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.
184 JH/01 Avoid costly startup code when not strictly needed. This reduces time
185 for some exim process initialisations. It does mean that the logging
186 of TLS configuration problems is only done for the daemon startup.
188 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
190 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
191 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
194 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
195 DKIM signing is being done. Previously a transport_filter would always
196 disable CHUNKING, falling back to traditional DATA.
198 JH/05 Regard command-line recipients as tainted.
200 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
202 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
203 PAM library frees one of the arguments given to it, despite the
204 documentation. Therefore a plain malloc must be used.
206 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
207 on-stack buffers were used, resulting in a taint trap when DSN information
208 copied from a received message was written into the buffer.
210 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
211 the ordering of its ARC headers. This caused a crash.
213 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
214 a new record was being constructed with information from the peer, a trap
217 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
218 installation would get error messages from DMARC verify, when it hit the
219 nonexistent file indicated by the default. Distros wanting DMARC enabled
220 should both provide the file and set the option.
221 Also enforce no DMARC verification for command-line sourced messages.
223 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
224 could, depending on the platform, hang at the STARTTLS response.
226 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
227 message on a connection. Previously if one message had ARC headers and
228 the following one did not, a crash could result when adding an
229 Authentication-Results: header.
231 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
232 Exim main code and Exim-related utities. The introduction of taint
233 tracking also did many adjustments to string handling. Since then, eximon
234 frequently terminated with an assert failure.
236 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
237 check for 452 responses. This slightly helps the inefficieny of doing
238 a large alias-expansion into a recipient-limited target. The max_rcpt
239 transport option still applies (and at the current default, will override
240 the new feature). The check is done for either cause of synch, and forces
241 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
242 connection. The new facility is not tunable at this time.
244 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
245 library live data was being used, so the results became garbage. Make
246 copies while it is still usable.
248 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
249 on delivery deferred (==) and failed (**) lines (if a delivery was
250 attemtped). Previously it was only on completion (=>) lines.
252 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
253 for the expansion of the server_scram_iter and server_scram_salt options.
255 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
256 are now specifically given a NO_DATA response without hitting the system
257 resolver. The library goes on to do the now-standard TXT lookup.
258 Use of dnsdb lookups is not affected.
260 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
261 only retrieve the errormessage once. Previously two calls to dlerror()
262 were used, and the second one (for mainlog/paniclog) retrieved null
265 JH/20 Taint checking: disallow use of tainted data for
266 - the appendfile transport file and directory options
267 - the pipe transport command
268 - the autoreply transport file, log and once options
269 - file names used by the redirect router (including filter files)
271 - paths used by single-key lookups
272 Previously this was permitted.
274 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
275 adjusted the size of a major service buffer; this failed because the
276 buffer was in use at the time. Change to a compile-time increase in the
277 buffer size, when this authenticator is compiled into exim.
279 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
280 previous fast-mode was untenable in the face of glibs using mmap to
281 support larger malloc requests.
283 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
284 New values supported, if defined on system where compiled:
285 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
286 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
288 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
289 running a limited number of proceses in parallel, a benefit is gained. The
290 amount varies with the platform hardware and load. The use of the option
291 queue_run_in_order means we cannot do this, as ordering becomes
294 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
295 had introduced a string-copy (for ensuring NUL-termination) which was not
296 appropriate for that case, which can include embedded NUL bytes in the
297 block of data. Investigation showed the copy to actually be needless, the
298 data being length-specified.
300 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
301 done during a receiving connection, and both used TLS, global info was
302 used rather than per-connection info for tracking the state of data
303 queued for transmission. This could result in a connection hang.
305 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
306 Previously, when delivering serveral messages down a single connection
307 only the first would provide a SIZE. This was due to the size information
308 not being properly tracked.
310 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
311 TAI (at 37 seconds currently), pretend to be in UTC for time-related
312 expansion and logging. Previously, spurious values such as a future
313 minute could be seen.
315 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
316 it could crash from a null-deref. This could also affect the
317 ${addresses: } operator and ${readsock } item.
319 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
320 message following a mime one, the variable was not reset.
322 JH/30 When an pipelined-connect fails at the first response, assume incorrect
323 cached capability (perhaps the peer reneged?) and immediately retry in
326 JH/31 Fix spurious detection of timeout while writing to transport filter.
328 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
329 an attempt to copy the string was made before checking it.
331 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
332 taint of the lookup key was maintained; we now regard the presence in the
333 filesystem as sufficient validation.
335 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
336 argument is supplied.
338 JH/35 The dsearch lookup now requires that the directory is an absolute path.
339 Previously this was not checked, and nonempty relative paths made an
340 access under Exim's current working directory.
342 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
343 Previously no event was raised.
345 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
346 parameter supplied by the sender MAIL FROM command. Previously it was
347 ignored, and only the check_spool_space option value for the required
350 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
351 the size of the signing public-key. Previously it was instead giving
352 the size of the signature hash.
354 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
355 the default. See the (new) dkim_verify_min_keysizes option.
357 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
358 and an ACL use a lookup for checking either the local_part or domain,
359 stale data could be accessed. Ensure that variable references are
360 dropped between messages.
362 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
363 by the client was not checked as pointing within response data before
364 being used. A malicious client could thus cause an out-of-bounds read and
365 possibly gain authentication. Fix by adding the check.
367 JH/42 Internationalisation: change the default for downconversion in the smtp
368 transport to be "if needed". Previously it was "as previously set" for
369 the message, which usually meant "if needed" for message-submission but
370 "no" for everything else. However, MTAs have been seen using SMTPUTF8
371 even when the envelope addresses did not need it, resulting in forwarding
372 failures to non-supporting MTAs. A downconvert in such cases will be
373 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
374 transport. The change does mean that addresses needing conversion will
375 be converted when previously a delivery failure would occur.
377 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
378 response was received it would be used unchecked in a fail-DSN, violating
379 standards on line-length limits. Truncate if needed.
381 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
382 doesn't work. (Additionally add a new main config option to configure the
389 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
390 side, to decode a TLS 1.3 packet capture.
392 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
393 Previously the default library behaviour applied, sending two, each in
396 JH/03 Debug output for ACL now gives the config file name and line number for
399 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
401 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
403 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
404 buffer overrun for (non-chunking) other transports.
406 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
407 TLS1.3, means that a server rejecting a client certificate is not visible
408 to the client until the first read of encrypted data (typically the
409 response to EHLO). Add detection for that case and treat it as a failed
410 TLS connection attempt, so that the normal retry-in-clear can work (if
411 suitably configured).
413 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
414 and/or domain. Found and fixed by Jason Betts.
416 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
417 configuration). If a CNAME target was not a wellformed name pattern, a
420 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
421 the OS reports them interleaved with other addresses.
423 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
424 used both for input and for a verify callout, both encrypted, SMTP
425 responses being sent by the server could be lost. This resulted in
426 dropped connections and sometimes bounces generated by a peer sending
429 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
430 string. Previously it was possible to cause undefined behaviour in a
431 library routine (usually a crash). Found by "zerons".
433 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
436 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
437 API was removed, so update to use the newer ones.
439 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
440 any timeout set, is taking a long time. Previously we would hang on to a
441 rotated logfile "forever" if the input was arriving with long gaps
442 (a previous attempt to fix addressed lack, for a long time, of initial
445 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
446 shared (NFS) environment. The length of the tempfile name is now
447 4 + 16 ("hdr.$message_exim_id") which might break on file
448 systems which restrict the file name length to lower values.
451 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
452 shared (NFS) environment.
454 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
455 did for all versions <4.90). Notably -M, -m, --invert, -I may be
458 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
459 on some platforms for bit 31.
461 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
462 to changes apparently associated with TLS1.3 handling some of the APIs
463 previously used were either nonfunctional or inappropriate. Strings
464 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
465 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
466 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
467 This affects log line X= elements, the $tls_{in,out}_cipher variables,
468 and the use of specific cipher names in the encrypted= ACL condition.
470 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
472 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
473 verification result was not updated unless hosts_require_ocsp applied.
475 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
476 queue_list_requires_admin set to false, non-admin users were denied the
479 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
480 directory-of-certs mode. Previously they were advertised despite the
483 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
484 A single TCP connection by a client will now hold a TLS connection open
485 for multiple message deliveries, by default. Previously the default was to
488 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
489 default. If built with the facility, DANE will be used. The facility
490 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
492 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
493 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
494 must be defined and you must still, unless you define DISABLE_TLS, manage
495 the the include-dir and library-file requirements that go with that
496 choice. Non-TLS builds are still supported.
498 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
501 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
502 default. If the platform supports and has the facility enabled, it will
503 be requested on all coneections.
505 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
506 controlled by the build-time option SUPPORT_PIPE_CONNECT.
508 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
510 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
511 success-DSN messages. Previously the From: header was always the default
512 one for these; the option was ignored.
514 JH/28 Fix the timeout on smtp response to apply to the whole response.
515 Previously it was reset for every read, so a teergrubing peer sending
516 single bytes within the time limit could extend the connection for a
517 long time. Credit to Qualsys Security Advisory Team for the discovery.
519 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
520 delivery address, which leaked information of the results of local
521 forwarding. Change to the original envelope recipient address, per
524 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
525 requested. Previously not bounce was generated and a log entry of
526 error ignored was made.
528 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
530 JH/32 Introduce a general tainting mechanism for values read from the input
531 channel, and values derived from them. Refuse to expand any tainted
532 values, to catch one form of exploit.
534 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
535 was unused and the unexpanded text used for the test. Found and
536 fixed by Ruben Jenster.
538 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
539 an attempt to use a TLS library read routine dereffed a nul pointer,
542 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
543 them in our smtp response.
545 JH/36 Have the general router option retry_use_local_part default to true when
546 any of the restrictive preconditions are set (to anything). Previously it
547 was only for check_local user. The change removes one item of manual
548 configuration which is required for proper retries when a remote router
549 handles a subset of addresses for a domain.
551 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
552 link count into consideration.
554 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
555 caused the extension of big_buffer, the following lines were ignored.
557 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
558 accordance with RFC 2308. Previously there was no expiry, so a longlived
559 receive process (eg. due to ACL delays) versus a short SOA value could
562 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
564 JH/39 Promote DMARC support to mainline.
566 JH/40 Bug 2452: Add a References: header to DSNs.
568 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
569 parameters. The relevant library call is documented as "Deprecated: This
570 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
571 3.6.0, DH parameters are negotiated following RFC7919."
573 HS/06 Change the default of dnssec_request_domains to "*"
575 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
576 carried on and emitted a BDAT command, even when PIPELINING was not
579 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
580 buffer was used for the filename, resulting in a trap when tainted
581 arguments (eg. $domain) were used.
583 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
584 recommended to avoid a possible server-load attack. The feature can be
585 re-enabled via the openssl_options main cofiguration option.
587 JH/45 local_scan API: documented the current smtp_printf() call. This changed
588 for version 4.90 - adding a "more data" boolean to the arguments.
589 Bumped the ABI version number also, this having been missed previously;
590 release versions 4.90 to 4.92.3 inclusive were effectively broken in
591 respect of usage of smtp_printf() by either local_scan code or libraries
592 accessed via the ${dlfunc } expansion item. Both will need coding
593 adjustment for any calls to smtp_printf() to match the new function
594 signature; a FALSE value for the new argument is always safe.
596 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
597 the file-offset (which the Linux syscall does, and exim expects); this
598 resulted in an indefinite loop.
600 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
601 to do ARC verification. The Authentication-Results: header line added
602 by the configuration then had no ARC item.
608 JH/01 Remove code calling the customisable local_scan function, unless a new
609 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
611 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
612 non-signal-safe functions being used.
614 JH/03 Bug 2269: When presented with a received message having a stupidly large
615 number of DKIM-Signature headers, disable DKIM verification to avoid
616 a resource-consumption attack. The limit is set at twenty.
618 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
619 report of oldest_pass in ${authres } in consequence, and separate out
620 some descriptions of reasons for verification fail.
622 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
623 files in the spool were present and unlocked. A queue-runner could spot
624 them, resulting in a duplicate delivery. Fix that by doing the unlock
625 after the unlink. Investigation by Tim Stewart. Take the opportunity to
626 add more error-checking on spoolfile handling while that code is being
629 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
630 No known attacks, no CVE, this is defensive hardening.
632 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
633 a queue-runner could start a delivery while other operations were ongoing.
634 Cutthrough delivery was a common victim, resulting in duplicate delivery.
635 Found and investigated by Tim Stewart. Fix by using the open message data
636 file handle rather than opening another, and not locally closing it (which
637 releases a lock) for that case, while creating the temporary .eml format
638 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
640 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
641 $sender_verify_failure/$recipient_verify_failure to "random".
643 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
646 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
647 Previously this would segfault.
649 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
652 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
653 like zero, since the resolver should be doing this for us, But we need one
654 as a CNAME but no MX presence gets the CNAME returned; we need to check
655 that doesn't point to an MX to declare it "no MX returned" rather than
656 "error, loop". A new main option is added so the older capability of
657 following some limited number of chain links is maintained.
659 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
661 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
662 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
663 now-deprecated earlier definitions used only specified the range up to TLS
664 1.2 (in the older-version library docs).
666 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
668 JH/15 Rework TLS client-side context management. Stop using a global, and
669 explicitly pass a context around. This enables future use of TLS for
670 connections to service-daemons (eg. malware scanning) while a client smtp
671 connection is using TLS; with cutthrough connections this is quite likely.
673 JH/16 Fix ARC verification to do AS checks in reverse order.
675 JH/17 Support a "tls" option on the ${readsocket } expansion item.
677 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
678 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
679 Previously the "utf8" would be re-prepended for every additional message.
681 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
682 Previously thery were accepted, resulting in issues when attempting to
683 forward messages to a non-supporting MTA.
685 PP/02 Let -n work with printing macros too, not just options.
687 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
688 one parent address was copied, and bogus data was used at delivery-logging
689 time. Either a crash (after delivery) or bogus log data could result.
690 Discovery and analysis by Tim Stewart.
692 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
693 Previously if the string ended mid-character, we did not insert the
694 promised '?' replacement.
696 PP/04 Documentation: current string operators work on bytes, not codepoints.
698 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
699 should pack well giving a smaller memory footprint so better caching and
700 therefore performance. Group the declarations where this can't be done so
701 that the byte-sized flag variables are not interspersed among pointer
702 variables, giving a better chance of good packing by the compiler.
704 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
705 non-null, to avoid issues with sites running BATV. Previously reports were
706 sent with an empty envelope sender so looked like bounces.
708 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
709 The ignore_error flag wasn't being returned from the filter subprocess so
710 was not set for later routers. Investigation and fix by Matthias Kurz.
712 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
713 and a msg:complete for the whole, when a message is manually removed using
714 -Mrm. Developement by Matthias Kurz, hacked on by JH.
716 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
717 a "Gnu special" function, asprintf() in the DB utility binary builds; I
718 hope that is portable enough.
720 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
721 requiring a known-CA anchor certificate; make it now rely entirely on the
722 TLSA as an anchor. Checking the name on the leaf cert against the name
723 on the A-record for the host is still done for TA (but not for EE mode).
725 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
726 included in delivery lines for non-proxied connections, when compiled with
727 SUPPORT_SOCKS and running with proxy logging enabled.
729 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
730 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
731 move the existing event to fire before the normal logging of message
732 failure so that custom logging is bracketed by normal logging.
734 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
735 msg:fail:internal event. Developement by Matthias Kurz.
737 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
738 far too small for todays use of crypto signatures stored there. Go all
739 the way to the max DNS message size of 64kB, even though this might be
740 overmuch for IOT constrained device use.
742 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
743 copy a string over itself. The library routine is documented as not
744 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
746 JH/32 For main options check_spool_space and check_inode_space, where the
747 platform supports 64b integers, support more than the previous 2^31 kB
748 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
749 the previous G, M, k.
751 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
752 $authenticated_fail_id variable on authentication failure. Previously
755 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
756 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
757 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
758 GNUTLS_SEC_PARAM_MEDIUM.
760 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
761 side. Previously we would continue as if no SNI had been received.
763 JH/36 Harden the handling of string-lists. When a list consisted of a sole
764 "<" character, which should be a list-separator specification, we walked
765 off past the nul-terimation.
767 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
768 causes) even when the retry time is not yet met. Previously they were
769 not, meaning that when (say) an account was over-quota and temp-rejecting,
770 and multiple senders' messages were queued, only one sender would get
771 notified on each configured delay_warning cycle.
773 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
775 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
776 cipher-suites, an error can be left on the stack even for a succeeding
777 accept; this results in impossible error messages when a later operation
780 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
781 return error codes indicating retry. Under TLS1.3 this becomes required.
783 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
784 it only wrote the new authenticators, resulting in a lack of tracking of
785 peer changes of ESMTP extensions until the next cache flush.
787 JH/41 Fix the loop reading a message header line to check for integer overflow,
788 and more-often against header_maxsize. Previously a crafted message could
789 induce a crash of the recive process; now the message is cleanly rejected.
791 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
792 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
798 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
799 When redis_servers is set to a list of > 1 element, and the Redis servers
800 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
801 case of MOVED into a DEFER case instead, thus moving the query onto the
802 next server in the list. For a cluster of N elements, all N servers must
803 be defined in redis_servers.
805 GF/02 Catch and remove uninitialized value warning in exiqsumm
806 Check for existence of @ARGV before looking at $ARGV[0]
808 JH/01 Replace the store_release() internal interface with store_newblock(),
809 which internalises the check required to safely use the old one, plus
810 the allocate and data copy operations duplicated in both (!) of the
811 extant use locations.
813 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
814 modifier. This matches the restriction on the commandline.
816 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
817 Previously only the last row was returned.
819 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
820 we assumed that tags in the header were well-formed, and parsed the
821 element content after inspecting only the first char of the tag.
822 Assumptions at that stage could crash the receive process on malformed
825 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
826 While running the DKIM ACL we operate on the Permanent memory pool so that
827 variables created with "set" persist to the DATA ACL. Also (at any time)
828 DNS lookups that fail create cache records using the Permanent pool. But
829 expansions release any allocations made on the current pool - so a dnsdb
830 lookup expansion done in the DKIM ACL releases the memory used for the
831 DNS negative-cache, and bad things result. Solution is to switch to the
832 Main pool for expansions.
833 While we're in that code, add checks on the DNS cache during store_reset,
834 active in the testsuite.
835 Problem spotted, and debugging aided, by Wolfgang Breyha.
837 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
838 When none of the hosts presented to a transport match an already-open
839 connection, close it and proceed with the list. Previously we would
840 queue the message. Spotted by Lena with Yahoo, probably involving
843 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
844 Previously a spurious "250 OK id=" response was appended to the proper
847 JH/08 The "support for" informational output now, which built with Content
848 Scanning support, has a line for the malware scanner interfaces compiled
849 in. Interface can be individually included or not at build time.
851 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
852 by the template makefile "src/EDITME". The "STREAM" support for an older
853 ClamAV interface method is removed.
855 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
856 rows affected is given instead).
858 JH/11 The runtime Berkeley DB library version is now additionally output by
859 "exim -d -bV". Previously only the compile-time version was shown.
861 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
862 SMTP connection. Previously, when one had more recipients than the
863 first, an abortive onward connection was made. Move to full support for
864 multiple onward connections in sequence, handling cutthrough connection
865 for all multi-message initiating connections.
867 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
868 routers. Previously, a multi-recipient message would fail to match the
869 onward-connection opened for the first recipient, and cause its closure.
871 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
872 a timeout on read on a GnuTLS initiating connection, resulting in the
873 initiating connection being dropped. This mattered most when the callout
874 was marked defer_ok. Fix to keep the two timeout-detection methods
877 JH/15 Relax results from ACL control request to enable cutthrough, in
878 unsupported situations, from error to silently (except under debug)
879 ignoring. This covers use with PRDR, frozen messages, queue-only and
882 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
884 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
885 metadata, resulting in a crash in free().
887 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
888 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
889 Broken also in d185889f4, with init system revamp.
891 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
892 Previously we abruptly closed the connection after reading a malware-
893 found indication; now we go on to read the "scan ok" response line,
896 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
897 ACL. Previously, a crash would result.
899 JH/19 Speed up macro lookups during configuration file read, by skipping non-
900 macro text after a replacement (previously it was only once per line) and
901 by skipping builtin macros when searching for an uppercase lead character.
903 JH/20 DANE support moved from Experimental to mainline. The Makefile control
904 for the build is renamed.
906 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
907 was allocated for every new TLS startup, meaning one per message. Fix
908 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
910 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
911 reported the original. Fix to report (as far as possible) the ACL
912 result replacing the original.
914 JH/23 Fix memory leak during multi-message connections using STARTTLS under
915 OpenSSL. Certificate information is loaded for every new TLS startup,
916 and the resources needed to be freed.
918 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
920 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
923 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
924 DATA response info to the (existing) per-recipient response info for
925 the "C=" log element. It can have useful tracking info from the
926 destination system. Patch from Simon Arlott.
928 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
929 length value. Previously this would segfault.
931 HS/02 Support Avast multiline protoocol, this allows passing flags to
932 newer versions of the scanner.
934 JH/28 Ensure that variables possibly set during message acceptance are marked
935 dead before release of memory in the daemon loop. This stops complaints
936 about them when the debug_store option is enabled. Discovered specifically
937 for sender_rate_period, but applies to a whole set of variables.
938 Do the same for the queue-runner and queue-list loops, for variables set
939 from spool message files. Do the same for the SMTP per-message loop, for
940 certain variables indirectly set in ACL operations.
942 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
943 as a multi-recipient message from a mailinglist manager). The coding had
944 an arbitrary cutoff number of characters while checking for more input;
945 enforced by writing a NUL into the buffer. This corrupted long / fast
946 input. The problem was exposed more widely when more pipelineing of SMTP
947 responses was introduced, and one Exim system was feeding another.
948 The symptom is log complaints of SMTP syntax error (NUL chars) on the
949 receiving system, and refused recipients seen by the sending system
950 (propating to people being dropped from mailing lists).
951 Discovered and pinpointed by David Carter.
953 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
954 replaced by the ${authresults } expansion.
956 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
958 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
959 allows proper process termination in container environments.
961 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
962 Previously the "final dot" had a newline after it; ensure it is CR,LF.
964 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
965 and "err_perm", deprecated since 4.83 when the RFC-defined words
966 "temperror" and "permerror" were introduced.
968 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
969 transport-filters or DKIM-signing. The restriction was lost in the
970 consolidation of verify-callout and delivery SMTP handling.
971 Extend the restriction to also cover ARC-signing.
973 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
974 in defer=pass mode supply a 450 to the initiator. Previously the message
977 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
978 tls_require_ciphers is used as before.
980 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
981 "pass_unscanned". Only tmpfails from the scanner are written to
982 the paniclog, as they may require admin intervention (permission
983 denied, license issues). Other scanner errors (like decompression
984 bombs) do not cause a paniclog entry.
986 JH/36 Fix reinitialisation of DKIM logging variable between messages.
987 Previously it was possible to log spurious information in receive log
990 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
991 triggered odd behaviour from Outlook Express clients.
993 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
996 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
997 since the IETF WG has not yet settled on that versus the original
998 "bare" representation.
1000 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1001 Previously the millisecond value corrupted the output.
1002 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1003 corrupted the output.
1009 JH/01 Rework error string handling in TLS interface so that the caller in
1010 more cases is responsible for logging. This permits library-sourced
1011 string to be attached to addresses during delivery, and collapses
1012 pairs of long lines into single ones.
1014 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1015 during configuration. Wildcards are allowed and expanded.
1017 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1018 This permits better logging.
1020 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1021 do not close it down and have the child transport start it up again on
1022 the passed-on TCP connection. Instead, proxy the child (and any
1023 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1024 affected: the continued delivery log lines do not have any DNSSEC, TLS
1025 Certificate or OCSP information. TLS cipher information is still logged.
1027 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1028 identical IP addresses on different listening ports. Will also affect
1031 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1032 add noisy ifdef guards to special-case this sillyness.
1033 Patch from Bernd Kuhls.
1035 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1036 than 255 are no longer allowed.
1038 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1039 Disable the session-cache too, which might reduce our load. Since we
1040 currrectly use a new context for every connection, both as server and
1041 client, there is no benefit for these.
1042 GnuTLS appears to not support tickets server-side by default (we don't
1043 call gnutls_session_ticket_enable_server()) but client side is enabled
1044 by default on recent versions (3.1.3 +) unless the PFS priority string
1047 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1048 <https://reproducible-builds.org/specs/source-date-epoch/>.
1050 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1051 the check for any unsuccessful recipients did not notice the limit, and
1052 erroneously found still-pending ones.
1054 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1055 MSG_MORE. Only in-clear (not on TLS connections).
1057 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1058 transport-filter is involved we can buffer the headers in memory for
1059 creating the signature, and read the spool data file once for the
1060 signature and again for transmission.
1062 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1063 4.77 as the kernel support then wasn't solid, having issues in 64bit
1064 mode. Now, it's been long enough. Add support for FreeBSD also.
1066 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1067 case where the routing stage had gathered several addresses to send to
1068 a host before calling the transport for the first, we previously failed
1069 to close down TLS in the old transport process before passing the TCP
1070 connection to the new process. The new one sent a STARTTLS command
1071 which naturally failed, giving a failed delivery and bloating the retry
1072 database. Investigation and fix prototype from Wolfgang Breyha.
1074 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1075 false-negatives in the check that the sender had not preempted a response
1076 or prompt from Exim (running as a server), due to that code's lack of
1077 awareness of the SMTP input buffering.
1079 PP/04 Add commandline_checks_require_admin option.
1080 Exim drops privileges sanely, various checks such as -be aren't a
1081 security problem, as long as you trust local users with access to their
1082 own account. When invoked by services which pass untrusted data to
1083 Exim, this might be an issue. Set this option in main configuration
1084 AND make fixes to the calling application, such as using `--` to stop
1087 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1088 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1089 responses to those, into a single TLS record each way (this usually means
1090 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1093 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1094 affects you only if you're dancing at the edge of the param size limits.
1095 If you are, and this message makes sense to you, then: raise the
1096 configured limit or use OpenSSL 1.1. Nothing we can do for older
1099 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1100 cmdline element to get the documented default one. Previously it was
1103 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1104 are unsuable for non-message-specific reasons - eg. connection timeout,
1105 banner-time rejection.
1107 JH/16 Fix logging of delivery remote port, when specified by router, under
1110 PP/06 Repair manualroute's ability to take options in any order, even if one
1111 is the name of a transport.
1114 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1116 JH/17 Change the list-building routines interface to use the expanding-string
1117 triplet model, for better allocation and copying behaviour.
1119 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1120 Previously it was constructed the first time a possibly-matching string
1121 was met in the configuration file input during startup; now it is done
1124 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1125 compatible one, to avoid the (poorly documented) possibility of a config
1126 file in the working directory redirecting the DB files, possibly correpting
1127 some existing file. CVE-2017-10140 assigned for BDB.
1129 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1130 cache-hot. Previously, although the result was properly cached, the
1131 initial verify call returned a defer.
1133 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1134 the main verify for receipient in uncached-mode.
1136 JH/22 Retire historical build files to an "unsupported" subdir. These are
1137 defined as "ones for which we have no current evidence of testing".
1139 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1140 if present. Previously it was ignored.
1142 JH/24 Start using specified-initialisers in C structure init coding. This is
1143 a C99 feature (it's 2017, so now considered safe).
1145 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1146 if was a fixed-sized field and bitmask ops via macros; it is now more
1149 PP/07 GitHub PR 56: Apply MariaDB build fix.
1150 Patch provided by Jaroslav Škarvada.
1152 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1153 during Coverity cleanups [4.87 JH/47]
1154 Diagnosis and fix provided by Michael Fischer v. Mollard.
1156 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1157 the right size to place the terminating semicolon on its own folded
1158 line, the header hash was calculated to an incorrect value thanks to
1159 the (relaxed) space the fold became.
1161 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1162 and confused the parent.
1164 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1165 which could crash as a result. This could lead to undeliverable messages.
1167 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1170 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1171 matters on fast-turnover and PID-randomising systems, which were getting
1172 out-of-order delivery.
1174 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1175 a possibly-overlapping copy. The symptom was that "Remote host closed
1176 connection in response to HELO" was logged instead of the actual 4xx
1179 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1180 Previously only that bufferd was discarded, resulting in SYMTP command
1183 JH/32 DKIM: when a message has multiple signatures matching an identity given
1184 in dkim_verify_signers, run the dkim acl once for each. Previously only
1185 one run was done. Bug 2189.
1187 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1188 "panic the current process" to "deliberately defer". The panic log is
1189 still written with the problem list name; the mail and reject logs now
1190 get a temp-reject line for the message that was being handled, saying
1191 something like "domains check lookup or other defer". The SMTP 451
1192 message is still "Temporary local problem".
1194 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1195 A crafted sequence of BDAT commands could result in in-use memory beeing
1196 freed. CVE-2017-16943.
1198 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1199 from SMTP input. Previously it was always done; now only done for DATA
1200 and not BDAT commands. CVE-2017-16944.
1202 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1203 to the message (such as an overlong header line). Previously this was
1204 not done and we did not exit BDAT mode. Followon from the previous item
1205 though a different problem.
1211 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1212 than -2003 did; needs libidn2 in addition to libidn.
1214 JH/02 The path option on a pipe transport is now expanded before use.
1216 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1217 Patch provided by "Björn", documentation fix added too.
1219 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1220 missing a wire-to-host endian conversion.
1222 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1223 close after a BDAT command line could be taken as a following command,
1224 giving a synch failure. Fix by only checking for synch immediately
1225 before acknowledging the chunk.
1227 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1228 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1229 macro. Patches provided by Josh Soref.
1231 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1232 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1233 by IANA as a service extension. However, John Klensin suggests that we
1236 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1237 the dkim code may be unix-mode line endings rather than smtp wire-format
1238 CRLF, so prepend a CR to any bare LF.
1240 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1241 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1243 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1244 the first were themselves being wrongly included in the feed into dkim
1245 processing; with most chunk sizes in use this resulted in an incorrect
1246 body hash calculated value.
1248 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1249 DKIM signature block, for verification. Although advised against by
1250 standards it is specifically not ruled illegal.
1252 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1254 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1255 missing a body hash (the bh= tag).
1257 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1258 It seems that HAProxy sends the Proxy Protocol information in clear and
1259 only then does a TLS startup, so do the same.
1261 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1262 TCP connections (such as for Spamd) unless the daemon successfully set
1263 Fast Open mode on its listening sockets. This fixes breakage seen on
1264 too-old kernels or those not configured for Fast Open, at the cost of
1265 requiring both directions being enabled for TFO, and TFO never being used
1266 by non-daemon-related Exim processes.
1268 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1269 endings, at least on the first header line. Try to canonify any that get
1270 past that check, despite the cost.
1272 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1273 now limited to an arbitrary five deep, while parsing addresses with the
1274 strip_excess_angle_brackets option enabled.
1276 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1277 instead leave the unprompted TLS handshake in socket buffer for the
1278 TLS library to consume.
1280 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1282 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1284 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1285 region in one operation, for speed, and this leaves assigned pointers
1286 dangling. Add checks run only under the testsuite which checks all
1287 variables at a store-reset and panics on a dangling pointer; add code
1288 explicitly nulling out all the variables discovered. Fixes one known
1289 bug: a transport crash, where a dangling pointer for $sending_ip_address
1290 originally assigned in a verify callout, is re-used.
1292 PP/06 Drop '.' from @INC in various Perl scripts.
1294 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1296 PP/08 Reduce a number of compilation warnings under clang; building with
1297 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1298 should be warning-free.
1300 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1302 HS/01 Fix portability problems introduced by PP/08 for platforms where
1303 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1305 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1306 chunk. This allows us to accept broken chunked messages. We need a more
1307 general solution here.
1309 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1310 already-broken messages in the queue.
1312 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1314 JH/19 Fix reference counting bug in routing-generated-address tracking.
1320 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1321 supports it and a size is available (ie. the sending peer gave us one).
1323 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1324 years of being deprecated). The replacements are the ACLs
1325 acl_smtp_mime and acl_not_smtp_mime.
1327 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1328 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1329 or even an in-clear connection were permitted. Now, if the host lookup
1330 was dnssec and dane was requested then the host is only used if the
1331 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1332 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1333 if one fails this test.
1334 This means that a poorly-configured remote DNS will make it incommunicado;
1335 but it protects against a DNS-interception attack on it.
1337 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1338 non-noisy when a race steals the message being considered.
1340 JH/05 If main configuration option tls_certificate is unset, generate a
1341 self-signed certificate for inbound TLS connections.
1343 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1344 in rewrites and routers.
1346 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1347 and logged a warning sing 4.83; now they are a configuration file error.
1349 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1350 (lacking @domain). Apply the same qualification processing as RCPT.
1352 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1354 JH/10 Support ${sha256:} applied to a string (as well as the previous
1357 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1358 a cutthrough deliver is pending, as we always want to make a connection.
1359 This also avoids re-routing the message when later placing the cutthrough
1360 connection after a verify cache hit.
1361 Do not update it with the verify result either.
1363 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1364 when routing results in more than one destination address.
1366 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1367 signing (which inhibits the cutthrough capability). Previously only
1368 the presence of an option was tested; now an expansion evaluating as
1369 empty is permissible (obviously it should depend only on data available
1370 when the cutthrough connection is made).
1372 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1373 the relevant preceding SMTP command did not note the pipelining mode.
1375 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1376 Previously they were not counted.
1378 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1379 as one having no matching records. Previously we deferred the message
1380 that needed the lookup.
1382 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1383 distinguished as "(=".
1385 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1386 for missing MX records. Previously it only worked for missing A records.
1388 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1390 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1391 after the data-go-ahead and data-ack. Patch from Jason Betts.
1393 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1394 even for a "none" policy. Patch from Tony Meyer.
1396 JH/22 Fix continued use of a connection for further deliveries. If a port was
1397 specified by a router, it must also match for the delivery to be
1400 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1401 When one of the recipients of a message was unsuitable for the connection
1402 (has no matching addresses), we lost track of needing to mark it
1403 deferred. As a result mail would be lost.
1405 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1407 JH/25 Decoding ACL controls is now done using a binary search; the source code
1408 takes up less space and should be simpler to maintain. Merge the ACL
1409 condition decode tables also, with similar effect.
1411 JH/26 Fix problem with one_time used on a redirect router which returned the
1412 parent address unchanged. A retry would see the parent address marked as
1413 delivered, so not attempt the (identical) child. As a result mail would
1416 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1417 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1418 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1421 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1422 Main config options check_{log,spool}_{inodes,space} are now
1423 100 inodes, 10MB unless set otherwise in the configuration.
1425 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1426 Previously it only applied to the main-section connection policy
1429 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1431 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1432 by me. Added RFC7919 DH primes as an alternative.
1434 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1435 are not in the system include path.
1437 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1438 GnuTLS, when a session startup failed (eg because the client disconnected)
1439 Exim did stdio operations after fclose. This was exposed by a recent
1440 change which nulled out the file handle after the fclose.
1442 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1443 signed directly by the cert-signing cert, rather than an intermediate
1444 OCSP-signing cert. This is the model used by LetsEncrypt.
1446 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1448 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1449 an incoming connection.
1451 HS/02 Bug 1802: Do not half-close the connection after sending a request
1454 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1455 fallback to "prime256v1".
1457 JH/34 SECURITY: Use proper copy of DATA command in error message.
1458 Could leak key material. Remotely exploitable. CVE-2016-9963.
1464 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1465 and 3.4.4 - once the server is enabled to respond to an OCSP request
1466 it does even when not requested, resulting in a stapling non-aware
1467 client dropping the TLS connection.
1469 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1470 support variable-length bit vectors. No functional change.
1472 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1473 The I= interface field on outgoing lines is now after the H= remote
1474 host field, same as incoming lines. There is a separate
1475 outgoing_interface log selector which allows you to disable the
1478 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1479 If not running log_selector +smtp_connection the mainlog would be held
1480 open indefinitely after a "too many connections" event, including to a
1481 deleted file after a log rotate. Leave the per net connection logging
1482 leaving it open for efficiency as that will be quickly detected by the
1483 check on the next write.
1485 HS/01 Bug 1671: Fix post transport crash.
1486 Processing the wait-<transport> messages could crash the delivery
1487 process if the message IDs didn't exist for some reason. When
1488 using 'split_spool_directory=yes' the construction of the spool
1489 file name failed already, exposing the same netto behaviour.
1491 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1492 mime_regex ACL conditions.
1494 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1495 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1496 to HELO, local diagnostic string.
1498 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1499 log line to debug. Even when configured with a tls authenticator many
1500 client connections are expected to not authenticate in this way, so
1501 an authenticate fail is not an error.
1503 HS/02 Add the Exim version string to the process info. This way exiwhat
1504 gives some more detail about the running daemon.
1506 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1507 matter for fast-change records such as DNSBLs.
1509 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1510 retry record, even if constant. There may be multiple transports with
1511 different interface settings and the retry behaviour needs to be kept
1514 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1516 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1518 JH/10 Bug 840: fix log_defer_output option of pipe transport
1520 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1521 hosts_randomize. This matters a lot when combined with mua_wrapper.
1523 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1524 ${quote_pgsql:<string>} operator.
1526 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1528 JH/14 Change header file name for freeradius-client. Relevant if compiling
1529 with Radius support; from the Gentoo tree and checked under Fedora.
1531 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1533 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1534 option transport_filter, meaning no filtering.
1536 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1538 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1539 defaults to "*" (all hosts). The variable is now available when not built
1540 with TLS, default unset, mainly to enable keeping the testsuite sane.
1541 If a server certificate is not supplied (via tls_certificate) an error is
1542 logged, and clients will find TLS connections fail on startup. Presumably
1543 they will retry in-clear.
1544 Packagers of Exim are strongly encouraged to create a server certificate
1545 at installation time.
1547 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1548 with the $config_file variable.
1550 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1551 in transport context, after the attempt, and per-recipient. The latter type
1552 is per host attempted. The event data is the error message, and the errno
1553 information encodes the lookup type (A vs. MX) used for the (first) host,
1554 and the trailing two digits of the smtp 4xx response.
1556 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1557 to write to mainlog (or rejectlog, paniclog) in the window between file
1558 creation and permissions/ownership being changed. Particularly affects
1559 installations where exicyclog is run as root, rather than exim user;
1560 result is that the running daemon panics and dies.
1562 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1564 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1565 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1566 "pri" and "weight". Note that the previous implicit priority given by the
1567 list order is no longer honoured.
1569 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1570 for DKIM processing.
1572 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1573 by defining SUPPORT_SOCKS.
1575 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1576 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1577 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1578 variables are renamed to proxy_{local,external}_{address,port}.
1580 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1581 for a build by defining SUPPORT_I18N
1583 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1584 of the query string, and make ${quote_redis:} do that quoting.
1586 JH/29 Move Events support from Experimental to mainline, enabled by default
1587 and removable for a build by defining DISABLE_EVENT.
1589 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1591 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1592 cached by the daemon.
1594 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1595 by defining LOOKUP_REDIS. The libhiredis library is required.
1597 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1598 keys are given for lookup.
1600 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1601 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1602 only supported when built with TLS support. The PolarSSL SHA routines
1603 are still used when the TLS library is too old for convenient support.
1605 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1606 openssl_options), for security. OpenSSL forces this from version 1.1.0
1607 server-side so match that on older versions.
1609 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1610 allocation for $value could be released as the expansion processing
1611 concluded, but leaving the global pointer active for it.
1613 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1614 and to use the domains and local_parts ACL conditions.
1616 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1617 incorrectly not doubled on cutthrough transmission, hence seen as a
1618 body-termination at the receiving system - resulting in truncated mails.
1619 Commonly the sender saw a TCP-level error, and retransmitted the message
1620 via the normal store-and-forward channel. This could result in duplicates
1621 received - but deduplicating mailstores were liable to retain only the
1622 initial truncated version.
1624 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1626 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1628 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1629 we're in there, support oversigning also; bug 1309.
1631 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1633 HS/04 Add support for keep_environment and add_environment options.
1635 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1636 either intentional arithmetic overflow during PRNG, or testing config-
1639 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1640 delivery resulted in actual delivery. Cancel cutthrough before DATA
1643 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1644 rejects a recipient: pass the reject to the originator.
1646 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1647 Many were false-positives and ignorable, but it's worth fixing the
1650 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1651 for the new environment-manipulation done at startup. Move the routines
1652 from being local to tls.c to being global via the os.c file.
1654 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1655 an extract embedded as result-arg for a map, the first arg for extract
1656 is unavailable so we cannot tell if this is a numbered or keyed
1657 extraction. Accept either.
1663 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1666 JH/02 The smtp transport option "multi_domain" is now expanded.
1668 JH/03 The smtp transport now requests PRDR by default, if the server offers
1671 JH/04 Certificate name checking on server certificates, when exim is a client,
1672 is now done by default. The transport option tls_verify_cert_hostnames
1673 can be used to disable this per-host. The build option
1674 EXPERIMENTAL_CERTNAMES is withdrawn.
1676 JH/05 The value of the tls_verify_certificates smtp transport and main options
1677 default to the word "system" to access the system default CA bundle.
1678 For GnuTLS, only version 3.0.20 or later.
1680 JH/06 Verification of the server certificate for a TLS connection is now tried
1681 (but not required) by default. The verification status is now logged by
1682 default, for both outbound TLS and client-certificate supplying inbound
1685 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1688 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1689 Status Notification (bounce) messages are now MIME format per RFC 3464.
1690 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1691 under the control of the dsn_advertise_hosts option, and routers may
1692 have a dsn_lasthop option.
1694 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1695 default, modifiable by a malware= option. The list separator for
1696 the options can now be changed in the usual way. Bug 68.
1698 JH/10 The smtp_receive_timeout main option is now expanded before use.
1700 JH/11 The incoming_interface log option now also enables logging of the
1701 local interface on delivery outgoing connections.
1703 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1704 if the interface and destination host and port all match.
1706 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1709 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1710 Patch from Andrew Lewis.
1712 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1713 now supports optional time-restrictions, weighting, and priority
1714 modifiers per server. Patch originally by <rommer@active.by>.
1716 JH/16 The spamd_address main option now supports a mixed list of local
1717 and remote servers. Remote servers can be IPv6 addresses, and
1718 specify a port-range.
1720 JH/17 Bug 68: The spamd_address main option now supports an optional
1721 timeout value per server.
1723 JH/18 Bug 1581: Router and transport options headers_add/remove can
1724 now have the list separator specified.
1726 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1729 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1732 JH/21 Support for the A6 type of dns record is withdrawn.
1734 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1735 rather than the verbs used.
1737 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1738 from 255 to 1024 chars.
1740 JH/24 Verification callouts now attempt to use TLS by default.
1742 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1743 are generic router options now. The defaults didn't change.
1745 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1746 Original patch from Alexander Shikoff, worked over by JH.
1748 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1749 files if ZCAT_COMMAND is not executable.
1751 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1753 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1755 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1756 Normally benign, it bites when the pair was led to by a CNAME;
1757 modern usage is to not canonicalize the domain to a CNAME target
1758 (and we were inconsistent anyway for A-only vs AAAA+A).
1760 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1762 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1763 when evaluating $sender_host_dnssec.
1765 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1766 $sender_helo_dnssec variable.
1768 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1770 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1772 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1774 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1775 documented as working, but never had. Support all but $spam_report.
1777 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1778 added for tls authenticator.
1780 HS/03 Add perl_taintmode main config option
1786 TL/01 When running the test suite, the README says that variables such as
1787 no_msglog_check are global and can be placed anywhere in a specific
1788 test's script, however it was observed that placement needed to be near
1789 the beginning for it to behave that way. Changed the runtest perl
1790 script to read through the entire script once to detect and set these
1791 variables, reset to the beginning of the script, and then run through
1792 the script parsing/test process like normal.
1794 TL/02 The BSD's have an arc4random API. One of the functions to induce
1795 adding randomness was arc4random_stir(), but it has been removed in
1796 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1797 function when detected.
1799 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1800 cause callback expansion.
1802 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1803 syntax errors in an expansion can be treated as a string instead of
1804 logging or causing an error, due to the internal use of bool_lax
1805 instead of bool when processing it.
1807 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1808 server certificates when making smtp deliveries.
1810 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1812 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1814 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1816 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1817 Merged patch from Sebastian Wiedenroth.
1819 JH/05 Fix results-pipe from transport process. Several recipients, combined
1820 with certificate use, exposed issues where response data items split
1821 over buffer boundaries were not parsed properly. This eventually
1822 resulted in duplicates being sent. This issue only became common enough
1823 to notice due to the introduction of connection certificate information,
1824 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1826 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1827 size buffer was used, resulting in syntax errors when an expansion
1830 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1831 version 3.3.6 or later.
1833 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1834 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1835 both become "event_action", the variables become $event_name, $event_data
1836 and $event_defer_errno. There is a new variable $verify_mode, usable in
1837 routers, transports and related events. The tls:cert event is now also
1838 raised for inbound connections, if the main configuration event_action
1841 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1842 early OCSP support, but no stapling (appears to be less than 1.0.0).
1844 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1845 server certificate names available under the smtp transport option
1846 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1849 JH/10 Time-related extraction expansions from certificates now use the main
1850 option "timezone" setting for output formatting, and are consistent
1851 between OpenSSL and GnuTLS compilations. Bug 1541.
1853 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1854 encoded parameter in the incoming message. Bug 1558.
1856 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1857 include certificate info, eximon was claiming there were spoolfile
1860 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1862 JH/14 Log delivery-related information more consistently, using the sequence
1863 "H=<name> [<ip>]" wherever possible.
1865 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1866 are problematic for Debian distribution, omit them from the release
1869 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1871 JH/16 Fix string representation of time values on 64bit time_t architectures.
1874 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1875 output list separator was used.
1880 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1881 checkers that were complaining about end of non-void function with no
1884 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1885 This was a regression introduced in 4.83 by another bugfix.
1887 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1889 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1890 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1896 TF/01 Correctly close the server side of TLS when forking for delivery.
1898 When a message was received over SMTP with TLS, Exim failed to clear up
1899 the incoming connection properly after forking off the child process to
1900 deliver the message. In some situations the subsequent outgoing
1901 delivery connection happened to have the same fd number as the incoming
1902 connection previously had. Exim would try to use TLS and fail, logging
1903 a "Bad file descriptor" error.
1905 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1906 utilities have not been installed.
1908 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1909 temporary space as the ACL may create new global variables.
1911 TL/01 LDAP support uses per connection or global context settings, depending
1912 upon the detected version of the libraries at build time.
1914 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1915 to extract and use the src ip:port in logging and expansions as if it
1916 were a direct connection from the outside internet. PPv2 support was
1917 updated based on HAProxy spec change in May 2014.
1919 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1921 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1922 Properly escape header and check for NULL return.
1924 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1927 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1929 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1930 characters in header names, implemented as a verify condition.
1931 Contributed by Michael Fischer v. Mollard.
1933 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1934 results permerror and temperror. Previous values are deprecated but
1935 still accepted. In a future release, err_perm and err_temp will be
1936 completely removed, which will be a backward incompatibility if the
1937 ACL tests for either of these two old results. Patch contributed by
1938 user bes-internal on the mailing list.
1940 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1942 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1943 selectors, in both main and reject logs.
1945 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1948 JH/07 Add malware type "sock" for talking to simple daemon.
1950 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1952 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1953 routers/transports under cutthrough routing.
1955 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1956 numbers. Touch up "bool" conditional to keep the same definition.
1958 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1960 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1962 JH/12 Expand items in router/transport headers_add or headers_remove lists
1963 individually rather than the list as a whole. Bug 1452.
1965 Required for reasonable handling of multiple headers_ options when
1966 they may be empty; requires that headers_remove items with embedded
1967 colons must have them doubled (or the list-separator changed).
1969 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1970 view the policy declared in the DMARC record. Currently, $dmarc_status
1971 is a combined value of both the record presence and the result of the
1974 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1976 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1977 dnslookup router and the smtp transport (applying to the forward
1980 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1981 of ldap servers used for a specific lookup. Patch provided by Heiko
1984 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1985 New variable $lookup_dnssec_authenticated for observability.
1987 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1988 Patch submitted by Lars Timman.
1990 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1992 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1993 Requires trusted mode and valid format message id, aborts otherwise.
1994 Patch contributed by Heiko Schlichting.
1996 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1997 certextract with support for various fields. Bug 1358.
1999 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2000 is requested by default, modifiable by smtp transport option
2003 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2004 operate on certificate variables to give certificate fingerprints
2005 Also new ${sha256:cert_variable}.
2007 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2009 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2010 Christian Aistleitner.
2012 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2014 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2015 file. Patch from Wolfgang Breyha.
2017 JH/25 Expand the coverage of the delivery $host and $host_address to
2018 client authenticators run in verify callout. Bug 1476.
2020 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2021 align with daemon_smtp_ports. Bug 72.
2023 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2024 support and error reporting did not work properly.
2026 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2027 and is readable. Patch from Andrew Colin Kissa.
2029 TL/14 Enhance documentation of ${run expansion and how it parses the
2030 commandline after expansion, particularly in the case when an
2031 unquoted variable expansion results in an empty value.
2033 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2035 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2036 by renaming away from it. Addresses GH issue 15, reported by
2039 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2041 TL/15 SECURITY: prevent double expansion in math comparison functions
2042 (can expand unsanitized data). Not remotely exploitable.
2049 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2051 PP/02 Make -n do something, by making it not do something.
2052 When combined with -bP, the name of an option is not output.
2054 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2057 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2058 $sender_host_name and config options to manage this, and basic check
2061 PP/05 DSCP support for outbound connections and control modifier for inbound.
2063 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2064 (Only plugin which currently uses this is kerberos4, which nobody should
2065 be using, but we should make it available and other future plugins might
2066 conceivably use it, even though it would break NAT; stuff *should* be
2067 using channel bindings instead).
2069 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2070 name; added for Sendmail compatibility; requires admin caller.
2071 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2072 just ignore it); requires trusted caller.
2073 Also parse but ignore: -Ac -Am -X<logfile>
2076 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2078 TL/02 Add +smtp_confirmation as a default logging option.
2080 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2081 Patch by Magnus Holmgren from 2007-02-20.
2083 TL/04 Bugzilla 1281 - Spec typo.
2084 Bugzilla 1283 - Spec typo.
2085 Bugzilla 1290 - Spec grammar fixes.
2087 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2089 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2091 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2092 mailing list by by Dmitry Isaikin.
2094 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2096 JH/02 Support "G" suffix to numbers in ${if comparisons.
2098 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2100 NM/01 Bugzilla 1197 - Spec typo
2101 Bugzilla 1196 - Spec examples corrections
2103 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2105 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2106 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2109 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2110 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2112 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2113 "acl {{name}{arg}...}", and optional args on acl condition
2116 JH/05 Permit multiple router/transport headers_add/remove lines.
2118 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2120 JH/07 Avoid using a waiting database for a single-message-only transport.
2121 Performance patch from Paul Fisher. Bugzilla 1262.
2123 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2126 JH/09 Add $headers_added variable, with content from use of ACL modifier
2127 add_header (but not yet added to the message). Bugzilla 199.
2129 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2130 Pulled from Bugzilla 817 by Wolfgang Breyha.
2132 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2134 (nb: this is the same fix as in Exim 4.80.1)
2136 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2139 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2142 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2144 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2145 advertise SMTP AUTH mechanism to us, instead of a generic
2146 protocol violation error. Also, make Exim more robust to bad
2147 data from the Dovecot auth socket.
2149 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2151 When a queue runner is handling a message, Exim first routes the
2152 recipient addresses, during which it prunes them based on the retry
2153 hints database. After that it attempts to deliver the message to
2154 any remaining recipients. It then updates the hints database using
2157 So if a recipient address works intermittently, it can get repeatedly
2158 deferred at routing time. The retry hints record remains fresh so the
2159 address never reaches the final cutoff time.
2161 This is a fairly common occurrence when a user is bumping up against
2162 their storage quota. Exim had some logic in its local delivery code
2163 to deal with this. However it did not apply to per-recipient defers
2164 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2166 This change adds a proper retry rule check during routing so that the
2167 final cutoff time is checked against the message's age. We only do
2168 this check if there is an address retry record and there is not a
2169 domain retry record; this implies that previous attempts to handle
2170 the address had the retry_use_local_parts option turned on. We use
2171 this as an approximation for the destination being like a local
2172 delivery, as in LMTP.
2174 I suspect this new check makes the old local delivery cutoff check
2175 redundant, but I have not verified this so I left the code in place.
2177 TF/02 Correct gecos expansion when From: is a prefix of the username.
2179 Test 0254 submits a message to Exim with the header
2183 When I ran the test suite under the user fanf2, Exim expanded
2184 the header to contain my full name, whereas it should have added
2185 a Resent-Sender: header. It erroneously treats any prefix of the
2186 username as equal to the username.
2188 This change corrects that bug.
2190 GF/01 DCC debug and logging tidyup
2191 Error conditions log to paniclog rather than rejectlog.
2192 Debug lines prefixed by "DCC: " to remove any ambiguity.
2194 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2196 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2197 Bug spotted by Jeremy Harris; was flawed since initial commit.
2198 Would have resulted in OCSP responses post-SNI triggering an Exim
2199 NULL dereference and crash.
2201 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2203 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2204 Bug detection, analysis and fix by Samuel Thibault.
2205 Bugzilla 1331, Debian bug #698092.
2207 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2209 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2210 Server implementation by Todd Lyons, client by JH.
2211 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2212 config variable "prdr_enable" controls whether the server
2213 advertises the facility. If the client requests PRDR a new
2214 acl_data_smtp_prdr ACL is called once for each recipient, after
2215 the body content is received and before the acl_smtp_data ACL.
2216 The client is controlled by both of: a hosts_try_prdr option
2217 on the smtp transport, and the server advertisement.
2218 Default client logging of deliveries and rejections involving
2219 PRDR are flagged with the string "PRDR".
2221 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2222 fclose(). Diagnosis by Todd Lyons.
2224 PP/17 Update configure.default to handle IPv6 localhost better.
2225 Patch by Alain Williams (plus minor tweaks).
2228 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2229 This is now consistent with GnuTLS, and is now documented: the
2230 previous undocumented portable approach to treating the option as
2231 unset was to force an expansion failure. That still works, and
2232 an empty string is now equivalent.
2234 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2235 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2236 not performing validation itself.
2238 PP/20 Added force_command boolean option to pipe transport.
2239 Patch from Nick Koston, of cPanel Inc.
2241 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2244 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2246 PP/21 Fix eximon continuous updating with timestamped log-files.
2247 Broken in a format-string cleanup in 4.80, missed when I repaired the
2248 other false fix of the same issue.
2249 Report and fix from Heiko Schlichting.
2252 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2253 Report from Prashanth Katuri.
2255 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2256 It's SecureTransport, so affects any MacOS clients which use the
2257 system-integrated TLS libraries, including email clients.
2259 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2260 using a MIME ACL for non-SMTP local injection.
2261 Report and assistance in diagnosis by Warren Baker.
2263 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2265 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2267 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2268 last id that failed so it may be referenced in subsequent ACL's.
2270 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2273 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2274 ldap library initialization, allowing self-signed CA's to be
2275 used. Also properly sets require_cert option later in code by
2276 using NULL (global ldap config) instead of ldap handle (per
2277 session). Bug diagnosis and testing by alxgomz.
2279 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2280 the src/util/ subdirectory.
2282 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2283 renamed to Transport Post Delivery Action by Jeremy Harris, as
2286 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2287 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2288 redis_servers = needs to be configured which will be used by the redis
2289 lookup. Patch from Warren Baker, of The Packet Hub.
2291 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2293 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2294 hostname or reverse DNS when processing a host list. Used suggestions
2295 from multiple comments on this bug.
2297 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2299 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2300 Missed a few lines, added it to make the runtest require no keyboard
2303 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2304 contains upper case chars. Make router use caseful_local_part.
2306 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2307 support when GnuTLS has been built with p11-kit.
2313 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2315 This, or similar/improved, will also be change PP/11 of 4.82.
2321 PP/01 Handle short writes when writing local log-files.
2322 In practice, only affects FreeBSD (8 onwards).
2323 Bugzilla 1053, with thanks to Dmitry Isaikin.
2325 NM/01 Bugzilla 949 - Documentation tweak
2327 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2330 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2332 PP/02 Implemented gsasl authenticator.
2334 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2336 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2337 `pkg-config foo` for cflags/libs.
2339 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2340 with rest of GSASL and with heimdal_gssapi.
2342 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2343 `pkg-config foo` for cflags/libs for the TLS implementation.
2345 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2346 properties get this fed in as external SSF. A number of robustness
2347 and debugging improvements to the cyrus_sasl authenticator.
2349 PP/08 cyrus_sasl server now expands the server_realm option.
2351 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2352 Patch by Jeremy Harris.
2354 PP/10 Added dbmjz lookup type.
2356 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2358 PP/12 MAIL args handles TAB as well as SP, for better interop with
2359 non-compliant senders.
2360 Analysis and variant patch by Todd Lyons.
2362 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2363 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2364 Patch from Dirk Mueller <dmueller@suse.com>
2366 PP/13 tls_peerdn now print-escaped for spool files.
2367 Observed some $tls_peerdn in wild which contained \n, which resulted
2368 in spool file corruption.
2370 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2371 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2372 or write after TLS renegotiation, which otherwise led to messages
2375 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2376 as a tracking header (ie: a signed header comes before the signature).
2377 Patch from Wolfgang Breyha.
2379 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2380 comma-sep list; embedded commas doubled.
2382 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2384 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2386 Report and patch from Dmitry Banschikov.
2388 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2389 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2390 failures. We appear to now support TLS1.1+ with Exim.
2392 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2393 lets Exim select keys and certificates based upon TLS SNI from client.
2394 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2395 before an outbound SMTP session. New log_selector, +tls_sni.
2397 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2398 NULL dereference. Report and patch from Alun Jones.
2400 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2401 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2402 Not seeing resolver debug output on NetBSD, but suspect this is a
2403 resolver implementation change.
2405 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2406 Left warnings. Added "eximon gdb" invocation mode.
2408 PP/21 Defaulting "accept_8bitmime" to true, not false.
2410 PP/22 Added -bw for inetd wait mode support.
2412 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2413 locate the relevant includes and libraries. Made this the default.
2415 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2416 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2418 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2419 This may cause build issues on older platforms.
2421 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2422 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2423 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2424 Added SNI support via GnuTLS too.
2425 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2427 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2429 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2430 Applied second patch from Janne, implementing suggestion to default
2431 multiple-strings-in-record handling to match SPF spec.
2433 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2435 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2436 read-only, out of scope).
2437 Patch from Wolfgang Breyha, report from Stuart Northfield.
2439 PP/29 Fix three issues highlighted by clang analyser static analysis.
2440 Only crash-plausible issue would require the Cambridge-specific
2441 iplookup router and a misconfiguration.
2442 Report from Marcin Mirosław.
2444 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2446 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2447 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2448 As part of this, removing so much warning spew let me fix some minor
2449 real issues in debug logging.
2451 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2452 assignment on my part. Fixed.
2454 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2455 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2456 Janne Snabb (who went above and beyond: thank you).
2458 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2459 string otherwise requires a connection and a bunch more work and it's
2460 relatively easy to get wrong. Should also expose TLS library linkage
2463 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2464 64-bit ${eval} (JH/03).
2466 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2467 GNU libc to support some of the 64-bit stuff, should not lead to
2468 conflicts. Defined before os.h is pulled in, so if a given platform
2469 needs to override this, it can.
2471 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2472 protection layer was required, which is not implemented.
2473 Bugzilla 1254, patch from Wolfgang Breyha.
2475 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2476 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2477 tls_dhparam take prime identifiers. Also unbreak combination of
2478 OpenSSL+DH_params+TLSSNI.
2480 PP/39 Disable SSLv2 by default in OpenSSL support.
2486 PP/01 Solaris build fix for Oracle's LDAP libraries.
2487 Bugzilla 1109, patch from Stephen Usher.
2489 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2491 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2494 TF/02 Fix a couple more cases where we did not log the error message
2495 when unlink() failed. See also change 4.74-TF/03.
2497 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2498 lock up or crash if it happened to be inside a call to libc when it
2499 got a SIGUSR1 from exiwhat.
2501 The SIGUSR1 handler appends the current process status to the process
2502 log which is later printed by exiwhat. It used to use the general
2503 purpose logging code to do this, but several functions it calls are
2504 not safe for signals.
2506 The new output code in the SIGUSR1 handler is specific to the process
2507 log, and simple enough that it's easy to inspect for signal safety.
2508 Removing some special cases also simplifies the general logging code.
2509 Removing the spurious timestamps from the process log simplifies
2512 TF/04 Improved ratelimit ACL condition.
2514 The /noupdate option has been deprecated in favour of /readonly which
2515 has clearer semantics. The /leaky, /strict, and /readonly update modes
2516 are mutually exclusive. The update mode is no longer included in the
2517 database key; it just determines when the database is updated. (This
2518 means that when you upgrade Exim will forget old rate measurements.)
2520 Exim now checks that the per_* options are used with an update mode that
2521 makes sense for the current ACL. For example, when Exim is processing a
2522 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2523 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2524 must specify per_mail/readonly. If you omit the update mode it defaults to
2525 /leaky where that makes sense (as before) or /readonly where required.
2527 The /noupdate option is now undocumented but still supported for
2528 backwards compatibility. It is equivalent to /readonly except that in
2529 ACLs where /readonly is required you may specify /leaky/noupdate or
2530 /strict/noupdate which are treated the same as /readonly.
2532 A useful new feature is the /count= option. This is a generalization
2533 of the per_byte option, so that you can measure the throughput of other
2534 aggregate values. For example, the per_byte option is now equivalent
2535 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2537 The per_rcpt option has been generalized using the /count= mechanism
2538 (though it's more complicated than the per_byte equivalence). When it is
2539 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2540 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2541 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2542 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2543 Note that using per_rcpt with a non-readonly update mode in more than
2544 one ACL will cause the recipients to be double-counted. (The per_mail
2545 and per_byte options don't have this problem.)
2547 The handling of very low rates has changed slightly. If the computed rate
2548 is less than the event's count (usually one) then this event is the first
2549 after a long gap. In this case the rate is set to the same as this event's
2550 count, so that the first message of a spam run is counted properly.
2552 The major new feature is a mechanism for counting the rate of unique
2553 events. The new per_addr option counts the number of different
2554 recipients that someone has sent messages to in the last time period. It
2555 behaves like per_rcpt if all the recipient addresses are different, but
2556 duplicate recipient addresses do not increase the measured rate. Like
2557 the /count= option this is a general mechanism, so the per_addr option
2558 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2559 example, measure the rate that a client uses different sender addresses
2560 with the options per_mail/unique=$sender_address. There are further
2561 details in the main documentation.
2563 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2565 TF/06 Removed a few PCRE remnants.
2567 TF/07 Automatically extract Exim's version number from tags in the git
2568 repository when doing development or release builds.
2570 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2571 Bugzilla 879. Patch from Paul Fisher.
2573 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2574 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2577 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2579 PP/05 Variable $av_failed, true if the AV scanner deferred.
2580 Bugzilla 1078. Patch from John Horne.
2582 PP/06 Stop make process more reliably on build failure.
2583 Bugzilla 1087. Patch from Heiko Schlittermann.
2585 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2586 Bugzilla 1089. Patch from Heiko Schlittermann.
2588 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2589 Bugzilla 1131. Patch from Holger Weiß.
2591 PP/09 Handle IPv6 addresses with SPF.
2592 Bugzilla 860. Patch from Wolfgang Breyha.
2594 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2596 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2599 PP/11 match_* no longer expand right-hand-side by default.
2600 New compile-time build option, EXPAND_LISTMATCH_RHS.
2601 New expansion conditions, "inlist", "inlisti".
2603 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2605 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2607 PP/14 fix log_write() format string regression from TF/03.
2608 Bugzilla 1152. Patch from Dmitry Isaikin.
2614 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2616 PP/02 Harmonised TLS library version reporting; only show if debugging.
2617 Layout now matches that introduced for other libraries in 4.74 PP/03.
2619 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2621 PP/04 New "dns_use_edns0" global option.
2623 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2626 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2627 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2629 TK/01 Updated PolarSSL code to 0.14.2.
2630 Bugzilla 1097. Patch from Andreas Metzler.
2632 PP/07 Catch divide-by-zero in ${eval:...}.
2633 Fixes bugzilla 1102.
2635 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2638 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2639 format-string attack -- SECURITY: remote arbitrary code execution.
2641 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2642 time unintentionally subject to list matching rules, letting the header
2643 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2644 arbitrary string expansion). This allowed for information disclosure.
2646 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2647 INT_MIN/-1 -- value coerced to INT_MAX.
2653 NM/01 Workaround for PCRE version dependency in version reporting
2656 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2657 This fixes portability to compilers other than gcc, notably
2658 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2660 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2661 makefiles for portability to HP-UX and POSIX correctness.
2663 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2664 Also via indented variable definition in the Makefile.
2665 (Debugging by Oliver Heesakkers).
2667 PP/02 Restore caching of spamd results with expanded spamd_address.
2668 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2670 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2671 Improves build reliability. Fix from: Frank Elsner
2673 NM/02 Fix wide character breakage in the rfc2047 coding
2674 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2676 NM/03 Allow underscore in dnslist lookups
2677 Fixes bug 1026. Patch from Graeme Fowler
2679 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2680 Code patches from Adam Ciarcinski of NetBSD.
2682 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2685 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2686 is logged, to avoid truncation. Patch from John Horne.
2688 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2689 Patch from Jakob Hirsch.
2691 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2692 SQL string expansion failure details.
2693 Patch from Andrey Oktyabrski.
2695 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2696 Patch from Simon Arlott.
2698 PP/09 New lookups functionality failed to compile on old gcc which rejects
2699 extern declarations in function scope.
2700 Patch from Oliver Fleischmann
2702 PP/10 Use sig_atomic_t for flags set from signal handlers.
2703 Check getgroups() return and improve debugging.
2704 Fixed developed for diagnosis in bug 927 (which turned out to be
2707 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2708 Patch from Mark Zealey.
2710 PP/12 Bugzilla 1056: Improved spamd server selection.
2711 Patch from Mark Zealey.
2713 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2714 Based on patch from Heiko Schlittermann.
2716 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2717 Patch from Uwe Doering, sign-off by Michael Haardt.
2719 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2720 variable declaration deep within a block. Bug and patch from
2723 PP/15 lookups-Makefile IRIX compatibility coercion.
2725 PP/16 Make DISABLE_DKIM build knob functional.
2727 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2728 Patch by Simon Arlott
2730 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2731 variable argument macros. Our copy now differs from upstream.
2737 TF/01 Failure to get a lock on a hints database can have serious
2738 consequences so log it to the panic log.
2740 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2741 controlled using the smtp_confirmation log selector.
2743 TF/03 Include the error message when we fail to unlink a spool file.
2745 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2746 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2747 for maintaining out-of-tree patches for some time.
2749 PP/01 Bugzilla 139: Documentation and portability issues.
2750 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2751 Handle per-OS dynamic-module compilation flags.
2753 PP/02 Let /dev/null have normal permissions.
2754 The 4.73 fixes were a little too stringent and complained about the
2755 permissions on /dev/null. Exempt it from some checks.
2756 Reported by Andreas M. Kirchwitz.
2758 PP/03 Report version information for many libraries, including
2759 Exim version information for dynamically loaded libraries. Created
2760 version.h, now support a version extension string for distributors
2761 who patch heavily. Dynamic module ABI change.
2763 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2764 privilege escalation vulnerability whereby the Exim run-time user
2765 can cause root to append content of the attacker's choosing to
2768 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2771 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2772 If dropping privileges for untrusted macros, we disabled normal logging
2773 on the basis that it would fail; for the Exim run-time user, this is not
2774 the case, and it resulted in successful deliveries going unlogged.
2775 Fixed. Reported by Andreas Metzler.
2781 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2782 only prepend for the Resent-* case. Fixes regression introduced in
2783 Exim 4.70 by NM/22 for Bugzilla 607.
2785 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2786 increasing numbers of administrators be bitten by this.
2788 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2790 PP/03 Bugzilla 994: added openssl_options main configuration option.
2792 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2794 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2796 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2798 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2799 without a peer certificate, leading to a segfault because of an
2800 assumption that peers always have certificates. Be a little more
2801 paranoid. Problem reported by Martin Tscholak.
2803 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2804 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2805 NB: ClamAV planning to remove STREAM in "middle of 2010".
2806 CL also introduces -bmalware, various -d+acl logging additions and
2807 more caution in buffer sizes.
2809 PP/09 Implemented reverse_ip expansion operator.
2811 PP/10 Bugzilla 937: provide a "debug" ACL control.
2813 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2815 PP/12 Bugzilla 973: Implement --version.
2817 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2819 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2821 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2823 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2824 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2825 ignore trailing whitespace.
2827 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2829 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2832 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2833 Notification from Dr Andrew Aitchison.
2835 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2836 ExtendedDetectionInfo response format.
2837 Notification from John Horne.
2839 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2842 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2843 XSL and documented dependency on system catalogs, with examples of how
2846 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2849 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2850 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2851 configuration file which is writeable by the Exim user or group.
2853 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2854 of configuration files to cover files specified with the -C option if
2855 they are going to be used with root privileges, not just the default
2858 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2859 option (effectively making it always true).
2861 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2862 files to be used while preserving root privileges.
2864 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2865 that rogue child processes cannot use them.
2867 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2868 run-time user, instead of root.
2870 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2871 Exim run-time user without dropping privileges.
2873 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2874 result string, instead of calling string_vformat() twice with the same
2877 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2878 for other users. Others should always drop root privileges if they use
2879 -C on the command line, even for a whitelisted configure file.
2881 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2883 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2889 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2890 $data_path, and $header_path variables; fixed documentation bugs and
2893 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2894 exipick to access non-standard spools, including the "frozen" queue
2897 NM/01 Bugzilla 965: Support mysql stored procedures.
2898 Patch from Alain Williams
2900 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2902 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2903 Patch from Andreas Metzler
2905 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2906 Patch from Kirill Miazine
2908 NM/05 Bugzilla 671: Added umask to procmail example.
2910 JJ/03 installed exipick 20100323.0, fixing doc bug
2912 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2913 directory. Notification and patch from Dan Rosenberg.
2915 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2917 TK/02 Improve log output when DKIM signing operation fails.
2919 MH/01 Treat the transport option dkim_domain as a colon separated
2920 list, not as a single string, and sign the message with each element,
2921 omitting multiple occurences of the same signer.
2923 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2924 Bugzilla 985, 986. Patch by Simon Arlott
2926 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2927 Patch by Simon Arlott
2929 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2930 MBX locking. Notification from Dan Rosenberg.
2936 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2938 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2940 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2942 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2944 NM/04 Bugzilla 915: Flag broken perl installation during build.
2950 TK/01 Added patch by Johannes Berg that expands the main option
2951 "spamd_address" if it starts with a dollar sign.
2953 TK/02 Write list of recipients to X-Envelope-Sender header when building
2954 the mbox-format spool file for content scanning (suggested by Jakob
2957 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2958 (http://www.dcc-servers.net/) support via dccifd. Activated by
2959 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2961 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2962 by Mark Daniel Reidel <mr@df.eu>.
2964 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2965 When building exim an external PCRE library is now needed -
2966 PCRE is a system library on the majority of modern systems.
2967 See entry on PCRE_LIBS in EDITME file.
2969 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2970 conversation. Added nologin parameter to request.
2971 Patch contributed by Kirill Miazine.
2973 TF/01 Do not log submission mode rewrites if they do not change the address.
2975 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2977 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2978 log files in place. Contributed by Roberto Lima.
2980 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2982 TF/03 Bugzilla 615: When checking the local_parts router precondition
2983 after a local_part_suffix or local_part_prefix option, Exim now
2984 does not use the address's named list lookup cache, since this
2985 contains cached lookups for the whole local part.
2987 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2988 Robert Millan. Documentation is in experimental-spec.txt.
2990 TF/04 Bugzilla 668: Fix parallel build (make -j).
2992 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2994 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2995 Patch provided by Jan Srzednicki.
2997 TF/05 Leading white space used to be stripped from $spam_report which
2998 wrecked the formatting. Now it is preserved.
3000 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3001 that they are available at delivery time.
3003 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3005 TF/08 TLS error reporting now respects the incoming_interface and
3006 incoming_port log selectors.
3008 TF/09 Produce a more useful error message if an SMTP transport's hosts
3009 setting expands to an empty string.
3011 NM/06 Bugzilla 744: EXPN did not work under TLS.
3012 Patch provided by Phil Pennock.
3014 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3015 Patch provided by Richard Godbee.
3017 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3018 acl_smtp_notquit, added index entry.
3020 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3021 Patch provided by Eugene Bujak.
3023 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3024 accept(). Patch provided by Maxim Dounin.
3026 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3027 Patch provided by Phil Pennock.
3029 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3031 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3032 Patch provided by Brad "anomie" Jorsch.
3034 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3035 Patch provided by Dean Brooks.
3037 TK/05 Add native DKIM support (does not depend on external libraries).
3039 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3040 Patch provided by Graeme Fowler.
3042 NM/16 Bugzilla 851: Documentation example syntax fix.
3044 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3046 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3049 NM/19 Bugzilla 745: TLS version reporting.
3050 Patch provided by Phil Pennock.
3052 NM/20 Bugzilla 167: bool: condition support.
3053 Patch provided by Phil Pennock.
3055 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3056 clients. Patch provided by Phil Pennock.
3058 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3059 Patch provided by Brad "anomie" Jorsch.
3061 NM/23 Bugzilla 687: Fix misparses in eximstats.
3062 Patch provided by Heiko Schlittermann.
3064 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3065 Patch provided by Heiko Schlittermann.
3067 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3068 plus update to original patch.
3070 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3072 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3073 Patch provided by David Brownlee.
3075 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3077 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3079 NM/30 Bugzilla 888: TLS documentation bugfixes.
3081 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3083 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3084 Unlike the original bugzilla I have changed all shell scripts in src tree.
3086 NM/33 Bugzilla 898: Transport filter timeout fix.
3087 Patch by Todd Rinaldo.
3089 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3090 Patch by Serge Demonchaux.
3092 NM/35 Bugzilla 39: Base64 decode bug fixes.
3093 Patch by Jakob Hirsch.
3095 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3097 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3099 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3101 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3107 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3108 ALT-N's libdkim that I have put here:
3109 http://duncanthrax.net/exim-experimental/
3111 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3112 were called 'true' and it seems that C99 defines that as a reserved
3113 keyword to be used with 'bool' variable types. That means you could
3114 not include C99-style headers which use bools without triggering
3115 build errors in sieve.c.
3117 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3118 as mailq or other aliases. Changed the --help handling significantly
3119 to do whats expected. exim_usage() emits usage/help information.
3121 SC/01 Added the -bylocaldomain option to eximstats.
3123 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3125 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3127 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3133 PH/01 Another patch from the Sieve maintainer.
3135 PH/02 When an IPv6 address is converted to a string for single-key lookup
3136 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3137 dots are used instead of colons so that keys in lsearch files need not
3138 contain colons. This was done some time before quoting was made available
3139 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3140 (notated using the quote facility) so as to distinguish them from IPv4
3141 keys. This meant that lookups for IP addresses in host lists did not work
3142 for iplsearch lookups.
3144 This has been fixed by arranging for IPv6 addresses to be expressed with
3145 colons if the lookup type is iplsearch. This is not incompatible, because
3146 previously such lookups could never work.
3148 The situation is now rather anomalous, since one *can* have colons in
3149 ordinary lsearch keys. However, making the change in all cases is
3150 incompatible and would probably break a number of configurations.
3152 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3155 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3156 conversion specification without a maximum field width, thereby enabling
3157 a rogue spamd server to cause a buffer overflow. While nobody in their
3158 right mind would setup Exim to query an untrusted spamd server, an
3159 attacker that gains access to a server running spamd could potentially
3160 exploit this vulnerability to run arbitrary code as the Exim user.
3162 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3163 $primary_hostname instead of what libspf2 thinks the hosts name is.
3165 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3166 a directory entry by the name of the lookup key. Previously, if a
3167 symlink pointed to a non-existing file or a file in a directory that
3168 Exim lacked permissions to read, a lookup for a key matching that
3169 symlink would fail. Now it is enough that a matching directory entry
3170 exists, symlink or not. (Bugzilla 503.)
3172 PH/03 The body_linecount and body_zerocount variables are now exported in the
3175 PH/04 Added the $dnslist_matched variable.
3177 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3178 This means they are set thereafter only if the connection becomes
3181 PH/06 Added the client_condition to authenticators so that some can be skipped
3182 by clients under certain conditions.
3184 PH/07 The error message for a badly-placed control=no_multiline_responses left
3185 "_responses" off the end of the name.
3187 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3189 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3190 (without spaces) instead of just copying the configuration text.
3192 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3194 PH/11 Added $max_received_linelength.
3196 PH/12 Added +ignore_defer and +include_defer to host lists.
3198 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3199 way in which PCRE > 7.0 is built.
3201 PH/14 Implemented queue_only_load_latch.
3203 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3204 MAIL command. The effect was to mangle the value on 64-bit systems.
3206 PH/16 Another patch from the Sieve maintainer.
3208 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3210 PH/18 If a system quota error occurred while trying to create the file for
3211 a maildir delivery, the message "Mailbox is full" was not appended to the
3212 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3213 only to a quota excession during the actual writing of the file.
3215 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3216 characters?) which causes problems in log lines. The DN values are now
3217 passed through string_printing() before being added to log lines.
3219 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3220 and InterBase are left for another time.)
3222 PH/21 Added message_body_newlines option.
3224 PH/22 Guard against possible overflow in moan_check_errorcopy().
3226 PH/23 POSIX allows open() to be a macro; guard against that.
3228 PH/24 If the recipient of an error message contained an @ in the local part
3229 (suitably quoted, of course), incorrect values were put in $domain and
3230 $local_part during the evaluation of errors_copy.
3236 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3237 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3240 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3241 issue a MAIL command.
3243 PH/02 In an ACL statement such as
3245 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3247 if a client was not listed at all, or was listed with a value other than
3248 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3249 the condition was not true (as it should be), so access was not denied.
3250 The bug was that the ! inversion was incorrectly passed on to the second
3251 item. This has been fixed.
3253 PH/03 Added additional dnslists conditions == and =& which are different from
3254 = and & when the dns lookup returns more than one IP address.
3256 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3257 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3259 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3260 FSYNC, which compiles an option called disable_fsync that allows for
3261 bypassing fsync(). The documentation is heavily laced with warnings.
3263 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3265 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3266 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3267 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3268 including adding "make clean"; (3) Added -fPIC when compiling the test
3269 dynamically loaded module, to get rid of a warning.
3271 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3272 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3273 = 0s. The bug is otherwise harmless.
3275 PH/07 There was a bug in the dovecot authenticator such that the value of
3276 $auth1 could be overwritten, and so not correctly preserved, after a
3277 successful authentication. This usually meant that the value preserved by
3278 the server_setid option was incorrect.
3280 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3282 PH/09 Installed PCRE release 7.0.
3284 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3285 run for batched SMTP input. It is now run at the start of every message
3286 in the batch. While fixing this I discovered that the process information
3287 (output by running exiwhat) was not always getting set for -bs and -bS
3288 input. This is fixed, and it now also says "batched" for BSMTP.
3290 PH/11 Added control=no_pipelining.
3292 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3293 patch, slightly modified), and move the expansion of helo_data till after
3294 the connection is made in the smtp transport (so it can use these
3297 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3299 PH/14 Added log_selector = +pid.
3301 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3303 PH/16 Add ${if forany and ${if forall.
3305 PH/17 Added dsn_from option to vary the From: line in DSNs.
3307 PH/18 Flush SMTP output before performing a callout, unless control =
3308 no_callout_flush is set.
3310 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3311 was true (the default) a successful delivery failed to delete the retry
3312 item, thus causing premature timeout of the address. The bug is now
3315 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3317 PH/21 Long custom messages for fakedefer and fakereject are now split up
3318 into multiline responses in the same way that messages for "deny" and
3319 other ACL rejections are.
3321 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3322 with slight modification.
3324 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3325 draft, changing the syntax and factoring some duplicate code".
3327 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3328 for deliveries of the second and subsequent messages over the same SMTP
3331 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3332 ${reduce, with only minor "tidies".
3334 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3336 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3337 expansion side effects.
3339 PH/27 When a message times out after an over-quota error from an Exim-imposed
3340 quota, the bounce message says "mailbox is full". This message was not
3341 being given when it was a system quota that was exceeded. It now should
3344 MH/03 Made $recipients available in local_scan(). local_scan() already has
3345 better access to the recipient list through recipients_list[], but
3346 $recipients can be useful in postmaster-provided expansion strings.
3348 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3349 in the case of a MAIL command with additional options following the
3350 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3351 were accidentally chopped off.
3353 PH/29 SMTP synchronization checks are implemented when a command is read -
3354 there is a check that no more input is waiting when there shouldn't be
3355 any. However, for some commands, a delay in an ACL can mean that it is
3356 some time before the response is written. In this time, more input might
3357 arrive, invalidly. So now there are extra checks after an ACL has run for
3358 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3359 pipelining has not been advertised.
3361 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3363 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3364 returned in the SMTP response when smtp_return_error_details was false.
3365 This has been fixed.
3367 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3368 instead of the C I/O that was originally supplied, because problems were
3369 reported on Solaris.
3371 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3372 Exim which did not show up earlier: it was assuming that a call to
3373 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3374 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3375 was a macro that became an assignment, so it seemed to work. This has
3376 changed to a proper function call with a void return, hence the compile
3377 error. Exim's code has been fixed.
3379 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3382 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3384 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3386 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3387 because it turns out to be too big or there is a timeout) there is no
3388 "Completed" line in the log. When some messages of this type were
3389 selected by exigrep, they were listed as "not completed". Others were
3390 picked up by some special patterns. I have improved the selection
3391 criteria to be more general.
3393 PH/38 The host_find_failed option in the manualroute router can now be set
3394 to "ignore", to completely ignore a host whose IP address cannot be
3395 found. If all hosts are ignored, the behaviour is controlled by the new
3396 host_all_ignored option.
3398 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3399 homing or because of multiple MX records with /mx) generated more than
3400 one IP address, and the following item turned out to be the local host,
3401 all the secondary addresses of the first item were incorrectly removed
3402 from the list, along with the local host and any following hosts (which
3403 is what is supposed to happen).
3405 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3406 whoever called Exim into the -H file. In the case of the daemon it was
3407 behaving confusingly. When first started, it used values for whoever
3408 started the daemon, but after a SIGHUP it used the Exim user (because it
3409 calls itself on a restart). I have changed the code so that it now always
3412 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3413 message are rejected with the same error (e.g. no authentication or bad
3414 sender address), and a DATA command is nevertheless sent (as can happen
3415 with PIPELINING or a stupid MUA), the error message that was given to the
3416 RCPT commands is included in the rejection of the DATA command. This is
3417 intended to be helpful for MUAs that show only the final error to their
3420 PH/42 Another patch from the Sieve maintainer.
3422 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3423 Eximstats - Fixed some broken HTML links and added missing column headers
3425 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3426 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3428 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3430 PH/43 Yet another patch from the Sieve maintainer.
3432 PH/44 I found a way to check for a TCP/IP connection going away before sending
3433 the response to the final '.' that terminates a message, but only in the
3434 case where the client has not sent further data following the '.'
3435 (unfortunately, this is allowed). However, in many cases there won't be
3436 any further data because there won't be any more messages to send. A call
3437 to select() can be used: if it shows that the input is "ready", there is
3438 either input waiting, or the socket has been closed. An attempt to read
3439 the next input character can distinguish the two cases. Previously, Exim
3440 would have sent an OK response which the client would never have see.
3441 This could lead to message repetition. This fix should cure that, at
3442 least in a lot of common cases.
3444 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3445 advertised in response to EHLO.
3451 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3452 fixed by 4.65/MH/01 (is this a record?) are fixed:
3454 (i) An empty string was always treated as zero by the numeric comparison
3455 operators. This behaviour has been restored.
3457 (ii) It is documented that the numeric comparison operators always treat
3458 their arguments as decimal numbers. This was broken in that numbers
3459 starting with 0 were being interpreted as octal.
3461 While fixing these problems I realized that there was another issue that
3462 hadn't been noticed. Values of message_size_limit (both the global option
3463 and the transport option) were treated as octal if they started with 0.
3464 The documentation was vague. These values are now always treated as
3465 decimal, and I will make that clear in the documentation.
3471 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3472 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3475 MH/01 Don't check that the operands of numeric comparison operators are
3476 integers when their expansion is in "skipping" mode (fixes bug
3477 introduced by 4.64-PH/07).
3479 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3480 child addresses, Exim now panics and dies. Previously, because the count
3481 is held in a short int, deliveries were likely to be lost. As such a
3482 large number of recipients for a single message is ridiculous
3483 (performance will be very, very poor), I have chosen to impose a limit
3484 rather than extend the field.
3490 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3491 leftover -K file (the existence of which was triggered by #402).
3492 While we were at it, introduced process PID as part of the -K
3493 filename. This should rule out race conditions when creating
3496 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3497 processing considerably. Previous code took too long for large mails,
3498 triggering a timeout which in turn triggers #401.
3500 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3501 in the DK code in transports.c. sendfile() is not really portable,
3502 hence the _LINUX specificness.
3504 TF/01 In the add_headers option to the mail command in an Exim filter,
3505 there was a bug that Exim would claim a syntax error in any
3506 header after the first one which had an odd number of characters
3509 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3510 callout verification, Exim cached a "reject" for the entire domain. This
3511 is correct for most verifications, but it is not correct for a recipient
3512 verification with use_sender or use_postmaster set, because in that case
3513 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3514 case of MAIL FROM:<> rejection from other early rejections (e.g.
3515 rejection of HELO). When verifying a recipient using a non-null MAIL
3516 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3517 Whatever the result of the callout, the value of the domain cache is
3518 left unchanged (for any other kind of callout, getting as far as trying
3519 RCPT means that the domain itself is ok).
3521 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3524 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3525 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3526 session, the daemon ignores ECONNECT errors and logs others; it now
3527 ignores EPIPE as well.
3529 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3530 (quoted-printable decoding).
3532 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3533 later the small subsequent patch to fix an introduced bug.
3535 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3537 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3539 PH/08 An error is now given if message_size_limit is specified negative.
3541 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3542 to be given (somewhat) arbitrary names.
3544 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3547 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3548 miscellaneous code fixes
3550 PH/10 Added the log_reject_target ACL modifier to specify where to log
3553 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3554 hostname. This is wrong, because it relates to the incoming message (and
3555 probably the interface on which it is arriving) and not to the outgoing
3556 callout (which could be using a different interface). This has been
3557 changed to use the value of the helo_data option from the smtp transport
3558 instead - this is what is used when a message is actually being sent. If
3559 there is no remote transport (possible with a router that sets up host
3560 addresses), $smtp_active_hostname is used.
3562 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3563 tweaks were necessary in order to get it to work (see also 21 below):
3564 (a) The code assumed that strncpy() returns a negative number on buffer
3565 overflow, which isn't the case. Replaced with Exim's string_format()
3567 (b) There were several signed/unsigned issues. I just did the minimum
3568 hacking in of casts. There is scope for a larger refactoring.
3569 (c) The code used strcasecmp() which is not a standard C function.
3570 Replaced with Exim's strcmpic() function.
3571 (d) The code set only $1; it now sets $auth1 as well.
3572 (e) A simple test gave the error "authentication client didn't specify
3573 service in request". It would seem that Dovecot has changed its
3574 interface. Fortunately there's a specification; I followed it and
3575 changed what the client sends and it appears to be working now.
3577 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3580 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3581 address A is aliased to B and C, where B exists and C does not. Without
3582 -v the output is "A verified" because verification stops after a
3583 successful redirection if more than one address is generated. However,
3584 with -v the child addresses are also verified. Exim was outputting "A
3585 failed to verify" and then showing the successful verification for C,
3586 with its parentage. It now outputs "B failed to verify", showing B's
3587 parentage before showing the successful verification of C.
3589 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3590 look up a TXT record in a specific list after matching in a combined
3593 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3594 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3595 they consult the DNS. I had assumed they would set it the way they
3596 wanted; and indeed my experiments on Linux seem to show that in some
3597 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3598 To be on the safe side, however, I have now made the interface to
3599 host_find_byname() similar to host_find_bydns(), with an argument
3600 containing the DNS resolver options. The host_find_byname() function now
3601 sets these options at its start, just as host_find_bydns() does. The smtp
3602 transport options dns_qualify_single and dns_search_parents are passed to
3603 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3604 of host_find_byname() use the default settings of RES_DEFNAMES
3605 (qualify_single) but not RES_DNSRCH (search_parents).
3607 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3608 spool_read_header() do less string testing, by means of a preliminary
3609 switch on the second character of optional "-foo" lines. (This is
3610 overdue, caused by the large number of possibilities that now exist.
3611 Originally there were few.) While I was there, I also converted the
3612 str(n)cmp tests so they don't re-test the leading "-" and the first
3613 character, in the hope this might squeeze out yet more improvement.
3615 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3616 flag allowing group syntax was set by the header_syntax check but not
3617 turned off, possible causing trouble later; (2) The flag was not being
3618 set at all for the header_verify test, causing "group"-style headers to
3619 be rejected. I have now set it in this case, and also caused header_
3620 verify to ignore an empty address taken from a group. While doing this, I
3621 came across some other cases where the code for allowing group syntax
3622 while scanning a header line wasn't quite right (mostly, not resetting
3623 the flag correctly in the right place). These bugs could have caused
3624 trouble for malformed header lines. I hope it is now all correct.
3626 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3627 with the "reply" argument non-NULL. The code, however (which originally
3628 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3629 but it didn't always do it. This confused somebody who was copying the
3630 code for some other use. I have removed all the tests.
3632 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3633 feature that was used to support insecure browsers during the U.S. crypto
3634 embargo. It requires special client support, and Exim is probably the
3635 only MTA that supported it -- and would never use it because real RSA is
3636 always available. This code has been removed, because it had the bad
3637 effect of slowing Exim down by computing (never used) parameters for the
3638 RSA_EXPORT functionality.
3640 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3641 authenticator to fail if there's a tab character in the incoming data
3642 (there should never be unless someone is messing about, as it's supposed
3643 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3644 if the connection is using TLS or if the remote IP is the same as the
3645 local IP, and the "valid-client-cert option" if a client certificate has
3648 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3649 authenticators. This can be used for authorization after authentication
3650 succeeds. (In the case of plaintext, it servers for both authentication
3653 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3654 if any retry times were supplied.
3656 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3657 connection, where there is no client IP address to check. In this
3658 situation, the verify now always succeeds.
3660 PH/25 Applied John Jetmore's -Mset patch.
3662 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3664 PH/27 In a string expansion for a processed (not raw) header when multiple
3665 headers of the same name were present, leading whitespace was being
3666 removed from all of them, but trailing whitespace was being removed only
3667 from the last one. Now trailing whitespace is removed from each header
3668 before concatenation. Completely empty headers in a concatenation (as
3669 before) are ignored.
3671 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3672 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3674 PH/29 [Removed. This was a change that I later backed out, and forgot to
3675 correct the ChangeLog entry (that I had efficiently created) before
3676 committing the later change.]
3678 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3679 address errors (4xx response to RCPT) over the same connection as other
3680 messages routed to the same hosts. Such deliveries are always "forced",
3681 so retry times are not inspected. This resulted in far too many retries
3682 for the affected addresses. The effect occurred only when there were more
3683 hosts than the hosts_max_try setting in the smtp transport when it had
3684 the 4xx errors. Those hosts that it had tried were not added to the list
3685 of hosts for which the message was waiting, so if all were tried, there
3686 was no problem. Two fixes have been applied:
3688 (i) If there are any address or message errors in an SMTP delivery, none
3689 of the hosts (tried or untried) are now added to the list of hosts
3690 for which the message is waiting, so the message should not be a
3691 candidate for sending over the same connection that was used for a
3692 successful delivery of some other message. This seems entirely
3693 reasonable: after all the message is NOT "waiting for some host".
3694 This is so "obvious" that I'm not sure why it wasn't done
3695 previously. Hope I haven't missed anything, but it can't do any
3696 harm, as the worst effect is to miss an optimization.
3698 (ii) If, despite (i), such a delivery is accidentally attempted, the
3699 routing retry time is respected, so at least it doesn't keep
3700 hammering the server.
3702 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3703 in ${readsocket because some servers need this prod.
3705 PH/32 Added some extra debug output when updating a wait-xxx database.
3707 PH/33 The hint "could be header name not terminated by colon", which has been
3708 given for certain expansion errors for a long time, was not being given
3709 for the ${if def:h_colon_omitted{... case.
3711 PH/34 The spec says: "With one important exception, whenever a domain list is
3712 being scanned, $domain contains the subject domain." There was at least
3713 one case where this was not true.
3715 PH/35 The error "getsockname() failed: connection reset by peer" was being
3716 written to the panic log as well as the main log, but it isn't really
3717 panic-worthy as it just means the connection died rather early on. I have
3718 removed the panic log writing for the ECONNRESET error when getsockname()
3721 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3722 runs only) independently of the message's sender address. This meant
3723 that, if the 4xx error was in fact related to the sender, a different
3724 message to the same recipient with a different sender could confuse
3725 things. In particular, this can happen when sending to a greylisting
3726 server, but other circumstances could also provoke similar problems.
3727 I have changed the default so that the retry time for these errors is now
3728 based a combination of the sender and recipient addresses. This change
3729 can be overridden by setting address_retry_include_sender=false in the
3732 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3733 remote server are returned as part of bounce messages. This was not
3734 happening for LMTP over a pipe (the lmtp transport), but now it is the
3735 same for both kinds of LMTP.
3737 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3738 in header lines that were in fact CNAMEs. This is no longer the case.
3740 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3741 and queue runs started by the daemon processed all messages. This has
3742 been fixed so that -R and -S can now usefully be given with -q<time>.
3744 PH/40 Import PCRE release 6.7 (fixes some bugs).
3746 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3748 PH/42 Give an error if -q is specified more than once.
3750 PH/43 Renamed the variables $interface_address and $interface_port as
3751 $received_ip_address and $received_port, to make it clear that these
3752 values apply to message reception, and not to the outgoing interface when
3753 a message is delivered. (The old names remain recognized, of course.)
3755 PH/44 There was no timeout on the connect() call when using a Unix domain
3756 socket in the ${readsocket expansion. There now is.
3758 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3759 be meaningful with "accept".
3761 SC/01 Eximstats V1.43
3762 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3764 SC/02 Eximstats V1.44
3765 Use a glob alias rather than an array ref in the generated
3766 parser. This improves both readability and performance.
3768 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3769 Collect SpamAssassin and rejection statistics.
3770 Don't display local sender or destination tables unless
3771 there is data to show.
3772 Added average volumes into the top table text output.
3774 SC/04 Eximstats V1.46
3775 Collect data on the number of addresses (recipients)
3776 as well as the number of messages.
3778 SC/05 Eximstats V1.47
3779 Added 'Message too big' to the list of mail rejection
3780 reasons (thanks to Marco Gaiarin).
3782 SC/06 Eximstats V1.48
3783 Mainlog lines which have GMT offsets and are too short to
3784 have a flag are now skipped.
3786 SC/07 Eximstats V1.49 (Alain Williams)
3787 Added the -emptyok flag.
3789 SC/08 Eximstats V1.50
3790 Fixes for obtaining the IP address from reject messages.
3792 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3793 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3794 whitespace changes from 4.64-PH/27
3796 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3799 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3800 are found, allow negative numbers in numeric criteria)
3802 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3804 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3805 to match changes made in 4.64-PH/43
3807 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3809 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3810 "domain" is actually an IP address, and force a failure. This locks out
3811 those revolvers/nameservers that support "A-for-A" lookups, in
3812 contravention of the specifications.
3814 PH/48 When a host name was looked up from an IP address, and the subsequent
3815 forward lookup of the name timed out, the host name was left in
3816 $sender_host_name, contrary to the specification.
3818 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3819 restricted to single-key lookups, Exim was not diagnosing an error if
3820 * or *@ was used with a query-style lookup.
3822 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3824 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3825 long ago, but noone interested enough thought of it. Let's just say that
3826 the "1.1" means that there are some new functions that weren't there at
3827 some point in the past.
3829 PH/51 Error processing for expansion failure of helo_data from an smtp
3830 transport during callout processing was broken.
3832 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3833 tested/used via the -bh/-bhc/-bs options.
3835 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3836 bug, fixed in subsequent PCRE releases).
3838 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3839 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3841 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3847 SC/01 Use a glob alias rather than an array ref in eximstats generated
3848 parser. This improves both readability and performance.
3850 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3851 Don't display local sender or destination tables in eximstats unless
3852 there is data to show.
3853 Added average volumes into the eximstats top table text output.
3855 SC/03 Collect data on the number of addresses (recipients) as well
3856 as the number of messages in eximstats.
3858 TF/01 Correct an error in the documentation for the redirect router. Exim
3859 does (usually) call initgroups() when daemonizing.
3861 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3862 with consistent privilege compared to when running as a daemon.
3864 TF/03 Note in the spec that $authenticated_id is not set for local
3865 submissions from trusted users.
3867 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3868 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3870 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3871 by adding some example configuration directives to the default
3872 configuration file. A little bit of work is required to uncomment the
3873 directives and define how usernames and passwords are checked, but
3874 there is now a framework to start from.
3876 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3877 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3878 without this. I don't know how relevant this is to other LDAP libraries.
3880 PH/02 Add the verb name to the "unknown ACL verb" error.
3882 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3884 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3886 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3887 directory not expanded when it should be if an expanded home directory
3888 was set for the address (which is overridden by the transport).
3890 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3893 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3894 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3895 because it is too late at that time, and has no effect.
3897 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3898 security issue with \' (bugzilla #107). I could not use the
3899 PQescapeStringConn() function, because it needs a PGconn value as one of
3902 PH/08 When testing addresses using -bt, indicate those final addresses that
3903 are duplicates that would not cause an additional delivery. At least one
3904 person was confused, thinking that -bt output corresponded to deliveries.
3905 (Suppressing duplicates isn't a good idea as you lose the information
3906 about possibly different redirections that led to the duplicates.)
3908 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3909 systems where poll() doesn't work, in particular OS X.
3911 PH/10 Added more information to debugging output for retry time not reached.
3913 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3914 operations in malware.c.
3916 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3919 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3920 syslog_duplication set false, log lines that would normally be written
3921 both the the main log and to the reject log were not written to syslog at
3924 PH/14 In the default configuration, change the use of "message" in ACL warn
3925 statements to "add_header".
3927 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3928 not followed by a command (e.g. "seen endif").
3930 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3931 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3934 PH/17 Added extra conditions to the default value of delay_warning_condition
3938 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3939 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3940 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3943 The Auto-Submitted: and various List- headers are standardised, whereas I
3944 don't think Precedence: ever was.
3946 PH/18 Refactored debugging code in route_finduser() to show more information,
3947 in particular, the error code if getpwnam() issues one.
3949 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3950 This is apparently needed in addition to the PH/07 change above to avoid
3951 any possible encoding problems.
3953 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3954 but not after initializing Perl.
3956 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3957 output them only if debugging. By default they are written stderr,
3958 apparently, which is not desirable.
3960 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3963 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3966 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3968 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3969 authenticated or an ident call has been made. Suppress the default
3970 values for $authenticated_id and $authenticated_sender (but permit -oMai
3971 and -oMas) when testing with -bh.
3973 PH/24 Re-jigged the order of the tests in the default configuration so that the
3974 tests for valid domains and recipients precede the DNS black list and CSA
3975 tests, on the grounds that those ones are more expensive.
3977 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3978 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3979 This bug exists in every version of Exim that I still have, right back to
3982 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3983 However, an attempt to turn on case-sensitivity in a regex key by
3984 including (?-i) didn't work because the subject string was already
3985 lowercased, and the effects were non-intuitive. It turns out that a
3986 one-line patch can be used to allow (?-i) to work as expected.
3992 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3993 other effects) broke the use of negated acl sub-conditions.
3995 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3998 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3999 "Deny" causes Exim to reject the incoming connection with a 554 error.
4000 Unfortunately, if there is a major crisis, such as a disk failure,
4001 tcp-wrappers gives "deny", whereas what one would like would be some
4002 kind of temporary error. A kludge has been added to help with this.
4003 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4004 554 error is used if errno is still zero or contains ENOENT (which occurs
4005 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4008 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4010 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4011 errors. Otherwise a message that provokes a temporary error (when other
4012 messages do not) can cause a whole host to time out.
4014 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4015 addresses were routed directly to files or pipes from a redirect router.
4016 File deliveries just didn't batch; pipe deliveries might have suffered
4019 PH/06 A failure to get a lock for a hints database would erroneously always say
4020 "Failed to get write lock", even when it was really a read lock.
4022 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4023 of 0600. This has been changed to use the value of the lockfile_mode
4024 option (which defaults to 0600).
4026 PH/08 Applied small patch from the Sieve maintainer.
4028 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4029 folder from quota calculations, a direct delivery into this folder messed
4030 up the contents of the maildirsize file. This was because the regex was
4031 used only to exclude .Trash (or whatever) when the size of the mailbox
4032 was calculated. There was no check that a delivery was happening into an
4033 excluded directory. This bug has been fixed by ignoring all quota
4034 processing for deliveries into excluded directories.
4036 PH/10 Added the maildirfolder_create_regex option to appendfile.
4042 PH/01 The code for finding all the local interface addresses on a FreeBSD
4043 system running IPv6 was broken. This may well have applied to all BSD
4044 systems, as well as to others that have similar system calls. The broken
4045 code found IPv4 interfaces correctly, but gave incorrect values for the
4046 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4047 that it would not match correctly against @[] and not recognize the IPv6
4050 PH/02 The ipliteral router was not recognizing addresses of the form user@
4051 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4053 PH/03 Added disable_ipv6.
4055 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4056 decoded form, because it is most often used to construct To: headers
4057 lines in autoreplies, and the decoded form may well be syntactically
4058 invalid. However, $reply_address has leading white space removed, and all
4059 newlines turned into spaces so that the autoreply transport does not
4062 PH/05 If group was specified without a user on a router, and no group or user
4063 was specified on a transport, the group from the router was ignored.
4065 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4066 for visible compile-time settings that can be used to change these
4067 numbers, for those that want even more. Backwards compatibility with old
4068 spool files has been maintained. However, going back to a previous Exim
4069 release will lost any variables that are in spool files.
4071 PH/07 Two small changes when running in the test harness: increase delay when
4072 passing a TCP/IP connection to a new process, in case the original
4073 process has to generate a bounce, and remove special handling of
4074 127.0.0.2 (sic), which is no longer necessary.
4076 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4077 be the same on different OS.
4079 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4082 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4083 whether --show-vars was specified or not
4085 JJ/02 exipick: Added support for new ACL variable spool format introduced
4088 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4089 syntactically invalid From: or Reply-to: line, and a filter used this to
4090 generate an autoreply, and therefore failed to obtain an address for the
4091 autoreply, Exim could try to deliver to a non-existent relative file
4092 name, causing unrelated and misleading errors. What now happens is that
4093 it logs this as a hard delivery error, but does not attempt to create a
4096 PH/11 The exinext utility has a -C option for testing purposes, but although
4097 the given file was scanned by exinext itself; it wasn't being passed on
4098 when Exim was called.
4100 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4101 an end-of-file indication when reading a command response.
4103 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4104 compiled. In many other places in Exim, IPv6 addresses are always
4105 recognized, so I have changed this. It also means that IPv4 domain
4106 literals of the form [IPV4:n.n.n.n] are now always recognized.
4108 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4109 used if the router is not running as root, for example, when verifying at
4110 ACL time, or when using -bh. The debugging output from this situation was
4111 non-existent - all you got was a failure to exec. I have made two
4114 (a) Failures to set uid/gid, the current directory, or a process leader
4115 in a subprocess such as that created by queryprogram now generate
4116 suitable debugging output when -d is set.
4118 (b) The queryprogram router detects when it is not running as root,
4119 outputs suitable debugging information if -d is set, and then runs
4120 the subprocess without attempting to change uid/gid.
4122 PH/15 Minor change to Makefile for building test_host (undocumented testing
4125 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4126 additional section of a DNS packet that returns MX or SRV records.
4127 Instead, it always explicitly searches for A/AAAA records. This avoids
4128 major problems that occur when a DNS server includes only records of one
4129 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4130 fixed another bug: if SRV records were looked up and the corresponding
4131 address records were *not* found in the additional section, the port
4132 values from the SRV records were lost.
4134 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4135 using the correct key (the original address) when searching the retry
4136 rules in order to find which one to use for generating the retry hint.
4138 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4139 adding the default one. Similarly, if it contains a Reply-To: header, the
4140 errors_reply_to option, if set, is not used.
4142 PH/19 When calculating a retry time, Exim used to measure the "time since
4143 failure" by looking at the "first failed" field in the retry record. Now
4144 it does not use this if it is later than than the arrival time of the
4145 message. Instead it uses the arrival time. This makes for better
4146 behaviour in cases where some deliveries succeed, thus re-setting the
4147 "first failed" field. An example is a quota failure for a huge message
4148 when small messages continue to be delivered. Without this change, the
4149 "time since failure" will always be short, possible causing more frequent
4150 delivery attempts for the huge message than are intended.
4151 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4153 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4154 $1, $2, $3) because the numerical variables can be reset during some
4155 expansion items (e.g. "match"), thereby losing the authentication data.
4157 PH/21 Make -bV show the size of off_t variables so that the test suite can
4158 decide whether to run tests for quotas > 2G.
4160 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4161 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4162 filecount value is greater than 2G or if a quota value is greater than 2G
4163 on a system where the size of off_t is not greater than 4, a panic error
4166 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4167 never match. The debug and -bh output now contains an explicit error
4168 message indicating a malformed IPv4 address or mask.
4170 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4171 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4172 PH/23 above applies.
4174 PH/25 Do not write to syslog when running in the test harness. The only
4175 occasion when this arises is a failure to open the main or panic logs
4176 (for which there is an explicit test).
4178 PH/26 Added the /no_tell option to "control=freeze".
4180 PH/27 If a host name lookup failed very early in a connection, for example, if
4181 the IP address matched host_lookup and the reverse lookup yielded a name
4182 that did not have a forward lookup, an error message of the form "no IP
4183 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4184 could be logged. Now it outputs the IP address instead of "NULL".
4186 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4187 allows the sender and the authenticated sender to be set when
4188 submitting a message from within Exim. Since child_open_exim() is
4189 documented for local_scan(), the new function should be too.
4191 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4192 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4193 results in an empty string is now treated as unset.
4195 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4197 PH/31 Added sender_verify_fail logging option.
4199 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4200 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4201 tidied the source and removed it altogether.
4203 PH/33 When a queue run was abandoned because the load average was too high, a
4204 log line was always written; now it is written only if the queue_run log
4205 selector is set. In addition, the log line for abandonment now contains
4206 information about the queue run such as the pid. This is always present
4207 in "start" and "stop" lines but was omitted from the "abandon" line.
4209 PH/34 Omit spaces between a header name and the colon in the error message that
4210 is given when verify = headers_syntax fails (if there are lots of them,
4211 the message gets confusing).
4213 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4214 names, as there are now some PTR records that contain slashes. This check
4215 is only to protect against broken name servers that fall over on strange
4216 characters, so the fact that it applies to all lookups doesn't matter.
4218 PH/36 Now that the new test suite is complete, we can remove some of the
4219 special code in Exim that was needed for the old test suite. For example,
4220 sorting DNS records because real resolvers return them in an arbitrary
4221 order. The new test suite's fake resolver always returns records in the
4224 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4225 bounces) except when queue_only is set, to avoid logging races between
4226 the different processes.
4228 PH/38 Panic-die if .include specifies a non-absolute path.
4230 PH/39 A tweak to the "H" retry rule from its user.
4232 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4233 a label. They prevented compilation on older perls.
4235 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4236 a warning to be raised on newish perls.
4238 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4239 on queue. Changes to match documented behaviour of showing count of
4240 messages matching specified criteria.
4242 PH/40 Changed the default ident timeout from 30s to 5s.
4244 PH/41 Added support for the use of login_cap features, on those BSD systems
4245 that have them, for controlling the resources used by pipe deliveries.
4247 PH/42 The content-scanning code uses fopen() to create files in which to put
4248 message data. Previously it was not paying any attention to the mode of
4249 the files. Exim runs with umask(0) because the rest of the code creates
4250 files with open(), and sets the required mode explicitly. Thus, these
4251 files were ending up world-writeable. This was not a big issue, because,
4252 being within the spool directory, they were not world-accessible. I have
4253 created a function called modefopen, which takes an additional mode
4254 argument. It sets umask(777), creates the file, chmods it to the required
4255 mode, then resets the umask. All the relevant calls to fopen() in the
4256 content scanning code have been changed to use this function.
4258 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4259 to 24 hours. This avoids potential overflow problems when processing G
4260 and H retry rules. I suspect nobody ever tinkers with this value.
4262 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4264 PH/45 When the plaintext authenticator is running as a client, the server's
4265 challenges are checked to ensure they are valid base64 strings. By
4266 default, the authentication attempt is cancelled if an invalid string is
4267 received. Setting client_ignore_invalid_base64 true ignores these errors.
4268 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4269 they are received. Thus, the responses can be made to depend on the
4270 challenges. If an invalid string is ignored, an empty string is placed in
4273 PH/46 Messages that are created by the autoreply transport now contains a
4274 References: header, in accordance with RFCs 2822 and 3834.
4276 PH/47 Added authenticated_sender_force to the smtp transport.
4278 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4280 PH/49 Installed latest patch from the Sieve maintainer.
4282 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4283 was also set, the appendfile transport was unnecessarily scanning a
4284 directory of message files (e.g. for maildir delivery) to find the count
4285 of files (along with the size), even though it did not need this
4286 information. It now does the scan only if it needs to find either the
4287 size of the count of files.
4289 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4291 PH/52 Two bugs concerned with error handling when the smtp transport is
4294 (i) Exim was not creating retry information for temporary errors given
4295 for individual recipients after the DATA command when the smtp transport
4296 was used in LMTP mode. This meant that they could be retried too
4297 frequently, and not timed out correctly.
4299 (ii) Exim was setting the flag that allows error details to be returned
4300 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4301 recipients that were returned after the DATA command.
4303 PH/53 This is related to PH/52, but is more general: for any failing address,
4304 when detailed error information was permitted to be returned to the
4305 sender, but the error was temporary, then after the final timeout, only
4306 "retry timeout exceeded" was returned. Now it returns the full error as
4307 well as "retry timeout exceeded".
4309 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4310 do this, and (what is worse) MTAs that accept it.
4312 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4313 will now be deprecated.
4315 PH/56 New os.c-cygwin from the Cygwin maintainer.
4317 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4318 formats (previously only available in exim formats via -bpr, -bpru,
4319 and -bpra. Now also available in native and exiqgrep formats)
4321 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4322 with very large, slow to parse queues
4324 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4326 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4328 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4329 responses to authentication challenges, though it was showing the
4330 challenges; (ii) I've removed the CR characters from the debug output for
4333 PH/58 Allow for the insertion of a newline as well as a space when a string
4334 is turned into more than one encoded-word during RFC 2047 encoding. The
4335 Sieve code now uses this.
4337 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4338 data_4xx, lost_connection, tls_required.
4340 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4341 message was being sent as an SMTP response.
4343 PH/61 Add -l and -k options to exicyclog.
4345 PH/62 When verifying, if an address was redirected to one new address, so that
4346 verification continued, and the new address failed or deferred after
4347 having set something in $address_data, the value of $address_data was not
4348 passed back to the ACL. This was different to the case when no
4349 redirection occurred. The value is now passed back in both cases.
4351 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4352 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4353 use login_cap.h, so on its own it isn't the distinguishing feature. The
4354 new name refers directly to the setclassresources() function.
4356 PH/65 Added configuration files for NetBSD3.
4358 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4360 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4361 is preferred over IPv4.
4363 PH/68 The bounce_return_message and bounce_return_body options were not being
4364 honoured for bounces generated during the reception of non-SMTP messages.
4365 In particular, this applied to messages rejected by the ACL. This bug has
4366 been fixed. However, if bounce_return_message is true and bounce_return_
4367 body is false, the headers that are returned for a non-SMTP message
4368 include only those that have been read before the error was detected.
4369 (In the case of an ACL rejection, they have all been read.)
4371 PH/69 The HTML version of the specification is now built in a directory called
4372 spec_html instead of spec.html, because the latter looks like a path with
4373 a MIME-type, and this confuses some software.
4375 PH/70 Catch two compiler warnings in sieve.c.
4377 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4378 function verify_get_ident() calls ip_connect() to connect a socket, but
4379 if the "connect()" function timed out, ip_connect() used to close the
4380 socket. However, verify_get_ident() also closes the socket later, and in
4381 between Exim writes to the log, which may get opened at this point. When
4382 the socket was closed in ip_connect(), the log could get the same file
4383 descriptor number as the socket. This naturally causes chaos. The fix is
4384 not to close the socket in ip_connect(); the socket should be closed by
4385 the function that creates it. There was only one place in the code where
4386 this was missing, in the iplookup router, which I don't think anybody now
4387 uses, but I've fixed it anyway.
4389 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4390 well as to direct DNS lookups. Otherwise the handling of names in host
4391 lists is inconsistent and therefore confusing.
4397 PH/01 Two changes to the default runtime configuration:
4399 (1) Move the checks for relay_from_hosts and authenticated clients from
4400 after to before the (commented out) DNS black list checks.
4402 (2) Add control=submission to the relay_from_hosts and authenticated
4403 clients checks, on the grounds that messages accepted by these
4404 statements are most likely to be submissions.
4406 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4408 (1) Generate an error if the third argument for the ${prvs expansion is
4411 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4414 (3) Reset the variables that are obtained from the first argument of
4415 ${prvscheck and used in the second argument before leaving the code,
4416 because their memory is reclaimed, so using them afterwards may do
4419 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4420 one (it's much easier than Tom thought :-).
4422 (5) Because of (4), we can now allow for the use of $prvscheck_result
4423 inside the third argument.
4425 PH/03 For some reason, the default setting of PATH when running a command from
4426 a pipe transport was just "/usr/bin". I have changed it to
4429 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4430 anything to be listed in the output from -bV.
4432 PH/05 When a filter generated an autoreply, the entire To: header line was
4433 quoted in the delivery log line, like this:
4435 => >A.N.Other <ano@some.domain> <original@ddress> ...
4437 This has been changed so that it extracts the operative address. There
4438 may be more than one such address. If so, they are comma-separated, like
4441 => >ano@some.domain,ona@other.domain <original@ddress> ...
4443 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4444 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4445 not being looked up in the rDNS to get a host name, Exim was showing the
4446 IP address twice in Received: lines, even though the IP addresses were
4447 identical. For example:
4449 Received: from [1.2.3.4] (helo=[1.2.3.4])
4451 However, if the real host name was known, it was omitting the HELO data
4452 if it matched the actual IP address. This has been tidied up so that it
4453 doesn't show the same IP address twice.
4455 PH/07 When both +timestamp and +memory debugging was on, the value given by
4456 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4457 called by the debug printing, thereby overwriting the timestamp buffer.
4458 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4460 PH/08 When the original message was included in an autoreply transport, it
4461 always said "this is a copy of the message, including all the headers",
4462 even if body_only or headers_only was set. It now gives an appropriate
4465 PH/09 Applied a patch from the Sieve maintainer which:
4467 o fixes some comments
4468 o adds the (disabled) notify extension core
4469 o adds some debug output for the result of if/elsif tests
4470 o points to the current vacation draft in the documentation
4471 and documents the missing references header update
4475 o fixes a bug in processing the envelope test (when testing
4476 multiple envelope elements, the last element determined the
4479 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4480 Electronic Mail") by including:
4482 Auto-submitted: auto-generated
4484 in the messages that it generates (bounce messages and others, such as
4485 warnings). In the case of bounce messages for non-SMTP messages, there was
4486 also a typo: it was using "Auto_submitted" (underscore instead of
4487 hyphen). Since every message generated by Exim is necessarily in response
4488 to another message, thes have all been changed to:
4490 Auto-Submitted: auto-replied
4492 in accordance with these statements in the RFC:
4494 The auto-replied keyword:
4496 - SHOULD be used on messages sent in direct response to another
4497 message by an automatic process,
4499 - MUST NOT be used on manually-generated messages,
4501 - MAY be used on Delivery Status Notifications (DSNs) and Message
4502 Disposition Notifications (MDNs),
4504 - MUST NOT be used on messages generated by automatic or periodic
4505 processes, except for messages which are automatic responses to
4508 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4509 to the default Received: header definition.
4511 PH/12 Added log selector acl_warn_skipped (default on).
4513 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4514 variables because (a) they are in the wrong storage pool and (b) even if
4515 they were copied, it wouldn't work properly because of the caching.
4517 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4518 checking when decoding. Apparently there are clients that generate
4519 overlong encoded strings. Why am I not surprised?
4521 PH/15 If the first argument of "${if match_address" was not empty, but did not
4522 contain an "@" character, Exim crashed. Now it writes a panic log message
4523 and treats the condition as false.
4525 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4527 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4528 extension "envelope-auth". The code is finished and in agreement with
4529 other implementations, but there is no documentation so far and in fact,
4530 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4531 not changing the active code.
4533 Print executed "if" and "elsif" statements when debugging is used. This
4534 helps a great deal to understand what a filter does.
4536 Document more things not specified clearly in RFC3028. I had all this
4537 sorted out, when out of a sudden new issues came to my mind. Oops."
4539 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4542 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4543 canonical form (as documented). However, after a host name lookup from
4544 the IP address, check_host() was doing a simple string comparison with
4545 addresses acquired from the DNS when checking that the found name did
4546 have the original IP as one of its addresses. Since any found IPv6
4547 addresses are likely to be in abbreviated form, the comparison could
4548 fail. Luckily, there already exists a function for doing the comparison
4549 by converting both addresses to binary, so now that is used instead of
4550 the text comparison.
4552 PH/20 There was another similar case to PH/19, when a complete host name was
4553 given in a host list; looking up its IP address could give an abbreviated
4554 form, whereas the current host's name might or might not be abbreviated.
4555 The same fix has been applied.
4561 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4562 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4565 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4566 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4568 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4570 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4571 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4572 Also, for "Auto-Submitted", treat anything other than "no" as
4573 non-personal, in accordance with RFC 3834. (Previously it treated
4574 anything starting "auto-" as non-personal.)
4576 TF/01 The control=submission/name=... option had a problem with syntax
4577 errors if the name included a slash character. The /name= option
4578 now slurps the rest of the string, so it can include any characters
4579 but it must come last in the list of options (after /sender_retain
4582 PH/05 Some modifications to the interface to the fake nameserver for the new
4590 TK/01 Added the "success_on_redirect" address verification option. See
4591 NewStuff for rationale and an example.
4593 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4595 PH/02 Patch to exigrep to allow it to work on syslog lines.
4597 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4598 fread() to skip over the body file's header line, because in Cygwin the
4599 header line is locked and is inaccessible.
4601 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4602 co-exist for some time) to make it clear that it is the Exim ID that is
4603 referenced, not the Message-ID: header line.
4605 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4606 string_format() function, because snprintf() does not exist on all
4609 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4610 ${stat: expansion item.
4612 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4613 protocol synchronization error", to keep the pedants happy.
4615 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4616 well as for IRIX systems, when gcc is being used. See the host.c source
4619 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4621 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4624 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4625 transport and to the smtp transport in LMTP mode.
4627 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4629 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4630 run a filter in a subprocess. This could lead to confusion in subsequent
4631 lookups in the parent process. There should also be a search_tidyup() at
4632 the end of the subprocess.
4634 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4635 only if the host matched helo_try_verify_hosts, which caused the
4636 verification to occur when the EHLO/HELO command was issued. The ACL just
4637 tested the remembered result. Now, if a previous verification attempt has
4638 not happened, "verify = helo" does it there and then.
4640 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4642 TK/03 Fix log output including CR from clamd.
4644 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4645 exist provoked a memory error which could cause a segfault.
4647 PH/15 Installed PCRE 6.2
4649 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4651 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4652 of the problem. Specifically, suggested +O2 rather than +O1 for the
4655 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4657 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4658 record, the port was not correctly logged when the outgoing_port log
4659 selector was set (it logged the transort's default port).
4661 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4662 fallback_hosts, and "hosts" in the smtp transport.
4664 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4665 host errors such as "Connection refused".
4667 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4668 authentication with radiusclient 0.4.9:
4670 - Error returned from rc_read_config was caught wrongly
4671 - Username/password not passed on to radius server due to wrong length.
4673 The presumption is that some radiusclient API changes for 4.51/PH/17
4674 were not taken care of correctly. The code is still untested by me (my
4675 Linux distribution still has 0.3.2 of radiusclient), but it was
4676 contributed by a Radius user.
4678 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4679 expanding the "port" option of the smtp transport.
4681 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4682 while reading a MIME header. Thanks to Tom Hughes for a patch.
4684 PH/24 Include config.h inside local_scan.h so that configuration settings are
4687 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4688 that in an ACL for RCPT (for example), you can examine exactly what was
4691 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4692 commands, but it was not correctly comparing the address with the actual
4693 client host address. Thus, it would show the EHLO address in Received:
4694 header lines when this was not necessary.
4696 PH/27 Added the % operator to ${eval:}.
4698 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4699 it should be ignoring failures (because with -C, for example, it has lost
4700 privilege). It wasn't ignoring creation failures other than "already
4703 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4706 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4707 because an input line was too long, either on its own, or by virtue of
4708 too many continuations, the temporary file was not being removed, and the
4709 return code was incorrect.
4711 PH/31 Missing "BOOL" in function definition in filtertest.c.
4713 PH/32 Applied Sieve patches from the maintainer.
4715 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4717 PH/33 Added "verify = not_blind".
4719 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4720 Local/Makefile (with some defaults set). These are used in built scripts
4721 such as exicyclog, but they have never been used in the exim_install
4722 script (though there are many overriding facilities there). I have
4723 arranged that the exim_install script now takes note of these two
4726 PH/35 Installed configuration files for Dragonfly.
4728 PH/36 When a locally submitted message by a trusted user did not contain a
4729 From: header, and the sender address was obtained from -f or from an SMTP
4730 MAIL command, and the trusted user did not use -F to supply a sender
4731 name, $originator_name was incorrectly used when constructing a From:
4732 header. Furthermore, $originator_name was used for submission mode
4733 messages from external hosts without From: headers in a similar way,
4734 which is clearly wrong.
4736 PH/37 Added control=suppress_local_fixups.
4738 PH/38 When log_selector = +received_sender was set, and the addition of the
4739 sender made the log line's construction buffer exactly full, or one byte
4740 less than full, an overflow happened when the terminating "\n" was
4743 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4744 when the result of a list match is failure because a DNS lookup failed.
4746 PH/40 RM_COMMAND is now used in the building process.
4748 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4749 the "build-* directories that it finds.
4751 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4752 address in a domain literal was a prefix of an interface address.
4754 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4755 when verifying a sender address, unless rewrite_headers is false.
4757 PH/44 Wrote a long comment about why errors_to addresses are verified as
4758 recipients, not senders.
4760 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4761 the ratelimit ACL was added.
4763 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4765 PH/46 Added extra information about PostgreSQL errors to the error string.
4767 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4768 avoiding the need to install special zones in a real server. This is
4769 backwards compatible; if it can't find the fake resolver, it drops back.
4770 Thus, both old and new test suites can be run.
4772 TF/02 Added util/ratelimit.pl
4774 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4775 clock is set back in time.
4777 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4778 Candler <B.Candler@pobox.com>.
4780 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4781 OK for addresses that are the result of redirections.
4783 PH/48 A number of further additions for the benefit of the new test suite,
4784 including a fake gethostbyname() that interfaces to the fake DNS resolver
4787 TF/06 The fix for widen_domains has also been applied to qualify_single and
4788 search_parents which are the other dnslookup options that can cause
4791 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4794 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4796 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4797 service" if it encountered an MX record with an empty target hostname.
4798 The message is now "an MX or SRV record indicated no SMTP service".
4800 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4801 if the DNS of the sending site is misconfigured. This is quite a
4802 common situation. This change restores the behaviour of treating a
4803 helo verification defer as a failure.
4805 PH/51 If self=fail was set on a router, the bounce message did not include the
4806 actual error message.
4812 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4814 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4815 command itself ended in error, the underlying message about the transport
4816 filter timeout was being overwritten with the pipe command error. Now the
4817 underlying error message should be appended to the second error message.
4819 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4821 PH/02 Reduce the amount of output that "make" produces by default. Full output
4822 can still be requested.
4824 PH/03 The warning log line about a condition test deferring for a "warn" verb
4825 was being output only once per connection, rather than after each
4826 occurrence (because it was using the same function as for successful
4827 "warn" verbs). This seems wrong, so I have changed it.
4829 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4830 it should not have, which might have caused a crash in the right
4831 circumstances, but probably never did.
4833 PH/04 Installed a modified version of Tony Finch's patch to make submission
4834 mode fix the return path as well as the Sender: header line, and to
4835 add a /name= option so that you can make the user's friendly name appear
4838 TF/03 Added the control = fakedefer ACL modifier.
4840 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4841 Mark Lowes for thorough testing.
4843 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4845 TK/03 Merged latest SRS patch from Miles Wilton.
4847 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4848 with the definition in sysexits.h (which is #included earlier).
4849 Fortunately, Exim does not actually use EX_OK. The code used to try to
4850 preserve the sysexits.h value, by assuming that macro definitions were
4851 scanned for macro replacements. I have been disabused of this notion,
4852 so now the code just undefines EX_OK before #including unistd.h.
4854 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4855 in the smtp transport. When a block could not be written in a single
4856 write() function, the timeout was being re-applied to each part-write.
4857 This seems wrong - if the receiver was accepting one byte at a time it
4858 would take for ever. The timeout is now adjusted when this happens. It
4859 doesn't have to be particularly precise.
4861 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4862 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4864 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4865 without a domain if the check to <postmaster@domain> fails.
4867 SC/01 Eximstats: added -xls and the ability to specify output files
4868 (patch written by Frank Heydlauf).
4870 SC/02 Eximstats: use FileHandles for outputting results.
4872 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4874 SC/04 Eximstats: fixed display of large numbers with -nvr option
4876 SC/05 Eximstats: fixed merging of reports with empty tables.
4878 SC/06 Eximstats: added the -include_original_destination flag
4880 SC/07 Eximstats: removed tabs and trailing whitespace.
4882 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4884 TK/06 MBOX spool code: Add real "From " MBOX separator line
4885 so the .eml file is really in mbox format (even though
4886 most programs do not really care). Patch from Alex Miller.
4888 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4889 The latter is generated from $received_to and is only set if the
4890 message has one envelope recipient. SA can use these headers,
4891 obviously out-of-the-box. Patch from Alex Miller.
4893 PH/08 The ${def test on a variable was returning false if the variable's
4894 value was "0", contrary to what the specification has always said!
4895 The result should be true unless the variable is empty.
4897 PH/09 The syntax error of a character other than { following "${if
4898 def:variable_name" (after optional whitespace) was not being diagnosed.
4899 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4900 accidental colon was present, for example, could give incorrect results.
4902 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4903 result is used (not including appendfile, where other changes are about
4906 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4907 This involved changing a lot of size variables from int to off_t. It
4908 should work with maildirs and everything.
4910 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4911 spamd dying while we are connected to it.
4913 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4916 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4917 function for BDB 4.3.
4919 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4921 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4922 the sources for PCRE have been re-arranged and more files are now
4925 PH/15 The code I had for printing potentially long long variables in PH/11
4926 above was not the best (it lost precision). The length of off_t variables
4927 is now inspected at build time, and an appropriate printing format (%ld
4928 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4929 to be "long long int" or "long int". This is needed for the internal
4930 formatting function string_vformat().
4932 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4933 the configuration file to be ":syslog", then the script "guesses" where
4934 the logs files are, rather than using the compiled in default. In our
4935 case the guess is not the same as the compiled default, so the script
4936 suddenly stopped working when I started to use syslog. The patch checks
4937 to see if log_file_path is "". If so, it attempts to read it from exim
4938 with no configuration file to get the compiled in version, before it
4939 falls back to the previous guessing code."
4941 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4942 implementing BATV in an Exim configuration. See NewStuff for the gory
4945 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4946 Makefile that are specific to HP-UX.
4948 PH/18 If the "use_postmaster" option was set for a recipient callout together
4949 with the "random" option, the postmaster address was used as the MAIL
4950 FROM address for the random test, but not for the subsequent recipient
4951 test. It is now used for both.
4953 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4954 patch removes a few documentation additions to RFC 3028, because the
4955 latest draft now contains them. It adds the new en;ascii-case comparator
4956 and a new error check for 8bit text in MIME parts. Comparator and
4957 require names are now matched exactly. I enabled the subaddress
4958 extension, but it is not well tested yet (read: it works for me)."
4960 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4961 rework some of the code of TK/09 above to avoid the hardwired use of
4962 "%lld" and "long long". Replaced the call to snprintf() with a call to
4965 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4966 records point to non-existent hosts", "retry timeout exceeded", and
4967 "retry time not reached for any host after a long failure period".
4969 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4970 experimental DomainKeys support:
4972 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4973 (2) On an error such as an illegally used "control", the wrong name for
4974 the control was given.
4976 These problems did NOT occur unless DomainKeys support was compiled.
4978 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4980 PH/24 Added ${if match_ip condition.
4982 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4983 databases so that it will be absolutely obvious if a crash occurs in the
4984 DB library. This is a regular occurrence (often caused by mis-matched
4987 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4988 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4989 picked up on a user's system that detects such things. There doesn't seem
4990 to be a gcc warning option for this - only an attribute that has to be
4991 put on the function's prototype. It seems that in Fedora Core 4 they have
4992 set this on a number of new functions. No doubt there will be more in due
4995 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4996 specify a transport. However, if an address that was verified by such a
4997 router was the subject of a callout, Exim crashed because it tried to
4998 read the rcpt_include_affixes from the non-existent transport. Now it
4999 just assumes that the setting of that option is false. This bug was
5000 introduced by 4.51/PH/31.
5002 PH/28 Changed -d+all to exclude +memory, because that information is very
5003 rarely of interest, but it makes the output a lot bigger. People tend to
5004 do -d+all out of habit.
5006 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5007 code in os-type was giving problems when libc.so lives in lib64, like on
5010 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5011 aren't the modern standard, and it seems that some systems' include files
5012 don't always have them. Exim was already checking for some of the newer
5013 ones like T_AAAA, and defining it itself. I've added checks for all the
5014 record types that Exim uses.
5016 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5017 not automatically generating a new one, as it is supposed to. This
5018 prevented TLS from working. If the file did exist, but contained invalid
5019 data, a new version was generated, as expected. It was only the case of a
5020 non-existent file that was broken.
5022 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5023 with a change in libdomainkeys > 0.64.
5025 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5026 from DNS. If the selector record carries the flag, it now has
5027 precedence over the domain-wide flag.
5029 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5031 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5032 the use of a port name that isn't defined in /etc/services) occurred, the
5033 message was deferred as in a normal delivery, and thus remained on the
5034 spool, instead of being failed because of the mua_wrapper setting. This
5035 is now fixed, and I tidied up some of the mua_wrapper messages at the
5038 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5039 the messages in a hash of arrays rather than using individual hashes.
5040 This is a bit cleaner and results in dramatic memory savings, albeit
5041 at a slight CPU cost.
5043 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5044 as requested by Marc Sherman.
5046 SC/10 Eximstats: added histograms for user specified patterns as requested
5049 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5051 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5052 fopen() in the content-scanning modules that did not already have it.
5058 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5059 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5061 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5063 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5065 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5066 to test Sieve filters that use "vacation".
5068 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5069 that changes the way the GnuTLS parameters are stored in the cache file.
5070 The new format can be generated externally. For backward compatibility,
5071 if the data in the cache doesn't make sense, Exim assumes it has read an
5072 old-format file, and it generates new data and writes a new file. This
5073 means that you can't go back to an older release without removing the
5076 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5077 work if there are any delivery delays because "one_time" forces the
5078 parent to be marked "delivered", so its unseen clone is never tried
5079 again. For this reason, Exim now forbids the simultaneous setting of
5082 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5083 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5084 entry if you want to know the details. The fix, however, neglected to
5085 consider the case when local delivery batching is involved. The test for
5086 "previously delivered" was not happening when checking to see if an
5087 address could be batched with a previous (undelivered) one; under
5088 certain circumstances this could lead to multiple deliveries to the same
5091 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5092 in its include files, and this causes problems building Exim.
5094 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5095 header_syntax) but Exim was just ignoring anything given after a slash.
5096 In particular, this caused confusion with an attempt to use "verify =
5097 reverse_host_lookup/defer_ok". An error is now given when options are
5098 supplied for verify items that do not have them. (Maybe reverse_host_
5099 lookup should have a defer_ok option, but that's a different point.)
5101 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5102 defined by RFC 821) to 2048, because there were problems with some AUTH
5103 commands, and RFC 1869 says the size should be increased for extended
5104 SMTP commands that take arguments.
5106 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5109 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5110 "unknown" error; now it says that the functionality isn't in the binary.
5112 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5113 an address' error message when a string expansion fails (syntax or
5114 whatever). Otherwise the password may appear in the log. Following change
5115 PH/42 below, there is no longer a chance of it appearing in a bounce
5118 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5120 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5121 of its addresses was ever tried. (Bugzilla bug #2.)
5123 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5124 the result incorrectly in the debug output. (It correctly added a newline
5125 to what was transported.)
5127 TF/01 Added $received_time.
5129 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5130 commented out examples of how to interface to a virus scanner and to
5131 SpamAssassin. Also added commented examples of av_scanner and
5132 spamd_address settings.
5134 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5135 and controls are allowed in which ACLs. There were a couple of minor
5136 errors. Some of the entries in the conditions table (which is a table of
5137 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5138 negation of where the condition IS allowed.
5140 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5142 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5143 header file does not have a version number, so I've had to invent a new
5144 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5145 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5146 radiusclient), but it was contributed by a Radius user.
5148 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5149 files or directories, for OpenSSL.
5151 PH/19 When an Exim process that is running as root has to create an Exim log
5152 file, it does so in a subprocess that runs as exim:exim so as to get the
5153 ownership right at creation (otherwise, other Exim processes might see
5154 the file with the wrong ownership). There was no test for failure of this
5155 fork() call, which would lead to the process getting stuck as it waited
5156 for a non-existent subprocess. Forks do occasionally fail when resources
5157 run out. I reviewed all the other calls to fork(); they all seem to check
5160 PH/20 When checking for unexpected SMTP input at connect time (before writing
5161 the banner), Exim was not dealing correctly with a non-positive return
5162 from the read() function. If the client had disconnected by this time,
5163 the result was a log entry for a synchronization error with an empty
5164 string after "input=" when read() returned zero. If read() returned -1
5165 (an event I could not check), uninitialized data bytes were printed.
5166 There were reports of junk text (parts of files, etc) appearing after
5169 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5171 PH/22 Added support for macro redefinition, and (re)definition in between
5172 driver and ACL definitions.
5174 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5175 forgetting to use the resulting value; it was using the unexpanded value.
5177 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5178 hadn't been configured. The fix is from Juergen Kreileder, who
5179 understands it better than I do:
5181 "Here's what I see happening with three configured cyrus_sasl
5182 authenticators configured (plain, login, cram-md5):
5184 On startup auth_cyrus_sasl_init() gets called for each of these.
5185 This means three calls to sasl_listmech() without a specified mech_list.
5186 => SASL tests which mechs of all available mechs actually work
5187 => three warnings about OTP not working
5188 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5190 With the patch, sasl_listmech() also gets called three times. But now
5191 SASL's mech_list option is set to the server_mech specified in the the
5192 authenticator. Or in other words, the answer from sasl_listmech()
5193 gets limited to just the mech you're testing for (which is different
5195 => the return list contains just 'plain' or 'login', 'cram-md5' or
5196 nothing depending on the value of ob->server_mech.
5198 I've just tested the patch: Authentication still works fine,
5199 unavailable mechs specified in the exim configuration are still
5200 caught, and the auth.log warnings about OTP are gone."
5202 PH/25 When debugging is enabled, the contents of the command line are added
5203 to the debugging output, even when log_selector=+arguments is not
5206 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5207 answer is "GNU", and only if the return is "GNU/something" is the answer
5210 PH/27 $acl_verify_message is now set immediately after the failure of a
5211 verification in an ACL, and so is available in subsequent modifiers. In
5212 particular, the message can be preserved by coding like this:
5214 warn !verify = sender
5215 set acl_m0 = $acl_verify_message
5217 Previously, $acl_verify_message was set only while expanding "message"
5218 and "log_message" when a very denied access.
5220 PH/28 Modified OS/os.c-Linux with
5222 -#ifndef OS_LOAD_AVERAGE
5223 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5225 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5226 nomenclature these days.)
5228 PH/29 Installed patch from the Sieve maintainer that adds the options
5229 sieve_useraddress and sieve_subaddress to the redirect router.
5231 PH/30 In these circumstances:
5232 . Two addresses routed to the same list of hosts;
5233 . First host does not offer TLS;
5234 . First host accepts first address;
5235 . First host gives temporary error to second address;
5236 . Second host offers TLS and a TLS session is established;
5237 . Second host accepts second address.
5238 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5239 and peerdn, if requested) that were in fact used only for the second
5242 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5243 attention to any local part prefix or suffix that was matched by the
5244 router that accepted the address. It now behaves in the same way as it
5245 does for delivery: the affixes are removed from the local part unless
5246 rcpt_include_affixes is set on the transport.
5248 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5249 timeout during the DATA phase of an incoming message.
5251 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5252 applied a patch sanctioned by the Sieve maintainer.
5254 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5255 the uid or gid is negative. A case of a negative gid caused this to be
5256 noticed. The fix allows for either to be negative.
5258 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5259 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5262 PH/36 The change PH/12 above was broken. Fixed it.
5264 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5265 the grounds that routing the same address twice would always produce the
5266 same answer. This might have been true once, but it is certainly no
5267 longer true now. Routing a child address may depend on the previous
5268 routing that produced that child. Some complicated redirection strategies
5269 went wrong when messages had multiple recipients, and made Exim's
5270 behaviour dependent on the order in which the addresses were given.
5272 I have moved the duplicate checking until after the routing is complete.
5273 Exim scans the addresses that are assigned to local and remote
5274 transports, and removes any duplicates. This means that more work will be
5275 done, as duplicates will always all be routed, but duplicates are
5276 presumably rare, so I don't expect this is of any significance.
5278 For deliveries to pipes, files, and autoreplies, the duplicate checking
5279 still happens during the routing process, since they are not going to be
5282 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5283 It corrects a timeout issue with spamd. This is Ian's comment: "The
5284 background is that sometimes spamd either never reads data from a
5285 connection it has accepted, or it never writes response data. The exiscan
5286 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5287 blindly assumes that writes won't block so it may never time out."
5289 PH/39 Allow G after quota size as well as K and M.
5291 PH/40 The value set for $authenticated_id in an authenticator may not contain
5292 binary zeroes or newlines because the value is written to log lines and
5293 to spool files. There was no check on this. Now the value is run through
5294 the string_printing() function so that such characters are converted to
5295 printable escape sequences.
5297 PH/41 $message_linecount is a new variable that contains the total number of
5298 lines in the message. Compare $body_linecount, which is the count for the
5301 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5302 bounce and delay warning messages, except in certain special cases, which
5305 (a) An SMTP error message from a remote host;
5306 (b) A message specified in a :fail: redirection;
5307 (c) A message specified in a "fail" command in a system filter;
5308 (d) A message specified in a FAIL return from the queryprogram router;
5309 (e) A message specified by the cannot_route_message router option.
5311 In these cases only, Exim does include the error details in bounce and
5312 warning messages. There are also a few cases where bland messages such
5313 as "unrouteable address" or "local delivery error" are given.
5315 PH/43 $value is now also set for the "else" part of a ${run expansion.
5317 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5318 being worked on, but at least Exim now implements the latest version to
5321 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5322 process to complete was treated as a delivery failure, a timeout while
5323 writing the message to the pipe was logged, but erroneously treated as a
5324 successful delivery. Such timeouts include transport filter timeouts. For
5325 consistency with the overall process timeout, these timeouts are now
5326 treated as errors, giving rise to delivery failures by default. However,
5327 there is now a new Boolean option for the pipe transport called
5328 timeout_defer, which, if set TRUE, converts the failures into defers for
5329 both kinds of timeout. A transport filter timeout is now identified in
5332 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5333 systems where "make" and "gmake" are different, calling "gmake" at top
5334 level broke things. I've arranged for the value of $(MAKE) to be passed
5335 from the Makefile to this script so that it can call the same version of
5339 A note about Exim versions 4.44 and 4.50
5340 ----------------------------------------
5342 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5343 changes of various kinds. As a consequence, a big documentation update was
5344 needed. This delayed the release for rather longer than seemed good, especially
5345 in the light of a couple of (minor) security issues. Therefore, the changes
5346 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5347 release. So 4.44 and 4.50 are in effect two different branches that both start
5350 I have left the 4.50 change log unchanged; it contains all the changes since
5351 4.43. The change log for 4.44 is below; many of its items are identical to
5352 those for 4.50. This seems to be the most sensible way to preserve the
5353 historical information.
5359 1. Minor wording change to the doc/README.SIEVE file.
5361 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5362 computation of the current number of files was incorrect.
5364 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5365 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5368 4. Give more explanation in the error message when the command for a transport
5369 filter fails to execute.
5371 5. There are several places where Exim runs a non-Exim command in a
5372 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5373 was being done only for the command run by the queryprogram router. It is
5374 now done for all such subprocesses. The other cases are: ${run, transport
5375 filters, and the commands run by the lmtp and pipe transports.
5377 6. Added CONFIGURE_GROUP build-time option.
5379 7. Some older OS have a limit of 256 on the maximum number of file
5380 descriptors. Exim was using setrlimit() to set 1000 as a large value
5381 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5382 systems. I've change it so that if it can't get 1000, it tries for 256.
5384 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5385 was an oversight, and furthermore, ever since the addition of extra
5386 controls (e.g. 4.43/32), the checks on when to allow different forms of
5387 "control" were broken. There should now be diagnostics for all cases when a
5388 control that does not make sense is encountered.
5390 9. Added the /retain_sender option to "control=submission".
5392 10. $recipients is now available in the predata ACL (oversight).
5394 11. Tidy the search cache before the fork to do a delivery from a message
5395 received from the command line. Otherwise the child will trigger a lookup
5396 failure and thereby defer the delivery if it tries to use (for example) a
5397 cached ldap connection that the parent has called unbind on.
5399 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5400 of $address_data from the recipient verification was clobbered by the
5401 sender verification.
5403 13. The value of address_data from a sender verification is now available in
5404 $sender_address_data in subsequent conditions in the ACL statement.
5406 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5408 15. Added a new option "connect=<time>" to callout options, to set a different
5411 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5412 was its contents. (It was OK if the option was not defined at all.)
5414 17. A "Completed" log line is now written for messages that are removed from
5415 the spool by the -Mrm option.
5417 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5418 information about exactly what failed.
5420 19. Added -dd to debug only the daemon process.
5422 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5423 handles timeouts, both on the server side and network timeouts. Renamed the
5424 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5426 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5427 It is now set to "smtps".
5429 22. $host_address is now set to the target address during the checking of
5430 ignore_target_hosts.
5432 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5433 being passed; this would have caused $sender_host_name to have been used if
5434 matching the list had actually called for a host name (not very likely,
5435 since this list is usually IP addresses). A host name is now passed as
5438 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5439 code that specifies a non-restarting handler (typically sigaction() in
5440 modern systems) in an attempt to fix a rare and obscure crash bug.
5442 25. Narrowed the window for a race in the daemon that could cause it to ignore
5443 SIGCHLD signals. This is not a major problem, because they are used only to
5444 wake it up if nothing else does.
5446 26. A malformed maildirsize file could cause Exim to calculate negative values
5447 for the mailbox size or file count. Odd effects could occur as a result.
5448 The maildirsize information is now recalculated if the size or filecount
5451 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5452 support for a long time. Removed HAVE_SYS_VFS_H.
5454 28. Installed the latest version of exipick from John Jetmore.
5456 29. In an address list, if the pattern was not a regular expression, an empty
5457 subject address (from a bounce message) matched only if the pattern was an
5458 empty string. Non-empty patterns were not even tested. This was the wrong
5459 because it is perfectly reasonable to use an empty address as part of a
5460 database query. An empty address is now tested by patterns that are
5461 lookups. However, all the other forms of pattern expect the subject to
5462 contain a local part and a domain, and therefore, for them, an empty
5463 address still always fails if the pattern is not itself empty.
5465 30. Exim went into a mad DNS loop when attempting to do a callout where the
5466 host was specified on an smtp transport, and looking it up yielded more
5467 than one IP address.
5469 31. Re-factored the code for checking spool and log partition space into a
5470 function that finds that data and another that does the check. The former
5471 is then used to implement four new variables: $spool_space, $log_space,
5472 $spool_inodes, and $log_inodes.
5474 32. The RFC2047 encoding function was originally intended for short strings
5475 such as real names; it was not keeping to the 75-character limit for
5476 encoded words that the RFC imposes. It now respects the limit, and
5477 generates multiple encoded words if necessary. To be on the safe side, I
5478 have increased the buffer size for the ${rfc2047: expansion operator from
5481 33. It is now permitted to omit both strings after an "if" condition; if the
5482 condition is true, the result is "true". As before, when the second string
5483 is omitted, a false condition yields an empty string. This makes it less
5484 cumbersome to write custom ACL and router conditions.
5486 34. Failure to deliver a bounce message always caused it to be frozen, even if
5487 there was an errors_to setting on the router. The errors_to setting is now
5490 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5491 canonical form (fully expanded) before being placed in
5492 $sender_host_address.
5494 36. The table in the code that translates DNS record types into text (T_A to
5495 "A" for instance) was missing entries for NS and CNAME. It is just possible
5496 that this could have caused confusion if both these types were looked up
5497 for the same domain, because the text type is used as part of Exim's
5498 per-process caching. But the chance of anyone hitting this buglet seems
5501 37. The dnsdb lookup has been extended in a number of ways.
5503 (1) There is a new type, "zns", which walks up the domain tree until it
5504 finds some nameserver records. It should be used with care.
5506 (2) There is a new type, "mxh", which is like "mx" except that it returns
5507 just the host names, not the priorities.
5509 (3) It is now possible to give a list of domains (or IP addresses) to be
5510 looked up. The behaviour when one of the lookups defers can be
5511 controlled by a keyword.
5513 (4) It is now possible to specify the separator character for use when
5514 multiple records are returned.
5516 38. The dnslists ACL condition has been extended: it is now possible to supply
5517 a list of IP addresses and/or domains to be looked up in a particular DNS
5520 39. Added log_selector=+queue_time_overall.
5522 40. When running the queue in the test harness, wait just a tad after forking a
5523 delivery process, to get repeatability of debugging output.
5525 41. Include certificate and key file names in error message when GnuTLS fails
5526 to set them up, because the GnuTLS error message doesn't include the name
5527 of the failing file when there is a problem reading it.
5529 42. Allow both -bf and -bF in the same test run.
5531 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5533 44. The "Exiscan patch" is now merged into the mainline Exim source.
5535 45. Sometimes the final signoff response after QUIT could fail to get
5536 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5537 < 0 before doing a fflush(). This bug looks as though it goes back to the
5538 introduction of TLS in release 3.20, but "sometimes" must have been rare
5539 because the tests only now provoked it.
5541 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5542 (this can affect the format of dates).
5544 47. exim_tidydb, when checking for the continued existence of a message for
5545 which it has found a message-specific retry record, was not finding
5546 messages that were in split spool directories. Consequently, it was
5547 deleting retry records that should have stayed in existence.
5549 48. Steve fixed some bugs in eximstats.
5551 49. The SPA authentication driver was not abandoning authentication and moving
5552 on to the next authenticator when an expansion was forced to fail,
5553 contradicting the general specification for all authenticators. Instead it
5554 was generating a temporary error. It now behaves as specified.
5556 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5557 (the order specifies the preference for clients). The order is now AES256,
5558 AES128, 3DES, ARCFOUR128.
5560 51. Small patch to Sieve code - explicitly set From: when generating an
5563 52. Exim crashed if a remote delivery caused a very long error message to be
5564 recorded - for instance if somebody sent an entire SpamAssassin report back
5565 as a large number of 550 error lines. This bug was coincidentally fixed by
5566 increasing the size of one of Exim's internal buffers (big_buffer) that
5567 happened as part of the Exiscan merge. However, to be on the safe side, I
5568 have made the code more robust (and fixed the comments that describe what
5571 53. Now that there can be additional text after "Completed" in log lines (if
5572 the queue_time_overall log selector is set), a one-byte patch to exigrep
5573 was needed to allow it to recognize "Completed" as not the last thing in
5576 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5577 patch that reportedly fixes this has been added. I am not expert enough to
5578 create a test for it. This is what the patch creator wrote:
5580 "I found a little strange behaviour of ldap code when working with
5581 Windows 2003 AD Domain, where users was placed in more than one
5582 Organization Units. When I tried to give exim partial DN, the exit code
5583 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5584 But simultaneously result of request was absolutely normal ldap result,
5585 so I produce this patch..."
5587 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5588 REFERENCE, so I have modified the code to exclude the patch when that macro
5591 55. Some experimental protocols are using DNS PTR records for new purposes. The
5592 keys for these records are domain names, not reversed IP addresses. The
5593 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5594 leaves it alone. Component reversal etc. now happens only for IP addresses.
5597 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5599 57. Double the size of the debug message buffer (to 2048) so that more of very
5600 long debug lines gets shown.
5602 58. The exicyclog utility now does better if the number of log files to keep
5603 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5605 59. Two changes related to the smtp_active_hostname option:
5607 (1) $smtp_active_hostname is now available as a variable.
5608 (2) The default for smtp_banner uses $smtp_active_hostname instead
5609 of $primary_hostname.
5611 60. The host_aton() function is supposed to be passed a string that is known
5612 to be a valid IP address. However, in the case of IPv6 addresses, it was
5613 not checking this. This is a hostage to fortune. Exim now panics and dies
5614 if the condition is not met. A case was found where this could be provoked
5615 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5616 components; fortuitously, this particular loophole had already been fixed
5617 by change 4.50/55 above.
5619 If there are any other similar loopholes, the new check in host_aton()
5620 itself should stop them being exploited. The report I received stated that
5621 data on the command line could provoke the exploit when Exim was running as
5622 exim, but did not say which command line option was involved. All I could
5623 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5624 running as the user.
5627 61. There was a buffer overflow vulnerability in the SPA authentication code
5628 (which came originally from the Samba project). I have added a test to the
5629 spa_base64_to_bits() function which I hope fixes it.
5632 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5633 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5635 63. The daemon start-up calls getloadavg() while still root for those OS that
5636 need the first call to be done as root, but it missed one case: when
5637 deliver_queue_load_max is set with deliver_drop_privilege. This is
5638 necessary for the benefit of the queue runner, because there is no re-exec
5639 when deliver_drop_privilege is set.
5641 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5642 This has been fixed.
5644 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5645 was in use, was not putting the data itself into the right store pool;
5646 consequently, it could be overwritten for a subsequent message in the same
5647 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5650 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5652 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5653 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5654 the function were treating the return as a boolean value, which happened to
5655 work because 0=false and not-0=true, but is not correct code.
5657 68. The host_aton() function was not handling scoped IPv6 addresses (those
5658 with, for example, "%eth0" on the end) correctly.
5660 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5661 negated items (that is, ~something) in unsigned ints. Some compilers
5662 apparently mutter when there is no cast.
5664 70. If an address verification called from an ACL failed, and did not produce a
5665 user-specific message (i.e. there was only a "system" message), nothing was
5666 put in $acl_verify_message. In this situation, it now puts the system
5669 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5670 session; change 4.31/43 added the unwanted input to the log line - except
5671 that it did not do this in the start of session case. It now does.
5673 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5674 This is wrong and can cause the other end to generate a synchronization
5675 error if it is another Exim or anything else that does the synchronization
5676 check. A QUIT command is no longer sent after a timeout.
5678 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5679 during host lookups.
5681 74. The defer_ok option of callout verification was not working if it was used
5682 when verifying addresses in header lines, that is, for this case:
5684 verify = header_sender/callout=defer_ok
5686 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5687 those file descriptors could be used for SMTP connections. If anything
5688 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5689 could be sent to the SMTP client, causing chaos. The daemon now opens
5690 stdin, stdout, and stderr to /dev/null when it puts itself into the
5693 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5694 log by default. The user can override this with suitable Perl magic.
5696 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5697 the log message when discard triggers, was not working for the DATA ACL or
5698 for the non-SMTP ACL.
5700 78. Error message wording change in sieve.c.
5702 79. If smtp_accept_max_per_host was set, the number of connections could be
5703 restricted to fewer than expected, because the daemon was trying to set up
5704 a new connection before checking whether the processes handling previous
5705 connections had finished. The check for completed processes is now done
5706 earlier. On busy systems, this bug wouldn't be noticed because something
5707 else would have woken the daemon, and it would have reaped the completed
5710 80. If a message was submitted locally by a user whose login name contained one
5711 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5712 It caused a spool format error. I have fixed the spool reading code. A
5713 related problem was that the "from" clause in the Received: line became
5714 illegal because of the space(s). It is now covered by ${quote_local_part.
5716 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5719 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5721 83. Patch from Sieve maintainer to fix unterminated string problem in
5722 "vacation" handling.
5724 84. Some minor changes to the Linux configuration files to help with other
5725 OS variants using glibc.
5727 85. One more patch for Sieve to update vacation handling to latest spec.
5730 ----------------------------------------------------
5731 See the note above about the 4.44 and 4.50 releases.
5732 ----------------------------------------------------
5738 1. Change 4.43/35 introduced a bug that caused file counts to be
5739 incorrectly computed when quota_filecount was set in an appendfile
5742 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5743 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5746 3. Give more explanation in the error message when the command for a transport
5747 filter fails to execute.
5749 4. There are several places where Exim runs a non-Exim command in a
5750 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5751 was being done only for the command run by the queryprogram router. It is
5752 now done for all such subprocesses. The other cases are: ${run, transport
5753 filters, and the commands run by the lmtp and pipe transports.
5755 5. Some older OS have a limit of 256 on the maximum number of file
5756 descriptors. Exim was using setrlimit() to set 1000 as a large value
5757 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5758 systems. I've change it so that if it can't get 1000, it tries for 256.
5760 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5761 was an oversight, and furthermore, ever since the addition of extra
5762 controls (e.g. 4.43/32), the checks on when to allow different forms of
5763 "control" were broken. There should now be diagnostics for all cases when a
5764 control that does not make sense is encountered.
5766 7. $recipients is now available in the predata ACL (oversight).
5768 8. Tidy the search cache before the fork to do a delivery from a message
5769 received from the command line. Otherwise the child will trigger a lookup
5770 failure and thereby defer the delivery if it tries to use (for example) a
5771 cached ldap connection that the parent has called unbind on.
5773 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5774 of $address_data from the recipient verification was clobbered by the
5775 sender verification.
5777 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5778 was its contents. (It was OK if the option was not defined at all.)
5780 11. A "Completed" log line is now written for messages that are removed from
5781 the spool by the -Mrm option.
5783 12. $host_address is now set to the target address during the checking of
5784 ignore_target_hosts.
5786 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5787 being passed; this would have caused $sender_host_name to have been used if
5788 matching the list had actually called for a host name (not very likely,
5789 since this list is usually IP addresses). A host name is now passed as
5792 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5793 code that specifies a non-restarting handler (typically sigaction() in
5794 modern systems) in an attempt to fix a rare and obscure crash bug.
5796 15. Narrowed the window for a race in the daemon that could cause it to ignore
5797 SIGCHLD signals. This is not a major problem, because they are used only to
5798 wake it up if nothing else does.
5800 16. A malformed maildirsize file could cause Exim to calculate negative values
5801 for the mailbox size or file count. Odd effects could occur as a result.
5802 The maildirsize information is now recalculated if the size or filecount
5805 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5806 support for a long time. Removed HAVE_SYS_VFS_H.
5808 18. Updated exipick to current release from John Jetmore.
5810 19. Allow an empty sender to be matched against a lookup in an address list.
5811 Previously the only cases considered were a regular expression, or an
5814 20. Exim went into a mad DNS lookup loop when doing a callout where the
5815 host was specified on the transport, if the DNS lookup yielded more than
5818 21. The RFC2047 encoding function was originally intended for short strings
5819 such as real names; it was not keeping to the 75-character limit for
5820 encoded words that the RFC imposes. It now respects the limit, and
5821 generates multiple encoded words if necessary. To be on the safe side, I
5822 have increased the buffer size for the ${rfc2047: expansion operator from
5825 22. Failure to deliver a bounce message always caused it to be frozen, even if
5826 there was an errors_to setting on the router. The errors_to setting is now
5829 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5830 canonical form (fully expanded) before being placed in
5831 $sender_host_address.
5833 24. Updated eximstats to version 1.33
5835 25. Include certificate and key file names in error message when GnuTLS fails
5836 to set them up, because the GnuTLS error message doesn't include the name
5837 of the failing file when there is a problem reading it.
5839 26. Expand error message when OpenSSL has problems setting up cert/key files.
5842 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5843 (this can affect the format of dates).
5845 28. exim_tidydb, when checking for the continued existence of a message for
5846 which it has found a message-specific retry record, was not finding
5847 messages that were in split spool directories. Consequently, it was
5848 deleting retry records that should have stayed in existence.
5850 29. eximstats updated to version 1.35
5851 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5852 1.35 - bugfix such that pie charts by volume are generated correctly
5854 30. The SPA authentication driver was not abandoning authentication and moving
5855 on to the next authenticator when an expansion was forced to fail,
5856 contradicting the general specification for all authenticators. Instead it
5857 was generating a temporary error. It now behaves as specified.
5859 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5860 (the order specifies the preference for clients). The order is now AES256,
5861 AES128, 3DES, ARCFOUR128.
5863 31. Small patch to Sieve code - explicitly set From: when generating an
5866 32. Exim crashed if a remote delivery caused a very long error message to be
5867 recorded - for instance if somebody sent an entire SpamAssassin report back
5868 as a large number of 550 error lines. This bug was coincidentally fixed by
5869 increasing the size of one of Exim's internal buffers (big_buffer) that
5870 happened as part of the Exiscan merge. However, to be on the safe side, I
5871 have made the code more robust (and fixed the comments that describe what
5874 33. Some experimental protocols are using DNS PTR records for new purposes. The
5875 keys for these records are domain names, not reversed IP addresses. The
5876 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5877 leaves it alone. Component reversal etc. now happens only for IP addresses.
5880 34. The host_aton() function is supposed to be passed a string that is known
5881 to be a valid IP address. However, in the case of IPv6 addresses, it was
5882 not checking this. This is a hostage to fortune. Exim now panics and dies
5883 if the condition is not met. A case was found where this could be provoked
5884 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5885 components; fortuitously, this particular loophole had already been fixed
5886 by change 4.50/55 or 4.44/33 above.
5888 If there are any other similar loopholes, the new check in host_aton()
5889 itself should stop them being exploited. The report I received stated that
5890 data on the command line could provoke the exploit when Exim was running as
5891 exim, but did not say which command line option was involved. All I could
5892 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5893 running as the user.
5896 35. There was a buffer overflow vulnerability in the SPA authentication code
5897 (which came originally from the Samba project). I have added a test to the
5898 spa_base64_to_bits() function which I hope fixes it.
5901 36. The daemon start-up calls getloadavg() while still root for those OS that
5902 need the first call to be done as root, but it missed one case: when
5903 deliver_queue_load_max is set with deliver_drop_privilege. This is
5904 necessary for the benefit of the queue runner, because there is no re-exec
5905 when deliver_drop_privilege is set.
5907 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5908 was in use, was not putting the data itself into the right store pool;
5909 consequently, it could be overwritten for a subsequent message in the same
5910 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5913 38. Sometimes the final signoff response after QUIT could fail to get
5914 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5915 < 0 before doing a fflush(). This bug looks as though it goes back to the
5916 introduction of TLS in release 3.20, but "sometimes" must have been rare
5917 because the tests only now provoked it.
5923 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5924 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5925 Now it returns an integer. A number of calls were still expecting a T/F
5926 return. Fortuitously, in all cases, the tests worked in OK situations,
5927 which is the norm. However, things would have gone wrong on any write
5928 failures on the smtp file descriptor. This function is used when sending
5929 messages over SMTP and also when doing verify callouts.
5931 2. When Exim is called to do synchronous delivery of a locally submitted
5932 message (the -odf or -odi options), it no longer closes stderr before doing
5935 3. Implemented the mua_wrapper option.
5937 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5939 5. Implemented the functions header_remove(), header_testname(),
5940 header_add_at_position(), and receive_remove_recipient(), and exported them
5943 6. If an ACL "warn" statement specified the addition of headers, Exim already
5944 inserted X-ACL-Warn: at the start if there was no header name. However, it
5945 was not making this test for the second and subsequent header lines if
5946 there were newlines in the string. This meant that an invalid header could
5947 be inserted if Exim was badly configured.
5949 7. Allow an ACL "warn" statement to add header lines at the start or after all
5950 the Received: headers, as well as at the end.
5952 8. Added the rcpt_4xx retry error code.
5954 9. Added postmaster_mailfrom=xxx to callout verification option.
5956 10. Added mailfrom=xxxx to the callout verification option, for verify=
5959 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5960 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5962 12. Inserted some casts to stop certain compilers warning when using pointer
5963 differences as field lengths or precisions in printf-type calls (mostly
5964 affecting debugging statements).
5966 13. Added optional readline() support for -be (dynamically loaded).
5968 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5969 same clock tick as a message's arrival, so that its received time was the
5970 same as the "first fail" time on the retry record, and that message
5971 remained on the queue past the ultimate address timeout, every queue runner
5972 would try a delivery (because it was past the ultimate address timeout) but
5973 after another failure, the ultimate address timeout, which should have then
5974 bounced the address, did not kick in. This was a "< instead of <=" error;
5975 in most cases the first failure would have been in the next clock tick
5976 after the received time, and all would be well.
5978 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5979 being recognized when the domain list was tested by the match_domain
5980 condition in an expansion string.
5982 16. Added the ${str2b64: operator.
5984 17. Exim was always calling setrlimit() to set a large limit for the number of
5985 processes, without checking whether the existing limit was already
5986 adequate. (It did check for the limit on file descriptors.) Furthermore,
5987 errors from getrlimit() and setrlimit() were being ignored. Now they are
5988 logged to the main and panic logs, but Exim does carry on, to try to do its
5989 job under whatever limits there are.
5991 18. Imported PCRE 5.0.
5993 19. Trivial typo in log message " temporarily refused connection" (the leading
5996 20. If the log selector return_path_on_delivery was set and an address was
5997 redirected to /dev/null, the delivery process crashed because it assumed
5998 that a return path would always be set for a "successful" delivery. In this
5999 case, the whole delivery is bypassed as an optimization, and therefore no
6002 21. Internal re-arrangement: the function for sending a challenge and reading
6003 a response while authentication was assuming a zero-terminated challenge
6004 string. It's now changed to take a pointer and a length, to allow for
6005 binary data in such strings.
6007 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6009 23. Exim was not respecting finduser_retries when seeking the login of the
6010 uid under which it was called; it was always trying 10 times. (The default
6011 setting of finduser_retries is zero.) Also, it was sleeping after the final
6012 failure, which is pointless.
6014 24. Implemented tls_on_connect_ports.
6016 25. Implemented acl_smtp_predata.
6018 26. If the domain in control=submission is set empty, Exim assumes that the
6019 authenticated id is a complete email address when it generates From: or
6020 Sender: header lines.
6022 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6023 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6024 chown and chgrp in /bin and hostname in /usr/bin.
6026 28. Exim was keeping the "process log" file open after each use, just as it
6027 does for the main log. This opens the possibility of it remaining open for
6028 long periods when the USR1 signal hits a daemon. Occasional processlog
6029 errors were reported, that could have been caused by this. Anyway, it seems
6030 much more sensible not to leave this file open at all, so that is what now
6033 29. The long-running daemon process does not normally write to the log once it
6034 has entered its main loop, and it closes the log before doing so. This is
6035 so that log files can straightforwardly be renamed and moved. However,
6036 there are a couple of unusual error situations where the daemon does write
6037 log entries, and I had neglected to close the log afterwards.
6039 30. The text of an SMTP error response that was received during a remote
6040 delivery was being truncated at 512 bytes. This is too short for some of
6041 the long messages that one sometimes sees. I've increased the limit to
6044 31. It is now possible to make retry rules that apply only when a message has a
6045 specific sender, in particular, an empty sender.
6047 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6048 it possible to be selective about when SMTP synchronization is enforced.
6050 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6052 32. Implemented hosts_connection_nolog.
6054 33. Added an ACL for QUIT.
6056 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6059 35. Added mailbox_size and mailbox_filecount to appendfile.
6061 36. Added control = no_multiline_responses to ACLs.
6063 37. There was a bug in the logic of the code that waits for the clock to tick
6064 in the case where the clock went backwards by a substantial amount such
6065 that the microsecond fraction of "now" was more than the microsecond
6066 fraction of "then" (but the whole seconds number was less).
6068 38. Added support for the libradius Radius client library this is found on
6069 FreeBSD (previously only the radiusclient library was supported).
6075 1. When certain lookups returned multiple values in the form name=value, the
6076 quoting of the values was not always being done properly. Specifically:
6077 (a) If the value started with a double quote, but contained no whitespace,
6079 (b) If the value contained whitespace other than a space character (i.e.
6080 tabs or newlines or carriage returns) it was not quoted.
6081 This fix has been applied to the mysql and pgsql lookups by writing a
6082 separate quoting function and calling it from the lookup code. The fix
6083 should probably also be applied to nisplus, ibase and oracle lookups, but
6084 since I cannot test any of those, I have not disturbed their existing code.
6086 2. A hit in the callout cache for a specific address caused a log line with no
6087 reason for rejecting RCPT. Now it says "Previous (cached) callout
6088 verification failure".
6090 3. There was an off-by-one bug in the queryprogram router. An over-long
6091 return line was truncated at 256 instead of 255 characters, thereby
6092 overflowing its buffer with the terminating zero. As well as fixing this, I
6093 have increased the buffer size to 1024 (and made a note to document this).
6095 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6096 when Exim is waiting for an SMTP response from a remote server, Exim
6097 restarts its select() call on the socket, thereby resetting its timeout.
6098 This is not a problem when such interrupts are rare. Somebody set up a cron
6099 job to run exiwhat every 2 minutes, which is less than the normal select()
6100 timeout (5 or 10 minutes). This meant that the select() timeout never
6101 kicked in because it was always reset. I have fixed this by comparing the
6102 time when an interrupt arrives with the time at the start of the first call
6103 to select(). If more time than the timeout has elapsed, the interrupt is
6104 treated as a timeout.
6106 5. Some internal re-factoring in preparation for the addition of Sieve
6107 extensions (by MH). In particular, the "personal" test is moved to a
6108 separate function, and given an option for scanning Cc: and Bcc: (which is
6109 not set for Exim filters).
6111 6. When Exim created an email address using the login of the caller as the
6112 local part (e.g. when creating a From: or Sender: header line), it was not
6113 quoting the local part when it contained special characters such as @.
6115 7. Installed new OpenBSD configuration files.
6117 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6118 try to make them clearer.
6120 9. Callout options, other than the timeout value, were being ignored when
6121 verifying sender addresses in header lines. For example, when using
6123 verify = header_sender/callout=no_cache
6125 the cache was (incorrectly) being used.
6127 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6128 only the Cygwin environment.
6130 11. When return_path_on_delivery was set as a log selector, if different remote
6131 addresses in the same message used different return paths and parallel
6132 remote delivery occurred, the wrong values would sometimes be logged.
6133 (Whenever a remote delivery process finished, the return path value from
6134 the most recently started remote delivery process was logged.)
6136 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6137 lines when AUTH and/or TLS are in use. This is the "received protocol"
6138 field. Exim used to use "asmtp" for authenticated SMTP, without any
6139 indication (in the protocol name) for TLS use. Now it follows the RFC and
6140 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6141 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6142 names appear in log lines as well as in Received: header lines.
6144 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6145 extensions, and comparison tests, and to fix some bugs.
6147 14. Changes to the "personal" filter test:
6149 (1) The test was buggy in that it was just doing the equivalent of
6150 "contains" tests on header lines. For example, if a user's address was
6151 anne@some.where, the "personal" test would incorrectly be true for
6153 To: susanne@some.where
6155 This test is now done by extracting each address from the header in turn,
6156 and checking the entire address. Other tests that are part of "personal"
6157 are now done using regular expressions (for example, to check local parts
6158 of addresses in From: header lines).
6160 (2) The list of non-personal local parts in From: addresses has been
6161 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6162 taken from the Sieve specification recommendations.
6164 (3) If the message contains any header line starting with "List-" it is
6165 treated as non-personal.
6167 (4) The test for "circular" in the Subject: header line has been removed
6168 because it now seems ill-conceived.
6170 15. Minor typos in src/EDITME comments corrected.
6172 16. Installed latest exipick from John Jetmore.
6174 17. If headers_add on a router specified a text string that was too long for
6175 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6176 of string_sprintf() is now avoided.
6178 18. $message_body_size was not set (it was always zero) when running the DATA
6179 ACL and the local_scan() function.
6181 19. For the "mail" command in an Exim filter, no default was being set for
6182 the once_repeat time, causing a random time value to be used if "once" was
6183 specified. (If the value happened to be <= 0, no repeat happened.) The
6184 default is now 0s, meaning "never repeat". The "vacation" command was OK
6185 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6186 (I found it when inspecting the code).
6188 20. There is now an overall timeout for performing a callout verification. It
6189 defaults to 4 times the callout timeout, which applies to individual SMTP
6190 commands during the callout. The overall timeout applies when there is more
6191 than one host that can be tried. The timeout is checked before trying the
6192 next host. This prevents very long delays if there are a large number of
6193 hosts and all are timing out (e.g. when the network connections are timing
6194 out). The value of the overall timeout can be changed by specifying an
6195 additional sub-option for "callout", called "maxwait". For example:
6197 verify = sender/callout=5s,maxwait=20s
6199 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6200 to the end before writing, but this should make it even safer).
6202 22. Exim was forgetting that it had advertised PIPELINING for the second and
6203 subsequent messages on an SMTP connection. It was also not resetting its
6204 memory on STARTTLS and an internal HELO.
6206 23. When Exim logs an SMTP synchronization error within a session, it now
6207 records whether PIPELINING has been advertised or not.
6209 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6210 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6211 rather than long int.
6213 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6215 26. Added the never_mail option to autoreply.
6221 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6222 crash if the getsockname() call failed; this can happen if a connection is
6223 closed very soon after it is established. The problem was simply in the
6224 order in which certain operations were done, causing Exim to try to write
6225 to the SMTP stream before it had set up the file descriptor. The bug has
6226 been fixed by making things happen in the correct order.
6232 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6233 before the connection was closed, thus losing the rejection response.
6235 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6236 some early Solaris releases, but causes trouble in current releases where
6237 socklen_t is defined.
6239 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6242 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6245 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6246 that releasing the top of it at the end releases what was used for sub-
6247 expansions (unless the block got too big). However, discard this block if
6248 the first thing is a variable or header, so that we can use its block when
6249 it is dynamic (useful for very large $message_headers, for example).
6251 6. Lookups now cache *every* query, not just the most recent. A new, separate
6252 store pool is used for this. It can be recovered when all lookup caches are
6253 flushed. Lookups now release memory at the end of their result strings.
6254 This has involved some general refactoring of the lookup sources.
6256 7. Some code has been added to the store_xxx() functions to reduce the amount
6257 of flapping under certain conditions.
6259 8. log_incoming_interface used to affect only the <= reception log lines. Now
6260 it causes the local interface and port to be added to several more SMTP log
6261 lines, for example "SMTP connection from", and rejection lines.
6263 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6265 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6267 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6268 could be overwritten at the end of the current message (or the start of a
6269 new message if it was set in a HELO ACL). The value is now preserved for
6270 the duration of the SMTP connection.
6272 12. If a transport had a headers_rewrite setting, and a matching header line
6273 contained an unqualified address, that address was qualified, even if it
6274 did not match any rewriting rules. The underlying bug was that the values
6275 of the flags that permit the existence of unqualified sender and recipient
6276 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6277 non-local messages, and by -bnq for local messages) were not being
6278 preserved with the message after it was received.
6280 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6281 "next input=" as part of the text comprising the host identity instead of
6282 the correct text. The code was using the same buffer for two different
6283 strings. However, depending on which order the printing function evaluated
6284 its arguments, the bug did not always show up. Under Linux, for example, my
6285 test suite worked just fine.
6287 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6288 doesn't work with some older versions of Perl. It has been changed to "my",
6289 which in any case is probably the better facility to use.
6291 15. A really picky compiler found some instances of statements for creating
6292 error messages that either had too many or two few arguments for the format
6295 16. The size of the buffer for calls to the DNS resolver has been increased
6296 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6297 for addresses that have a lot of PTR records. This alleviates a problem; it
6298 does not fully solve it.
6300 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6301 buffer now truncates the list and logs the incident, which is the same
6302 action as happens when Exim is looking up a host name and its aliases.
6303 Previously in this situation something unpredictable would happen;
6304 sometimes it was "internal error: store_reset failed".
6306 18. If a server dropped the connection unexpectedly when an Exim client was
6307 using GnuTLS and trying to read a response, the client delivery process
6308 crashed while trying to generate an error log message.
6310 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6311 string, for example:
6313 warn message = H1: something\nH2: something
6315 the text was added as a single header line from Exim's point of view
6316 though it ended up OK in the delivered message. However, searching for the
6317 second and subsequent header lines using $h_h2: did not work. This has been
6318 fixed. Similarly, if a system filter added multiple headers in this way,
6319 the routers could not see them.
6321 20. Expanded the error message when iplsearch is called with an invalid key to
6322 suggest using net-iplsearch in a host list.
6324 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6325 ACLs are no longer actually imposed (and a message to that effect is
6328 22. If a "gecos" field in a passwd entry contained escaped characters, in
6329 particular, if it contained a \" sequence, Exim got it wrong when building
6330 a From: or a Sender: header from that name. A second bug also caused
6331 incorrect handling when an unquoted " was present following a character
6332 that needed quoting.
6334 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6335 was not being matched caselessly.
6337 24. Arranged for all hyphens in the exim.8 source to be escaped with
6340 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6341 properly. Recipient callout cache records were still being keyed to include
6342 the sender, even when use_sender was set false. This led to far more
6343 callouts that were necessary. The sender is no longer included in the key
6344 when use_sender is false.
6346 26. Added "control = submission" modifier to ACLs.
6348 27. Added the ${base62d: operator to decode base 62 numbers.
6350 28. dnsdb lookups can now access SRV records.
6352 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6353 the configuration file.
6355 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6356 (-v) mode. This makes the output for a verbose queue run more intelligible.
6358 31. Added a use_postmaster feature to recipient callouts.
6360 32. Added the $body_zerocount variable, containing the number of binary zero
6361 bytes in the message body.
6363 33. The time of last modification of the "new" subdirectory is now used as the
6364 "mailbox time last read" when there is a quota error for a maildir
6367 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6369 35. Added +ignore_unknown as a special item in host lists.
6371 36. Code for decoding IPv6 addresses in host lists is now included, even if
6372 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6373 address was recognized as an IP address, but was then not correctly decoded
6374 into binary, causing unexpected and incorrect effects when compared with
6381 1. Very minor rewording of debugging text in manualroute to say "list of
6382 hosts" instead of "hostlist".
6384 2. If verify=header_syntax was set, and a header line with an unqualified
6385 address (no domain) and a large number of spaces between the end of the
6386 name and the colon was received, the reception process suffered a buffer
6387 overflow, and (when I tested it) crashed. This was caused by some obsolete
6388 code that should have been removed. The fix is to remove it!
6390 3. When running in the test harness, delay a bit after writing a bounce
6391 message to get a bit more predictability in the log output.
6393 4. Added a call to search_tidyup() just before forking a reception process. In
6394 theory, someone could use a lookup in the expansion of smtp_accept_max_
6395 per_host which, without the tidyup, could leave open a database connection.
6397 5. Added the variables $recipient_data and $sender_data which get set from a
6398 lookup success in an ACL "recipients" or "senders" condition, or a router
6399 "senders" option, similar to $domain_data and $local_part_data.
6401 6. Moved the writing of debug_print from before to after the "senders" test
6404 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6405 problems for message scanning, either using a data ACL, or using
6406 local_scan() because the Received: header was not generated till after they
6407 were called (in order to set the time as the time of reception completion).
6408 I have revised the way this works. The header is now generated after the
6409 body is received, but before the ACL or local_scan() are called. After they
6410 are run, the timestamp in the header is updated.
6416 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6417 before starting a queue runner without re-exec. This happened only when
6418 deliver_drop_privilege was set or when the Exim user was set to root. The
6419 effect of the bug was that timeouts during subsequent deliveries caused
6420 crashes instead of being properly handled. The handler is now left at its
6421 default (and expected) setting.
6423 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6424 message, again when deliver_drop_privilege is set or Exim is run as root.
6425 The bug described in (1) was not present in this case, but the tidying up
6426 of the other signals was missing. I have made the two cases consistent.
6428 3. The ignore_target_hosts setting on a manualroute router was being ignored
6429 for hosts that were looked up using the /MX notation.
6431 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6434 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6435 operated on the sender address. After changing the $sender_address to <>
6436 for the sender address verify, Exim was re-instated it as the original
6437 (before rewriting) address, but remembering that it had rewritten it, so it
6438 wasn't rewriting it again. This bug also had the effect of breaking the
6439 sender address verification caching when the sender address was rewritten.
6441 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6442 This has been changed so that if the ip literal address matches
6443 ignore_target_hosts, the router declines.
6445 7. Added expansion conditions match_domain, match_address, and match_local_
6446 part (NOT match_host).
6448 8. The placeholder for the Received: header didn't have a length field set.
6450 9. Added code to Exim itself and to exim_lock to test for a specific race
6451 condition that could lead to file corruption when using MBX delivery. The
6452 issue is with the lockfile that is created in /tmp. If this file is removed
6453 after a process has opened it but before that process has acquired a lock,
6454 there is the potential for a second process to recreate the file and also
6455 acquire a lock. This could lead to two Exim processes writing to the file
6456 at the same time. The added code performs the same test as UW imapd; it
6457 checks after acquiring the lock that its file descriptor still refers to
6458 the same named file.
6460 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6461 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6464 11. Added the smtp_active_hostname option. If used, this will typically be made
6465 to depend on the incoming interface address. Because $interface_address is
6466 not set up until the daemon has forked a reception process, error responses
6467 that can happen earlier (such as "too many connections") no longer contain
6470 12. If an expansion in a condition on a "warn" statement fails because a lookup
6471 defers, the "warn" statement is abandoned, and the next ACL statement is
6472 processed. Previously this caused the whole ACL to be aborted.
6474 13. Added the iplsearch lookup type.
6476 14. Added ident_timeout as a log selector.
6478 15. Added tls_certificate_verified as a log selector.
6480 16. Added a global option tls_require_ciphers (compare the smtp transport
6481 option of the same name). This controls incoming TLS connections.
6483 17. I finally figured out how to make tls_require_ciphers do a similar thing
6484 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6485 before starting the TLS session.
6487 18. Tabs are now shown as \t in -bP output.
6489 19. If the log selector return_path_on_delivery was set, Exim crashed when
6490 bouncing a message because it had too many Received: header lines.
6492 20. If two routers both had headers_remove settings, and the first one included
6493 a superfluous trailing colon, the final name in the first list and the
6494 first name in the second list were incorrectly joined into one item (with a
6495 colon in the middle).
6501 1. Added -C and -D options to the exinext utility, mainly to make it easier
6502 to include in the automated testing, but these could be helpful when
6503 multiple configurations are in use.
6505 2. The exinext utility was not formatting the output nicely when there was
6506 an alternate port involved in the retry record key, nor when there was a
6507 message id as well (for retries that were specific to a specific message
6508 and a specific host). It was also confused by IPv6 addresses, because of
6509 the additional colons they contain. I have fixed the IPv4 problem, and
6510 patched it up to do a reasonable job for IPv6.
6512 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6513 delivery, the log line now contains "pipelined" if PIPELINING was used.
6515 4. An SMTP transport process used to panic and die if the bind() call to set
6516 an explicit outgoing interface failed. This has been changed; it is now
6517 treated in the same way as a connect() failure.
6519 5. A reference to $sender_host_name in the part of a conditional expansion
6520 that was being skipped was still causing a DNS lookup. This no longer
6523 6. The def: expansion condition was not recognizing references to header lines
6524 that used bh_ and bheader_.
6526 7. Added the _cache feature to named lists.
6528 8. The code for checking quota_filecount in the appendfile transport was
6529 allowing one more file than it should have been.
6531 9. For compatibility with Sendmail, the command line option
6539 and sets the incoming protocol and host name (for trusted callers). The
6540 host name and its colon can be omitted when only the protocol is to be set.
6541 Note the Exim already has two private options, -pd and -ps, that refer to
6542 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6543 "s", but I don't think that's a major issue.
6545 10. A number of refactoring changes to the code, none of which should affect
6548 (a) The number of logging options was getting close to filling up the
6549 32-bit word that was used as a bit map. I have split them into two classes:
6550 those that are passed in the argument to log_write(), and those that are
6551 only ever tested independently outside of that function. These are now in
6552 separate 32-bit words, so there is plenty of room for expansion again.
6553 There is no change in the user interface or the logging behaviour.
6555 (b) When building, for example, log lines, the code previously used a
6556 macro that called string_cat() twice, in order to add two strings. This is
6557 not really sufficiently general. Furthermore, there was one instance where
6558 it was actually wrong because one of the argument was used twice, and in
6559 one call a function was used. (As it happened, calling the function twice
6560 did not affect the overall behaviour.) The macro has been replaced by a
6561 function that can join an arbitrary number of extra strings onto a growing
6564 (c) The code for expansion conditions now uses a table and a binary chop
6565 instead of a serial search (which was left over from when there were very
6566 few conditions). Also, it now recognizes conditions like "pam" even when
6567 the relevant support is not compiled in: a suitably worded error message is
6568 given if an attempt is made to use such a condition.
6570 11. Added ${time_interval:xxxxx}.
6572 12. A bug was causing one of the ddress fields not to be passed back correctly
6573 from remote delivery subprocesses. The field in question was not being
6574 subsequently used, so this caused to problems in practice.
6576 13. Added new log selectors queue_time and deliver_time.
6578 14. Might have fixed a bug in maildirsizefile handling that threw up
6579 "unexpected character" debug warnings, and recalculated the data
6580 unnecessarily. In any case, I expanded the warning message to give more
6583 15. Added the message "Restricted characters in address" to the statements in
6584 the default ACL that block characters like @ and % in local parts.
6586 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6587 Three changes have been made:
6589 (a) There was a serious bug; a negative response to MAIL caused the whole
6590 recipient domain to be cached as invalid, thereby blocking all messages
6591 to all local parts at the same domain, from all senders. This bug has
6592 been fixed. The domain is no longer cached after a negative response to
6593 MAIL if the sender used is not empty.
6595 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6598 (c) A new callout option, "use_sender" has been added for people who want
6599 the modified behaviour.
6605 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6608 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6609 indeed breaks things for older releases.
6611 3. Added additional logging to the case where there is a problem reading data
6612 from a filter that is running in a subprocess using a pipe, in order to
6613 try to track down a specific problem.
6615 4. Testing facility fudge: when running in the test harness and attempting
6616 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6617 getting "No route to host". Convert this to a timeout.
6619 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6622 6. Some OS don't have socklen_t but use size_t instead. This affects the
6623 fifth argument of getsockopt() amongst other things. This is now
6624 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6625 can be set for individual OS. I have set it for SunOS5, OSF1, and
6626 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6627 some earlier ones do not.
6629 7. Change 4.30/15 was not doing the test caselessly.
6631 8. The standard form for an IPv6 address literal was being rejected by address
6632 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6633 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6634 this, as well as the form without the "IPv6" on the front (but only when
6635 address literals are enabled, of course).
6637 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6639 10. Exim crashed if a message with an empty sender address specified by -f
6640 encountered a router with an errors_to setting. This could be provoked only
6641 by a command such as
6645 where an empty string was supplied; "<>" did not hit this bug.
6647 11. Installed PCRE release 4.5.
6649 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6650 remained set. It is now erased.
6652 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6653 times from message ids (which are base 36 rather than the normal 62).
6655 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6656 were being counted as actual protocol errors, and logged if the log
6657 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6658 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6659 and DATA following a set of rejected RCPTs do not count as protocol errors.
6660 In other words, Exim assumes they were pipelined, though this may not
6661 actually be the case. Of course, in all cases the client gets an
6662 appropriate error code.
6664 15. If a lookup fails in an ACL condition, a message about the failure may
6665 be available; it is used if testing the ACL cannot continue, because most
6666 such messages specify what the cause of the deferral is. However, some
6667 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6668 that caused an old message to be retained and used if a later statement
6669 caused a defer, replacing the real cause of the deferral.
6671 16. If an IP address had so many PTR records that the DNS lookup buffer
6672 was not large enough to hold them, Exim could crash while trying to process
6673 the truncated data. It now detects and logs this case.
6675 17. Further to 4.21/58, another change has been made: if (and only if) the
6676 first line of a message (the first header line) ends with CRLF, a bare LF
6677 in a subsequent header line has a space inserted after it, so as not to
6678 terminate the header.
6680 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6681 unnecessarily, used atoi() instead of strtol(), and didn't check the
6682 termination when getting file sizes from file names by regex.
6684 19. Completely re-implemented the support for maildirsize files, in the light
6685 of a number of problems with the previous contributed implementation
6686 (4.30/29). In particular:
6688 . If the quota is zero, the maildirsize file is maintained, but no quota is
6691 . If the maildir directory does not exist, it is created before any attempt
6692 to write a maildirsize file.
6694 . The quota value in the file is just a cache; if the quota is changed in
6695 the transport, the new value overrides.
6697 . A regular expression is available for excluding directories from the
6700 20. The autoreply transport checks the characters in options that define the
6701 message's headers; it allows continued headers, but it was checking with
6702 isspace() after an embedded newline instead of explicitly looking for a
6705 21. If all the "regular" hosts to which an address was routed had passed their
6706 expiry times, and had not reached their retry times, the address was
6707 bounced, even if fallback hosts were defined. Now Exim should go on to try
6710 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6711 equivalent code in the SMTP transport. Some hosts send humungous responses
6712 to HELO/EHLO, more than 1024 it seems.
6714 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6715 gives compiler warnings in some environments. I've now done it "properly",
6718 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6719 (because of problems with the built-in one) was declared to return uschar *
6720 instead of char *, causing compiler failure.
6722 25. Fixed a file descriptor leak when processing alias/forward files.
6724 26. Fixed a minor format string issue in dbfn.c.
6726 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6728 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6729 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6732 29. When a sender address is verified, it is cached, to save repeating the test
6733 when there is more than one recipient in a message. However, when the
6734 verification involves a callout, it is possible for different callout
6735 options to be set for different recipients. It is too complicated to keep
6736 track of this in the cache, so now Exim always runs a verification when a
6737 callout is required, relying on the callout cache for the optimization.
6738 The overhead is duplication of the address routing, but this should not be
6741 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6742 to be verified with callout=postmaster, and the main callout worked but the
6743 postmaster check failed, the verification correctly failed. However, if a
6744 subsequent RCPT command asked for sender verification *without* the
6745 postmaster check, incorrect caching caused this verification also to fail,
6748 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6749 it was not caching the DNS options (qualify_single, search_parents) that
6750 were used when the lookup failed. A subsequent lookup with different
6751 options therefore always gave the same answer, though there were cases
6752 where it should not have. (Example: a "domains = !$mx_any" option on a
6753 dnslookup router: the "domains" option is always processed without any
6754 widening, but the router might have qualify_single set.) Now Exim uses the
6755 cached value only when the same options are set.
6757 32. Added John Jetmore's "exipick" utility to the distribution.
6759 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6760 than a timeout (e.g. a certificate is required, and is not provided), an
6761 Exim server now closes the connection immediately. Previously it waited for
6762 the client to close - but if the client is SSL, it seems that they each
6763 wait for each other, leading to a delay before one of them times out.
6765 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6766 maintained 0.8.x compatibility because I don't think many are using it, and
6767 it is clearly obsolete.
6769 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6772 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6773 client certificate was expired. A simple patch fixes this, though I don't
6774 understand the full logic of why the verify callback is called multiple
6777 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6778 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6779 which causes problems with some clients (such as the Certicom SSL Plus
6780 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6781 disables the coutermeasure allowing Eudora to connect."
6783 38. Exim was not checking that a write() to a log file succeeded. This could
6784 lead to Bad Things if a log got too big, in particular if it hit a file
6785 size limit. Exim now panics and dies if it cannot write to a log file, just
6786 as it does if it cannot open a log file.
6788 39. Modified OS/Makefile-Linux so that it now contains
6790 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6792 The two -D definitions ensure that Exim is compiled with large file
6793 support, which makes it possible to handle log files that are bigger than
6796 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6797 instance) a domain was checked against a named list that involved a lookup,
6798 causing $domain_data to be set, then another domain was checked against the
6799 same list, then the first domain was re-checked, the value of $domain_data
6800 after the final check could be wrong. In particular, if the second check
6801 failed, it could be set empty. This bug probably also applied to
6804 41. The strip_trailing_dot option was not being applied to the address given
6805 with the -f command-line option.
6807 42. The code for reading a message's header from the spool was incrementing
6808 $received_count, but never initializing it. This meant that the value was
6809 incorrect (doubled) while delivering a message in the same process in which
6810 it was received. In the most common configuration of Exim, this never
6811 happens - a fresh exec is done - but it can happen when
6812 deliver_drop_privilege is set.
6814 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6815 it now includes up to 150 characters of the unexpected data in the log
6818 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6819 and building data strings. The size of both of these buffers was 10 000
6820 bytes - far larger than anybody would *ever* want, thought I. Needless to
6821 say, somebody hit the limit. I have increased the maximum line length to
6822 20 000 and the maximum data length of concatenated lines to 100 000. I have
6823 also fixed two bugs, because there was no checking on these buffers. Tsk,
6824 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6825 buffer is too small.
6827 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6828 lsearch lookups. Now it does.
6830 46. When parsing a route_list item in a manualroute router, a fixed-length
6831 buffer was used for the list of hosts. I made this 1024 bytes long,
6832 thinking that nobody would ever have a list of hosts that long. Wrong.
6833 Somebody had a whole pile of complicated expansion conditions, and the
6834 string was silently truncated, leading to an expansion error. It turns out
6835 that it is easier to change to an unlimited length (owing to other changes
6836 that have happened since this code was originally written) than to build
6837 structure for giving a limitation error. The length of the item that
6838 expands into the list of hosts is now unlimited.
6840 47. The lsearch lookup could not handle data where the length of text line was
6841 more than 4095 characters. Such lines were truncated, leading to shortened
6842 data being returned. It should now handle lines of any length.
6844 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6845 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6848 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6849 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6850 no longer happens in comment lines. A list of replacements is now placed
6851 at the head of all of the source files, except those whose only change is
6852 to replace PERL_COMMAND in the very first #! line.
6854 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6855 messages on the queue, with a bottom-up merge sort, using code contributed
6856 by Michael Haardt. This should make operations like -bp somewhat faster on
6857 large queues. It won't affect queue runners, except when queue_run_in_order
6860 51. Installed eximstats 1.31 in the distribution.
6862 52. Added support for SRV lookups to the dnslookup router.
6864 53. If an ACL referred to $message_body or $message_body_end, the value was not
6865 reset for any messages that followed in the same SMTP session.
6867 54. The store-handling optimization for building very long strings was not
6868 differentiating between the different store pools. I don't think this
6869 actually made any difference in practice, but I've tidied it.
6871 55. While running the routers to verify a sender address, $sender_address
6872 was still set to the sender address. This is wrong, because when routing to
6873 send a bounce to the sender, it would be empty. Therefore, I have changed
6874 it so that, while verifying a sender address, $sender_address is set to <>.
6875 (There is no change to what happens when verifying a recipient address.)
6877 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6878 target A or AAAA records (if not already returned) without resetting the
6879 qualify_single or search_parents options of the DNS resolver. These are
6880 inappropriate in this case because the targets of MX and SRV records must
6881 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6882 target that, when qualified, matched something in the local domain. These
6883 two options are now turned off when doing these lookups.
6885 57. It seems that at least some releases of Reiserfs (which does not have the
6886 concept of a fixed number of inodes) returns zero and not -1 for the
6887 number of available inodes. This interacted badly with check_spool_inodes,
6888 which assumed that -1 was the "no such thing" setting. What I have done is
6889 to check that the total number of inodes is greater than zero before doing
6890 the test of how many are available.
6892 58. When a "warn" ACL statement has a log_message modifier, the message is
6893 remembered, and not repeated. This is to avoid a lot of repetition when a
6894 message has many recipients that cause the same warning to be written.
6895 However, Exim was preserving the list of already written lines for an
6896 entire SMTP session, which doesn't seem right. The memory is now reset if a
6897 new message is started.
6899 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6900 debugging output unless log_rewrite was also set.
6902 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6903 of (int)(handle) when we know that handle contains (void *)(-1).
6905 61. The Exim daemon panic-logs an error return when it closes the incoming
6906 connection. However "connection reset by peer" seems to be common, and
6907 isn't really an error worthy of noting specially, so that particular error
6910 62. When Exim is trying to find all the local interfaces, it used to panic and
6911 die if the ioctl to get the interface flags failed. However, it seems that
6912 on at least one OS (Solaris 9) it is possible to have an interface that is
6913 included in the list of interfaces, but for which you get a failure error
6914 for this call. This happens when the interface is not "plumbed" into a
6915 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6916 failure of the "get flags" call assumes that the interface is down.
6918 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6919 makes life easier for people who are doing arithmetic on fields extracted
6920 from dates, where you often get leading zeros that should not be
6921 interpreted as octal.
6923 64. Added qualify_domain to the redirect router, to override the global
6926 65. If a pathologically long header line contained very many addresses (the
6927 report of this problem mentioned 10 000) and each of them was rewritten,
6928 Exim could use up a very large amount of memory. (It kept on making new
6929 copies of the header line as it rewrote, and never released the old ones.)
6930 At the expense of a bit more processing, the header rewriting function has
6931 been changed so that it no longer eats memory in this way.
6933 66. The generation of the Received: header has been moved from the time that a
6934 message starts to be received, to the time that it finishes. The timestamp
6935 in the Received: header should now be very close to that of the <= log
6936 line. There are two side-effects of this change:
6938 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6939 logged header lines no longer include the local Received: line, because
6940 it has not yet been created. The same applies to a copy of the message
6941 that is returned to a non-SMTP sender when a message is rejected.
6943 (b) When a filter file is tested using -bf, no additional Received: header
6944 is added to the test message. After some thought, I decided that this
6947 This change does not affect the value of $received_for. It is still set
6948 after address rewriting, but before local_scan() is called.
6950 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6952 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6953 gave an unhelpful panic error message, and a defer error. I have managed to
6954 change this behaviour so that it now rejects any supplied certificate,
6955 which seems right, as the list of acceptable certificates is empty.
6957 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6958 gave an unhelpful defer error. I have not managed to make this reject any
6959 supplied certificates, but the error message it gives is "no certificate
6960 supplied", which is not helpful.
6962 70. exigrep's output now also includes lines that are not associated with any
6963 message, but which match the given pattern. Implemented by a patch from
6964 Martin Sluka, which also tidied up the Perl a bit.
6966 71. Recipient callout verification, like sender verification, was using <> in
6967 the MAIL FROM command. This isn't really the right thing, since the actual
6968 sender may affect whether the remote host accepts the recipient or not. I
6969 have changed it to use the actual sender in the callout; this means that
6970 the cache record is now keyed on a recipient/sender pair, not just the
6971 recipient address. There doesn't seem to be a real danger of callout loops,
6972 since a callout by the remote host to check the sender would use <>.
6973 [SEE ABOVE: changed after hitting problems.]
6975 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6976 temporary errors. However, in the case of such a code being given after
6977 the end of a data transmission (i.e. after ".") Exim was failing to write
6978 a retry record for the message. (Yes, there was some broken host that was
6979 actually sending 8xx at this point.)
6981 73. An unknown lookup type in a host list could cause Exim to panic-die when
6982 the list was checked. (An example that provoked this was putting <; in the
6983 middle of a list instead of at the start.) If this happened during a DATA
6984 ACL check, a -D file could be left lying around. This kind of configuration
6985 error no longer causes Exim to die; instead it causes a defer error. The
6986 incident is still logged to the main and panic logs.
6988 74. Buglet left over from Exim 3 conversion. The message "too many messages
6989 in one connection" was written to the rejectlog but not the mainlog, except
6990 when address rewriting (yes!) was being logged.
6992 75. Added write_rejectlog option.
6994 76. When a system filter was run not as root (that is, when system_filter_user
6995 was set), the values of the $n variables were not being returned to the
6996 main process; thus, they were not subsequently available in the $sn
6999 77. Added +return_path_on_delivery log selector.
7001 78. A connection timeout was being treated differently from recipients deferred
7002 when testing hosts_max_try with a message that was older than the host's
7003 retry timeout. (The host should not be counted, thus allowing all hosts to
7004 be tried at least once before bouncing.) This may have been the cause of an
7005 occasionally reported bug whereby a message would remain on the queue
7006 longer than the retry timeout, but would be bounced if a delivery was
7007 forced. I say "may" because I never totally pinned down the problem;
7008 setting up timeout/retry tests is difficult. See also the next item.
7010 79. The ultimate address timeout was not being applied to errors that involved
7011 a combination of host plus message (for example, a timeout on a MAIL
7012 command). When an address resolved to a number of possible hosts, and they
7013 were not all tried for each delivery (e.g. because of hosts_max_try), a
7014 message could remain on the queue longer than the retry timeout.
7016 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7019 81. Fixed an obscure SMTP outgoing bug which required at least the following
7020 conditions: (a) there was another message waiting for the same server;
7021 (b) the server returned 5xx to all RCPT commands in the first message so
7022 that the message was not completed; (c) the server dropped the connection
7023 or gave a negative response to the RSET that Exim sends to abort the
7024 transaction. The observed case was a dropped connection after DATA that had
7025 been sent in pipelining mode. That is, the server had advertised PIPELINING
7026 but was not implementing it correctly. The effect of the bug was incorrect
7027 behaviour, such as trying another host, and this could lead to a crash.
7033 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7034 and daemon.c were passed as pointers to ints; they should have been
7035 pointers to socklen_t variables (which are typically unsigned ints).
7037 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7040 3. Fixed a really odd bug that affected only the testing scheme; patching a
7041 certain fixed string in the binary changed the value of another string that
7042 happened to be identical to the end of the original first string.
7044 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7045 name", it returns that address as the IP address. On some operating
7046 systems (e.g. Solaris), it also passes back the IP address string as the
7047 "host name". However, on others (e.g. Linux), it passes back an empty
7048 string. Exim wasn't checking for this, and was changing the host name to an
7049 empty string, assuming it had been canonicalized.
7051 5. Although rare, it is permitted to have more than one PTR record for a given
7052 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7053 all the names associated with an address, because they do in Solaris.
7054 However, it seems that they do not in Linux for data that comes from the
7055 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7056 I found this out when I moved to a new Linux workstation and tried to run
7057 the Exim test suite.
7059 To get round this problem I have changed the code so that it now does its
7060 own call to the DNS to look up PTR records when searching for a host name.
7061 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7062 addresses that are only in /etc/hosts are still found.
7064 This behaviour is, however, controlled by an option called host_lookup_
7065 order, which defaults to "bydns:byaddr". If people want to use the other
7066 order, or indeed, just use one or the other means of lookup, they can
7067 specify it in this variable.
7069 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7070 some operating systems, this comes back from gethostbyaddr() as an empty
7071 string, and this is what Exim used to test for. However, it seems that in
7072 other systems, "." is yielded. Exim now tests for this case too.
7074 7. The values of check_spool_space and check_log_space are now held internally
7075 as a number of kilobytes instead of an absolute number of bytes. If a
7076 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7077 kilobyte. This means that much larger values can be stored.
7079 8. Exim monitor: an attempt to get the action menu when not actually pointing
7080 at a message produces an empty menu entitled "No message selected". This
7081 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7082 no entries in it ("Shell widget menu has zero width and/or height"). So I
7083 have added a single, blank menu entry in this case.
7085 9. Added ${quote_local_part.
7087 10. MIME decoding is now applied to the contents of Subject: header lines when
7090 11. Now that a reference to $sender_host_address automatically causes a reverse
7091 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7092 host lookup before query-style lookups in lists that might use this
7093 variable. This has therefore been abolished, and the "net-" prefix is no
7094 longer necessary for query-style lookups.
7096 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7097 have been a typo for LFLAGS, so it has been changed.
7099 13. The install script calls Exim with "-C /dev/null" in order to find the
7100 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7101 to be output. However, since Exim outputs its version number before the
7102 error, it didn't break the script. It just looked ugly. I fixed this by
7103 always allowing "-C /dev/null" if the caller is root.
7105 14. Ignore overlarge ACL variable number when reading spool file - insurance
7106 against a later release with more variables having written the file.
7108 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7109 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7110 this, as well as the form without the "IPv6" on the front.
7112 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7113 OS/Makefile-Darwin file.
7115 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7116 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7118 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7119 variable. However, some OS use ints for inodes. I've added cast to long int
7120 to get rid of the compiler warning.
7122 19. I had forgotten to lock out "/../" in configuration file names when
7123 ALT_CONFIG_PREFIX was set.
7125 20. Routers used for verification do not need to specify transports. However,
7126 if such a router generated a host list, and callout was configured, Exim
7127 crashed, because it could not find a port number from the (non-existent)
7128 transport. It now assumes port 25 in this circumstance.
7130 21. Added the -t option to exigrep.
7132 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7133 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7134 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7135 them is set, LOOKUP_LSEARCH is forced.
7137 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7139 24. Added sender and host information to the "rejected by local_scan()" log
7140 line; previously there was no indication of these.
7142 25. Added .include_if_exists.
7144 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7145 Linux. It turns out that not all file systems support this. Apparently some
7146 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7147 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7148 not supported on the file descriptor, is now ignored when Exim is trying to
7149 sync a directory. This applies only to Linux.
7151 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7153 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7154 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7157 29. Added support for maildirsize files from supplied patch (modified a bit).
7159 30. The use of :fail: followed by an empty string could lead Exim to respond to
7160 sender verification failures with (e.g.):
7162 550 Verification failed for <xxx>
7163 550 Sender verify failed
7165 where the first response line was missing the '-' that indicates it is not
7166 the final line of the response.
7168 31. The loop for finding the name of the user that called Exim had a hardwired
7169 limit of 10; it now uses the value of finduser_retries, which is used for
7170 all other user lookups.
7172 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7175 33. Exim was neglecting to zero errno before one call of strtol() when
7176 expanding a string and expecting an integer value. On some systems this
7177 resulted in spurious "integer overflow" errors. Also, it was casting the
7178 result into an int without checking.
7180 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7181 did not work. The code looks as if it has *never* worked, though it appears
7182 to have been documented since at least release 1.62. I have made it work.
7184 35. The "timeout_DNS" error in retry rules, also documented since at least
7185 1.62, also never worked. As it isn't clear exactly what this means, and
7186 clearly it isn't a major issue, I have abolished the feature by treating it
7187 as "timeout", and writing a warning to the main and panic logs.
7189 36. The display of retry rules for -brt wasn't always showing the error code
7192 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7193 timeout_connect_A, timeout_connect_MX.
7195 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7196 to the empty sender.
7198 39. The daemon was not analysing the content of -oX till after it had closed
7199 stderr and disconnected from the controlling terminal. This meant that any
7200 syntax errors were only noted on the panic log, and the return code from
7201 the command was 0. By re-arranging the code a little, I've made the
7202 decoding happen first, so such errors now appear on stderr, and the return
7203 code is 1. However, the actual setting up of the sockets still happens in
7204 the disconnected process, so errors there are still only recorded on the
7207 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7208 connections (as happens on some IP stacks) was logged at start up time as
7209 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7210 it from "IPv6 and IPv4", which means that two separate sockets are being
7213 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7214 says whether AF_INET or AF_INET6 was passed as an argument.
7216 42. Exiwhat output was messed up when time zones were included in log
7219 43. Exiwhat now gives more information about the daemon's listening ports,
7220 and whether -tls-on-connect was used.
7222 44. The "port" option of the smtp transport is now expanded.
7224 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7225 silently ignored. Now an error message is written to the main and panic
7228 46. There's a new ACL modifier called "logwrite" which writes to a log file
7229 as soon as it is encountered.
7231 47. Added $local_user_uid and $local_user_gid at routing time.
7233 48. Exim crashed when trying to verify a sender address that was being
7236 49. Exim was recognizing only a space character after ".include". It now also
7237 recognizes a tab character.
7239 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7240 extracting the relevant information from the specification. The man page no
7241 longer contains scrambled data for the -d option, and I've added a section
7242 at the front about calling Exim under different names.
7244 51. Added "extra_headers" argument to the "mail" command in filter files.
7246 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7249 53. Installed eximstats 1.29.
7251 54. Added transport_filter_timeout as a generic transport option.
7253 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7254 Cc: header lines. This was required by RFC 822, but it not required by RFC
7257 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7258 incoming messages that did not have them. Now it does so only if the
7259 message originates locally, that is, if there is no associated remote host
7260 address. When Resent- header lines are present, this applies to the Resent-
7261 lines rather than the non-Resent- lines.
7263 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7264 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7266 58. Messages for configuration errors now include the name of the main
7267 configuration file - useful now that there may be more than one file in a
7268 list (.included file names were always shown).
7270 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7271 for those rare installations that do not start the daemon as root or run it
7272 setuid root. I've cut out the call to initgroups() if the daemon is not
7275 60. The Exim user and group can now be bound into the binary as text strings
7276 that are looked up at the start of Exim's processing.
7278 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7280 62. Added $mailstore_basename variable.
7282 63. Installed patch to sieve.c from Michael Haardt.
7284 64. When Exim failed to open the panic log after failing to open the main log,
7285 the original message it was trying to log was written to stderr and debug
7286 output, but if they were not available (the usual case in production), it
7287 was lost. Now it is written to syslog before the two lines that record the
7288 failures to open the logs.
7290 65. Users' Exim filters run in subprocesses under the user's uid. It is
7291 possible for a "deliver" command or an alias in a "personal" command to
7292 provoke an address rewrite. If logging of address rewriting is configured,
7293 this fails because the process is not running as root or exim. There may be
7294 a better way of dealing with this, but for the moment (because 4.30 needs
7295 to be released), I have disabled address rewrite logging when running a
7296 filter in a non-root, non-exim process.
7302 1. The buildconfig auxiliary program wasn't quoting the value set for
7303 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7304 not defined. This bug was masked in 4.22 by the effect that was fixed in
7307 2. Some messages that were rejected after a message id was allocated were
7308 shown as "incomplete" by exigrep. It no longer does this for messages that
7309 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7311 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7312 have allow_domain_literals set, the ID did not get logged in the <= line.
7313 Domain literals are now always recognized in Message-ID: header lines.
7315 4. The first argument for a ${extract expansion item is the key name or field
7316 number. Leading and trailing spaces in this item were not being ignored,
7317 causing some misleading effects.
7319 5. When deliver_drop_privilege was set, single queue runner processes started
7320 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7321 same command in the process it spins off) were not dropping privilege.
7323 6. When the daemon running as "exim" started a queue runner, it always
7324 re-executed Exim in the spun-off process. This is a waste of effort when
7325 deliver_drop_privilege is set. The new process now just calls the
7326 queue-runner function directly.
7332 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7335 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7336 ignored. Though the use of -oP was forcing the writing of a pid file, it
7337 was always written to the default place.
7339 3. If the message "no IP address found for host xxxx" is generated during
7340 incoming verification, it is now followed by identification of the incoming
7341 connection (so you can more easily find what provoked it).
7343 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7345 5. Added some features to "harden" Exim a bit more against certain attacks:
7347 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7348 be put in Local/Makefile. This is like the never_users runtime option,
7349 but it cannot be overridden. The default setting is "root".
7351 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7352 prefix string with which any file named in a -C command line option
7355 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7356 is retained for -C and -D only if the caller of Exim is root. Without
7357 it, the exim user may also use -C and -D and retain privilege.
7359 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7360 command line option is disabled.
7362 6. Macro names set by the -D option must start with an upper case letter, just
7363 like macro names defined in the configuration file.
7365 7. Added "dereference=" facility to LDAP.
7367 8. Two instances of the typo "uknown" in the source files are fixed.
7369 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7370 the Configure-Makefile script screwed up while processing it.
7372 10. Incorporated PCRE 4.4.
7374 11. The SMTP synchronization check was not operating right at the start of an
7375 SMTP session. For example, it could not catch a HELO sent before the client
7376 waited for the greeting. There is now a check for outstanding input at the
7377 point when the greeting is written. Because of the duplex, asynchronous
7378 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7379 way, but not yet received, when the check is performed.
7381 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7382 on TCP/IP sockets, because this apparently causes some broken clients to
7385 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7386 unchanged) from the Cygwin maintainer.
7388 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7389 was supported instead of testing for mbx. Effectively a typo.
7391 15. The spa authenticator server code was not checking that the input it
7392 received was valid base64.
7394 16. The debug output line for the "set" modifier in ACLs was not showing the
7395 name of the variable that was being set.
7397 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7399 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7400 lookup on its own. Something else was needed to trigger the lookup. For
7401 example, a match in host_lookup or the need for a host name in a host list.
7402 Now, if $sender_host_name is referenced and the host name has not yet been
7403 looked up, a lookup is performed. If the lookup fails, the variable remains
7404 empty, and $host_lookup_failed is set to "1".
7406 19. Added "eqi" as a case-independent comparison operator.
7408 20. The saslauthd authentication condition could segfault if neither service
7409 nor realm was specified.
7411 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7412 error message that was logged was misleading, and incoming SMTP
7413 connections were dropped. The message is now more accurate, and temporary
7414 errors are given to SMTP connections.
7416 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7417 (except RSET) with a 421 error, until QUIT is received. However, it was
7418 failing to send a response to QUIT.
7420 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7421 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7422 ACL is now run after the helo_try_verify_hosts code.
7424 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7425 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7426 case-independent, so other case variants are also recognized). Apparently
7427 some systems use these upper case variants.
7429 25. If more than two messages were waiting for the same host, and a transport
7430 filter was specified for the transport, Exim sent two messages over the
7431 same TCP/IP connection, and then failed with "socket operation on non-
7432 socket" when it tried to send the third.
7434 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7436 27. The extern definition of crypt16() in expand.c was not being excluded when
7437 the OS had its own crypt16() function.
7439 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7440 as a preferred synonym for return_size_limit, both as an option and as an
7443 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7445 30. Changed the default configuration ACL to relax the local part checking rule
7446 for addresses that are not in any local domains. For these addresses,
7447 slashes and pipe symbols are allowed within local parts, but the sequence
7448 /../ is explicitly forbidden.
7450 31. SPA server authentication was not clearing the challenge buffer before
7453 32. log_message in a "warn" ACL statement was writing to the reject log as
7454 well as to the main log, which contradicts the documentation and doesn't
7455 seem right (because no rejection is happening). So I have stopped it.
7457 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7458 However, I am unable to do any testing of this.
7460 34. Fixed an infelicity in the appendfile transport. When checking directories
7461 for a mailbox, to see if any needed to be created, it was accidentally
7462 using path names with one or more superfluous leading slashes; tracing
7463 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7465 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7466 contents are added to the log line that is written for every discarded
7467 recipient. (Previously a log_message setting was ignored.)
7469 36. The ${quote: operator now quotes the string if it is empty.
7471 37. The install script runs exim in order to find its version number. If for
7472 some reason other than non-existence or emptiness, which it checks, it
7473 could not run './exim', it was installing it with an empty version number,
7474 i.e. as "exim-". This error state is now caught, and the installation is
7477 38. An argument was missing from the function that creates an error message
7478 when Exim fails to connect to the socket for saslauthd authentication.
7479 This could cause Exim to crash, or give a corrupted message.
7481 39. Added isip, isip4, and isip6 to ${if conditions.
7483 40. The ACL variables $acl_xx are now saved with the message, and can be
7484 accessed later in routers, transports, and filters.
7486 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7487 strings in the file are not string-expanded.
7489 42. If a MAIL command specified a SIZE value that was too large to fit into an
7490 int variable, the check against message_size_limit failed. Such values are
7491 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7492 day this will have to be increased, but I don't think I want to be around
7493 when emails are that large.
7500 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7501 iconv() is not standard in FreeBSD.
7503 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7504 IPv6 enabled. The observed symptom was a segmentation fault on return from
7505 the function os_common_find_running_interfaces() in src/os.c.
7507 3. In the check_special_case() function in daemon.c I had used "errno" as an
7508 argument name, which causes warnings on some systems. This was basically a
7509 typo, since it was named "eno" in the comments!
7511 4. The code that waits for the clock to tick (at a resolution of some fraction
7512 of a second) so as to ensure message-id uniqueness was always waiting for
7513 at least one whole tick, when it could have waited for less. [This is
7514 almost certainly not relevant at current processor speeds, where it is
7515 unlikely to ever wait at all. But we try to future-proof.]
7517 5. The function that sleeps for a time interval that includes fractions of a
7518 second contained a race. It did not block SIGALRM between setting the
7519 timer, and suspending (a couple of lines later). If the interval was short
7520 and the sigsuspend() was delayed until after it had expired, the suspension
7521 never ended. On busy systems this could lead to processes getting stuck for
7524 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7525 process, before it forks any delivery processes. The open lookup caching
7526 mechanism meant that the open file or database connection was passed into
7527 the delivery process. The problem was that delivery processes always tidy
7528 up cached lookup data. This could cause a problem for the next delivery
7529 process started by the queue runner, because the external queue runner
7530 process does not know about the closure. So the next delivery process
7531 still has data in the lookup cache. In the case of a file lookup, there was
7532 no problem because closing a file descriptor in a subprocess doesn't affect
7533 the parent. However, if the lookup was caching a connection to a database,
7534 the connection was closed, and the second delivery process was likely to
7535 see errors such as "PGSQL: query failed: server closed the connection
7536 unexpectedly". The problem has been fixed by closing all cached lookups
7537 in a queue runner before running a delivery process.
7539 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7540 seem to have the "const" qualifier which it has on other OS. I've
7543 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7544 *of the same type* (client or server) with the same public name that an
7545 error should be diagnosed.
7547 9. When Exim looked up a host name for an IP address, but failed to find the
7548 original IP address when looking up the host name (a safety check), it
7549 output the message "<ip address> does not match any IP for NULL", which was
7550 confusing, to say the least. The bug was that the host name should have
7551 appeared instead of "NULL".
7553 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7554 user that is built into the binary, and the -C or -D options is used, root
7555 privilege is dropped before the configuration file is read. In addition,
7556 logging is switched to stderr instead of the normal log files. If the
7557 configuration then re-defines the Exim user, the unprivileged environment
7558 is probably not what is expected, so Exim logs a panic warning message (but
7561 However, if deliver_drop_privilege is set, the unprivileged state may well
7562 be exactly what is intended, so the warning has been cut out in that case,
7563 and Exim is allowed to try to write to its normal log files.
7569 1. smtp_return_error_details was not giving details for temporary sender
7570 or receiver verification errors.
7572 2. Diagnose a configuration error if two authenticators have the same public
7575 3. Exim used not to create the message log file for a message until the first
7576 delivery attempt. This could be confusing when incoming messages were held
7577 for policy or load reasons. The message log file is now created at the time
7578 the message is received, and an initial "Received" line is written to it.
7580 4. The automatically generated man page for command line options had a minor
7581 bug that caused no ill effects; however, a more serious problem was that
7582 the procedure for building the man page automatically didn't always
7583 operate. Consequently, release 4.20 contains an out-of-date version. This
7584 shouldn't happen again.
7586 5. When building Exim with embedded Perl support, the script that builds the
7587 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7588 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7590 6. The freeze_tell option was not being used for messages that were frozen on
7591 arrival, either by an ACL or by local_scan().
7593 7. Added the smtp_incomplete_transaction log selector.
7595 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7596 was accepting AUTH without a new EHLO.
7598 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7599 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7602 10. Logging of TCP/IP connections (when configured) now happens in the main
7603 daemon process instead of the child process, so that the TCP/IP connection
7604 count is more accurate (but it can never be perfect).
7606 11. The use of "drop" in a nested ACL was not being handled correctly in the
7607 outer ACL. Now, if condition failure induced by the nested "drop" causes
7608 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7609 or "require"), the connection is dropped.
7611 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7612 that yield "discard" can now be used with an "accept" or a "discard" verb,
7613 but an error is generated for any others (because I can't see a useful way
7614 to define what should happen).
7616 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7617 are now processed in the same way as lines in the Exim configuration file.
7618 In particular, continuation lines are supported.
7620 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7622 15. Added -ti meaning -t -i.
7624 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7625 domains, and warn by logging if others are found.
7627 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7628 returned from GIFCONF when Exim is trying to find the list of interfaces on
7629 a host. The code in os.c has been modified to copy each ifreq to an aligned
7630 structure in all cases.
7632 Also, in some cases, the returned ifreq's were being copied to a 'struct
7633 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7634 means the last couple of bytes of an IPv6 address could be chopped if the
7635 ifreq contained only a normal sockaddr (14 bytes storage).
7637 18. Named domain lists were not supported in the hosts_treat_as_local option.
7638 An entry such as +xxxx was not recognized, and was treated as a literal
7641 19. Ensure that header lines added by a DATA ACL are included in the reject log
7642 if the ACL subsequently rejects the message.
7644 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7645 MD5 (which is deprecated).
7647 21. When testing a filter file using -bf, Exim was writing a message when it
7648 took the sender from a "From " line in the message, but it was not doing so
7649 when it took $return_path from a Return-Path: header line. It now does.
7651 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7652 with a valid header line field name (a series of printing characters
7653 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7655 23. Changed "disc" in the source to "disk" to conform to the documentation and
7656 the book and for uniformity.
7658 24. Ignore Sendmail's -Ooption=value command line item.
7660 25. When execve() failed while trying to run a command in a pipe transport,
7661 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7662 could be confused with a return value of 69 from the command itself. This
7663 has been changed to 127, the value the shell returns if it is asked to run
7664 a non-existent command. The wording for the related log line suggests a
7665 non-existent command as the problem.
7667 26. If received_header_text expands to an empty string, do not add a Received:
7668 header line to the message. (Well, it adds a token one on the spool, but
7669 marks it "old" so that it doesn't get used or transmitted.)
7671 27. Installed eximstats 1.28 (addition of -nt option).
7673 28. There was no check for failure on the call to getsockname() in the daemon
7674 code. This can fail if there is a shortage of resources on the system, with
7675 ENOMEM, for example. A temporary error is now given on failure.
7677 29. Contrary to the C standard, it seems that in some environments, the
7678 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7679 program. Exim now does this explicitly; it affects the formatting of
7680 timestamps using strftime().
7682 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7683 complaints. I've now initialized all the variables, to avoid this.
7685 32. Header lines added by a system filter were not being "seen" during
7686 transport-time rewrites.
7688 33. The info_callback() function passed to OpenSSL is set up with type void
7689 (*)(SSL *, int, int), as described somewhere. However, when calling the
7690 function (actually a macro) that sets it up, the type void(*)() is
7691 expected. I've put in a cast to prevent warnings from picky compilers.
7693 34. If a DNS black list lookup found a CNAME record, but there were no A
7694 records associated with the domain it pointed at, Exim crashed.
7696 35. If a DNS black list lookup returned more than one A record, Exim ignored
7697 all but the first. It now scans all returned addresses if a particular IP
7698 value is being sought. In this situation, the contents of the
7699 $dnslist_value variable are a list of all the addresses, separated by a
7702 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7703 to accept a host name and all its aliases if the forward lookup for any of
7704 them yielded the IP address of the incoming connection. Now it accepts only
7705 those names whose forward lookup yields the correct IP address. Any other
7706 names are discarded. This closes a loophole whereby a rogue DNS
7707 administrator could create reverse DNS records to break through a
7708 wildcarded host restriction in an ACL.
7710 37. If a user filter or a system filter that ran in a subprocess used any of
7711 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7712 the wrong values were passed to the pipe command ($thisaddress had the
7713 value of $0, $0 had the value of $1, etc). This bug was introduced by
7714 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7716 38. Improved the line breaking for long SMTP error messages from ACLs.
7717 Previously, if there was no break point between 40 and 75 characters, Exim
7718 left the rest of the message alone. Two changes have been made: (a) I've
7719 reduced the minimum length to 35 characters; (b) if it can't find a break
7720 point between 35 and 75 characters, it looks ahead and uses the first one
7721 that it finds. This may give the occasional overlong line, but at least the
7722 remaining text gets split now.
7724 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7725 file descriptors might be low, and that setting 1000 would always raise it.
7726 It turns out that in some environments, the limit is already over 1000 and
7727 that lowering it causes trouble. So now Exim takes care not to decrease it.
7729 40. When delivering a message, the value of $return_path is set to $sender_
7730 address at the start of routing (routers may change the value). By an
7731 oversight, this default was not being set up when an address was tested by
7732 -bt or -bv, which affected the outcome if any router or filter referred to
7735 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7736 log only when "message" or "log_message" are set. However, if one of the
7737 conditions was an address verification, or a call to a nested ACL, the
7738 messages generated by the underlying test were being passed through. This
7739 no longer happens. The underlying message is available in $acl_verify_
7740 message for both "message" and "log_message" expansions, so it can be
7741 passed through if needed.
7743 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7744 new expansion $bh_ to give the encoded byte string without charset
7745 translation. Translation happens only if iconv() is available; HAVE_ICONV
7746 indicates this at build time. HEADERS_CHARSET gives the charset to
7747 translate to; headers_charset can change it in the configuration, and
7748 "headers charset" can change it in an individual filter file.
7750 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7751 that creates RFC 2047 encoded "words" labels them as that charset instead
7752 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7753 expansion operator; (ii) when Exim creates a From: line for a local
7754 message; (iii) when a header line is rewritten to include a "phrase" part.
7756 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7757 buggy, causing it to skip the first lines of messages whose message ID
7758 ended in 'D'. This would not have bitten before Exim release 4.14, because
7759 message IDs were unlikely to end in 'D' before then. The effect was to have
7760 incorrect size information for certain domains.
7762 45. #include "config.h" was missing at the start of the crypt16.c module. This
7763 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7766 46. If there was a timeout during a "random" callout check, Exim treated it as
7767 a failure of the random address, and carried on sending RSET and the real
7768 address. If the delay was just some slowness somewhere, the response to the
7769 original RCPT would be taken as a response to RSET and so on, causing
7770 mayhem of various kinds.
7772 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7773 when I implemented it. It didn't allow for the fact that some option values
7774 may legitimately be negative (e.g. size_addition), and it didn't even do
7775 the right test for positive values.
7777 48. Domain names in DNS records are case-independent. Exim always looks them up
7778 in lower case. Some resolvers return domain names in exactly the case they
7779 appear in the zone file, that is, they may contain uppercase letters. Not
7780 all resolvers do this - some return always lower case. Exim was treating a
7781 change of case by a resolver as a change of domain, similar to a widening
7782 of a domain abbreviation. This triggered its re-routing code and so it was
7783 trying to route what was effectively the same domain again. This normally
7784 caused routing to fail (because the router wouldn't handle the domain
7785 twice). Now Exim checks for this case specially, and just changes the
7786 casing of the domain that it ultimately uses when it transmits the message
7789 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7792 50. If a filter generated a file delivery with a non-absolute name (possible if
7793 no home directory exists for the router), the forbid_file option was not
7796 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7797 the existing equality matching.
7799 52. Exim was using ints instead of ino_t variables in some places where it was
7800 dealing with inode numbers.
7802 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7803 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7804 called TMPDIR, and if it finds it is different, it changes its value.
7806 54. The smtp_printf() function is now made available to local_scan() so
7807 additional output lines can be written before returning. There is also an
7808 smtp_fflush() function to enable the detection of a dropped connection.
7809 The variables smtp_input and smtp_batched_input are exported to
7812 55. Changed the default runtime configuration: the message "Unknown user"
7813 has been removed from the ACL, and instead placed on the localuser router,
7814 using the cannot_route_message feature. This means that any verification
7815 failures that generate their own messages won't get overridden. Similarly,
7816 the "Unrouteable address" message that was in the ACL for unverifiable
7817 relay addresses has also been removed.
7819 56. Added hosts_avoid_esmtp to the smtp transport.
7821 57. The exicyclog script was not checking for the esoteric option
7822 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7823 will work only if exicyclog is run under the appropriate euid.
7825 58. Following a discussion on the list, the rules by which Exim recognises line
7826 endings on incoming messages have been changed. The -dropcr and drop_cr
7827 options are now no-ops, retained only for backwards compatibility. The
7828 following line terminators are recognized: LF CRLF CR. However, special
7829 processing applies to CR:
7831 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7832 nor a local message in the state where . is a terminator.
7834 (ii) If a bare CR is encountered in a header line, an extra space is added
7835 after the line terminator so as not to end the header. The reasoning
7836 behind this is that bare CRs in header lines are most likely either
7837 to be mistakes, or people trying to play silly games.
7839 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7840 was being incorrectly given as 18 bytes larger than it should have been.
7841 This is a VOB (very old bug).
7843 60. This may never have affected anything current, but just in case it has:
7844 When the local host is found other than at the start of a list of hosts,
7845 the local host, those with the same MX, and any that follow, are discarded.
7846 When the list in question was part of a longer list of hosts, the following
7847 hosts (not currently being processed) were also being discarded. This no
7848 longer happens. I'm not sure if this situation could ever has previously
7851 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7854 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7855 header. This is something that is recommended in a new Internet Draft, and
7856 is something that is documented as being done by Sendmail. There are two
7857 possible values. For messages generated by the autoreply transport, Exim
7860 Auto-Submitted: auto-replied
7862 whereas for all other generated messages (e.g. bounces) it adds
7864 Auto-Submitted: auto-generated
7866 63. The "personal" condition in filters now includes a test for the
7867 Auto-Submitted: header. If it contains the string "auto-" the message it
7868 not considered personal.
7870 64. Added rcpt_include_affixes as a generic transport option.
7872 65. Added queue_only_override (default true).
7874 66. Added the syslog_duplication option.
7876 67. If what should have been the first header line of a message consisted of
7877 a space followed by a colon, Exim was mis-interpreting it as a header line.
7878 It isn't of course - it is syntactically invalid and should therefore be
7879 treated as the start of the message body. The misbehaviour could have
7880 caused a number of strange effects, including loss of data in subsequent
7881 header lines, and spool format errors.
7883 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7884 client host had authenticated. This control can now be exercised by an ACL
7885 for more flexibility.
7887 69. By default, callouts do not happen when testing with -bh. There is now a
7888 variant, -bhc, which does actually run the callout code, including
7889 consulting and updating the callout cache.
7891 70. Added support for saslauthd authentication, courtesy of Alexander
7894 71. If statvfs() failed on the spool or log directories while checking their
7895 size for availability, Exim confusingly gave the error "space shortage".
7896 Furthermore, in debugging mode it crashed with a floating point exception.
7897 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7898 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7899 serious problem, Exim now writes to the main and panic logs when this
7900 happens, with details of the failure. It then refuses to accept the
7901 incoming message, giving the message "spool directory problem" or "log
7902 directory problem" with a 421 code for SMTP messages.
7904 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7905 0, 1, and 2 exist, because some OS complain for execs without them (see
7906 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7907 descriptors. However, the code omitted to check that the open succeeded,
7908 causing mysterious errors if for some reason the permissions on /dev/null
7909 got screwed. Now Exim writes a message to the main and panic logs, and
7910 bombs out if it can't open /dev/null.
7912 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7913 interact so that it is all more flexible. It is supposed to remain
7914 backwards compatible. Also added extra_local_interfaces.
7916 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7917 to bomb out with an assertion failure - to the client this appears as a
7918 connection drop. This problem occurs in the part of the code that was taken
7919 from the Samba project. Fortunately, the assertion is in a very simple
7920 function, so I have fixed this by reproducing the function inline in the
7921 one place where it is called, and arranging for authentication to fail
7922 instead of killing the process with assert().
7924 75. The SPA client code was not working when the server requested OEM rather
7925 than Unicode encoding.
7927 76. Added code to make require_files with a specific uid setting more usable in
7928 the case where statting the file as root fails - usually a non-root-mounted
7929 NFS file system. When this happens and the failure is EACCES, Exim now
7930 forks a subprocess and does the per-uid checking as the relevant uid.
7932 77. Added process_log_path.
7934 78. If log_file_path was not explicitly set, a setting of check_log_space or
7935 check_log_inodes was ignored.
7937 79. If a space check for the spool or log partitions fails, the incident is now
7938 logged. Of course, in the latter case the data may get lost...
7940 80. Added the %p formatting code to string_format() so that it can be used to
7941 print addresses in debug_print(). Adjusted all the address printing in the
7942 debugging in store.c to use %p rather than %d.
7944 81. There was a concern that a line of code in smtp_in.c could overflow a
7945 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7946 initially expressed, the concern was not well-founded, because trailing
7947 spaces are removed early. However, if the trailing spaces were followed by
7948 a NULL, they did not get removed, so the overflow was possible. Two fixes
7951 (a) I re-wrote the offending code in a cleaner fashion.
7952 (b) If an incoming SMTP command contains a NULL character, it is rejected
7955 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7956 runs initgroups(), so that if the Exim user is in any additional groups,
7957 they will be used during message reception.
7963 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob