing these problems I realized that there was another issue that
3470 hadn't been noticed. Values of message_size_limit (both the global option
3471 and the transport option) were treated as octal if they started with 0.
3472 The documentation was vague. These values are now always treated as
3473 decimal, and I will make that clear in the documentation.
3479 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3480 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3483 MH/01 Don't check that the operands of numeric comparison operators are
3484 integers when their expansion is in "skipping" mode (fixes bug
3485 introduced by 4.64-PH/07).
3487 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3488 child addresses, Exim now panics and dies. Previously, because the count
3489 is held in a short int, deliveries were likely to be lost. As such a
3490 large number of recipients for a single message is ridiculous
3491 (performance will be very, very poor), I have chosen to impose a limit
3492 rather than extend the field.
3498 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3499 leftover -K file (the existence of which was triggered by #402).
3500 While we were at it, introduced process PID as part of the -K
3501 filename. This should rule out race conditions when creating
3504 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3505 processing considerably. Previous code took too long for large mails,
3506 triggering a timeout which in turn triggers #401.
3508 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3509 in the DK code in transports.c. sendfile() is not really portable,
3510 hence the _LINUX specificness.
3512 TF/01 In the add_headers option to the mail command in an Exim filter,
3513 there was a bug that Exim would claim a syntax error in any
3514 header after the first one which had an odd number of characters
3517 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3518 callout verification, Exim cached a "reject" for the entire domain. This
3519 is correct for most verifications, but it is not correct for a recipient
3520 verification with use_sender or use_postmaster set, because in that case
3521 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3522 case of MAIL FROM:<> rejection from other early rejections (e.g.
3523 rejection of HELO). When verifying a recipient using a non-null MAIL
3524 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3525 Whatever the result of the callout, the value of the domain cache is
3526 left unchanged (for any other kind of callout, getting as far as trying
3527 RCPT means that the domain itself is ok).
3529 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3532 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3533 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3534 session, the daemon ignores ECONNECT errors and logs others; it now
3535 ignores EPIPE as well.
3537 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3538 (quoted-printable decoding).
3540 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3541 later the small subsequent patch to fix an introduced bug.
3543 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3545 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3547 PH/08 An error is now given if message_size_limit is specified negative.
3549 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3550 to be given (somewhat) arbitrary names.
3552 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3555 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3556 miscellaneous code fixes
3558 PH/10 Added the log_reject_target ACL modifier to specify where to log
3561 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3562 hostname. This is wrong, because it relates to the incoming message (and
3563 probably the interface on which it is arriving) and not to the outgoing
3564 callout (which could be using a different interface). This has been
3565 changed to use the value of the helo_data option from the smtp transport
3566 instead - this is what is used when a message is actually being sent. If
3567 there is no remote transport (possible with a router that sets up host
3568 addresses), $smtp_active_hostname is used.
3570 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3571 tweaks were necessary in order to get it to work (see also 21 below):
3572 (a) The code assumed that strncpy() returns a negative number on buffer
3573 overflow, which isn't the case. Replaced with Exim's string_format()
3575 (b) There were several signed/unsigned issues. I just did the minimum
3576 hacking in of casts. There is scope for a larger refactoring.
3577 (c) The code used strcasecmp() which is not a standard C function.
3578 Replaced with Exim's strcmpic() function.
3579 (d) The code set only $1; it now sets $auth1 as well.
3580 (e) A simple test gave the error "authentication client didn't specify
3581 service in request". It would seem that Dovecot has changed its
3582 interface. Fortunately there's a specification; I followed it and
3583 changed what the client sends and it appears to be working now.
3585 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3588 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3589 address A is aliased to B and C, where B exists and C does not. Without
3590 -v the output is "A verified" because verification stops after a
3591 successful redirection if more than one address is generated. However,
3592 with -v the child addresses are also verified. Exim was outputting "A
3593 failed to verify" and then showing the successful verification for C,
3594 with its parentage. It now outputs "B failed to verify", showing B's
3595 parentage before showing the successful verification of C.
3597 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3598 look up a TXT record in a specific list after matching in a combined
3601 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3602 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3603 they consult the DNS. I had assumed they would set it the way they
3604 wanted; and indeed my experiments on Linux seem to show that in some
3605 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3606 To be on the safe side, however, I have now made the interface to
3607 host_find_byname() similar to host_find_bydns(), with an argument
3608 containing the DNS resolver options. The host_find_byname() function now
3609 sets these options at its start, just as host_find_bydns() does. The smtp
3610 transport options dns_qualify_single and dns_search_parents are passed to
3611 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3612 of host_find_byname() use the default settings of RES_DEFNAMES
3613 (qualify_single) but not RES_DNSRCH (search_parents).
3615 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3616 spool_read_header() do less string testing, by means of a preliminary
3617 switch on the second character of optional "-foo" lines. (This is
3618 overdue, caused by the large number of possibilities that now exist.
3619 Originally there were few.) While I was there, I also converted the
3620 str(n)cmp tests so they don't re-test the leading "-" and the first
3621 character, in the hope this might squeeze out yet more improvement.
3623 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3624 flag allowing group syntax was set by the header_syntax check but not
3625 turned off, possible causing trouble later; (2) The flag was not being
3626 set at all for the header_verify test, causing "group"-style headers to
3627 be rejected. I have now set it in this case, and also caused header_
3628 verify to ignore an empty address taken from a group. While doing this, I
3629 came across some other cases where the code for allowing group syntax
3630 while scanning a header line wasn't quite right (mostly, not resetting
3631 the flag correctly in the right place). These bugs could have caused
3632 trouble for malformed header lines. I hope it is now all correct.
3634 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3635 with the "reply" argument non-NULL. The code, however (which originally
3636 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3637 but it didn't always do it. This confused somebody who was copying the
3638 code for some other use. I have removed all the tests.
3640 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3641 feature that was used to support insecure browsers during the U.S. crypto
3642 embargo. It requires special client support, and Exim is probably the
3643 only MTA that supported it -- and would never use it because real RSA is
3644 always available. This code has been removed, because it had the bad
3645 effect of slowing Exim down by computing (never used) parameters for the
3646 RSA_EXPORT functionality.
3648 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3649 authenticator to fail if there's a tab character in the incoming data
3650 (there should never be unless someone is messing about, as it's supposed
3651 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3652 if the connection is using TLS or if the remote IP is the same as the
3653 local IP, and the "valid-client-cert option" if a client certificate has
3656 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3657 authenticators. This can be used for authorization after authentication
3658 succeeds. (In the case of plaintext, it servers for both authentication
3661 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3662 if any retry times were supplied.
3664 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3665 connection, where there is no client IP address to check. In this
3666 situation, the verify now always succeeds.
3668 PH/25 Applied John Jetmore's -Mset patch.
3670 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3672 PH/27 In a string expansion for a processed (not raw) header when multiple
3673 headers of the same name were present, leading whitespace was being
3674 removed from all of them, but trailing whitespace was being removed only
3675 from the last one. Now trailing whitespace is removed from each header
3676 before concatenation. Completely empty headers in a concatenation (as
3677 before) are ignored.
3679 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3680 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3682 PH/29 [Removed. This was a change that I later backed out, and forgot to
3683 correct the ChangeLog entry (that I had efficiently created) before
3684 committing the later change.]
3686 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3687 address errors (4xx response to RCPT) over the same connection as other
3688 messages routed to the same hosts. Such deliveries are always "forced",
3689 so retry times are not inspected. This resulted in far too many retries
3690 for the affected addresses. The effect occurred only when there were more
3691 hosts than the hosts_max_try setting in the smtp transport when it had
3692 the 4xx errors. Those hosts that it had tried were not added to the list
3693 of hosts for which the message was waiting, so if all were tried, there
3694 was no problem. Two fixes have been applied:
3696 (i) If there are any address or message errors in an SMTP delivery, none
3697 of the hosts (tried or untried) are now added to the list of hosts
3698 for which the message is waiting, so the message should not be a
3699 candidate for sending over the same connection that was used for a
3700 successful delivery of some other message. This seems entirely
3701 reasonable: after all the message is NOT "waiting for some host".
3702 This is so "obvious" that I'm not sure why it wasn't done
3703 previously. Hope I haven't missed anything, but it can't do any
3704 harm, as the worst effect is to miss an optimization.
3706 (ii) If, despite (i), such a delivery is accidentally attempted, the
3707 routing retry time is respected, so at least it doesn't keep
3708 hammering the server.
3710 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3711 in ${readsocket because some servers need this prod.
3713 PH/32 Added some extra debug output when updating a wait-xxx database.
3715 PH/33 The hint "could be header name not terminated by colon", which has been
3716 given for certain expansion errors for a long time, was not being given
3717 for the ${if def:h_colon_omitted{... case.
3719 PH/34 The spec says: "With one important exception, whenever a domain list is
3720 being scanned, $domain contains the subject domain." There was at least
3721 one case where this was not true.
3723 PH/35 The error "getsockname() failed: connection reset by peer" was being
3724 written to the panic log as well as the main log, but it isn't really
3725 panic-worthy as it just means the connection died rather early on. I have
3726 removed the panic log writing for the ECONNRESET error when getsockname()
3729 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3730 runs only) independently of the message's sender address. This meant
3731 that, if the 4xx error was in fact related to the sender, a different
3732 message to the same recipient with a different sender could confuse
3733 things. In particular, this can happen when sending to a greylisting
3734 server, but other circumstances could also provoke similar problems.
3735 I have changed the default so that the retry time for these errors is now
3736 based a combination of the sender and recipient addresses. This change
3737 can be overridden by setting address_retry_include_sender=false in the
3740 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3741 remote server are returned as part of bounce messages. This was not
3742 happening for LMTP over a pipe (the lmtp transport), but now it is the
3743 same for both kinds of LMTP.
3745 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3746 in header lines that were in fact CNAMEs. This is no longer the case.
3748 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3749 and queue runs started by the daemon processed all messages. This has
3750 been fixed so that -R and -S can now usefully be given with -q<time>.
3752 PH/40 Import PCRE release 6.7 (fixes some bugs).
3754 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3756 PH/42 Give an error if -q is specified more than once.
3758 PH/43 Renamed the variables $interface_address and $interface_port as
3759 $received_ip_address and $received_port, to make it clear that these
3760 values apply to message reception, and not to the outgoing interface when
3761 a message is delivered. (The old names remain recognized, of course.)
3763 PH/44 There was no timeout on the connect() call when using a Unix domain
3764 socket in the ${readsocket expansion. There now is.
3766 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3767 be meaningful with "accept".
3769 SC/01 Eximstats V1.43
3770 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3772 SC/02 Eximstats V1.44
3773 Use a glob alias rather than an array ref in the generated
3774 parser. This improves both readability and performance.
3776 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3777 Collect SpamAssassin and rejection statistics.
3778 Don't display local sender or destination tables unless
3779 there is data to show.
3780 Added average volumes into the top table text output.
3782 SC/04 Eximstats V1.46
3783 Collect data on the number of addresses (recipients)
3784 as well as the number of messages.
3786 SC/05 Eximstats V1.47
3787 Added 'Message too big' to the list of mail rejection
3788 reasons (thanks to Marco Gaiarin).
3790 SC/06 Eximstats V1.48
3791 Mainlog lines which have GMT offsets and are too short to
3792 have a flag are now skipped.
3794 SC/07 Eximstats V1.49 (Alain Williams)
3795 Added the -emptyok flag.
3797 SC/08 Eximstats V1.50
3798 Fixes for obtaining the IP address from reject messages.
3800 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3801 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3802 whitespace changes from 4.64-PH/27
3804 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3807 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3808 are found, allow negative numbers in numeric criteria)
3810 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3812 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3813 to match changes made in 4.64-PH/43
3815 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3817 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3818 "domain" is actually an IP address, and force a failure. This locks out
3819 those revolvers/nameservers that support "A-for-A" lookups, in
3820 contravention of the specifications.
3822 PH/48 When a host name was looked up from an IP address, and the subsequent
3823 forward lookup of the name timed out, the host name was left in
3824 $sender_host_name, contrary to the specification.
3826 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3827 restricted to single-key lookups, Exim was not diagnosing an error if
3828 * or *@ was used with a query-style lookup.
3830 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3832 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3833 long ago, but noone interested enough thought of it. Let's just say that
3834 the "1.1" means that there are some new functions that weren't there at
3835 some point in the past.
3837 PH/51 Error processing for expansion failure of helo_data from an smtp
3838 transport during callout processing was broken.
3840 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3841 tested/used via the -bh/-bhc/-bs options.
3843 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3844 bug, fixed in subsequent PCRE releases).
3846 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3847 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3849 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3855 SC/01 Use a glob alias rather than an array ref in eximstats generated
3856 parser. This improves both readability and performance.
3858 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3859 Don't display local sender or destination tables in eximstats unless
3860 there is data to show.
3861 Added average volumes into the eximstats top table text output.
3863 SC/03 Collect data on the number of addresses (recipients) as well
3864 as the number of messages in eximstats.
3866 TF/01 Correct an error in the documentation for the redirect router. Exim
3867 does (usually) call initgroups() when daemonizing.
3869 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3870 with consistent privilege compared to when running as a daemon.
3872 TF/03 Note in the spec that $authenticated_id is not set for local
3873 submissions from trusted users.
3875 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3876 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3878 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3879 by adding some example configuration directives to the default
3880 configuration file. A little bit of work is required to uncomment the
3881 directives and define how usernames and passwords are checked, but
3882 there is now a framework to start from.
3884 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3885 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3886 without this. I don't know how relevant this is to other LDAP libraries.
3888 PH/02 Add the verb name to the "unknown ACL verb" error.
3890 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3892 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3894 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3895 directory not expanded when it should be if an expanded home directory
3896 was set for the address (which is overridden by the transport).
3898 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3901 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3902 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3903 because it is too late at that time, and has no effect.
3905 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3906 security issue with \' (bugzilla #107). I could not use the
3907 PQescapeStringConn() function, because it needs a PGconn value as one of
3910 PH/08 When testing addresses using -bt, indicate those final addresses that
3911 are duplicates that would not cause an additional delivery. At least one
3912 person was confused, thinking that -bt output corresponded to deliveries.
3913 (Suppressing duplicates isn't a good idea as you lose the information
3914 about possibly different redirections that led to the duplicates.)
3916 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3917 systems where poll() doesn't work, in particular OS X.
3919 PH/10 Added more information to debugging output for retry time not reached.
3921 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3922 operations in malware.c.
3924 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3927 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3928 syslog_duplication set false, log lines that would normally be written
3929 both the the main log and to the reject log were not written to syslog at
3932 PH/14 In the default configuration, change the use of "message" in ACL warn
3933 statements to "add_header".
3935 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3936 not followed by a command (e.g. "seen endif").
3938 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3939 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3942 PH/17 Added extra conditions to the default value of delay_warning_condition
3946 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3947 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3948 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3951 The Auto-Submitted: and various List- headers are standardised, whereas I
3952 don't think Precedence: ever was.
3954 PH/18 Refactored debugging code in route_finduser() to show more information,
3955 in particular, the error code if getpwnam() issues one.
3957 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3958 This is apparently needed in addition to the PH/07 change above to avoid
3959 any possible encoding problems.
3961 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3962 but not after initializing Perl.
3964 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3965 output them only if debugging. By default they are written stderr,
3966 apparently, which is not desirable.
3968 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3971 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3974 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3976 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3977 authenticated or an ident call has been made. Suppress the default
3978 values for $authenticated_id and $authenticated_sender (but permit -oMai
3979 and -oMas) when testing with -bh.
3981 PH/24 Re-jigged the order of the tests in the default configuration so that the
3982 tests for valid domains and recipients precede the DNS black list and CSA
3983 tests, on the grounds that those ones are more expensive.
3985 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3986 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3987 This bug exists in every version of Exim that I still have, right back to
3990 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3991 However, an attempt to turn on case-sensitivity in a regex key by
3992 including (?-i) didn't work because the subject string was already
3993 lowercased, and the effects were non-intuitive. It turns out that a
3994 one-line patch can be used to allow (?-i) to work as expected.
4000 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4001 other effects) broke the use of negated acl sub-conditions.
4003 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4006 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4007 "Deny" causes Exim to reject the incoming connection with a 554 error.
4008 Unfortunately, if there is a major crisis, such as a disk failure,
4009 tcp-wrappers gives "deny", whereas what one would like would be some
4010 kind of temporary error. A kludge has been added to help with this.
4011 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4012 554 error is used if errno is still zero or contains ENOENT (which occurs
4013 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4016 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4018 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4019 errors. Otherwise a message that provokes a temporary error (when other
4020 messages do not) can cause a whole host to time out.
4022 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4023 addresses were routed directly to files or pipes from a redirect router.
4024 File deliveries just didn't batch; pipe deliveries might have suffered
4027 PH/06 A failure to get a lock for a hints database would erroneously always say
4028 "Failed to get write lock", even when it was really a read lock.
4030 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4031 of 0600. This has been changed to use the value of the lockfile_mode
4032 option (which defaults to 0600).
4034 PH/08 Applied small patch from the Sieve maintainer.
4036 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4037 folder from quota calculations, a direct delivery into this folder messed
4038 up the contents of the maildirsize file. This was because the regex was
4039 used only to exclude .Trash (or whatever) when the size of the mailbox
4040 was calculated. There was no check that a delivery was happening into an
4041 excluded directory. This bug has been fixed by ignoring all quota
4042 processing for deliveries into excluded directories.
4044 PH/10 Added the maildirfolder_create_regex option to appendfile.
4050 PH/01 The code for finding all the local interface addresses on a FreeBSD
4051 system running IPv6 was broken. This may well have applied to all BSD
4052 systems, as well as to others that have similar system calls. The broken
4053 code found IPv4 interfaces correctly, but gave incorrect values for the
4054 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4055 that it would not match correctly against @[] and not recognize the IPv6
4058 PH/02 The ipliteral router was not recognizing addresses of the form user@
4059 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4061 PH/03 Added disable_ipv6.
4063 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4064 decoded form, because it is most often used to construct To: headers
4065 lines in autoreplies, and the decoded form may well be syntactically
4066 invalid. However, $reply_address has leading white space removed, and all
4067 newlines turned into spaces so that the autoreply transport does not
4070 PH/05 If group was specified without a user on a router, and no group or user
4071 was specified on a transport, the group from the router was ignored.
4073 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4074 for visible compile-time settings that can be used to change these
4075 numbers, for those that want even more. Backwards compatibility with old
4076 spool files has been maintained. However, going back to a previous Exim
4077 release will lost any variables that are in spool files.
4079 PH/07 Two small changes when running in the test harness: increase delay when
4080 passing a TCP/IP connection to a new process, in case the original
4081 process has to generate a bounce, and remove special handling of
4082 127.0.0.2 (sic), which is no longer necessary.
4084 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4085 be the same on different OS.
4087 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4090 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4091 whether --show-vars was specified or not
4093 JJ/02 exipick: Added support for new ACL variable spool format introduced
4096 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4097 syntactically invalid From: or Reply-to: line, and a filter used this to
4098 generate an autoreply, and therefore failed to obtain an address for the
4099 autoreply, Exim could try to deliver to a non-existent relative file
4100 name, causing unrelated and misleading errors. What now happens is that
4101 it logs this as a hard delivery error, but does not attempt to create a
4104 PH/11 The exinext utility has a -C option for testing purposes, but although
4105 the given file was scanned by exinext itself; it wasn't being passed on
4106 when Exim was called.
4108 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4109 an end-of-file indication when reading a command response.
4111 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4112 compiled. In many other places in Exim, IPv6 addresses are always
4113 recognized, so I have changed this. It also means that IPv4 domain
4114 literals of the form [IPV4:n.n.n.n] are now always recognized.
4116 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4117 used if the router is not running as root, for example, when verifying at
4118 ACL time, or when using -bh. The debugging output from this situation was
4119 non-existent - all you got was a failure to exec. I have made two
4122 (a) Failures to set uid/gid, the current directory, or a process leader
4123 in a subprocess such as that created by queryprogram now generate
4124 suitable debugging output when -d is set.
4126 (b) The queryprogram router detects when it is not running as root,
4127 outputs suitable debugging information if -d is set, and then runs
4128 the subprocess without attempting to change uid/gid.
4130 PH/15 Minor change to Makefile for building test_host (undocumented testing
4133 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4134 additional section of a DNS packet that returns MX or SRV records.
4135 Instead, it always explicitly searches for A/AAAA records. This avoids
4136 major problems that occur when a DNS server includes only records of one
4137 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4138 fixed another bug: if SRV records were looked up and the corresponding
4139 address records were *not* found in the additional section, the port
4140 values from the SRV records were lost.
4142 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4143 using the correct key (the original address) when searching the retry
4144 rules in order to find which one to use for generating the retry hint.
4146 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4147 adding the default one. Similarly, if it contains a Reply-To: header, the
4148 errors_reply_to option, if set, is not used.
4150 PH/19 When calculating a retry time, Exim used to measure the "time since
4151 failure" by looking at the "first failed" field in the retry record. Now
4152 it does not use this if it is later than than the arrival time of the
4153 message. Instead it uses the arrival time. This makes for better
4154 behaviour in cases where some deliveries succeed, thus re-setting the
4155 "first failed" field. An example is a quota failure for a huge message
4156 when small messages continue to be delivered. Without this change, the
4157 "time since failure" will always be short, possible causing more frequent
4158 delivery attempts for the huge message than are intended.
4159 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4161 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4162 $1, $2, $3) because the numerical variables can be reset during some
4163 expansion items (e.g. "match"), thereby losing the authentication data.
4165 PH/21 Make -bV show the size of off_t variables so that the test suite can
4166 decide whether to run tests for quotas > 2G.
4168 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4169 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4170 filecount value is greater than 2G or if a quota value is greater than 2G
4171 on a system where the size of off_t is not greater than 4, a panic error
4174 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4175 never match. The debug and -bh output now contains an explicit error
4176 message indicating a malformed IPv4 address or mask.
4178 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4179 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4180 PH/23 above applies.
4182 PH/25 Do not write to syslog when running in the test harness. The only
4183 occasion when this arises is a failure to open the main or panic logs
4184 (for which there is an explicit test).
4186 PH/26 Added the /no_tell option to "control=freeze".
4188 PH/27 If a host name lookup failed very early in a connection, for example, if
4189 the IP address matched host_lookup and the reverse lookup yielded a name
4190 that did not have a forward lookup, an error message of the form "no IP
4191 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4192 could be logged. Now it outputs the IP address instead of "NULL".
4194 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4195 allows the sender and the authenticated sender to be set when
4196 submitting a message from within Exim. Since child_open_exim() is
4197 documented for local_scan(), the new function should be too.
4199 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4200 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4201 results in an empty string is now treated as unset.
4203 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4205 PH/31 Added sender_verify_fail logging option.
4207 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4208 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4209 tidied the source and removed it altogether.
4211 PH/33 When a queue run was abandoned because the load average was too high, a
4212 log line was always written; now it is written only if the queue_run log
4213 selector is set. In addition, the log line for abandonment now contains
4214 information about the queue run such as the pid. This is always present
4215 in "start" and "stop" lines but was omitted from the "abandon" line.
4217 PH/34 Omit spaces between a header name and the colon in the error message that
4218 is given when verify = headers_syntax fails (if there are lots of them,
4219 the message gets confusing).
4221 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4222 names, as there are now some PTR records that contain slashes. This check
4223 is only to protect against broken name servers that fall over on strange
4224 characters, so the fact that it applies to all lookups doesn't matter.
4226 PH/36 Now that the new test suite is complete, we can remove some of the
4227 special code in Exim that was needed for the old test suite. For example,
4228 sorting DNS records because real resolvers return them in an arbitrary
4229 order. The new test suite's fake resolver always returns records in the
4232 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4233 bounces) except when queue_only is set, to avoid logging races between
4234 the different processes.
4236 PH/38 Panic-die if .include specifies a non-absolute path.
4238 PH/39 A tweak to the "H" retry rule from its user.
4240 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4241 a label. They prevented compilation on older perls.
4243 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4244 a warning to be raised on newish perls.
4246 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4247 on queue. Changes to match documented behaviour of showing count of
4248 messages matching specified criteria.
4250 PH/40 Changed the default ident timeout from 30s to 5s.
4252 PH/41 Added support for the use of login_cap features, on those BSD systems
4253 that have them, for controlling the resources used by pipe deliveries.
4255 PH/42 The content-scanning code uses fopen() to create files in which to put
4256 message data. Previously it was not paying any attention to the mode of
4257 the files. Exim runs with umask(0) because the rest of the code creates
4258 files with open(), and sets the required mode explicitly. Thus, these
4259 files were ending up world-writeable. This was not a big issue, because,
4260 being within the spool directory, they were not world-accessible. I have
4261 created a function called modefopen, which takes an additional mode
4262 argument. It sets umask(777), creates the file, chmods it to the required
4263 mode, then resets the umask. All the relevant calls to fopen() in the
4264 content scanning code have been changed to use this function.
4266 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4267 to 24 hours. This avoids potential overflow problems when processing G
4268 and H retry rules. I suspect nobody ever tinkers with this value.
4270 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4272 PH/45 When the plaintext authenticator is running as a client, the server's
4273 challenges are checked to ensure they are valid base64 strings. By
4274 default, the authentication attempt is cancelled if an invalid string is
4275 received. Setting client_ignore_invalid_base64 true ignores these errors.
4276 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4277 they are received. Thus, the responses can be made to depend on the
4278 challenges. If an invalid string is ignored, an empty string is placed in
4281 PH/46 Messages that are created by the autoreply transport now contains a
4282 References: header, in accordance with RFCs 2822 and 3834.
4284 PH/47 Added authenticated_sender_force to the smtp transport.
4286 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4288 PH/49 Installed latest patch from the Sieve maintainer.
4290 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4291 was also set, the appendfile transport was unnecessarily scanning a
4292 directory of message files (e.g. for maildir delivery) to find the count
4293 of files (along with the size), even though it did not need this
4294 information. It now does the scan only if it needs to find either the
4295 size of the count of files.
4297 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4299 PH/52 Two bugs concerned with error handling when the smtp transport is
4302 (i) Exim was not creating retry information for temporary errors given
4303 for individual recipients after the DATA command when the smtp transport
4304 was used in LMTP mode. This meant that they could be retried too
4305 frequently, and not timed out correctly.
4307 (ii) Exim was setting the flag that allows error details to be returned
4308 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4309 recipients that were returned after the DATA command.
4311 PH/53 This is related to PH/52, but is more general: for any failing address,
4312 when detailed error information was permitted to be returned to the
4313 sender, but the error was temporary, then after the final timeout, only
4314 "retry timeout exceeded" was returned. Now it returns the full error as
4315 well as "retry timeout exceeded".
4317 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4318 do this, and (what is worse) MTAs that accept it.
4320 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4321 will now be deprecated.
4323 PH/56 New os.c-cygwin from the Cygwin maintainer.
4325 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4326 formats (previously only available in exim formats via -bpr, -bpru,
4327 and -bpra. Now also available in native and exiqgrep formats)
4329 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4330 with very large, slow to parse queues
4332 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4334 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4336 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4337 responses to authentication challenges, though it was showing the
4338 challenges; (ii) I've removed the CR characters from the debug output for
4341 PH/58 Allow for the insertion of a newline as well as a space when a string
4342 is turned into more than one encoded-word during RFC 2047 encoding. The
4343 Sieve code now uses this.
4345 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4346 data_4xx, lost_connection, tls_required.
4348 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4349 message was being sent as an SMTP response.
4351 PH/61 Add -l and -k options to exicyclog.
4353 PH/62 When verifying, if an address was redirected to one new address, so that
4354 verification continued, and the new address failed or deferred after
4355 having set something in $address_data, the value of $address_data was not
4356 passed back to the ACL. This was different to the case when no
4357 redirection occurred. The value is now passed back in both cases.
4359 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4360 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4361 use login_cap.h, so on its own it isn't the distinguishing feature. The
4362 new name refers directly to the setclassresources() function.
4364 PH/65 Added configuration files for NetBSD3.
4366 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4368 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4369 is preferred over IPv4.
4371 PH/68 The bounce_return_message and bounce_return_body options were not being
4372 honoured for bounces generated during the reception of non-SMTP messages.
4373 In particular, this applied to messages rejected by the ACL. This bug has
4374 been fixed. However, if bounce_return_message is true and bounce_return_
4375 body is false, the headers that are returned for a non-SMTP message
4376 include only those that have been read before the error was detected.
4377 (In the case of an ACL rejection, they have all been read.)
4379 PH/69 The HTML version of the specification is now built in a directory called
4380 spec_html instead of spec.html, because the latter looks like a path with
4381 a MIME-type, and this confuses some software.
4383 PH/70 Catch two compiler warnings in sieve.c.
4385 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4386 function verify_get_ident() calls ip_connect() to connect a socket, but
4387 if the "connect()" function timed out, ip_connect() used to close the
4388 socket. However, verify_get_ident() also closes the socket later, and in
4389 between Exim writes to the log, which may get opened at this point. When
4390 the socket was closed in ip_connect(), the log could get the same file
4391 descriptor number as the socket. This naturally causes chaos. The fix is
4392 not to close the socket in ip_connect(); the socket should be closed by
4393 the function that creates it. There was only one place in the code where
4394 this was missing, in the iplookup router, which I don't think anybody now
4395 uses, but I've fixed it anyway.
4397 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4398 well as to direct DNS lookups. Otherwise the handling of names in host
4399 lists is inconsistent and therefore confusing.
4405 PH/01 Two changes to the default runtime configuration:
4407 (1) Move the checks for relay_from_hosts and authenticated clients from
4408 after to before the (commented out) DNS black list checks.
4410 (2) Add control=submission to the relay_from_hosts and authenticated
4411 clients checks, on the grounds that messages accepted by these
4412 statements are most likely to be submissions.
4414 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4416 (1) Generate an error if the third argument for the ${prvs expansion is
4419 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4422 (3) Reset the variables that are obtained from the first argument of
4423 ${prvscheck and used in the second argument before leaving the code,
4424 because their memory is reclaimed, so using them afterwards may do
4427 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4428 one (it's much easier than Tom thought :-).
4430 (5) Because of (4), we can now allow for the use of $prvscheck_result
4431 inside the third argument.
4433 PH/03 For some reason, the default setting of PATH when running a command from
4434 a pipe transport was just "/usr/bin". I have changed it to
4437 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4438 anything to be listed in the output from -bV.
4440 PH/05 When a filter generated an autoreply, the entire To: header line was
4441 quoted in the delivery log line, like this:
4443 => >A.N.Other <ano@some.domain> <original@ddress> ...
4445 This has been changed so that it extracts the operative address. There
4446 may be more than one such address. If so, they are comma-separated, like
4449 => >ano@some.domain,ona@other.domain <original@ddress> ...
4451 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4452 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4453 not being looked up in the rDNS to get a host name, Exim was showing the
4454 IP address twice in Received: lines, even though the IP addresses were
4455 identical. For example:
4457 Received: from [1.2.3.4] (helo=[1.2.3.4])
4459 However, if the real host name was known, it was omitting the HELO data
4460 if it matched the actual IP address. This has been tidied up so that it
4461 doesn't show the same IP address twice.
4463 PH/07 When both +timestamp and +memory debugging was on, the value given by
4464 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4465 called by the debug printing, thereby overwriting the timestamp buffer.
4466 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4468 PH/08 When the original message was included in an autoreply transport, it
4469 always said "this is a copy of the message, including all the headers",
4470 even if body_only or headers_only was set. It now gives an appropriate
4473 PH/09 Applied a patch from the Sieve maintainer which:
4475 o fixes some comments
4476 o adds the (disabled) notify extension core
4477 o adds some debug output for the result of if/elsif tests
4478 o points to the current vacation draft in the documentation
4479 and documents the missing references header update
4483 o fixes a bug in processing the envelope test (when testing
4484 multiple envelope elements, the last element determined the
4487 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4488 Electronic Mail") by including:
4490 Auto-submitted: auto-generated
4492 in the messages that it generates (bounce messages and others, such as
4493 warnings). In the case of bounce messages for non-SMTP messages, there was
4494 also a typo: it was using "Auto_submitted" (underscore instead of
4495 hyphen). Since every message generated by Exim is necessarily in response
4496 to another message, thes have all been changed to:
4498 Auto-Submitted: auto-replied
4500 in accordance with these statements in the RFC:
4502 The auto-replied keyword:
4504 - SHOULD be used on messages sent in direct response to another
4505 message by an automatic process,
4507 - MUST NOT be used on manually-generated messages,
4509 - MAY be used on Delivery Status Notifications (DSNs) and Message
4510 Disposition Notifications (MDNs),
4512 - MUST NOT be used on messages generated by automatic or periodic
4513 processes, except for messages which are automatic responses to
4516 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4517 to the default Received: header definition.
4519 PH/12 Added log selector acl_warn_skipped (default on).
4521 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4522 variables because (a) they are in the wrong storage pool and (b) even if
4523 they were copied, it wouldn't work properly because of the caching.
4525 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4526 checking when decoding. Apparently there are clients that generate
4527 overlong encoded strings. Why am I not surprised?
4529 PH/15 If the first argument of "${if match_address" was not empty, but did not
4530 contain an "@" character, Exim crashed. Now it writes a panic log message
4531 and treats the condition as false.
4533 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4535 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4536 extension "envelope-auth". The code is finished and in agreement with
4537 other implementations, but there is no documentation so far and in fact,
4538 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4539 not changing the active code.
4541 Print executed "if" and "elsif" statements when debugging is used. This
4542 helps a great deal to understand what a filter does.
4544 Document more things not specified clearly in RFC3028. I had all this
4545 sorted out, when out of a sudden new issues came to my mind. Oops."
4547 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4550 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4551 canonical form (as documented). However, after a host name lookup from
4552 the IP address, check_host() was doing a simple string comparison with
4553 addresses acquired from the DNS when checking that the found name did
4554 have the original IP as one of its addresses. Since any found IPv6
4555 addresses are likely to be in abbreviated form, the comparison could
4556 fail. Luckily, there already exists a function for doing the comparison
4557 by converting both addresses to binary, so now that is used instead of
4558 the text comparison.
4560 PH/20 There was another similar case to PH/19, when a complete host name was
4561 given in a host list; looking up its IP address could give an abbreviated
4562 form, whereas the current host's name might or might not be abbreviated.
4563 The same fix has been applied.
4569 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4570 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4573 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4574 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4576 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4578 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4579 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4580 Also, for "Auto-Submitted", treat anything other than "no" as
4581 non-personal, in accordance with RFC 3834. (Previously it treated
4582 anything starting "auto-" as non-personal.)
4584 TF/01 The control=submission/name=... option had a problem with syntax
4585 errors if the name included a slash character. The /name= option
4586 now slurps the rest of the string, so it can include any characters
4587 but it must come last in the list of options (after /sender_retain
4590 PH/05 Some modifications to the interface to the fake nameserver for the new
4598 TK/01 Added the "success_on_redirect" address verification option. See
4599 NewStuff for rationale and an example.
4601 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4603 PH/02 Patch to exigrep to allow it to work on syslog lines.
4605 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4606 fread() to skip over the body file's header line, because in Cygwin the
4607 header line is locked and is inaccessible.
4609 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4610 co-exist for some time) to make it clear that it is the Exim ID that is
4611 referenced, not the Message-ID: header line.
4613 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4614 string_format() function, because snprintf() does not exist on all
4617 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4618 ${stat: expansion item.
4620 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4621 protocol synchronization error", to keep the pedants happy.
4623 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4624 well as for IRIX systems, when gcc is being used. See the host.c source
4627 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4629 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4632 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4633 transport and to the smtp transport in LMTP mode.
4635 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4637 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4638 run a filter in a subprocess. This could lead to confusion in subsequent
4639 lookups in the parent process. There should also be a search_tidyup() at
4640 the end of the subprocess.
4642 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4643 only if the host matched helo_try_verify_hosts, which caused the
4644 verification to occur when the EHLO/HELO command was issued. The ACL just
4645 tested the remembered result. Now, if a previous verification attempt has
4646 not happened, "verify = helo" does it there and then.
4648 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4650 TK/03 Fix log output including CR from clamd.
4652 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4653 exist provoked a memory error which could cause a segfault.
4655 PH/15 Installed PCRE 6.2
4657 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4659 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4660 of the problem. Specifically, suggested +O2 rather than +O1 for the
4663 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4665 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4666 record, the port was not correctly logged when the outgoing_port log
4667 selector was set (it logged the transort's default port).
4669 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4670 fallback_hosts, and "hosts" in the smtp transport.
4672 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4673 host errors such as "Connection refused".
4675 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4676 authentication with radiusclient 0.4.9:
4678 - Error returned from rc_read_config was caught wrongly
4679 - Username/password not passed on to radius server due to wrong length.
4681 The presumption is that some radiusclient API changes for 4.51/PH/17
4682 were not taken care of correctly. The code is still untested by me (my
4683 Linux distribution still has 0.3.2 of radiusclient), but it was
4684 contributed by a Radius user.
4686 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4687 expanding the "port" option of the smtp transport.
4689 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4690 while reading a MIME header. Thanks to Tom Hughes for a patch.
4692 PH/24 Include config.h inside local_scan.h so that configuration settings are
4695 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4696 that in an ACL for RCPT (for example), you can examine exactly what was
4699 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4700 commands, but it was not correctly comparing the address with the actual
4701 client host address. Thus, it would show the EHLO address in Received:
4702 header lines when this was not necessary.
4704 PH/27 Added the % operator to ${eval:}.
4706 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4707 it should be ignoring failures (because with -C, for example, it has lost
4708 privilege). It wasn't ignoring creation failures other than "already
4711 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4714 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4715 because an input line was too long, either on its own, or by virtue of
4716 too many continuations, the temporary file was not being removed, and the
4717 return code was incorrect.
4719 PH/31 Missing "BOOL" in function definition in filtertest.c.
4721 PH/32 Applied Sieve patches from the maintainer.
4723 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4725 PH/33 Added "verify = not_blind".
4727 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4728 Local/Makefile (with some defaults set). These are used in built scripts
4729 such as exicyclog, but they have never been used in the exim_install
4730 script (though there are many overriding facilities there). I have
4731 arranged that the exim_install script now takes note of these two
4734 PH/35 Installed configuration files for Dragonfly.
4736 PH/36 When a locally submitted message by a trusted user did not contain a
4737 From: header, and the sender address was obtained from -f or from an SMTP
4738 MAIL command, and the trusted user did not use -F to supply a sender
4739 name, $originator_name was incorrectly used when constructing a From:
4740 header. Furthermore, $originator_name was used for submission mode
4741 messages from external hosts without From: headers in a similar way,
4742 which is clearly wrong.
4744 PH/37 Added control=suppress_local_fixups.
4746 PH/38 When log_selector = +received_sender was set, and the addition of the
4747 sender made the log line's construction buffer exactly full, or one byte
4748 less than full, an overflow happened when the terminating "\n" was
4751 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4752 when the result of a list match is failure because a DNS lookup failed.
4754 PH/40 RM_COMMAND is now used in the building process.
4756 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4757 the "build-* directories that it finds.
4759 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4760 address in a domain literal was a prefix of an interface address.
4762 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4763 when verifying a sender address, unless rewrite_headers is false.
4765 PH/44 Wrote a long comment about why errors_to addresses are verified as
4766 recipients, not senders.
4768 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4769 the ratelimit ACL was added.
4771 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4773 PH/46 Added extra information about PostgreSQL errors to the error string.
4775 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4776 avoiding the need to install special zones in a real server. This is
4777 backwards compatible; if it can't find the fake resolver, it drops back.
4778 Thus, both old and new test suites can be run.
4780 TF/02 Added util/ratelimit.pl
4782 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4783 clock is set back in time.
4785 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4786 Candler <B.Candler@pobox.com>.
4788 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4789 OK for addresses that are the result of redirections.
4791 PH/48 A number of further additions for the benefit of the new test suite,
4792 including a fake gethostbyname() that interfaces to the fake DNS resolver
4795 TF/06 The fix for widen_domains has also been applied to qualify_single and
4796 search_parents which are the other dnslookup options that can cause
4799 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4802 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4804 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4805 service" if it encountered an MX record with an empty target hostname.
4806 The message is now "an MX or SRV record indicated no SMTP service".
4808 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4809 if the DNS of the sending site is misconfigured. This is quite a
4810 common situation. This change restores the behaviour of treating a
4811 helo verification defer as a failure.
4813 PH/51 If self=fail was set on a router, the bounce message did not include the
4814 actual error message.
4820 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4822 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4823 command itself ended in error, the underlying message about the transport
4824 filter timeout was being overwritten with the pipe command error. Now the
4825 underlying error message should be appended to the second error message.
4827 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4829 PH/02 Reduce the amount of output that "make" produces by default. Full output
4830 can still be requested.
4832 PH/03 The warning log line about a condition test deferring for a "warn" verb
4833 was being output only once per connection, rather than after each
4834 occurrence (because it was using the same function as for successful
4835 "warn" verbs). This seems wrong, so I have changed it.
4837 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4838 it should not have, which might have caused a crash in the right
4839 circumstances, but probably never did.
4841 PH/04 Installed a modified version of Tony Finch's patch to make submission
4842 mode fix the return path as well as the Sender: header line, and to
4843 add a /name= option so that you can make the user's friendly name appear
4846 TF/03 Added the control = fakedefer ACL modifier.
4848 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4849 Mark Lowes for thorough testing.
4851 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4853 TK/03 Merged latest SRS patch from Miles Wilton.
4855 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4856 with the definition in sysexits.h (which is #included earlier).
4857 Fortunately, Exim does not actually use EX_OK. The code used to try to
4858 preserve the sysexits.h value, by assuming that macro definitions were
4859 scanned for macro replacements. I have been disabused of this notion,
4860 so now the code just undefines EX_OK before #including unistd.h.
4862 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4863 in the smtp transport. When a block could not be written in a single
4864 write() function, the timeout was being re-applied to each part-write.
4865 This seems wrong - if the receiver was accepting one byte at a time it
4866 would take for ever. The timeout is now adjusted when this happens. It
4867 doesn't have to be particularly precise.
4869 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4870 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4872 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4873 without a domain if the check to <postmaster@domain> fails.
4875 SC/01 Eximstats: added -xls and the ability to specify output files
4876 (patch written by Frank Heydlauf).
4878 SC/02 Eximstats: use FileHandles for outputting results.
4880 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4882 SC/04 Eximstats: fixed display of large numbers with -nvr option
4884 SC/05 Eximstats: fixed merging of reports with empty tables.
4886 SC/06 Eximstats: added the -include_original_destination flag
4888 SC/07 Eximstats: removed tabs and trailing whitespace.
4890 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4892 TK/06 MBOX spool code: Add real "From " MBOX separator line
4893 so the .eml file is really in mbox format (even though
4894 most programs do not really care). Patch from Alex Miller.
4896 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4897 The latter is generated from $received_to and is only set if the
4898 message has one envelope recipient. SA can use these headers,
4899 obviously out-of-the-box. Patch from Alex Miller.
4901 PH/08 The ${def test on a variable was returning false if the variable's
4902 value was "0", contrary to what the specification has always said!
4903 The result should be true unless the variable is empty.
4905 PH/09 The syntax error of a character other than { following "${if
4906 def:variable_name" (after optional whitespace) was not being diagnosed.
4907 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4908 accidental colon was present, for example, could give incorrect results.
4910 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4911 result is used (not including appendfile, where other changes are about
4914 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4915 This involved changing a lot of size variables from int to off_t. It
4916 should work with maildirs and everything.
4918 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4919 spamd dying while we are connected to it.
4921 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4924 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4925 function for BDB 4.3.
4927 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4929 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4930 the sources for PCRE have been re-arranged and more files are now
4933 PH/15 The code I had for printing potentially long long variables in PH/11
4934 above was not the best (it lost precision). The length of off_t variables
4935 is now inspected at build time, and an appropriate printing format (%ld
4936 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4937 to be "long long int" or "long int". This is needed for the internal
4938 formatting function string_vformat().
4940 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4941 the configuration file to be ":syslog", then the script "guesses" where
4942 the logs files are, rather than using the compiled in default. In our
4943 case the guess is not the same as the compiled default, so the script
4944 suddenly stopped working when I started to use syslog. The patch checks
4945 to see if log_file_path is "". If so, it attempts to read it from exim
4946 with no configuration file to get the compiled in version, before it
4947 falls back to the previous guessing code."
4949 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4950 implementing BATV in an Exim configuration. See NewStuff for the gory
4953 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4954 Makefile that are specific to HP-UX.
4956 PH/18 If the "use_postmaster" option was set for a recipient callout together
4957 with the "random" option, the postmaster address was used as the MAIL
4958 FROM address for the random test, but not for the subsequent recipient
4959 test. It is now used for both.
4961 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4962 patch removes a few documentation additions to RFC 3028, because the
4963 latest draft now contains them. It adds the new en;ascii-case comparator
4964 and a new error check for 8bit text in MIME parts. Comparator and
4965 require names are now matched exactly. I enabled the subaddress
4966 extension, but it is not well tested yet (read: it works for me)."
4968 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4969 rework some of the code of TK/09 above to avoid the hardwired use of
4970 "%lld" and "long long". Replaced the call to snprintf() with a call to
4973 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4974 records point to non-existent hosts", "retry timeout exceeded", and
4975 "retry time not reached for any host after a long failure period".
4977 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4978 experimental DomainKeys support:
4980 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4981 (2) On an error such as an illegally used "control", the wrong name for
4982 the control was given.
4984 These problems did NOT occur unless DomainKeys support was compiled.
4986 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4988 PH/24 Added ${if match_ip condition.
4990 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4991 databases so that it will be absolutely obvious if a crash occurs in the
4992 DB library. This is a regular occurrence (often caused by mis-matched
4995 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4996 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4997 picked up on a user's system that detects such things. There doesn't seem
4998 to be a gcc warning option for this - only an attribute that has to be
4999 put on the function's prototype. It seems that in Fedora Core 4 they have
5000 set this on a number of new functions. No doubt there will be more in due
5003 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5004 specify a transport. However, if an address that was verified by such a
5005 router was the subject of a callout, Exim crashed because it tried to
5006 read the rcpt_include_affixes from the non-existent transport. Now it
5007 just assumes that the setting of that option is false. This bug was
5008 introduced by 4.51/PH/31.
5010 PH/28 Changed -d+all to exclude +memory, because that information is very
5011 rarely of interest, but it makes the output a lot bigger. People tend to
5012 do -d+all out of habit.
5014 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5015 code in os-type was giving problems when libc.so lives in lib64, like on
5018 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5019 aren't the modern standard, and it seems that some systems' include files
5020 don't always have them. Exim was already checking for some of the newer
5021 ones like T_AAAA, and defining it itself. I've added checks for all the
5022 record types that Exim uses.
5024 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5025 not automatically generating a new one, as it is supposed to. This
5026 prevented TLS from working. If the file did exist, but contained invalid
5027 data, a new version was generated, as expected. It was only the case of a
5028 non-existent file that was broken.
5030 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5031 with a change in libdomainkeys > 0.64.
5033 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5034 from DNS. If the selector record carries the flag, it now has
5035 precedence over the domain-wide flag.
5037 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5039 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5040 the use of a port name that isn't defined in /etc/services) occurred, the
5041 message was deferred as in a normal delivery, and thus remained on the
5042 spool, instead of being failed because of the mua_wrapper setting. This
5043 is now fixed, and I tidied up some of the mua_wrapper messages at the
5046 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5047 the messages in a hash of arrays rather than using individual hashes.
5048 This is a bit cleaner and results in dramatic memory savings, albeit
5049 at a slight CPU cost.
5051 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5052 as requested by Marc Sherman.
5054 SC/10 Eximstats: added histograms for user specified patterns as requested
5057 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5059 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5060 fopen() in the content-scanning modules that did not already have it.
5066 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5067 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5069 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5071 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5073 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5074 to test Sieve filters that use "vacation".
5076 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5077 that changes the way the GnuTLS parameters are stored in the cache file.
5078 The new format can be generated externally. For backward compatibility,
5079 if the data in the cache doesn't make sense, Exim assumes it has read an
5080 old-format file, and it generates new data and writes a new file. This
5081 means that you can't go back to an older release without removing the
5084 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5085 work if there are any delivery delays because "one_time" forces the
5086 parent to be marked "delivered", so its unseen clone is never tried
5087 again. For this reason, Exim now forbids the simultaneous setting of
5090 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5091 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5092 entry if you want to know the details. The fix, however, neglected to
5093 consider the case when local delivery batching is involved. The test for
5094 "previously delivered" was not happening when checking to see if an
5095 address could be batched with a previous (undelivered) one; under
5096 certain circumstances this could lead to multiple deliveries to the same
5099 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5100 in its include files, and this causes problems building Exim.
5102 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5103 header_syntax) but Exim was just ignoring anything given after a slash.
5104 In particular, this caused confusion with an attempt to use "verify =
5105 reverse_host_lookup/defer_ok". An error is now given when options are
5106 supplied for verify items that do not have them. (Maybe reverse_host_
5107 lookup should have a defer_ok option, but that's a different point.)
5109 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5110 defined by RFC 821) to 2048, because there were problems with some AUTH
5111 commands, and RFC 1869 says the size should be increased for extended
5112 SMTP commands that take arguments.
5114 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5117 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5118 "unknown" error; now it says that the functionality isn't in the binary.
5120 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5121 an address' error message when a string expansion fails (syntax or
5122 whatever). Otherwise the password may appear in the log. Following change
5123 PH/42 below, there is no longer a chance of it appearing in a bounce
5126 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5128 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5129 of its addresses was ever tried. (Bugzilla bug #2.)
5131 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5132 the result incorrectly in the debug output. (It correctly added a newline
5133 to what was transported.)
5135 TF/01 Added $received_time.
5137 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5138 commented out examples of how to interface to a virus scanner and to
5139 SpamAssassin. Also added commented examples of av_scanner and
5140 spamd_address settings.
5142 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5143 and controls are allowed in which ACLs. There were a couple of minor
5144 errors. Some of the entries in the conditions table (which is a table of
5145 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5146 negation of where the condition IS allowed.
5148 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5150 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5151 header file does not have a version number, so I've had to invent a new
5152 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5153 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5154 radiusclient), but it was contributed by a Radius user.
5156 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5157 files or directories, for OpenSSL.
5159 PH/19 When an Exim process that is running as root has to create an Exim log
5160 file, it does so in a subprocess that runs as exim:exim so as to get the
5161 ownership right at creation (otherwise, other Exim processes might see
5162 the file with the wrong ownership). There was no test for failure of this
5163 fork() call, which would lead to the process getting stuck as it waited
5164 for a non-existent subprocess. Forks do occasionally fail when resources
5165 run out. I reviewed all the other calls to fork(); they all seem to check
5168 PH/20 When checking for unexpected SMTP input at connect time (before writing
5169 the banner), Exim was not dealing correctly with a non-positive return
5170 from the read() function. If the client had disconnected by this time,
5171 the result was a log entry for a synchronization error with an empty
5172 string after "input=" when read() returned zero. If read() returned -1
5173 (an event I could not check), uninitialized data bytes were printed.
5174 There were reports of junk text (parts of files, etc) appearing after
5177 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5179 PH/22 Added support for macro redefinition, and (re)definition in between
5180 driver and ACL definitions.
5182 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5183 forgetting to use the resulting value; it was using the unexpanded value.
5185 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5186 hadn't been configured. The fix is from Juergen Kreileder, who
5187 understands it better than I do:
5189 "Here's what I see happening with three configured cyrus_sasl
5190 authenticators configured (plain, login, cram-md5):
5192 On startup auth_cyrus_sasl_init() gets called for each of these.
5193 This means three calls to sasl_listmech() without a specified mech_list.
5194 => SASL tests which mechs of all available mechs actually work
5195 => three warnings about OTP not working
5196 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5198 With the patch, sasl_listmech() also gets called three times. But now
5199 SASL's mech_list option is set to the server_mech specified in the the
5200 authenticator. Or in other words, the answer from sasl_listmech()
5201 gets limited to just the mech you're testing for (which is different
5203 => the return list contains just 'plain' or 'login', 'cram-md5' or
5204 nothing depending on the value of ob->server_mech.
5206 I've just tested the patch: Authentication still works fine,
5207 unavailable mechs specified in the exim configuration are still
5208 caught, and the auth.log warnings about OTP are gone."
5210 PH/25 When debugging is enabled, the contents of the command line are added
5211 to the debugging output, even when log_selector=+arguments is not
5214 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5215 answer is "GNU", and only if the return is "GNU/something" is the answer
5218 PH/27 $acl_verify_message is now set immediately after the failure of a
5219 verification in an ACL, and so is available in subsequent modifiers. In
5220 particular, the message can be preserved by coding like this:
5222 warn !verify = sender
5223 set acl_m0 = $acl_verify_message
5225 Previously, $acl_verify_message was set only while expanding "message"
5226 and "log_message" when a very denied access.
5228 PH/28 Modified OS/os.c-Linux with
5230 -#ifndef OS_LOAD_AVERAGE
5231 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5233 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5234 nomenclature these days.)
5236 PH/29 Installed patch from the Sieve maintainer that adds the options
5237 sieve_useraddress and sieve_subaddress to the redirect router.
5239 PH/30 In these circumstances:
5240 . Two addresses routed to the same list of hosts;
5241 . First host does not offer TLS;
5242 . First host accepts first address;
5243 . First host gives temporary error to second address;
5244 . Second host offers TLS and a TLS session is established;
5245 . Second host accepts second address.
5246 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5247 and peerdn, if requested) that were in fact used only for the second
5250 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5251 attention to any local part prefix or suffix that was matched by the
5252 router that accepted the address. It now behaves in the same way as it
5253 does for delivery: the affixes are removed from the local part unless
5254 rcpt_include_affixes is set on the transport.
5256 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5257 timeout during the DATA phase of an incoming message.
5259 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5260 applied a patch sanctioned by the Sieve maintainer.
5262 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5263 the uid or gid is negative. A case of a negative gid caused this to be
5264 noticed. The fix allows for either to be negative.
5266 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5267 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5270 PH/36 The change PH/12 above was broken. Fixed it.
5272 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5273 the grounds that routing the same address twice would always produce the
5274 same answer. This might have been true once, but it is certainly no
5275 longer true now. Routing a child address may depend on the previous
5276 routing that produced that child. Some complicated redirection strategies
5277 went wrong when messages had multiple recipients, and made Exim's
5278 behaviour dependent on the order in which the addresses were given.
5280 I have moved the duplicate checking until after the routing is complete.
5281 Exim scans the addresses that are assigned to local and remote
5282 transports, and removes any duplicates. This means that more work will be
5283 done, as duplicates will always all be routed, but duplicates are
5284 presumably rare, so I don't expect this is of any significance.
5286 For deliveries to pipes, files, and autoreplies, the duplicate checking
5287 still happens during the routing process, since they are not going to be
5290 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5291 It corrects a timeout issue with spamd. This is Ian's comment: "The
5292 background is that sometimes spamd either never reads data from a
5293 connection it has accepted, or it never writes response data. The exiscan
5294 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5295 blindly assumes that writes won't block so it may never time out."
5297 PH/39 Allow G after quota size as well as K and M.
5299 PH/40 The value set for $authenticated_id in an authenticator may not contain
5300 binary zeroes or newlines because the value is written to log lines and
5301 to spool files. There was no check on this. Now the value is run through
5302 the string_printing() function so that such characters are converted to
5303 printable escape sequences.
5305 PH/41 $message_linecount is a new variable that contains the total number of
5306 lines in the message. Compare $body_linecount, which is the count for the
5309 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5310 bounce and delay warning messages, except in certain special cases, which
5313 (a) An SMTP error message from a remote host;
5314 (b) A message specified in a :fail: redirection;
5315 (c) A message specified in a "fail" command in a system filter;
5316 (d) A message specified in a FAIL return from the queryprogram router;
5317 (e) A message specified by the cannot_route_message router option.
5319 In these cases only, Exim does include the error details in bounce and
5320 warning messages. There are also a few cases where bland messages such
5321 as "unrouteable address" or "local delivery error" are given.
5323 PH/43 $value is now also set for the "else" part of a ${run expansion.
5325 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5326 being worked on, but at least Exim now implements the latest version to
5329 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5330 process to complete was treated as a delivery failure, a timeout while
5331 writing the message to the pipe was logged, but erroneously treated as a
5332 successful delivery. Such timeouts include transport filter timeouts. For
5333 consistency with the overall process timeout, these timeouts are now
5334 treated as errors, giving rise to delivery failures by default. However,
5335 there is now a new Boolean option for the pipe transport called
5336 timeout_defer, which, if set TRUE, converts the failures into defers for
5337 both kinds of timeout. A transport filter timeout is now identified in
5340 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5341 systems where "make" and "gmake" are different, calling "gmake" at top
5342 level broke things. I've arranged for the value of $(MAKE) to be passed
5343 from the Makefile to this script so that it can call the same version of
5347 A note about Exim versions 4.44 and 4.50
5348 ----------------------------------------
5350 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5351 changes of various kinds. As a consequence, a big documentation update was
5352 needed. This delayed the release for rather longer than seemed good, especially
5353 in the light of a couple of (minor) security issues. Therefore, the changes
5354 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5355 release. So 4.44 and 4.50 are in effect two different branches that both start
5358 I have left the 4.50 change log unchanged; it contains all the changes since
5359 4.43. The change log for 4.44 is below; many of its items are identical to
5360 those for 4.50. This seems to be the most sensible way to preserve the
5361 historical information.
5367 1. Minor wording change to the doc/README.SIEVE file.
5369 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5370 computation of the current number of files was incorrect.
5372 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5373 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5376 4. Give more explanation in the error message when the command for a transport
5377 filter fails to execute.
5379 5. There are several places where Exim runs a non-Exim command in a
5380 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5381 was being done only for the command run by the queryprogram router. It is
5382 now done for all such subprocesses. The other cases are: ${run, transport
5383 filters, and the commands run by the lmtp and pipe transports.
5385 6. Added CONFIGURE_GROUP build-time option.
5387 7. Some older OS have a limit of 256 on the maximum number of file
5388 descriptors. Exim was using setrlimit() to set 1000 as a large value
5389 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5390 systems. I've change it so that if it can't get 1000, it tries for 256.
5392 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5393 was an oversight, and furthermore, ever since the addition of extra
5394 controls (e.g. 4.43/32), the checks on when to allow different forms of
5395 "control" were broken. There should now be diagnostics for all cases when a
5396 control that does not make sense is encountered.
5398 9. Added the /retain_sender option to "control=submission".
5400 10. $recipients is now available in the predata ACL (oversight).
5402 11. Tidy the search cache before the fork to do a delivery from a message
5403 received from the command line. Otherwise the child will trigger a lookup
5404 failure and thereby defer the delivery if it tries to use (for example) a
5405 cached ldap connection that the parent has called unbind on.
5407 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5408 of $address_data from the recipient verification was clobbered by the
5409 sender verification.
5411 13. The value of address_data from a sender verification is now available in
5412 $sender_address_data in subsequent conditions in the ACL statement.
5414 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5416 15. Added a new option "connect=<time>" to callout options, to set a different
5419 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5420 was its contents. (It was OK if the option was not defined at all.)
5422 17. A "Completed" log line is now written for messages that are removed from
5423 the spool by the -Mrm option.
5425 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5426 information about exactly what failed.
5428 19. Added -dd to debug only the daemon process.
5430 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5431 handles timeouts, both on the server side and network timeouts. Renamed the
5432 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5434 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5435 It is now set to "smtps".
5437 22. $host_address is now set to the target address during the checking of
5438 ignore_target_hosts.
5440 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5441 being passed; this would have caused $sender_host_name to have been used if
5442 matching the list had actually called for a host name (not very likely,
5443 since this list is usually IP addresses). A host name is now passed as
5446 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5447 code that specifies a non-restarting handler (typically sigaction() in
5448 modern systems) in an attempt to fix a rare and obscure crash bug.
5450 25. Narrowed the window for a race in the daemon that could cause it to ignore
5451 SIGCHLD signals. This is not a major problem, because they are used only to
5452 wake it up if nothing else does.
5454 26. A malformed maildirsize file could cause Exim to calculate negative values
5455 for the mailbox size or file count. Odd effects could occur as a result.
5456 The maildirsize information is now recalculated if the size or filecount
5459 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5460 support for a long time. Removed HAVE_SYS_VFS_H.
5462 28. Installed the latest version of exipick from John Jetmore.
5464 29. In an address list, if the pattern was not a regular expression, an empty
5465 subject address (from a bounce message) matched only if the pattern was an
5466 empty string. Non-empty patterns were not even tested. This was the wrong
5467 because it is perfectly reasonable to use an empty address as part of a
5468 database query. An empty address is now tested by patterns that are
5469 lookups. However, all the other forms of pattern expect the subject to
5470 contain a local part and a domain, and therefore, for them, an empty
5471 address still always fails if the pattern is not itself empty.
5473 30. Exim went into a mad DNS loop when attempting to do a callout where the
5474 host was specified on an smtp transport, and looking it up yielded more
5475 than one IP address.
5477 31. Re-factored the code for checking spool and log partition space into a
5478 function that finds that data and another that does the check. The former
5479 is then used to implement four new variables: $spool_space, $log_space,
5480 $spool_inodes, and $log_inodes.
5482 32. The RFC2047 encoding function was originally intended for short strings
5483 such as real names; it was not keeping to the 75-character limit for
5484 encoded words that the RFC imposes. It now respects the limit, and
5485 generates multiple encoded words if necessary. To be on the safe side, I
5486 have increased the buffer size for the ${rfc2047: expansion operator from
5489 33. It is now permitted to omit both strings after an "if" condition; if the
5490 condition is true, the result is "true". As before, when the second string
5491 is omitted, a false condition yields an empty string. This makes it less
5492 cumbersome to write custom ACL and router conditions.
5494 34. Failure to deliver a bounce message always caused it to be frozen, even if
5495 there was an errors_to setting on the router. The errors_to setting is now
5498 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5499 canonical form (fully expanded) before being placed in
5500 $sender_host_address.
5502 36. The table in the code that translates DNS record types into text (T_A to
5503 "A" for instance) was missing entries for NS and CNAME. It is just possible
5504 that this could have caused confusion if both these types were looked up
5505 for the same domain, because the text type is used as part of Exim's
5506 per-process caching. But the chance of anyone hitting this buglet seems
5509 37. The dnsdb lookup has been extended in a number of ways.
5511 (1) There is a new type, "zns", which walks up the domain tree until it
5512 finds some nameserver records. It should be used with care.
5514 (2) There is a new type, "mxh", which is like "mx" except that it returns
5515 just the host names, not the priorities.
5517 (3) It is now possible to give a list of domains (or IP addresses) to be
5518 looked up. The behaviour when one of the lookups defers can be
5519 controlled by a keyword.
5521 (4) It is now possible to specify the separator character for use when
5522 multiple records are returned.
5524 38. The dnslists ACL condition has been extended: it is now possible to supply
5525 a list of IP addresses and/or domains to be looked up in a particular DNS
5528 39. Added log_selector=+queue_time_overall.
5530 40. When running the queue in the test harness, wait just a tad after forking a
5531 delivery process, to get repeatability of debugging output.
5533 41. Include certificate and key file names in error message when GnuTLS fails
5534 to set them up, because the GnuTLS error message doesn't include the name
5535 of the failing file when there is a problem reading it.
5537 42. Allow both -bf and -bF in the same test run.
5539 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5541 44. The "Exiscan patch" is now merged into the mainline Exim source.
5543 45. Sometimes the final signoff response after QUIT could fail to get
5544 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5545 < 0 before doing a fflush(). This bug looks as though it goes back to the
5546 introduction of TLS in release 3.20, but "sometimes" must have been rare
5547 because the tests only now provoked it.
5549 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5550 (this can affect the format of dates).
5552 47. exim_tidydb, when checking for the continued existence of a message for
5553 which it has found a message-specific retry record, was not finding
5554 messages that were in split spool directories. Consequently, it was
5555 deleting retry records that should have stayed in existence.
5557 48. Steve fixed some bugs in eximstats.
5559 49. The SPA authentication driver was not abandoning authentication and moving
5560 on to the next authenticator when an expansion was forced to fail,
5561 contradicting the general specification for all authenticators. Instead it
5562 was generating a temporary error. It now behaves as specified.
5564 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5565 (the order specifies the preference for clients). The order is now AES256,
5566 AES128, 3DES, ARCFOUR128.
5568 51. Small patch to Sieve code - explicitly set From: when generating an
5571 52. Exim crashed if a remote delivery caused a very long error message to be
5572 recorded - for instance if somebody sent an entire SpamAssassin report back
5573 as a large number of 550 error lines. This bug was coincidentally fixed by
5574 increasing the size of one of Exim's internal buffers (big_buffer) that
5575 happened as part of the Exiscan merge. However, to be on the safe side, I
5576 have made the code more robust (and fixed the comments that describe what
5579 53. Now that there can be additional text after "Completed" in log lines (if
5580 the queue_time_overall log selector is set), a one-byte patch to exigrep
5581 was needed to allow it to recognize "Completed" as not the last thing in
5584 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5585 patch that reportedly fixes this has been added. I am not expert enough to
5586 create a test for it. This is what the patch creator wrote:
5588 "I found a little strange behaviour of ldap code when working with
5589 Windows 2003 AD Domain, where users was placed in more than one
5590 Organization Units. When I tried to give exim partial DN, the exit code
5591 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5592 But simultaneously result of request was absolutely normal ldap result,
5593 so I produce this patch..."
5595 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5596 REFERENCE, so I have modified the code to exclude the patch when that macro
5599 55. Some experimental protocols are using DNS PTR records for new purposes. The
5600 keys for these records are domain names, not reversed IP addresses. The
5601 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5602 leaves it alone. Component reversal etc. now happens only for IP addresses.
5605 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5607 57. Double the size of the debug message buffer (to 2048) so that more of very
5608 long debug lines gets shown.
5610 58. The exicyclog utility now does better if the number of log files to keep
5611 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5613 59. Two changes related to the smtp_active_hostname option:
5615 (1) $smtp_active_hostname is now available as a variable.
5616 (2) The default for smtp_banner uses $smtp_active_hostname instead
5617 of $primary_hostname.
5619 60. The host_aton() function is supposed to be passed a string that is known
5620 to be a valid IP address. However, in the case of IPv6 addresses, it was
5621 not checking this. This is a hostage to fortune. Exim now panics and dies
5622 if the condition is not met. A case was found where this could be provoked
5623 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5624 components; fortuitously, this particular loophole had already been fixed
5625 by change 4.50/55 above.
5627 If there are any other similar loopholes, the new check in host_aton()
5628 itself should stop them being exploited. The report I received stated that
5629 data on the command line could provoke the exploit when Exim was running as
5630 exim, but did not say which command line option was involved. All I could
5631 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5632 running as the user.
5635 61. There was a buffer overflow vulnerability in the SPA authentication code
5636 (which came originally from the Samba project). I have added a test to the
5637 spa_base64_to_bits() function which I hope fixes it.
5640 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5641 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5643 63. The daemon start-up calls getloadavg() while still root for those OS that
5644 need the first call to be done as root, but it missed one case: when
5645 deliver_queue_load_max is set with deliver_drop_privilege. This is
5646 necessary for the benefit of the queue runner, because there is no re-exec
5647 when deliver_drop_privilege is set.
5649 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5650 This has been fixed.
5652 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5653 was in use, was not putting the data itself into the right store pool;
5654 consequently, it could be overwritten for a subsequent message in the same
5655 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5658 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5660 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5661 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5662 the function were treating the return as a boolean value, which happened to
5663 work because 0=false and not-0=true, but is not correct code.
5665 68. The host_aton() function was not handling scoped IPv6 addresses (those
5666 with, for example, "%eth0" on the end) correctly.
5668 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5669 negated items (that is, ~something) in unsigned ints. Some compilers
5670 apparently mutter when there is no cast.
5672 70. If an address verification called from an ACL failed, and did not produce a
5673 user-specific message (i.e. there was only a "system" message), nothing was
5674 put in $acl_verify_message. In this situation, it now puts the system
5677 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5678 session; change 4.31/43 added the unwanted input to the log line - except
5679 that it did not do this in the start of session case. It now does.
5681 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5682 This is wrong and can cause the other end to generate a synchronization
5683 error if it is another Exim or anything else that does the synchronization
5684 check. A QUIT command is no longer sent after a timeout.
5686 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5687 during host lookups.
5689 74. The defer_ok option of callout verification was not working if it was used
5690 when verifying addresses in header lines, that is, for this case:
5692 verify = header_sender/callout=defer_ok
5694 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5695 those file descriptors could be used for SMTP connections. If anything
5696 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5697 could be sent to the SMTP client, causing chaos. The daemon now opens
5698 stdin, stdout, and stderr to /dev/null when it puts itself into the
5701 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5702 log by default. The user can override this with suitable Perl magic.
5704 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5705 the log message when discard triggers, was not working for the DATA ACL or
5706 for the non-SMTP ACL.
5708 78. Error message wording change in sieve.c.
5710 79. If smtp_accept_max_per_host was set, the number of connections could be
5711 restricted to fewer than expected, because the daemon was trying to set up
5712 a new connection before checking whether the processes handling previous
5713 connections had finished. The check for completed processes is now done
5714 earlier. On busy systems, this bug wouldn't be noticed because something
5715 else would have woken the daemon, and it would have reaped the completed
5718 80. If a message was submitted locally by a user whose login name contained one
5719 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5720 It caused a spool format error. I have fixed the spool reading code. A
5721 related problem was that the "from" clause in the Received: line became
5722 illegal because of the space(s). It is now covered by ${quote_local_part.
5724 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5727 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5729 83. Patch from Sieve maintainer to fix unterminated string problem in
5730 "vacation" handling.
5732 84. Some minor changes to the Linux configuration files to help with other
5733 OS variants using glibc.
5735 85. One more patch for Sieve to update vacation handling to latest spec.
5738 ----------------------------------------------------
5739 See the note above about the 4.44 and 4.50 releases.
5740 ----------------------------------------------------
5746 1. Change 4.43/35 introduced a bug that caused file counts to be
5747 incorrectly computed when quota_filecount was set in an appendfile
5750 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5751 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5754 3. Give more explanation in the error message when the command for a transport
5755 filter fails to execute.
5757 4. There are several places where Exim runs a non-Exim command in a
5758 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5759 was being done only for the command run by the queryprogram router. It is
5760 now done for all such subprocesses. The other cases are: ${run, transport
5761 filters, and the commands run by the lmtp and pipe transports.
5763 5. Some older OS have a limit of 256 on the maximum number of file
5764 descriptors. Exim was using setrlimit() to set 1000 as a large value
5765 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5766 systems. I've change it so that if it can't get 1000, it tries for 256.
5768 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5769 was an oversight, and furthermore, ever since the addition of extra
5770 controls (e.g. 4.43/32), the checks on when to allow different forms of
5771 "control" were broken. There should now be diagnostics for all cases when a
5772 control that does not make sense is encountered.
5774 7. $recipients is now available in the predata ACL (oversight).
5776 8. Tidy the search cache before the fork to do a delivery from a message
5777 received from the command line. Otherwise the child will trigger a lookup
5778 failure and thereby defer the delivery if it tries to use (for example) a
5779 cached ldap connection that the parent has called unbind on.
5781 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5782 of $address_data from the recipient verification was clobbered by the
5783 sender verification.
5785 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5786 was its contents. (It was OK if the option was not defined at all.)
5788 11. A "Completed" log line is now written for messages that are removed from
5789 the spool by the -Mrm option.
5791 12. $host_address is now set to the target address during the checking of
5792 ignore_target_hosts.
5794 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5795 being passed; this would have caused $sender_host_name to have been used if
5796 matching the list had actually called for a host name (not very likely,
5797 since this list is usually IP addresses). A host name is now passed as
5800 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5801 code that specifies a non-restarting handler (typically sigaction() in
5802 modern systems) in an attempt to fix a rare and obscure crash bug.
5804 15. Narrowed the window for a race in the daemon that could cause it to ignore
5805 SIGCHLD signals. This is not a major problem, because they are used only to
5806 wake it up if nothing else does.
5808 16. A malformed maildirsize file could cause Exim to calculate negative values
5809 for the mailbox size or file count. Odd effects could occur as a result.
5810 The maildirsize information is now recalculated if the size or filecount
5813 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5814 support for a long time. Removed HAVE_SYS_VFS_H.
5816 18. Updated exipick to current release from John Jetmore.
5818 19. Allow an empty sender to be matched against a lookup in an address list.
5819 Previously the only cases considered were a regular expression, or an
5822 20. Exim went into a mad DNS lookup loop when doing a callout where the
5823 host was specified on the transport, if the DNS lookup yielded more than
5826 21. The RFC2047 encoding function was originally intended for short strings
5827 such as real names; it was not keeping to the 75-character limit for
5828 encoded words that the RFC imposes. It now respects the limit, and
5829 generates multiple encoded words if necessary. To be on the safe side, I
5830 have increased the buffer size for the ${rfc2047: expansion operator from
5833 22. Failure to deliver a bounce message always caused it to be frozen, even if
5834 there was an errors_to setting on the router. The errors_to setting is now
5837 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5838 canonical form (fully expanded) before being placed in
5839 $sender_host_address.
5841 24. Updated eximstats to version 1.33
5843 25. Include certificate and key file names in error message when GnuTLS fails
5844 to set them up, because the GnuTLS error message doesn't include the name
5845 of the failing file when there is a problem reading it.
5847 26. Expand error message when OpenSSL has problems setting up cert/key files.
5850 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5851 (this can affect the format of dates).
5853 28. exim_tidydb, when checking for the continued existence of a message for
5854 which it has found a message-specific retry record, was not finding
5855 messages that were in split spool directories. Consequently, it was
5856 deleting retry records that should have stayed in existence.
5858 29. eximstats updated to version 1.35
5859 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5860 1.35 - bugfix such that pie charts by volume are generated correctly
5862 30. The SPA authentication driver was not abandoning authentication and moving
5863 on to the next authenticator when an expansion was forced to fail,
5864 contradicting the general specification for all authenticators. Instead it
5865 was generating a temporary error. It now behaves as specified.
5867 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5868 (the order specifies the preference for clients). The order is now AES256,
5869 AES128, 3DES, ARCFOUR128.
5871 31. Small patch to Sieve code - explicitly set From: when generating an
5874 32. Exim crashed if a remote delivery caused a very long error message to be
5875 recorded - for instance if somebody sent an entire SpamAssassin report back
5876 as a large number of 550 error lines. This bug was coincidentally fixed by
5877 increasing the size of one of Exim's internal buffers (big_buffer) that
5878 happened as part of the Exiscan merge. However, to be on the safe side, I
5879 have made the code more robust (and fixed the comments that describe what
5882 33. Some experimental protocols are using DNS PTR records for new purposes. The
5883 keys for these records are domain names, not reversed IP addresses. The
5884 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5885 leaves it alone. Component reversal etc. now happens only for IP addresses.
5888 34. The host_aton() function is supposed to be passed a string that is known
5889 to be a valid IP address. However, in the case of IPv6 addresses, it was
5890 not checking this. This is a hostage to fortune. Exim now panics and dies
5891 if the condition is not met. A case was found where this could be provoked
5892 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5893 components; fortuitously, this particular loophole had already been fixed
5894 by change 4.50/55 or 4.44/33 above.
5896 If there are any other similar loopholes, the new check in host_aton()
5897 itself should stop them being exploited. The report I received stated that
5898 data on the command line could provoke the exploit when Exim was running as
5899 exim, but did not say which command line option was involved. All I could
5900 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5901 running as the user.
5904 35. There was a buffer overflow vulnerability in the SPA authentication code
5905 (which came originally from the Samba project). I have added a test to the
5906 spa_base64_to_bits() function which I hope fixes it.
5909 36. The daemon start-up calls getloadavg() while still root for those OS that
5910 need the first call to be done as root, but it missed one case: when
5911 deliver_queue_load_max is set with deliver_drop_privilege. This is
5912 necessary for the benefit of the queue runner, because there is no re-exec
5913 when deliver_drop_privilege is set.
5915 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5916 was in use, was not putting the data itself into the right store pool;
5917 consequently, it could be overwritten for a subsequent message in the same
5918 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5921 38. Sometimes the final signoff response after QUIT could fail to get
5922 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5923 < 0 before doing a fflush(). This bug looks as though it goes back to the
5924 introduction of TLS in release 3.20, but "sometimes" must have been rare
5925 because the tests only now provoked it.
5931 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5932 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5933 Now it returns an integer. A number of calls were still expecting a T/F
5934 return. Fortuitously, in all cases, the tests worked in OK situations,
5935 which is the norm. However, things would have gone wrong on any write
5936 failures on the smtp file descriptor. This function is used when sending
5937 messages over SMTP and also when doing verify callouts.
5939 2. When Exim is called to do synchronous delivery of a locally submitted
5940 message (the -odf or -odi options), it no longer closes stderr before doing
5943 3. Implemented the mua_wrapper option.
5945 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5947 5. Implemented the functions header_remove(), header_testname(),
5948 header_add_at_position(), and receive_remove_recipient(), and exported them
5951 6. If an ACL "warn" statement specified the addition of headers, Exim already
5952 inserted X-ACL-Warn: at the start if there was no header name. However, it
5953 was not making this test for the second and subsequent header lines if
5954 there were newlines in the string. This meant that an invalid header could
5955 be inserted if Exim was badly configured.
5957 7. Allow an ACL "warn" statement to add header lines at the start or after all
5958 the Received: headers, as well as at the end.
5960 8. Added the rcpt_4xx retry error code.
5962 9. Added postmaster_mailfrom=xxx to callout verification option.
5964 10. Added mailfrom=xxxx to the callout verification option, for verify=
5967 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5968 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5970 12. Inserted some casts to stop certain compilers warning when using pointer
5971 differences as field lengths or precisions in printf-type calls (mostly
5972 affecting debugging statements).
5974 13. Added optional readline() support for -be (dynamically loaded).
5976 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5977 same clock tick as a message's arrival, so that its received time was the
5978 same as the "first fail" time on the retry record, and that message
5979 remained on the queue past the ultimate address timeout, every queue runner
5980 would try a delivery (because it was past the ultimate address timeout) but
5981 after another failure, the ultimate address timeout, which should have then
5982 bounced the address, did not kick in. This was a "< instead of <=" error;
5983 in most cases the first failure would have been in the next clock tick
5984 after the received time, and all would be well.
5986 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5987 being recognized when the domain list was tested by the match_domain
5988 condition in an expansion string.
5990 16. Added the ${str2b64: operator.
5992 17. Exim was always calling setrlimit() to set a large limit for the number of
5993 processes, without checking whether the existing limit was already
5994 adequate. (It did check for the limit on file descriptors.) Furthermore,
5995 errors from getrlimit() and setrlimit() were being ignored. Now they are
5996 logged to the main and panic logs, but Exim does carry on, to try to do its
5997 job under whatever limits there are.
5999 18. Imported PCRE 5.0.
6001 19. Trivial typo in log message " temporarily refused connection" (the leading
6004 20. If the log selector return_path_on_delivery was set and an address was
6005 redirected to /dev/null, the delivery process crashed because it assumed
6006 that a return path would always be set for a "successful" delivery. In this
6007 case, the whole delivery is bypassed as an optimization, and therefore no
6010 21. Internal re-arrangement: the function for sending a challenge and reading
6011 a response while authentication was assuming a zero-terminated challenge
6012 string. It's now changed to take a pointer and a length, to allow for
6013 binary data in such strings.
6015 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6017 23. Exim was not respecting finduser_retries when seeking the login of the
6018 uid under which it was called; it was always trying 10 times. (The default
6019 setting of finduser_retries is zero.) Also, it was sleeping after the final
6020 failure, which is pointless.
6022 24. Implemented tls_on_connect_ports.
6024 25. Implemented acl_smtp_predata.
6026 26. If the domain in control=submission is set empty, Exim assumes that the
6027 authenticated id is a complete email address when it generates From: or
6028 Sender: header lines.
6030 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6031 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6032 chown and chgrp in /bin and hostname in /usr/bin.
6034 28. Exim was keeping the "process log" file open after each use, just as it
6035 does for the main log. This opens the possibility of it remaining open for
6036 long periods when the USR1 signal hits a daemon. Occasional processlog
6037 errors were reported, that could have been caused by this. Anyway, it seems
6038 much more sensible not to leave this file open at all, so that is what now
6041 29. The long-running daemon process does not normally write to the log once it
6042 has entered its main loop, and it closes the log before doing so. This is
6043 so that log files can straightforwardly be renamed and moved. However,
6044 there are a couple of unusual error situations where the daemon does write
6045 log entries, and I had neglected to close the log afterwards.
6047 30. The text of an SMTP error response that was received during a remote
6048 delivery was being truncated at 512 bytes. This is too short for some of
6049 the long messages that one sometimes sees. I've increased the limit to
6052 31. It is now possible to make retry rules that apply only when a message has a
6053 specific sender, in particular, an empty sender.
6055 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6056 it possible to be selective about when SMTP synchronization is enforced.
6058 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6060 32. Implemented hosts_connection_nolog.
6062 33. Added an ACL for QUIT.
6064 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6067 35. Added mailbox_size and mailbox_filecount to appendfile.
6069 36. Added control = no_multiline_responses to ACLs.
6071 37. There was a bug in the logic of the code that waits for the clock to tick
6072 in the case where the clock went backwards by a substantial amount such
6073 that the microsecond fraction of "now" was more than the microsecond
6074 fraction of "then" (but the whole seconds number was less).
6076 38. Added support for the libradius Radius client library this is found on
6077 FreeBSD (previously only the radiusclient library was supported).
6083 1. When certain lookups returned multiple values in the form name=value, the
6084 quoting of the values was not always being done properly. Specifically:
6085 (a) If the value started with a double quote, but contained no whitespace,
6087 (b) If the value contained whitespace other than a space character (i.e.
6088 tabs or newlines or carriage returns) it was not quoted.
6089 This fix has been applied to the mysql and pgsql lookups by writing a
6090 separate quoting function and calling it from the lookup code. The fix
6091 should probably also be applied to nisplus, ibase and oracle lookups, but
6092 since I cannot test any of those, I have not disturbed their existing code.
6094 2. A hit in the callout cache for a specific address caused a log line with no
6095 reason for rejecting RCPT. Now it says "Previous (cached) callout
6096 verification failure".
6098 3. There was an off-by-one bug in the queryprogram router. An over-long
6099 return line was truncated at 256 instead of 255 characters, thereby
6100 overflowing its buffer with the terminating zero. As well as fixing this, I
6101 have increased the buffer size to 1024 (and made a note to document this).
6103 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6104 when Exim is waiting for an SMTP response from a remote server, Exim
6105 restarts its select() call on the socket, thereby resetting its timeout.
6106 This is not a problem when such interrupts are rare. Somebody set up a cron
6107 job to run exiwhat every 2 minutes, which is less than the normal select()
6108 timeout (5 or 10 minutes). This meant that the select() timeout never
6109 kicked in because it was always reset. I have fixed this by comparing the
6110 time when an interrupt arrives with the time at the start of the first call
6111 to select(). If more time than the timeout has elapsed, the interrupt is
6112 treated as a timeout.
6114 5. Some internal re-factoring in preparation for the addition of Sieve
6115 extensions (by MH). In particular, the "personal" test is moved to a
6116 separate function, and given an option for scanning Cc: and Bcc: (which is
6117 not set for Exim filters).
6119 6. When Exim created an email address using the login of the caller as the
6120 local part (e.g. when creating a From: or Sender: header line), it was not
6121 quoting the local part when it contained special characters such as @.
6123 7. Installed new OpenBSD configuration files.
6125 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6126 try to make them clearer.
6128 9. Callout options, other than the timeout value, were being ignored when
6129 verifying sender addresses in header lines. For example, when using
6131 verify = header_sender/callout=no_cache
6133 the cache was (incorrectly) being used.
6135 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6136 only the Cygwin environment.
6138 11. When return_path_on_delivery was set as a log selector, if different remote
6139 addresses in the same message used different return paths and parallel
6140 remote delivery occurred, the wrong values would sometimes be logged.
6141 (Whenever a remote delivery process finished, the return path value from
6142 the most recently started remote delivery process was logged.)
6144 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6145 lines when AUTH and/or TLS are in use. This is the "received protocol"
6146 field. Exim used to use "asmtp" for authenticated SMTP, without any
6147 indication (in the protocol name) for TLS use. Now it follows the RFC and
6148 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6149 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6150 names appear in log lines as well as in Received: header lines.
6152 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6153 extensions, and comparison tests, and to fix some bugs.
6155 14. Changes to the "personal" filter test:
6157 (1) The test was buggy in that it was just doing the equivalent of
6158 "contains" tests on header lines. For example, if a user's address was
6159 anne@some.where, the "personal" test would incorrectly be true for
6161 To: susanne@some.where
6163 This test is now done by extracting each address from the header in turn,
6164 and checking the entire address. Other tests that are part of "personal"
6165 are now done using regular expressions (for example, to check local parts
6166 of addresses in From: header lines).
6168 (2) The list of non-personal local parts in From: addresses has been
6169 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6170 taken from the Sieve specification recommendations.
6172 (3) If the message contains any header line starting with "List-" it is
6173 treated as non-personal.
6175 (4) The test for "circular" in the Subject: header line has been removed
6176 because it now seems ill-conceived.
6178 15. Minor typos in src/EDITME comments corrected.
6180 16. Installed latest exipick from John Jetmore.
6182 17. If headers_add on a router specified a text string that was too long for
6183 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6184 of string_sprintf() is now avoided.
6186 18. $message_body_size was not set (it was always zero) when running the DATA
6187 ACL and the local_scan() function.
6189 19. For the "mail" command in an Exim filter, no default was being set for
6190 the once_repeat time, causing a random time value to be used if "once" was
6191 specified. (If the value happened to be <= 0, no repeat happened.) The
6192 default is now 0s, meaning "never repeat". The "vacation" command was OK
6193 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6194 (I found it when inspecting the code).
6196 20. There is now an overall timeout for performing a callout verification. It
6197 defaults to 4 times the callout timeout, which applies to individual SMTP
6198 commands during the callout. The overall timeout applies when there is more
6199 than one host that can be tried. The timeout is checked before trying the
6200 next host. This prevents very long delays if there are a large number of
6201 hosts and all are timing out (e.g. when the network connections are timing
6202 out). The value of the overall timeout can be changed by specifying an
6203 additional sub-option for "callout", called "maxwait". For example:
6205 verify = sender/callout=5s,maxwait=20s
6207 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6208 to the end before writing, but this should make it even safer).
6210 22. Exim was forgetting that it had advertised PIPELINING for the second and
6211 subsequent messages on an SMTP connection. It was also not resetting its
6212 memory on STARTTLS and an internal HELO.
6214 23. When Exim logs an SMTP synchronization error within a session, it now
6215 records whether PIPELINING has been advertised or not.
6217 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6218 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6219 rather than long int.
6221 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6223 26. Added the never_mail option to autoreply.
6229 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6230 crash if the getsockname() call failed; this can happen if a connection is
6231 closed very soon after it is established. The problem was simply in the
6232 order in which certain operations were done, causing Exim to try to write
6233 to the SMTP stream before it had set up the file descriptor. The bug has
6234 been fixed by making things happen in the correct order.
6240 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6241 before the connection was closed, thus losing the rejection response.
6243 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6244 some early Solaris releases, but causes trouble in current releases where
6245 socklen_t is defined.
6247 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6250 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6253 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6254 that releasing the top of it at the end releases what was used for sub-
6255 expansions (unless the block got too big). However, discard this block if
6256 the first thing is a variable or header, so that we can use its block when
6257 it is dynamic (useful for very large $message_headers, for example).
6259 6. Lookups now cache *every* query, not just the most recent. A new, separate
6260 store pool is used for this. It can be recovered when all lookup caches are
6261 flushed. Lookups now release memory at the end of their result strings.
6262 This has involved some general refactoring of the lookup sources.
6264 7. Some code has been added to the store_xxx() functions to reduce the amount
6265 of flapping under certain conditions.
6267 8. log_incoming_interface used to affect only the <= reception log lines. Now
6268 it causes the local interface and port to be added to several more SMTP log
6269 lines, for example "SMTP connection from", and rejection lines.
6271 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6273 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6275 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6276 could be overwritten at the end of the current message (or the start of a
6277 new message if it was set in a HELO ACL). The value is now preserved for
6278 the duration of the SMTP connection.
6280 12. If a transport had a headers_rewrite setting, and a matching header line
6281 contained an unqualified address, that address was qualified, even if it
6282 did not match any rewriting rules. The underlying bug was that the values
6283 of the flags that permit the existence of unqualified sender and recipient
6284 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6285 non-local messages, and by -bnq for local messages) were not being
6286 preserved with the message after it was received.
6288 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6289 "next input=" as part of the text comprising the host identity instead of
6290 the correct text. The code was using the same buffer for two different
6291 strings. However, depending on which order the printing function evaluated
6292 its arguments, the bug did not always show up. Under Linux, for example, my
6293 test suite worked just fine.
6295 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6296 doesn't work with some older versions of Perl. It has been changed to "my",
6297 which in any case is probably the better facility to use.
6299 15. A really picky compiler found some instances of statements for creating
6300 error messages that either had too many or two few arguments for the format
6303 16. The size of the buffer for calls to the DNS resolver has been increased
6304 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6305 for addresses that have a lot of PTR records. This alleviates a problem; it
6306 does not fully solve it.
6308 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6309 buffer now truncates the list and logs the incident, which is the same
6310 action as happens when Exim is looking up a host name and its aliases.
6311 Previously in this situation something unpredictable would happen;
6312 sometimes it was "internal error: store_reset failed".
6314 18. If a server dropped the connection unexpectedly when an Exim client was
6315 using GnuTLS and trying to read a response, the client delivery process
6316 crashed while trying to generate an error log message.
6318 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6319 string, for example:
6321 warn message = H1: something\nH2: something
6323 the text was added as a single header line from Exim's point of view
6324 though it ended up OK in the delivered message. However, searching for the
6325 second and subsequent header lines using $h_h2: did not work. This has been
6326 fixed. Similarly, if a system filter added multiple headers in this way,
6327 the routers could not see them.
6329 20. Expanded the error message when iplsearch is called with an invalid key to
6330 suggest using net-iplsearch in a host list.
6332 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6333 ACLs are no longer actually imposed (and a message to that effect is
6336 22. If a "gecos" field in a passwd entry contained escaped characters, in
6337 particular, if it contained a \" sequence, Exim got it wrong when building
6338 a From: or a Sender: header from that name. A second bug also caused
6339 incorrect handling when an unquoted " was present following a character
6340 that needed quoting.
6342 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6343 was not being matched caselessly.
6345 24. Arranged for all hyphens in the exim.8 source to be escaped with
6348 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6349 properly. Recipient callout cache records were still being keyed to include
6350 the sender, even when use_sender was set false. This led to far more
6351 callouts that were necessary. The sender is no longer included in the key
6352 when use_sender is false.
6354 26. Added "control = submission" modifier to ACLs.
6356 27. Added the ${base62d: operator to decode base 62 numbers.
6358 28. dnsdb lookups can now access SRV records.
6360 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6361 the configuration file.
6363 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6364 (-v) mode. This makes the output for a verbose queue run more intelligible.
6366 31. Added a use_postmaster feature to recipient callouts.
6368 32. Added the $body_zerocount variable, containing the number of binary zero
6369 bytes in the message body.
6371 33. The time of last modification of the "new" subdirectory is now used as the
6372 "mailbox time last read" when there is a quota error for a maildir
6375 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6377 35. Added +ignore_unknown as a special item in host lists.
6379 36. Code for decoding IPv6 addresses in host lists is now included, even if
6380 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6381 address was recognized as an IP address, but was then not correctly decoded
6382 into binary, causing unexpected and incorrect effects when compared with
6389 1. Very minor rewording of debugging text in manualroute to say "list of
6390 hosts" instead of "hostlist".
6392 2. If verify=header_syntax was set, and a header line with an unqualified
6393 address (no domain) and a large number of spaces between the end of the
6394 name and the colon was received, the reception process suffered a buffer
6395 overflow, and (when I tested it) crashed. This was caused by some obsolete
6396 code that should have been removed. The fix is to remove it!
6398 3. When running in the test harness, delay a bit after writing a bounce
6399 message to get a bit more predictability in the log output.
6401 4. Added a call to search_tidyup() just before forking a reception process. In
6402 theory, someone could use a lookup in the expansion of smtp_accept_max_
6403 per_host which, without the tidyup, could leave open a database connection.
6405 5. Added the variables $recipient_data and $sender_data which get set from a
6406 lookup success in an ACL "recipients" or "senders" condition, or a router
6407 "senders" option, similar to $domain_data and $local_part_data.
6409 6. Moved the writing of debug_print from before to after the "senders" test
6412 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6413 problems for message scanning, either using a data ACL, or using
6414 local_scan() because the Received: header was not generated till after they
6415 were called (in order to set the time as the time of reception completion).
6416 I have revised the way this works. The header is now generated after the
6417 body is received, but before the ACL or local_scan() are called. After they
6418 are run, the timestamp in the header is updated.
6424 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6425 before starting a queue runner without re-exec. This happened only when
6426 deliver_drop_privilege was set or when the Exim user was set to root. The
6427 effect of the bug was that timeouts during subsequent deliveries caused
6428 crashes instead of being properly handled. The handler is now left at its
6429 default (and expected) setting.
6431 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6432 message, again when deliver_drop_privilege is set or Exim is run as root.
6433 The bug described in (1) was not present in this case, but the tidying up
6434 of the other signals was missing. I have made the two cases consistent.
6436 3. The ignore_target_hosts setting on a manualroute router was being ignored
6437 for hosts that were looked up using the /MX notation.
6439 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6442 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6443 operated on the sender address. After changing the $sender_address to <>
6444 for the sender address verify, Exim was re-instated it as the original
6445 (before rewriting) address, but remembering that it had rewritten it, so it
6446 wasn't rewriting it again. This bug also had the effect of breaking the
6447 sender address verification caching when the sender address was rewritten.
6449 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6450 This has been changed so that if the ip literal address matches
6451 ignore_target_hosts, the router declines.
6453 7. Added expansion conditions match_domain, match_address, and match_local_
6454 part (NOT match_host).
6456 8. The placeholder for the Received: header didn't have a length field set.
6458 9. Added code to Exim itself and to exim_lock to test for a specific race
6459 condition that could lead to file corruption when using MBX delivery. The
6460 issue is with the lockfile that is created in /tmp. If this file is removed
6461 after a process has opened it but before that process has acquired a lock,
6462 there is the potential for a second process to recreate the file and also
6463 acquire a lock. This could lead to two Exim processes writing to the file
6464 at the same time. The added code performs the same test as UW imapd; it
6465 checks after acquiring the lock that its file descriptor still refers to
6466 the same named file.
6468 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6469 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6472 11. Added the smtp_active_hostname option. If used, this will typically be made
6473 to depend on the incoming interface address. Because $interface_address is
6474 not set up until the daemon has forked a reception process, error responses
6475 that can happen earlier (such as "too many connections") no longer contain
6478 12. If an expansion in a condition on a "warn" statement fails because a lookup
6479 defers, the "warn" statement is abandoned, and the next ACL statement is
6480 processed. Previously this caused the whole ACL to be aborted.
6482 13. Added the iplsearch lookup type.
6484 14. Added ident_timeout as a log selector.
6486 15. Added tls_certificate_verified as a log selector.
6488 16. Added a global option tls_require_ciphers (compare the smtp transport
6489 option of the same name). This controls incoming TLS connections.
6491 17. I finally figured out how to make tls_require_ciphers do a similar thing
6492 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6493 before starting the TLS session.
6495 18. Tabs are now shown as \t in -bP output.
6497 19. If the log selector return_path_on_delivery was set, Exim crashed when
6498 bouncing a message because it had too many Received: header lines.
6500 20. If two routers both had headers_remove settings, and the first one included
6501 a superfluous trailing colon, the final name in the first list and the
6502 first name in the second list were incorrectly joined into one item (with a
6503 colon in the middle).
6509 1. Added -C and -D options to the exinext utility, mainly to make it easier
6510 to include in the automated testing, but these could be helpful when
6511 multiple configurations are in use.
6513 2. The exinext utility was not formatting the output nicely when there was
6514 an alternate port involved in the retry record key, nor when there was a
6515 message id as well (for retries that were specific to a specific message
6516 and a specific host). It was also confused by IPv6 addresses, because of
6517 the additional colons they contain. I have fixed the IPv4 problem, and
6518 patched it up to do a reasonable job for IPv6.
6520 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6521 delivery, the log line now contains "pipelined" if PIPELINING was used.
6523 4. An SMTP transport process used to panic and die if the bind() call to set
6524 an explicit outgoing interface failed. This has been changed; it is now
6525 treated in the same way as a connect() failure.
6527 5. A reference to $sender_host_name in the part of a conditional expansion
6528 that was being skipped was still causing a DNS lookup. This no longer
6531 6. The def: expansion condition was not recognizing references to header lines
6532 that used bh_ and bheader_.
6534 7. Added the _cache feature to named lists.
6536 8. The code for checking quota_filecount in the appendfile transport was
6537 allowing one more file than it should have been.
6539 9. For compatibility with Sendmail, the command line option
6547 and sets the incoming protocol and host name (for trusted callers). The
6548 host name and its colon can be omitted when only the protocol is to be set.
6549 Note the Exim already has two private options, -pd and -ps, that refer to
6550 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6551 "s", but I don't think that's a major issue.
6553 10. A number of refactoring changes to the code, none of which should affect
6556 (a) The number of logging options was getting close to filling up the
6557 32-bit word that was used as a bit map. I have split them into two classes:
6558 those that are passed in the argument to log_write(), and those that are
6559 only ever tested independently outside of that function. These are now in
6560 separate 32-bit words, so there is plenty of room for expansion again.
6561 There is no change in the user interface or the logging behaviour.
6563 (b) When building, for example, log lines, the code previously used a
6564 macro that called string_cat() twice, in order to add two strings. This is
6565 not really sufficiently general. Furthermore, there was one instance where
6566 it was actually wrong because one of the argument was used twice, and in
6567 one call a function was used. (As it happened, calling the function twice
6568 did not affect the overall behaviour.) The macro has been replaced by a
6569 function that can join an arbitrary number of extra strings onto a growing
6572 (c) The code for expansion conditions now uses a table and a binary chop
6573 instead of a serial search (which was left over from when there were very
6574 few conditions). Also, it now recognizes conditions like "pam" even when
6575 the relevant support is not compiled in: a suitably worded error message is
6576 given if an attempt is made to use such a condition.
6578 11. Added ${time_interval:xxxxx}.
6580 12. A bug was causing one of the ddress fields not to be passed back correctly
6581 from remote delivery subprocesses. The field in question was not being
6582 subsequently used, so this caused to problems in practice.
6584 13. Added new log selectors queue_time and deliver_time.
6586 14. Might have fixed a bug in maildirsizefile handling that threw up
6587 "unexpected character" debug warnings, and recalculated the data
6588 unnecessarily. In any case, I expanded the warning message to give more
6591 15. Added the message "Restricted characters in address" to the statements in
6592 the default ACL that block characters like @ and % in local parts.
6594 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6595 Three changes have been made:
6597 (a) There was a serious bug; a negative response to MAIL caused the whole
6598 recipient domain to be cached as invalid, thereby blocking all messages
6599 to all local parts at the same domain, from all senders. This bug has
6600 been fixed. The domain is no longer cached after a negative response to
6601 MAIL if the sender used is not empty.
6603 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6606 (c) A new callout option, "use_sender" has been added for people who want
6607 the modified behaviour.
6613 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6616 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6617 indeed breaks things for older releases.
6619 3. Added additional logging to the case where there is a problem reading data
6620 from a filter that is running in a subprocess using a pipe, in order to
6621 try to track down a specific problem.
6623 4. Testing facility fudge: when running in the test harness and attempting
6624 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6625 getting "No route to host". Convert this to a timeout.
6627 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6630 6. Some OS don't have socklen_t but use size_t instead. This affects the
6631 fifth argument of getsockopt() amongst other things. This is now
6632 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6633 can be set for individual OS. I have set it for SunOS5, OSF1, and
6634 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6635 some earlier ones do not.
6637 7. Change 4.30/15 was not doing the test caselessly.
6639 8. The standard form for an IPv6 address literal was being rejected by address
6640 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6641 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6642 this, as well as the form without the "IPv6" on the front (but only when
6643 address literals are enabled, of course).
6645 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6647 10. Exim crashed if a message with an empty sender address specified by -f
6648 encountered a router with an errors_to setting. This could be provoked only
6649 by a command such as
6653 where an empty string was supplied; "<>" did not hit this bug.
6655 11. Installed PCRE release 4.5.
6657 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6658 remained set. It is now erased.
6660 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6661 times from message ids (which are base 36 rather than the normal 62).
6663 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6664 were being counted as actual protocol errors, and logged if the log
6665 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6666 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6667 and DATA following a set of rejected RCPTs do not count as protocol errors.
6668 In other words, Exim assumes they were pipelined, though this may not
6669 actually be the case. Of course, in all cases the client gets an
6670 appropriate error code.
6672 15. If a lookup fails in an ACL condition, a message about the failure may
6673 be available; it is used if testing the ACL cannot continue, because most
6674 such messages specify what the cause of the deferral is. However, some
6675 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6676 that caused an old message to be retained and used if a later statement
6677 caused a defer, replacing the real cause of the deferral.
6679 16. If an IP address had so many PTR records that the DNS lookup buffer
6680 was not large enough to hold them, Exim could crash while trying to process
6681 the truncated data. It now detects and logs this case.
6683 17. Further to 4.21/58, another change has been made: if (and only if) the
6684 first line of a message (the first header line) ends with CRLF, a bare LF
6685 in a subsequent header line has a space inserted after it, so as not to
6686 terminate the header.
6688 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6689 unnecessarily, used atoi() instead of strtol(), and didn't check the
6690 termination when getting file sizes from file names by regex.
6692 19. Completely re-implemented the support for maildirsize files, in the light
6693 of a number of problems with the previous contributed implementation
6694 (4.30/29). In particular:
6696 . If the quota is zero, the maildirsize file is maintained, but no quota is
6699 . If the maildir directory does not exist, it is created before any attempt
6700 to write a maildirsize file.
6702 . The quota value in the file is just a cache; if the quota is changed in
6703 the transport, the new value overrides.
6705 . A regular expression is available for excluding directories from the
6708 20. The autoreply transport checks the characters in options that define the
6709 message's headers; it allows continued headers, but it was checking with
6710 isspace() after an embedded newline instead of explicitly looking for a
6713 21. If all the "regular" hosts to which an address was routed had passed their
6714 expiry times, and had not reached their retry times, the address was
6715 bounced, even if fallback hosts were defined. Now Exim should go on to try
6718 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6719 equivalent code in the SMTP transport. Some hosts send humungous responses
6720 to HELO/EHLO, more than 1024 it seems.
6722 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6723 gives compiler warnings in some environments. I've now done it "properly",
6726 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6727 (because of problems with the built-in one) was declared to return uschar *
6728 instead of char *, causing compiler failure.
6730 25. Fixed a file descriptor leak when processing alias/forward files.
6732 26. Fixed a minor format string issue in dbfn.c.
6734 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6736 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6737 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6740 29. When a sender address is verified, it is cached, to save repeating the test
6741 when there is more than one recipient in a message. However, when the
6742 verification involves a callout, it is possible for different callout
6743 options to be set for different recipients. It is too complicated to keep
6744 track of this in the cache, so now Exim always runs a verification when a
6745 callout is required, relying on the callout cache for the optimization.
6746 The overhead is duplication of the address routing, but this should not be
6749 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6750 to be verified with callout=postmaster, and the main callout worked but the
6751 postmaster check failed, the verification correctly failed. However, if a
6752 subsequent RCPT command asked for sender verification *without* the
6753 postmaster check, incorrect caching caused this verification also to fail,
6756 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6757 it was not caching the DNS options (qualify_single, search_parents) that
6758 were used when the lookup failed. A subsequent lookup with different
6759 options therefore always gave the same answer, though there were cases
6760 where it should not have. (Example: a "domains = !$mx_any" option on a
6761 dnslookup router: the "domains" option is always processed without any
6762 widening, but the router might have qualify_single set.) Now Exim uses the
6763 cached value only when the same options are set.
6765 32. Added John Jetmore's "exipick" utility to the distribution.
6767 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6768 than a timeout (e.g. a certificate is required, and is not provided), an
6769 Exim server now closes the connection immediately. Previously it waited for
6770 the client to close - but if the client is SSL, it seems that they each
6771 wait for each other, leading to a delay before one of them times out.
6773 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6774 maintained 0.8.x compatibility because I don't think many are using it, and
6775 it is clearly obsolete.
6777 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6780 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6781 client certificate was expired. A simple patch fixes this, though I don't
6782 understand the full logic of why the verify callback is called multiple
6785 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6786 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6787 which causes problems with some clients (such as the Certicom SSL Plus
6788 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6789 disables the coutermeasure allowing Eudora to connect."
6791 38. Exim was not checking that a write() to a log file succeeded. This could
6792 lead to Bad Things if a log got too big, in particular if it hit a file
6793 size limit. Exim now panics and dies if it cannot write to a log file, just
6794 as it does if it cannot open a log file.
6796 39. Modified OS/Makefile-Linux so that it now contains
6798 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6800 The two -D definitions ensure that Exim is compiled with large file
6801 support, which makes it possible to handle log files that are bigger than
6804 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6805 instance) a domain was checked against a named list that involved a lookup,
6806 causing $domain_data to be set, then another domain was checked against the
6807 same list, then the first domain was re-checked, the value of $domain_data
6808 after the final check could be wrong. In particular, if the second check
6809 failed, it could be set empty. This bug probably also applied to
6812 41. The strip_trailing_dot option was not being applied to the address given
6813 with the -f command-line option.
6815 42. The code for reading a message's header from the spool was incrementing
6816 $received_count, but never initializing it. This meant that the value was
6817 incorrect (doubled) while delivering a message in the same process in which
6818 it was received. In the most common configuration of Exim, this never
6819 happens - a fresh exec is done - but it can happen when
6820 deliver_drop_privilege is set.
6822 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6823 it now includes up to 150 characters of the unexpected data in the log
6826 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6827 and building data strings. The size of both of these buffers was 10 000
6828 bytes - far larger than anybody would *ever* want, thought I. Needless to
6829 say, somebody hit the limit. I have increased the maximum line length to
6830 20 000 and the maximum data length of concatenated lines to 100 000. I have
6831 also fixed two bugs, because there was no checking on these buffers. Tsk,
6832 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6833 buffer is too small.
6835 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6836 lsearch lookups. Now it does.
6838 46. When parsing a route_list item in a manualroute router, a fixed-length
6839 buffer was used for the list of hosts. I made this 1024 bytes long,
6840 thinking that nobody would ever have a list of hosts that long. Wrong.
6841 Somebody had a whole pile of complicated expansion conditions, and the
6842 string was silently truncated, leading to an expansion error. It turns out
6843 that it is easier to change to an unlimited length (owing to other changes
6844 that have happened since this code was originally written) than to build
6845 structure for giving a limitation error. The length of the item that
6846 expands into the list of hosts is now unlimited.
6848 47. The lsearch lookup could not handle data where the length of text line was
6849 more than 4095 characters. Such lines were truncated, leading to shortened
6850 data being returned. It should now handle lines of any length.
6852 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6853 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6856 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6857 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6858 no longer happens in comment lines. A list of replacements is now placed
6859 at the head of all of the source files, except those whose only change is
6860 to replace PERL_COMMAND in the very first #! line.
6862 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6863 messages on the queue, with a bottom-up merge sort, using code contributed
6864 by Michael Haardt. This should make operations like -bp somewhat faster on
6865 large queues. It won't affect queue runners, except when queue_run_in_order
6868 51. Installed eximstats 1.31 in the distribution.
6870 52. Added support for SRV lookups to the dnslookup router.
6872 53. If an ACL referred to $message_body or $message_body_end, the value was not
6873 reset for any messages that followed in the same SMTP session.
6875 54. The store-handling optimization for building very long strings was not
6876 differentiating between the different store pools. I don't think this
6877 actually made any difference in practice, but I've tidied it.
6879 55. While running the routers to verify a sender address, $sender_address
6880 was still set to the sender address. This is wrong, because when routing to
6881 send a bounce to the sender, it would be empty. Therefore, I have changed
6882 it so that, while verifying a sender address, $sender_address is set to <>.
6883 (There is no change to what happens when verifying a recipient address.)
6885 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6886 target A or AAAA records (if not already returned) without resetting the
6887 qualify_single or search_parents options of the DNS resolver. These are
6888 inappropriate in this case because the targets of MX and SRV records must
6889 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6890 target that, when qualified, matched something in the local domain. These
6891 two options are now turned off when doing these lookups.
6893 57. It seems that at least some releases of Reiserfs (which does not have the
6894 concept of a fixed number of inodes) returns zero and not -1 for the
6895 number of available inodes. This interacted badly with check_spool_inodes,
6896 which assumed that -1 was the "no such thing" setting. What I have done is
6897 to check that the total number of inodes is greater than zero before doing
6898 the test of how many are available.
6900 58. When a "warn" ACL statement has a log_message modifier, the message is
6901 remembered, and not repeated. This is to avoid a lot of repetition when a
6902 message has many recipients that cause the same warning to be written.
6903 However, Exim was preserving the list of already written lines for an
6904 entire SMTP session, which doesn't seem right. The memory is now reset if a
6905 new message is started.
6907 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6908 debugging output unless log_rewrite was also set.
6910 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6911 of (int)(handle) when we know that handle contains (void *)(-1).
6913 61. The Exim daemon panic-logs an error return when it closes the incoming
6914 connection. However "connection reset by peer" seems to be common, and
6915 isn't really an error worthy of noting specially, so that particular error
6918 62. When Exim is trying to find all the local interfaces, it used to panic and
6919 die if the ioctl to get the interface flags failed. However, it seems that
6920 on at least one OS (Solaris 9) it is possible to have an interface that is
6921 included in the list of interfaces, but for which you get a failure error
6922 for this call. This happens when the interface is not "plumbed" into a
6923 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6924 failure of the "get flags" call assumes that the interface is down.
6926 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6927 makes life easier for people who are doing arithmetic on fields extracted
6928 from dates, where you often get leading zeros that should not be
6929 interpreted as octal.
6931 64. Added qualify_domain to the redirect router, to override the global
6934 65. If a pathologically long header line contained very many addresses (the
6935 report of this problem mentioned 10 000) and each of them was rewritten,
6936 Exim could use up a very large amount of memory. (It kept on making new
6937 copies of the header line as it rewrote, and never released the old ones.)
6938 At the expense of a bit more processing, the header rewriting function has
6939 been changed so that it no longer eats memory in this way.
6941 66. The generation of the Received: header has been moved from the time that a
6942 message starts to be received, to the time that it finishes. The timestamp
6943 in the Received: header should now be very close to that of the <= log
6944 line. There are two side-effects of this change:
6946 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6947 logged header lines no longer include the local Received: line, because
6948 it has not yet been created. The same applies to a copy of the message
6949 that is returned to a non-SMTP sender when a message is rejected.
6951 (b) When a filter file is tested using -bf, no additional Received: header
6952 is added to the test message. After some thought, I decided that this
6955 This change does not affect the value of $received_for. It is still set
6956 after address rewriting, but before local_scan() is called.
6958 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6960 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6961 gave an unhelpful panic error message, and a defer error. I have managed to
6962 change this behaviour so that it now rejects any supplied certificate,
6963 which seems right, as the list of acceptable certificates is empty.
6965 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6966 gave an unhelpful defer error. I have not managed to make this reject any
6967 supplied certificates, but the error message it gives is "no certificate
6968 supplied", which is not helpful.
6970 70. exigrep's output now also includes lines that are not associated with any
6971 message, but which match the given pattern. Implemented by a patch from
6972 Martin Sluka, which also tidied up the Perl a bit.
6974 71. Recipient callout verification, like sender verification, was using <> in
6975 the MAIL FROM command. This isn't really the right thing, since the actual
6976 sender may affect whether the remote host accepts the recipient or not. I
6977 have changed it to use the actual sender in the callout; this means that
6978 the cache record is now keyed on a recipient/sender pair, not just the
6979 recipient address. There doesn't seem to be a real danger of callout loops,
6980 since a callout by the remote host to check the sender would use <>.
6981 [SEE ABOVE: changed after hitting problems.]
6983 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6984 temporary errors. However, in the case of such a code being given after
6985 the end of a data transmission (i.e. after ".") Exim was failing to write
6986 a retry record for the message. (Yes, there was some broken host that was
6987 actually sending 8xx at this point.)
6989 73. An unknown lookup type in a host list could cause Exim to panic-die when
6990 the list was checked. (An example that provoked this was putting <; in the
6991 middle of a list instead of at the start.) If this happened during a DATA
6992 ACL check, a -D file could be left lying around. This kind of configuration
6993 error no longer causes Exim to die; instead it causes a defer error. The
6994 incident is still logged to the main and panic logs.
6996 74. Buglet left over from Exim 3 conversion. The message "too many messages
6997 in one connection" was written to the rejectlog but not the mainlog, except
6998 when address rewriting (yes!) was being logged.
7000 75. Added write_rejectlog option.
7002 76. When a system filter was run not as root (that is, when system_filter_user
7003 was set), the values of the $n variables were not being returned to the
7004 main process; thus, they were not subsequently available in the $sn
7007 77. Added +return_path_on_delivery log selector.
7009 78. A connection timeout was being treated differently from recipients deferred
7010 when testing hosts_max_try with a message that was older than the host's
7011 retry timeout. (The host should not be counted, thus allowing all hosts to
7012 be tried at least once before bouncing.) This may have been the cause of an
7013 occasionally reported bug whereby a message would remain on the queue
7014 longer than the retry timeout, but would be bounced if a delivery was
7015 forced. I say "may" because I never totally pinned down the problem;
7016 setting up timeout/retry tests is difficult. See also the next item.
7018 79. The ultimate address timeout was not being applied to errors that involved
7019 a combination of host plus message (for example, a timeout on a MAIL
7020 command). When an address resolved to a number of possible hosts, and they
7021 were not all tried for each delivery (e.g. because of hosts_max_try), a
7022 message could remain on the queue longer than the retry timeout.
7024 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7027 81. Fixed an obscure SMTP outgoing bug which required at least the following
7028 conditions: (a) there was another message waiting for the same server;
7029 (b) the server returned 5xx to all RCPT commands in the first message so
7030 that the message was not completed; (c) the server dropped the connection
7031 or gave a negative response to the RSET that Exim sends to abort the
7032 transaction. The observed case was a dropped connection after DATA that had
7033 been sent in pipelining mode. That is, the server had advertised PIPELINING
7034 but was not implementing it correctly. The effect of the bug was incorrect
7035 behaviour, such as trying another host, and this could lead to a crash.
7041 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7042 and daemon.c were passed as pointers to ints; they should have been
7043 pointers to socklen_t variables (which are typically unsigned ints).
7045 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7048 3. Fixed a really odd bug that affected only the testing scheme; patching a
7049 certain fixed string in the binary changed the value of another string that
7050 happened to be identical to the end of the original first string.
7052 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7053 name", it returns that address as the IP address. On some operating
7054 systems (e.g. Solaris), it also passes back the IP address string as the
7055 "host name". However, on others (e.g. Linux), it passes back an empty
7056 string. Exim wasn't checking for this, and was changing the host name to an
7057 empty string, assuming it had been canonicalized.
7059 5. Although rare, it is permitted to have more than one PTR record for a given
7060 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7061 all the names associated with an address, because they do in Solaris.
7062 However, it seems that they do not in Linux for data that comes from the
7063 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7064 I found this out when I moved to a new Linux workstation and tried to run
7065 the Exim test suite.
7067 To get round this problem I have changed the code so that it now does its
7068 own call to the DNS to look up PTR records when searching for a host name.
7069 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7070 addresses that are only in /etc/hosts are still found.
7072 This behaviour is, however, controlled by an option called host_lookup_
7073 order, which defaults to "bydns:byaddr". If people want to use the other
7074 order, or indeed, just use one or the other means of lookup, they can
7075 specify it in this variable.
7077 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7078 some operating systems, this comes back from gethostbyaddr() as an empty
7079 string, and this is what Exim used to test for. However, it seems that in
7080 other systems, "." is yielded. Exim now tests for this case too.
7082 7. The values of check_spool_space and check_log_space are now held internally
7083 as a number of kilobytes instead of an absolute number of bytes. If a
7084 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7085 kilobyte. This means that much larger values can be stored.
7087 8. Exim monitor: an attempt to get the action menu when not actually pointing
7088 at a message produces an empty menu entitled "No message selected". This
7089 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7090 no entries in it ("Shell widget menu has zero width and/or height"). So I
7091 have added a single, blank menu entry in this case.
7093 9. Added ${quote_local_part.
7095 10. MIME decoding is now applied to the contents of Subject: header lines when
7098 11. Now that a reference to $sender_host_address automatically causes a reverse
7099 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7100 host lookup before query-style lookups in lists that might use this
7101 variable. This has therefore been abolished, and the "net-" prefix is no
7102 longer necessary for query-style lookups.
7104 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7105 have been a typo for LFLAGS, so it has been changed.
7107 13. The install script calls Exim with "-C /dev/null" in order to find the
7108 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7109 to be output. However, since Exim outputs its version number before the
7110 error, it didn't break the script. It just looked ugly. I fixed this by
7111 always allowing "-C /dev/null" if the caller is root.
7113 14. Ignore overlarge ACL variable number when reading spool file - insurance
7114 against a later release with more variables having written the file.
7116 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7117 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7118 this, as well as the form without the "IPv6" on the front.
7120 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7121 OS/Makefile-Darwin file.
7123 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7124 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7126 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7127 variable. However, some OS use ints for inodes. I've added cast to long int
7128 to get rid of the compiler warning.
7130 19. I had forgotten to lock out "/../" in configuration file names when
7131 ALT_CONFIG_PREFIX was set.
7133 20. Routers used for verification do not need to specify transports. However,
7134 if such a router generated a host list, and callout was configured, Exim
7135 crashed, because it could not find a port number from the (non-existent)
7136 transport. It now assumes port 25 in this circumstance.
7138 21. Added the -t option to exigrep.
7140 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7141 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7142 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7143 them is set, LOOKUP_LSEARCH is forced.
7145 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7147 24. Added sender and host information to the "rejected by local_scan()" log
7148 line; previously there was no indication of these.
7150 25. Added .include_if_exists.
7152 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7153 Linux. It turns out that not all file systems support this. Apparently some
7154 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7155 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7156 not supported on the file descriptor, is now ignored when Exim is trying to
7157 sync a directory. This applies only to Linux.
7159 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7161 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7162 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7165 29. Added support for maildirsize files from supplied patch (modified a bit).
7167 30. The use of :fail: followed by an empty string could lead Exim to respond to
7168 sender verification failures with (e.g.):
7170 550 Verification failed for <xxx>
7171 550 Sender verify failed
7173 where the first response line was missing the '-' that indicates it is not
7174 the final line of the response.
7176 31. The loop for finding the name of the user that called Exim had a hardwired
7177 limit of 10; it now uses the value of finduser_retries, which is used for
7178 all other user lookups.
7180 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7183 33. Exim was neglecting to zero errno before one call of strtol() when
7184 expanding a string and expecting an integer value. On some systems this
7185 resulted in spurious "integer overflow" errors. Also, it was casting the
7186 result into an int without checking.
7188 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7189 did not work. The code looks as if it has *never* worked, though it appears
7190 to have been documented since at least release 1.62. I have made it work.
7192 35. The "timeout_DNS" error in retry rules, also documented since at least
7193 1.62, also never worked. As it isn't clear exactly what this means, and
7194 clearly it isn't a major issue, I have abolished the feature by treating it
7195 as "timeout", and writing a warning to the main and panic logs.
7197 36. The display of retry rules for -brt wasn't always showing the error code
7200 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7201 timeout_connect_A, timeout_connect_MX.
7203 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7204 to the empty sender.
7206 39. The daemon was not analysing the content of -oX till after it had closed
7207 stderr and disconnected from the controlling terminal. This meant that any
7208 syntax errors were only noted on the panic log, and the return code from
7209 the command was 0. By re-arranging the code a little, I've made the
7210 decoding happen first, so such errors now appear on stderr, and the return
7211 code is 1. However, the actual setting up of the sockets still happens in
7212 the disconnected process, so errors there are still only recorded on the
7215 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7216 connections (as happens on some IP stacks) was logged at start up time as
7217 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7218 it from "IPv6 and IPv4", which means that two separate sockets are being
7221 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7222 says whether AF_INET or AF_INET6 was passed as an argument.
7224 42. Exiwhat output was messed up when time zones were included in log
7227 43. Exiwhat now gives more information about the daemon's listening ports,
7228 and whether -tls-on-connect was used.
7230 44. The "port" option of the smtp transport is now expanded.
7232 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7233 silently ignored. Now an error message is written to the main and panic
7236 46. There's a new ACL modifier called "logwrite" which writes to a log file
7237 as soon as it is encountered.
7239 47. Added $local_user_uid and $local_user_gid at routing time.
7241 48. Exim crashed when trying to verify a sender address that was being
7244 49. Exim was recognizing only a space character after ".include". It now also
7245 recognizes a tab character.
7247 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7248 extracting the relevant information from the specification. The man page no
7249 longer contains scrambled data for the -d option, and I've added a section
7250 at the front about calling Exim under different names.
7252 51. Added "extra_headers" argument to the "mail" command in filter files.
7254 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7257 53. Installed eximstats 1.29.
7259 54. Added transport_filter_timeout as a generic transport option.
7261 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7262 Cc: header lines. This was required by RFC 822, but it not required by RFC
7265 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7266 incoming messages that did not have them. Now it does so only if the
7267 message originates locally, that is, if there is no associated remote host
7268 address. When Resent- header lines are present, this applies to the Resent-
7269 lines rather than the non-Resent- lines.
7271 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7272 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7274 58. Messages for configuration errors now include the name of the main
7275 configuration file - useful now that there may be more than one file in a
7276 list (.included file names were always shown).
7278 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7279 for those rare installations that do not start the daemon as root or run it
7280 setuid root. I've cut out the call to initgroups() if the daemon is not
7283 60. The Exim user and group can now be bound into the binary as text strings
7284 that are looked up at the start of Exim's processing.
7286 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7288 62. Added $mailstore_basename variable.
7290 63. Installed patch to sieve.c from Michael Haardt.
7292 64. When Exim failed to open the panic log after failing to open the main log,
7293 the original message it was trying to log was written to stderr and debug
7294 output, but if they were not available (the usual case in production), it
7295 was lost. Now it is written to syslog before the two lines that record the
7296 failures to open the logs.
7298 65. Users' Exim filters run in subprocesses under the user's uid. It is
7299 possible for a "deliver" command or an alias in a "personal" command to
7300 provoke an address rewrite. If logging of address rewriting is configured,
7301 this fails because the process is not running as root or exim. There may be
7302 a better way of dealing with this, but for the moment (because 4.30 needs
7303 to be released), I have disabled address rewrite logging when running a
7304 filter in a non-root, non-exim process.
7310 1. The buildconfig auxiliary program wasn't quoting the value set for
7311 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7312 not defined. This bug was masked in 4.22 by the effect that was fixed in
7315 2. Some messages that were rejected after a message id was allocated were
7316 shown as "incomplete" by exigrep. It no longer does this for messages that
7317 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7319 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7320 have allow_domain_literals set, the ID did not get logged in the <= line.
7321 Domain literals are now always recognized in Message-ID: header lines.
7323 4. The first argument for a ${extract expansion item is the key name or field
7324 number. Leading and trailing spaces in this item were not being ignored,
7325 causing some misleading effects.
7327 5. When deliver_drop_privilege was set, single queue runner processes started
7328 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7329 same command in the process it spins off) were not dropping privilege.
7331 6. When the daemon running as "exim" started a queue runner, it always
7332 re-executed Exim in the spun-off process. This is a waste of effort when
7333 deliver_drop_privilege is set. The new process now just calls the
7334 queue-runner function directly.
7340 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7343 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7344 ignored. Though the use of -oP was forcing the writing of a pid file, it
7345 was always written to the default place.
7347 3. If the message "no IP address found for host xxxx" is generated during
7348 incoming verification, it is now followed by identification of the incoming
7349 connection (so you can more easily find what provoked it).
7351 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7353 5. Added some features to "harden" Exim a bit more against certain attacks:
7355 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7356 be put in Local/Makefile. This is like the never_users runtime option,
7357 but it cannot be overridden. The default setting is "root".
7359 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7360 prefix string with which any file named in a -C command line option
7363 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7364 is retained for -C and -D only if the caller of Exim is root. Without
7365 it, the exim user may also use -C and -D and retain privilege.
7367 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7368 command line option is disabled.
7370 6. Macro names set by the -D option must start with an upper case letter, just
7371 like macro names defined in the configuration file.
7373 7. Added "dereference=" facility to LDAP.
7375 8. Two instances of the typo "uknown" in the source files are fixed.
7377 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7378 the Configure-Makefile script screwed up while processing it.
7380 10. Incorporated PCRE 4.4.
7382 11. The SMTP synchronization check was not operating right at the start of an
7383 SMTP session. For example, it could not catch a HELO sent before the client
7384 waited for the greeting. There is now a check for outstanding input at the
7385 point when the greeting is written. Because of the duplex, asynchronous
7386 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7387 way, but not yet received, when the check is performed.
7389 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7390 on TCP/IP sockets, because this apparently causes some broken clients to
7393 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7394 unchanged) from the Cygwin maintainer.
7396 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7397 was supported instead of testing for mbx. Effectively a typo.
7399 15. The spa authenticator server code was not checking that the input it
7400 received was valid base64.
7402 16. The debug output line for the "set" modifier in ACLs was not showing the
7403 name of the variable that was being set.
7405 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7407 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7408 lookup on its own. Something else was needed to trigger the lookup. For
7409 example, a match in host_lookup or the need for a host name in a host list.
7410 Now, if $sender_host_name is referenced and the host name has not yet been
7411 looked up, a lookup is performed. If the lookup fails, the variable remains
7412 empty, and $host_lookup_failed is set to "1".
7414 19. Added "eqi" as a case-independent comparison operator.
7416 20. The saslauthd authentication condition could segfault if neither service
7417 nor realm was specified.
7419 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7420 error message that was logged was misleading, and incoming SMTP
7421 connections were dropped. The message is now more accurate, and temporary
7422 errors are given to SMTP connections.
7424 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7425 (except RSET) with a 421 error, until QUIT is received. However, it was
7426 failing to send a response to QUIT.
7428 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7429 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7430 ACL is now run after the helo_try_verify_hosts code.
7432 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7433 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7434 case-independent, so other case variants are also recognized). Apparently
7435 some systems use these upper case variants.
7437 25. If more than two messages were waiting for the same host, and a transport
7438 filter was specified for the transport, Exim sent two messages over the
7439 same TCP/IP connection, and then failed with "socket operation on non-
7440 socket" when it tried to send the third.
7442 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7444 27. The extern definition of crypt16() in expand.c was not being excluded when
7445 the OS had its own crypt16() function.
7447 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7448 as a preferred synonym for return_size_limit, both as an option and as an
7451 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7453 30. Changed the default configuration ACL to relax the local part checking rule
7454 for addresses that are not in any local domains. For these addresses,
7455 slashes and pipe symbols are allowed within local parts, but the sequence
7456 /../ is explicitly forbidden.
7458 31. SPA server authentication was not clearing the challenge buffer before
7461 32. log_message in a "warn" ACL statement was writing to the reject log as
7462 well as to the main log, which contradicts the documentation and doesn't
7463 seem right (because no rejection is happening). So I have stopped it.
7465 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7466 However, I am unable to do any testing of this.
7468 34. Fixed an infelicity in the appendfile transport. When checking directories
7469 for a mailbox, to see if any needed to be created, it was accidentally
7470 using path names with one or more superfluous leading slashes; tracing
7471 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7473 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7474 contents are added to the log line that is written for every discarded
7475 recipient. (Previously a log_message setting was ignored.)
7477 36. The ${quote: operator now quotes the string if it is empty.
7479 37. The install script runs exim in order to find its version number. If for
7480 some reason other than non-existence or emptiness, which it checks, it
7481 could not run './exim', it was installing it with an empty version number,
7482 i.e. as "exim-". This error state is now caught, and the installation is
7485 38. An argument was missing from the function that creates an error message
7486 when Exim fails to connect to the socket for saslauthd authentication.
7487 This could cause Exim to crash, or give a corrupted message.
7489 39. Added isip, isip4, and isip6 to ${if conditions.
7491 40. The ACL variables $acl_xx are now saved with the message, and can be
7492 accessed later in routers, transports, and filters.
7494 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7495 strings in the file are not string-expanded.
7497 42. If a MAIL command specified a SIZE value that was too large to fit into an
7498 int variable, the check against message_size_limit failed. Such values are
7499 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7500 day this will have to be increased, but I don't think I want to be around
7501 when emails are that large.
7508 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7509 iconv() is not standard in FreeBSD.
7511 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7512 IPv6 enabled. The observed symptom was a segmentation fault on return from
7513 the function os_common_find_running_interfaces() in src/os.c.
7515 3. In the check_special_case() function in daemon.c I had used "errno" as an
7516 argument name, which causes warnings on some systems. This was basically a
7517 typo, since it was named "eno" in the comments!
7519 4. The code that waits for the clock to tick (at a resolution of some fraction
7520 of a second) so as to ensure message-id uniqueness was always waiting for
7521 at least one whole tick, when it could have waited for less. [This is
7522 almost certainly not relevant at current processor speeds, where it is
7523 unlikely to ever wait at all. But we try to future-proof.]
7525 5. The function that sleeps for a time interval that includes fractions of a
7526 second contained a race. It did not block SIGALRM between setting the
7527 timer, and suspending (a couple of lines later). If the interval was short
7528 and the sigsuspend() was delayed until after it had expired, the suspension
7529 never ended. On busy systems this could lead to processes getting stuck for
7532 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7533 process, before it forks any delivery processes. The open lookup caching
7534 mechanism meant that the open file or database connection was passed into
7535 the delivery process. The problem was that delivery processes always tidy
7536 up cached lookup data. This could cause a problem for the next delivery
7537 process started by the queue runner, because the external queue runner
7538 process does not know about the closure. So the next delivery process
7539 still has data in the lookup cache. In the case of a file lookup, there was
7540 no problem because closing a file descriptor in a subprocess doesn't affect
7541 the parent. However, if the lookup was caching a connection to a database,
7542 the connection was closed, and the second delivery process was likely to
7543 see errors such as "PGSQL: query failed: server closed the connection
7544 unexpectedly". The problem has been fixed by closing all cached lookups
7545 in a queue runner before running a delivery process.
7547 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7548 seem to have the "const" qualifier which it has on other OS. I've
7551 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7552 *of the same type* (client or server) with the same public name that an
7553 error should be diagnosed.
7555 9. When Exim looked up a host name for an IP address, but failed to find the
7556 original IP address when looking up the host name (a safety check), it
7557 output the message "<ip address> does not match any IP for NULL", which was
7558 confusing, to say the least. The bug was that the host name should have
7559 appeared instead of "NULL".
7561 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7562 user that is built into the binary, and the -C or -D options is used, root
7563 privilege is dropped before the configuration file is read. In addition,
7564 logging is switched to stderr instead of the normal log files. If the
7565 configuration then re-defines the Exim user, the unprivileged environment
7566 is probably not what is expected, so Exim logs a panic warning message (but
7569 However, if deliver_drop_privilege is set, the unprivileged state may well
7570 be exactly what is intended, so the warning has been cut out in that case,
7571 and Exim is allowed to try to write to its normal log files.
7577 1. smtp_return_error_details was not giving details for temporary sender
7578 or receiver verification errors.
7580 2. Diagnose a configuration error if two authenticators have the same public
7583 3. Exim used not to create the message log file for a message until the first
7584 delivery attempt. This could be confusing when incoming messages were held
7585 for policy or load reasons. The message log file is now created at the time
7586 the message is received, and an initial "Received" line is written to it.
7588 4. The automatically generated man page for command line options had a minor
7589 bug that caused no ill effects; however, a more serious problem was that
7590 the procedure for building the man page automatically didn't always
7591 operate. Consequently, release 4.20 contains an out-of-date version. This
7592 shouldn't happen again.
7594 5. When building Exim with embedded Perl support, the script that builds the
7595 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7596 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7598 6. The freeze_tell option was not being used for messages that were frozen on
7599 arrival, either by an ACL or by local_scan().
7601 7. Added the smtp_incomplete_transaction log selector.
7603 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7604 was accepting AUTH without a new EHLO.
7606 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7607 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7610 10. Logging of TCP/IP connections (when configured) now happens in the main
7611 daemon process instead of the child process, so that the TCP/IP connection
7612 count is more accurate (but it can never be perfect).
7614 11. The use of "drop" in a nested ACL was not being handled correctly in the
7615 outer ACL. Now, if condition failure induced by the nested "drop" causes
7616 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7617 or "require"), the connection is dropped.
7619 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7620 that yield "discard" can now be used with an "accept" or a "discard" verb,
7621 but an error is generated for any others (because I can't see a useful way
7622 to define what should happen).
7624 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7625 are now processed in the same way as lines in the Exim configuration file.
7626 In particular, continuation lines are supported.
7628 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7630 15. Added -ti meaning -t -i.
7632 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7633 domains, and warn by logging if others are found.
7635 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7636 returned from GIFCONF when Exim is trying to find the list of interfaces on
7637 a host. The code in os.c has been modified to copy each ifreq to an aligned
7638 structure in all cases.
7640 Also, in some cases, the returned ifreq's were being copied to a 'struct
7641 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7642 means the last couple of bytes of an IPv6 address could be chopped if the
7643 ifreq contained only a normal sockaddr (14 bytes storage).
7645 18. Named domain lists were not supported in the hosts_treat_as_local option.
7646 An entry such as +xxxx was not recognized, and was treated as a literal
7649 19. Ensure that header lines added by a DATA ACL are included in the reject log
7650 if the ACL subsequently rejects the message.
7652 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7653 MD5 (which is deprecated).
7655 21. When testing a filter file using -bf, Exim was writing a message when it
7656 took the sender from a "From " line in the message, but it was not doing so
7657 when it took $return_path from a Return-Path: header line. It now does.
7659 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7660 with a valid header line field name (a series of printing characters
7661 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7663 23. Changed "disc" in the source to "disk" to conform to the documentation and
7664 the book and for uniformity.
7666 24. Ignore Sendmail's -Ooption=value command line item.
7668 25. When execve() failed while trying to run a command in a pipe transport,
7669 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7670 could be confused with a return value of 69 from the command itself. This
7671 has been changed to 127, the value the shell returns if it is asked to run
7672 a non-existent command. The wording for the related log line suggests a
7673 non-existent command as the problem.
7675 26. If received_header_text expands to an empty string, do not add a Received:
7676 header line to the message. (Well, it adds a token one on the spool, but
7677 marks it "old" so that it doesn't get used or transmitted.)
7679 27. Installed eximstats 1.28 (addition of -nt option).
7681 28. There was no check for failure on the call to getsockname() in the daemon
7682 code. This can fail if there is a shortage of resources on the system, with
7683 ENOMEM, for example. A temporary error is now given on failure.
7685 29. Contrary to the C standard, it seems that in some environments, the
7686 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7687 program. Exim now does this explicitly; it affects the formatting of
7688 timestamps using strftime().
7690 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7691 complaints. I've now initialized all the variables, to avoid this.
7693 32. Header lines added by a system filter were not being "seen" during
7694 transport-time rewrites.
7696 33. The info_callback() function passed to OpenSSL is set up with type void
7697 (*)(SSL *, int, int), as described somewhere. However, when calling the
7698 function (actually a macro) that sets it up, the type void(*)() is
7699 expected. I've put in a cast to prevent warnings from picky compilers.
7701 34. If a DNS black list lookup found a CNAME record, but there were no A
7702 records associated with the domain it pointed at, Exim crashed.
7704 35. If a DNS black list lookup returned more than one A record, Exim ignored
7705 all but the first. It now scans all returned addresses if a particular IP
7706 value is being sought. In this situation, the contents of the
7707 $dnslist_value variable are a list of all the addresses, separated by a
7710 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7711 to accept a host name and all its aliases if the forward lookup for any of
7712 them yielded the IP address of the incoming connection. Now it accepts only
7713 those names whose forward lookup yields the correct IP address. Any other
7714 names are discarded. This closes a loophole whereby a rogue DNS
7715 administrator could create reverse DNS records to break through a
7716 wildcarded host restriction in an ACL.
7718 37. If a user filter or a system filter that ran in a subprocess used any of
7719 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7720 the wrong values were passed to the pipe command ($thisaddress had the
7721 value of $0, $0 had the value of $1, etc). This bug was introduced by
7722 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7724 38. Improved the line breaking for long SMTP error messages from ACLs.
7725 Previously, if there was no break point between 40 and 75 characters, Exim
7726 left the rest of the message alone. Two changes have been made: (a) I've
7727 reduced the minimum length to 35 characters; (b) if it can't find a break
7728 point between 35 and 75 characters, it looks ahead and uses the first one
7729 that it finds. This may give the occasional overlong line, but at least the
7730 remaining text gets split now.
7732 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7733 file descriptors might be low, and that setting 1000 would always raise it.
7734 It turns out that in some environments, the limit is already over 1000 and
7735 that lowering it causes trouble. So now Exim takes care not to decrease it.
7737 40. When delivering a message, the value of $return_path is set to $sender_
7738 address at the start of routing (routers may change the value). By an
7739 oversight, this default was not being set up when an address was tested by
7740 -bt or -bv, which affected the outcome if any router or filter referred to
7743 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7744 log only when "message" or "log_message" are set. However, if one of the
7745 conditions was an address verification, or a call to a nested ACL, the
7746 messages generated by the underlying test were being passed through. This
7747 no longer happens. The underlying message is available in $acl_verify_
7748 message for both "message" and "log_message" expansions, so it can be
7749 passed through if needed.
7751 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7752 new expansion $bh_ to give the encoded byte string without charset
7753 translation. Translation happens only if iconv() is available; HAVE_ICONV
7754 indicates this at build time. HEADERS_CHARSET gives the charset to
7755 translate to; headers_charset can change it in the configuration, and
7756 "headers charset" can change it in an individual filter file.
7758 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7759 that creates RFC 2047 encoded "words" labels them as that charset instead
7760 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7761 expansion operator; (ii) when Exim creates a From: line for a local
7762 message; (iii) when a header line is rewritten to include a "phrase" part.
7764 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7765 buggy, causing it to skip the first lines of messages whose message ID
7766 ended in 'D'. This would not have bitten before Exim release 4.14, because
7767 message IDs were unlikely to end in 'D' before then. The effect was to have
7768 incorrect size information for certain domains.
7770 45. #include "config.h" was missing at the start of the crypt16.c module. This
7771 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7774 46. If there was a timeout during a "random" callout check, Exim treated it as
7775 a failure of the random address, and carried on sending RSET and the real
7776 address. If the delay was just some slowness somewhere, the response to the
7777 original RCPT would be taken as a response to RSET and so on, causing
7778 mayhem of various kinds.
7780 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7781 when I implemented it. It didn't allow for the fact that some option values
7782 may legitimately be negative (e.g. size_addition), and it didn't even do
7783 the right test for positive values.
7785 48. Domain names in DNS records are case-independent. Exim always looks them up
7786 in lower case. Some resolvers return domain names in exactly the case they
7787 appear in the zone file, that is, they may contain uppercase letters. Not
7788 all resolvers do this - some return always lower case. Exim was treating a
7789 change of case by a resolver as a change of domain, similar to a widening
7790 of a domain abbreviation. This triggered its re-routing code and so it was
7791 trying to route what was effectively the same domain again. This normally
7792 caused routing to fail (because the router wouldn't handle the domain
7793 twice). Now Exim checks for this case specially, and just changes the
7794 casing of the domain that it ultimately uses when it transmits the message
7797 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7800 50. If a filter generated a file delivery with a non-absolute name (possible if
7801 no home directory exists for the router), the forbid_file option was not
7804 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7805 the existing equality matching.
7807 52. Exim was using ints instead of ino_t variables in some places where it was
7808 dealing with inode numbers.
7810 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7811 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7812 called TMPDIR, and if it finds it is different, it changes its value.
7814 54. The smtp_printf() function is now made available to local_scan() so
7815 additional output lines can be written before returning. There is also an
7816 smtp_fflush() function to enable the detection of a dropped connection.
7817 The variables smtp_input and smtp_batched_input are exported to
7820 55. Changed the default runtime configuration: the message "Unknown user"
7821 has been removed from the ACL, and instead placed on the localuser router,
7822 using the cannot_route_message feature. This means that any verification
7823 failures that generate their own messages won't get overridden. Similarly,
7824 the "Unrouteable address" message that was in the ACL for unverifiable
7825 relay addresses has also been removed.
7827 56. Added hosts_avoid_esmtp to the smtp transport.
7829 57. The exicyclog script was not checking for the esoteric option
7830 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7831 will work only if exicyclog is run under the appropriate euid.
7833 58. Following a discussion on the list, the rules by which Exim recognises line
7834 endings on incoming messages have been changed. The -dropcr and drop_cr
7835 options are now no-ops, retained only for backwards compatibility. The
7836 following line terminators are recognized: LF CRLF CR. However, special
7837 processing applies to CR:
7839 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7840 nor a local message in the state where . is a terminator.
7842 (ii) If a bare CR is encountered in a header line, an extra space is added
7843 after the line terminator so as not to end the header. The reasoning
7844 behind this is that bare CRs in header lines are most likely either
7845 to be mistakes, or people trying to play silly games.
7847 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7848 was being incorrectly given as 18 bytes larger than it should have been.
7849 This is a VOB (very old bug).
7851 60. This may never have affected anything current, but just in case it has:
7852 When the local host is found other than at the start of a list of hosts,
7853 the local host, those with the same MX, and any that follow, are discarded.
7854 When the list in question was part of a longer list of hosts, the following
7855 hosts (not currently being processed) were also being discarded. This no
7856 longer happens. I'm not sure if this situation could ever has previously
7859 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7862 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7863 header. This is something that is recommended in a new Internet Draft, and
7864 is something that is documented as being done by Sendmail. There are two
7865 possible values. For messages generated by the autoreply transport, Exim
7868 Auto-Submitted: auto-replied
7870 whereas for all other generated messages (e.g. bounces) it adds
7872 Auto-Submitted: auto-generated
7874 63. The "personal" condition in filters now includes a test for the
7875 Auto-Submitted: header. If it contains the string "auto-" the message it
7876 not considered personal.
7878 64. Added rcpt_include_affixes as a generic transport option.
7880 65. Added queue_only_override (default true).
7882 66. Added the syslog_duplication option.
7884 67. If what should have been the first header line of a message consisted of
7885 a space followed by a colon, Exim was mis-interpreting it as a header line.
7886 It isn't of course - it is syntactically invalid and should therefore be
7887 treated as the start of the message body. The misbehaviour could have
7888 caused a number of strange effects, including loss of data in subsequent
7889 header lines, and spool format errors.
7891 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7892 client host had authenticated. This control can now be exercised by an ACL
7893 for more flexibility.
7895 69. By default, callouts do not happen when testing with -bh. There is now a
7896 variant, -bhc, which does actually run the callout code, including
7897 consulting and updating the callout cache.
7899 70. Added support for saslauthd authentication, courtesy of Alexander
7902 71. If statvfs() failed on the spool or log directories while checking their
7903 size for availability, Exim confusingly gave the error "space shortage".
7904 Furthermore, in debugging mode it crashed with a floating point exception.
7905 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7906 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7907 serious problem, Exim now writes to the main and panic logs when this
7908 happens, with details of the failure. It then refuses to accept the
7909 incoming message, giving the message "spool directory problem" or "log
7910 directory problem" with a 421 code for SMTP messages.
7912 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7913 0, 1, and 2 exist, because some OS complain for execs without them (see
7914 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7915 descriptors. However, the code omitted to check that the open succeeded,
7916 causing mysterious errors if for some reason the permissions on /dev/null
7917 got screwed. Now Exim writes a message to the main and panic logs, and
7918 bombs out if it can't open /dev/null.
7920 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7921 interact so that it is all more flexible. It is supposed to remain
7922 backwards compatible. Also added extra_local_interfaces.
7924 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7925 to bomb out with an assertion failure - to the client this appears as a
7926 connection drop. This problem occurs in the part of the code that was taken
7927 from the Samba project. Fortunately, the assertion is in a very simple
7928 function, so I have fixed this by reproducing the function inline in the
7929 one place where it is called, and arranging for authentication to fail
7930 instead of killing the process with assert().
7932 75. The SPA client code was not working when the server requested OEM rather
7933 than Unicode encoding.
7935 76. Added code to make require_files with a specific uid setting more usable in
7936 the case where statting the file as root fails - usually a non-root-mounted
7937 NFS file system. When this happens and the failure is EACCES, Exim now
7938 forks a subprocess and does the per-uid checking as the relevant uid.
7940 77. Added process_log_path.
7942 78. If log_file_path was not explicitly set, a setting of check_log_space or
7943 check_log_inodes was ignored.
7945 79. If a space check for the spool or log partitions fails, the incident is now
7946 logged. Of course, in the latter case the data may get lost...
7948 80. Added the %p formatting code to string_format() so that it can be used to
7949 print addresses in debug_print(). Adjusted all the address printing in the
7950 debugging in store.c to use %p rather than %d.
7952 81. There was a concern that a line of code in smtp_in.c could overflow a
7953 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7954 initially expressed, the concern was not well-founded, because trailing
7955 spaces are removed early. However, if the trailing spaces were followed by
7956 a NULL, they did not get removed, so the overflow was possible. Two fixes
7959 (a) I re-wrote the offending code in a cleaner fashion.
7960 (b) If an incoming SMTP command contains a NULL character, it is rejected
7963 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7964 runs initgroups(), so that if the Exim user is in any additional groups,
7965 they will be used during message reception.
7971 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob