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.
6 Since Exim version 4.94
7 -----------------------
9 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
10 as arguments, so an implementation trying to copy these into a local
11 buffer was taking a taint-enforcement trap. Fix by using dynamically
12 created buffers. Similar fix for radius expansion condition.
14 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
15 reasonable, eg. to count headers. Fix by using dynamically created
16 buffers rather than a local. Do similar fixes for ACL actions "dcc",
17 "log_reject_target", "malware" and "spam"; the arguments are expanded
18 so could be handling tainted values.
20 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
21 broken the (no-op) support for this sendmail command. Restore it
22 to doing nothing, silently, and returning good status.
24 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
25 record path was given (or the default used) without a leading directory
26 path, an error occurred on trying to open it. Use the transport's working
29 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
30 transport. Previously it was the name on the DNS A-record; use instead
31 the head of the CNAME chain leading there (if there is one). This seems
32 to align better with RFC 6125.
34 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
35 smtp_accept_max_per_host allocated resources which were not released
36 when the limit was exceeded. This eventually crashed the daemon. Fix
37 by adding a relase action in that path.
39 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
40 expanded; previously using tainted values was rejected. Fix by using
41 dynamically-created buffers.
43 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
44 Previously a macro used one argument twice; when called with the
45 argument as an expression having side-effects, incorrect operation
46 resulted. Use an inlineable function.
48 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
49 held open for a verify callout. Previously this wan not accounted for
50 and a corrupt onward SMTP conversation resulted.
52 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
53 excluded, not matching the documentation.
55 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
56 was given for the sqlite_dbfile a trap resulted.
58 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
59 "name" argument resulted in a trap. There is no reason to disallow such;
60 this was a coding error.
62 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
63 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
64 spent suspended, ignoring the Posix definition. Previously we assumed
65 it did and a constant offset from real time could be used as a correction.
66 Change to using the same clock source for the start-of-message and the
67 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
68 exists, just to get a clock slightly more aligned to reality.
70 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
71 RFC says it is optional some validators care. The missing char was not
72 intended but triggered by a line-wrap alignement. Discovery and fix by
73 Guillaume Outters, hacked on by JH.
75 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
76 name being quoted was tainted a trap would be taken. Fix by using
77 dynamicaly created buffers. The routine could have been called by a
78 rewrite with the "h" flag, by using the "-F" command-line option, or
79 by using a "name=" option on a control=submission ACL modifier.
81 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
82 Previously when a whitespace character was specified it was not inserted
83 after removing the newline.
85 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
86 is_tainted() had an off-by-one error in the overenthusiastic direction.
87 Find and fix by Gavan. Although NetBSD is not a supported platform for
88 4.94 this bug could affect other platforms.
90 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
91 is_tainted() had an off-by-one error in the overenthusiastic direction.
92 Find and fix by Gavan. Although NetBSD is not a supported platform for
93 4.94 this bug could affect other platforms.
94 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
95 Previously when a whitespace character was specified it was not inserted
96 after removing the newline.
98 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
99 the domain part of the recipient address. This overrides any tls_sni
100 option set, which was previously used.
102 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
105 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
106 than one server was defined and depending on the platform memory layout
107 details, an internal consistency trap could be hit while walking the list
110 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
111 files. The value is used by the authresults expansion item. Previously
112 if this was used in a router or transport, a crash could result.
114 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
115 applied. This resulted, if any header-line rewrite rules were configured,
116 in a panic-log trigerrable by sending a message with a long address in
117 a header. Fix by increaing the arbitrary limit to larger than a single
118 (dewrapped) 5322 header line maximum size.
120 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
121 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
122 RFC 6648 which deprecates X- options in protocols as a general practice.
123 Changeover between the implementations is handled by the mechanisms
130 JH/01 Avoid costly startup code when not strictly needed. This reduces time
131 for some exim process initialisations. It does mean that the logging
132 of TLS configuration problems is only done for the daemon startup.
134 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
136 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
137 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
140 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
141 DKIM signing is being done. Previously a transport_filter would always
142 disable CHUNKING, falling back to traditional DATA.
144 JH/05 Regard command-line receipients as tainted.
146 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
148 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
149 PAM library frees one of the arguments given to it, despite the
150 documentation. Therefore a plain malloc must be used.
152 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
153 on-stack buffers were used, resulting in a taint trap when DSN information
154 copied from a received message was written into the buffer.
156 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
157 the ordering of its ARC headers. This caused a crash.
159 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
160 a new record was being constructed with information from the peer, a trap
163 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
164 installation would get error messages from DMARC verify, when it hit the
165 nonexistent file indicated by the default. Distros wanting DMARC enabled
166 should both provide the file and set the option.
167 Also enforce no DMARC verification for command-line sourced messages.
169 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
170 could, depending on the platform, hang at the STARTTLS response.
172 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
173 message on a connection. Previously if one message had ARC headers and
174 the following one did not, a crash could result when adding an
175 Authentication-Results: header.
177 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
178 Exim main code and Exim-related utities. The introduction of taint
179 tracking also did many adjustments to string handling. Since then, eximon
180 frequently terminated with an assert failure.
182 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
183 check for 452 responses. This slightly helps the inefficieny of doing
184 a large alias-expansion into a recipient-limited target. The max_rcpt
185 transport option still applies (and at the current default, will override
186 the new feature). The check is done for either cause of synch, and forces
187 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
188 connection. The new facility is not tunable at this time.
190 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
191 library live data was being used, so the results became garbage. Make
192 copies while it is still usable.
194 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
195 on delivery deferred (==) and failed (**) lines (if a delivery was
196 attemtped). Previously it was only on completion (=>) lines.
198 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
199 for the expansion of the server_scram_iter and server_scram_salt options.
201 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
202 are now specifically given a NO_DATA response without hitting the system
203 resolver. The library goes on to do the now-standard TXT lookup.
204 Use of dnsdb lookups is not affected.
206 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
207 only retrieve the errormessage once. Previously two calls to dlerror()
208 were used, and the second one (for mainlog/paniclog) retrieved null
211 JH/20 Taint checking: disallow use of tainted data for
212 - the appendfile transport file and directory options
213 - the pipe transport command
214 - the autoreply transport file, log and once options
215 - file names used by the redirect router (including filter files)
217 - paths used by single-key lookups
218 Previously this was permitted.
220 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
221 adjusted the size of a major service buffer; this failed because the
222 buffer was in use at the time. Change to a compile-time increase in the
223 buffer size, when this authenticator is compiled into exim.
225 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
226 previous fast-mode was untenable in the face of glibs using mmap to
227 support larger malloc requests.
229 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
230 New values supported, if defined on system where compiled:
231 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
232 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
234 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
235 running a limited number of proceses in parallel, a benefit is gained. The
236 amount varies with the platform hardware and load. The use of the option
237 queue_run_in_order means we cannot do this, as ordering becomes
240 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
241 had introduced a string-copy (for ensuring NUL-termination) which was not
242 appropriate for that case, which can include embedded NUL bytes in the
243 block of data. Investigation showed the copy to actually be needless, the
244 data being length-specified.
246 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
247 done during a receiving connection, and both used TLS, global info was
248 used rather than per-connection info for tracking the state of data
249 queued for transmission. This could result in a connection hang.
251 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
252 Previously, when delivering serveral messages down a single connection
253 only the first would provide a SIZE. This was due to the size information
254 not being properly tracked.
256 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
257 TAI (at 37 seconds currently), pretend to be in UTC for time-related
258 expansion and logging. Previously, spurious values such as a future
259 minute could be seen.
261 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
262 it could crash from a null-deref. This could also affect the
263 ${addresses: } operator and ${readsock } item.
265 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
266 message following a mime one, the variable was not reset.
268 JH/30 When an pipelined-connect fails at the first response, assume incorrect
269 cached capability (perhaps the peer reneged?) and immediately retry in
272 JH/31 Fix spurious detection of timeout while writing to transport filter.
274 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
275 an attempt to copy the string was made before checking it.
277 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
278 taint of the lookup key was maintained; we now regard the presence in the
279 filesystem as sufficient validation.
281 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
282 argument is supplied.
284 JH/35 The dsearch lookup now requires that the directory is an absolute path.
285 Previously this was not checked, and nonempty relative paths made an
286 access under Exim's current working directory.
288 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
289 Previously no event was raised.
291 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
292 parameter supplied by the sender MAIL FROM command. Previously it was
293 ignored, and only the check_spool_space option value for the required
296 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
297 the size of the signing public-key. Previously it was instead giving
298 the size of the signature hash.
300 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
301 the default. See the (new) dkim_verify_min_keysizes option.
303 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
304 and an ACL use a lookup for checking either the local_part or domain,
305 stale data could be accessed. Ensure that variable references are
306 dropped between messages.
308 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
309 by the client was not checked as pointing within response data before
310 being used. A malicious client could thus cause an out-of-bounds read and
311 possibly gain authentication. Fix by adding the check.
313 JH/42 Internationalisation: change the default for downconversion in the smtp
314 transport to be "if needed". Previously it was "as previously set" for
315 the message, which usually meant "if needed" for message-submission but
316 "no" for everything else. However, MTAs have been seen using SMTPUTF8
317 even when the envelope addresses did not need it, resulting in forwarding
318 failures to non-supporting MTAs. A downconvert in such cases will be
319 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
320 transport. The change does mean that addresses needing conversion will
321 be converted when previously a delivery failure would occur.
323 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
324 response was received it would be used unchecked in a fail-DSN, violating
325 standards on line-length limits. Truncate if needed.
327 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
328 doesn't work. (Additionally add a new main config option to configure the
335 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
336 side, to decode a TLS 1.3 packet capture.
338 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
339 Previously the default library behaviour applied, sending two, each in
342 JH/03 Debug output for ACL now gives the config file name and line number for
345 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
347 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
349 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
350 buffer overrun for (non-chunking) other transports.
352 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
353 TLS1.3, means that a server rejecting a client certificate is not visible
354 to the client until the first read of encrypted data (typically the
355 response to EHLO). Add detection for that case and treat it as a failed
356 TLS connection attempt, so that the normal retry-in-clear can work (if
357 suitably configured).
359 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
360 and/or domain. Found and fixed by Jason Betts.
362 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
363 configuration). If a CNAME target was not a wellformed name pattern, a
366 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
367 the OS reports them interleaved with other addresses.
369 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
370 used both for input and for a verify callout, both encrypted, SMTP
371 responses being sent by the server could be lost. This resulted in
372 dropped connections and sometimes bounces generated by a peer sending
375 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
376 string. Previously it was possible to cause undefined behaviour in a
377 library routine (usually a crash). Found by "zerons".
379 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
382 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
383 API was removed, so update to use the newer ones.
385 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
386 any timeout set, is taking a long time. Previously we would hang on to a
387 rotated logfile "forever" if the input was arriving with long gaps
388 (a previous attempt to fix addressed lack, for a long time, of initial
391 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
392 shared (NFS) environment. The length of the tempfile name is now
393 4 + 16 ("hdr.$message_exim_id") which might break on file
394 systems which restrict the file name length to lower values.
397 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
398 shared (NFS) environment.
400 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
401 did for all versions <4.90). Notably -M, -m, --invert, -I may be
404 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
405 on some platforms for bit 31.
407 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
408 to changes apparently associated with TLS1.3 handling some of the APIs
409 previously used were either nonfunctional or inappropriate. Strings
410 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
411 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
412 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
413 This affects log line X= elements, the $tls_{in,out}_cipher variables,
414 and the use of specific cipher names in the encrypted= ACL condition.
416 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
418 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
419 verification result was not updated unless hosts_require_ocsp applied.
421 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
422 queue_list_requires_admin set to false, non-admin users were denied the
425 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
426 directory-of-certs mode. Previously they were advertised despite the
429 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
430 A single TCP connection by a client will now hold a TLS connection open
431 for multiple message deliveries, by default. Previously the default was to
434 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
435 default. If built with the facility, DANE will be used. The facility
436 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
438 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
439 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
440 must be defined and you must still, unless you define DISABLE_TLS, manage
441 the the include-dir and library-file requirements that go with that
442 choice. Non-TLS builds are still supported.
444 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
447 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
448 default. If the platform supports and has the facility enabled, it will
449 be requested on all coneections.
451 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
452 controlled by the build-time option SUPPORT_PIPE_CONNECT.
454 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
456 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
457 success-DSN messages. Previously the From: header was always the default
458 one for these; the option was ignored.
460 JH/28 Fix the timeout on smtp response to apply to the whole response.
461 Previously it was reset for every read, so a teergrubing peer sending
462 single bytes within the time limit could extend the connection for a
463 long time. Credit to Qualsys Security Advisory Team for the discovery.
465 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
466 delivery address, which leaked information of the results of local
467 forwarding. Change to the original envelope recipient address, per
470 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
471 requested. Previously not bounce was generated and a log entry of
472 error ignored was made.
474 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
476 JH/32 Introduce a general tainting mechanism for values read from the input
477 channel, and values derived from them. Refuse to expand any tainted
478 values, to catch one form of exploit.
480 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
481 was unused and the unexpanded text used for the test. Found and
482 fixed by Ruben Jenster.
484 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
485 an attempt to use a TLS library read routine dereffed a nul pointer,
488 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
489 them in our smtp response.
491 JH/36 Have the general router option retry_use_local_part default to true when
492 any of the restrictive preconditions are set (to anything). Previously it
493 was only for check_local user. The change removes one item of manual
494 configuration which is required for proper retries when a remote router
495 handles a subset of addresses for a domain.
497 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
498 link count into consideration.
500 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
501 caused the extension of big_buffer, the following lines were ignored.
503 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
504 accordance with RFC 2308. Previously there was no expiry, so a longlived
505 receive process (eg. due to ACL delays) versus a short SOA value could
508 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
510 JH/39 Promote DMARC support to mainline.
512 JH/40 Bug 2452: Add a References: header to DSNs.
514 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
515 parameters. The relevant library call is documented as "Deprecated: This
516 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
517 3.6.0, DH parameters are negotiated following RFC7919."
519 HS/06 Change the default of dnssec_request_domains to "*"
521 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
522 carried on and emitted a BDAT command, even when PIPELINING was not
525 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
526 buffer was used for the filename, resulting in a trap when tainted
527 arguments (eg. $domain) were used.
529 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
530 recommended to avoid a possible server-load attack. The feature can be
531 re-enabled via the openssl_options main cofiguration option.
533 JH/45 local_scan API: documented the current smtp_printf() call. This changed
534 for version 4.90 - adding a "more data" boolean to the arguments.
535 Bumped the ABI version number also, this having been missed previously;
536 release versions 4.90 to 4.92.3 inclusive were effectively broken in
537 respect of usage of smtp_printf() by either local_scan code or libraries
538 accessed via the ${dlfunc } expansion item. Both will need coding
539 adjustment for any calls to smtp_printf() to match the new function
540 signature; a FALSE value for the new argument is always safe.
542 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
543 the file-offset (which the Linux syscall does, and exim expects); this
544 resulted in an indefinite loop.
546 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
547 to do ARC verification. The Authentication-Results: header line added
548 by the configuration then had no ARC item.
554 JH/01 Remove code calling the customisable local_scan function, unless a new
555 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
557 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
558 non-signal-safe functions being used.
560 JH/03 Bug 2269: When presented with a received message having a stupidly large
561 number of DKIM-Signature headers, disable DKIM verification to avoid
562 a resource-consumption attack. The limit is set at twenty.
564 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
565 report of oldest_pass in ${authres } in consequence, and separate out
566 some descriptions of reasons for verification fail.
568 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
569 files in the spool were present and unlocked. A queue-runner could spot
570 them, resulting in a duplicate delivery. Fix that by doing the unlock
571 after the unlink. Investigation by Tim Stewart. Take the opportunity to
572 add more error-checking on spoolfile handling while that code is being
575 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
576 No known attacks, no CVE, this is defensive hardening.
578 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
579 a queue-runner could start a delivery while other operations were ongoing.
580 Cutthrough delivery was a common victim, resulting in duplicate delivery.
581 Found and investigated by Tim Stewart. Fix by using the open message data
582 file handle rather than opening another, and not locally closing it (which
583 releases a lock) for that case, while creating the temporary .eml format
584 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
586 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
587 $sender_verify_failure/$recipient_verify_failure to "random".
589 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
592 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
593 Previously this would segfault.
595 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
598 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
599 like zero, since the resolver should be doing this for us, But we need one
600 as a CNAME but no MX presence gets the CNAME returned; we need to check
601 that doesn't point to an MX to declare it "no MX returned" rather than
602 "error, loop". A new main option is added so the older capability of
603 following some limited number of chain links is maintained.
605 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
607 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
608 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
609 now-deprecated earlier definitions used only specified the range up to TLS
610 1.2 (in the older-version library docs).
612 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
614 JH/15 Rework TLS client-side context management. Stop using a global, and
615 explicitly pass a context around. This enables future use of TLS for
616 connections to service-daemons (eg. malware scanning) while a client smtp
617 connection is using TLS; with cutthrough connections this is quite likely.
619 JH/16 Fix ARC verification to do AS checks in reverse order.
621 JH/17 Support a "tls" option on the ${readsocket } expansion item.
623 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
624 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
625 Previously the "utf8" would be re-prepended for every additional message.
627 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
628 Previously thery were accepted, resulting in issues when attempting to
629 forward messages to a non-supporting MTA.
631 PP/02 Let -n work with printing macros too, not just options.
633 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
634 one parent address was copied, and bogus data was used at delivery-logging
635 time. Either a crash (after delivery) or bogus log data could result.
636 Discovery and analysis by Tim Stewart.
638 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
639 Previously if the string ended mid-character, we did not insert the
640 promised '?' replacement.
642 PP/04 Documentation: current string operators work on bytes, not codepoints.
644 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
645 should pack well giving a smaller memory footprint so better caching and
646 therefore performance. Group the declarations where this can't be done so
647 that the byte-sized flag variables are not interspersed among pointer
648 variables, giving a better chance of good packing by the compiler.
650 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
651 non-null, to avoid issues with sites running BATV. Previously reports were
652 sent with an empty envelope sender so looked like bounces.
654 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
655 The ignore_error flag wasn't being returned from the filter subprocess so
656 was not set for later routers. Investigation and fix by Matthias Kurz.
658 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
659 and a msg:complete for the whole, when a message is manually removed using
660 -Mrm. Developement by Matthias Kurz, hacked on by JH.
662 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
663 a "Gnu special" function, asprintf() in the DB utility binary builds; I
664 hope that is portable enough.
666 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
667 requiring a known-CA anchor certificate; make it now rely entirely on the
668 TLSA as an anchor. Checking the name on the leaf cert against the name
669 on the A-record for the host is still done for TA (but not for EE mode).
671 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
672 included in delivery lines for non-proxied connections, when compiled with
673 SUPPORT_SOCKS and running with proxy logging enabled.
675 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
676 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
677 move the existing event to fire before the normal logging of message
678 failure so that custom logging is bracketed by normal logging.
680 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
681 msg:fail:internal event. Developement by Matthias Kurz.
683 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
684 far too small for todays use of crypto signatures stored there. Go all
685 the way to the max DNS message size of 64kB, even though this might be
686 overmuch for IOT constrained device use.
688 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
689 copy a string over itself. The library routine is documented as not
690 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
692 JH/32 For main options check_spool_space and check_inode_space, where the
693 platform supports 64b integers, support more than the previous 2^31 kB
694 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
695 the previous G, M, k.
697 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
698 $authenticated_fail_id variable on authentication failure. Previously
701 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
702 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
703 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
704 GNUTLS_SEC_PARAM_MEDIUM.
706 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
707 side. Previously we would continue as if no SNI had been received.
709 JH/36 Harden the handling of string-lists. When a list consisted of a sole
710 "<" character, which should be a list-separator specification, we walked
711 off past the nul-terimation.
713 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
714 causes) even when the retry time is not yet met. Previously they were
715 not, meaning that when (say) an account was over-quota and temp-rejecting,
716 and multiple senders' messages were queued, only one sender would get
717 notified on each configured delay_warning cycle.
719 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
721 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
722 cipher-suites, an error can be left on the stack even for a succeeding
723 accept; this results in impossible error messages when a later operation
726 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
727 return error codes indicating retry. Under TLS1.3 this becomes required.
729 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
730 it only wrote the new authenticators, resulting in a lack of tracking of
731 peer changes of ESMTP extensions until the next cache flush.
733 JH/41 Fix the loop reading a message header line to check for integer overflow,
734 and more-often against header_maxsize. Previously a crafted message could
735 induce a crash of the recive process; now the message is cleanly rejected.
737 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
738 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
744 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
745 When redis_servers is set to a list of > 1 element, and the Redis servers
746 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
747 case of MOVED into a DEFER case instead, thus moving the query onto the
748 next server in the list. For a cluster of N elements, all N servers must
749 be defined in redis_servers.
751 GF/02 Catch and remove uninitialized value warning in exiqsumm
752 Check for existence of @ARGV before looking at $ARGV[0]
754 JH/01 Replace the store_release() internal interface with store_newblock(),
755 which internalises the check required to safely use the old one, plus
756 the allocate and data copy operations duplicated in both (!) of the
757 extant use locations.
759 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
760 modifier. This matches the restriction on the commandline.
762 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
763 Previously only the last row was returned.
765 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
766 we assumed that tags in the header were well-formed, and parsed the
767 element content after inspecting only the first char of the tag.
768 Assumptions at that stage could crash the receive process on malformed
771 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
772 While running the DKIM ACL we operate on the Permanent memory pool so that
773 variables created with "set" persist to the DATA ACL. Also (at any time)
774 DNS lookups that fail create cache records using the Permanent pool. But
775 expansions release any allocations made on the current pool - so a dnsdb
776 lookup expansion done in the DKIM ACL releases the memory used for the
777 DNS negative-cache, and bad things result. Solution is to switch to the
778 Main pool for expansions.
779 While we're in that code, add checks on the DNS cache during store_reset,
780 active in the testsuite.
781 Problem spotted, and debugging aided, by Wolfgang Breyha.
783 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
784 When none of the hosts presented to a transport match an already-open
785 connection, close it and proceed with the list. Previously we would
786 queue the message. Spotted by Lena with Yahoo, probably involving
789 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
790 Previously a spurious "250 OK id=" response was appended to the proper
793 JH/08 The "support for" informational output now, which built with Content
794 Scanning support, has a line for the malware scanner interfaces compiled
795 in. Interface can be individually included or not at build time.
797 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
798 by the template makefile "src/EDITME". The "STREAM" support for an older
799 ClamAV interface method is removed.
801 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
802 rows affected is given instead).
804 JH/11 The runtime Berkeley DB library version is now additionally output by
805 "exim -d -bV". Previously only the compile-time version was shown.
807 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
808 SMTP connection. Previously, when one had more receipients than the
809 first, an abortive onward connection was made. Move to full support for
810 multiple onward connections in sequence, handling cutthrough connection
811 for all multi-message initiating connections.
813 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
814 routers. Previously, a multi-recipient message would fail to match the
815 onward-connection opened for the first recipient, and cause its closure.
817 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
818 a timeout on read on a GnuTLS initiating connection, resulting in the
819 initiating connection being dropped. This mattered most when the callout
820 was marked defer_ok. Fix to keep the two timeout-detection methods
823 JH/15 Relax results from ACL control request to enable cutthrough, in
824 unsupported situations, from error to silently (except under debug)
825 ignoring. This covers use with PRDR, frozen messages, queue-only and
828 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
830 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
831 metadata, resulting in a crash in free().
833 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
834 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
835 Broken also in d185889f4, with init system revamp.
837 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
838 Previously we abruptly closed the connection after reading a malware-
839 found indication; now we go on to read the "scan ok" response line,
842 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
843 ACL. Previously, a crash would result.
845 JH/19 Speed up macro lookups during configuration file read, by skipping non-
846 macro text after a replacement (previously it was only once per line) and
847 by skipping builtin macros when searching for an uppercase lead character.
849 JH/20 DANE support moved from Experimental to mainline. The Makefile control
850 for the build is renamed.
852 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
853 was allocated for every new TLS startup, meaning one per message. Fix
854 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
856 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
857 reported the original. Fix to report (as far as possible) the ACL
858 result replacing the original.
860 JH/23 Fix memory leak during multi-message connections using STARTTLS under
861 OpenSSL. Certificate information is loaded for every new TLS startup,
862 and the resources needed to be freed.
864 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
866 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
869 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
870 DATA response info to the (existing) per-recipient response info for
871 the "C=" log element. It can have useful tracking info from the
872 destination system. Patch from Simon Arlott.
874 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
875 length value. Previously this would segfault.
877 HS/02 Support Avast multiline protoocol, this allows passing flags to
878 newer versions of the scanner.
880 JH/28 Ensure that variables possibly set during message acceptance are marked
881 dead before release of memory in the daemon loop. This stops complaints
882 about them when the debug_store option is enabled. Discovered specifically
883 for sender_rate_period, but applies to a whole set of variables.
884 Do the same for the queue-runner and queue-list loops, for variables set
885 from spool message files. Do the same for the SMTP per-message loop, for
886 certain variables indirectly set in ACL operations.
888 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
889 as a multi-recipient message from a mailinglist manager). The coding had
890 an arbitrary cutoff number of characters while checking for more input;
891 enforced by writing a NUL into the buffer. This corrupted long / fast
892 input. The problem was exposed more widely when more pipelineing of SMTP
893 responses was introduced, and one Exim system was feeding another.
894 The symptom is log complaints of SMTP syntax error (NUL chars) on the
895 receiving system, and refused recipients seen by the sending system
896 (propating to people being dropped from mailing lists).
897 Discovered and pinpointed by David Carter.
899 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
900 replaced by the ${authresults } expansion.
902 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
904 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
905 allows proper process termination in container environments.
907 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
908 Previously the "final dot" had a newline after it; ensure it is CR,LF.
910 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
911 and "err_perm", deprecated since 4.83 when the RFC-defined words
912 "temperror" and "permerror" were introduced.
914 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
915 transport-filters or DKIM-signing. The restriction was lost in the
916 consolidation of verify-callout and delivery SMTP handling.
917 Extend the restriction to also cover ARC-signing.
919 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
920 in defer=pass mode supply a 450 to the initiator. Previously the message
923 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
924 tls_require_ciphers is used as before.
926 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
927 "pass_unscanned". Only tmpfails from the scanner are written to
928 the paniclog, as they may require admin intervention (permission
929 denied, license issues). Other scanner errors (like decompression
930 bombs) do not cause a paniclog entry.
932 JH/36 Fix reinitialisation of DKIM logging variable between messages.
933 Previously it was possible to log spurious information in receive log
936 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
937 triggered odd behaviour from Outlook Express clients.
939 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
942 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
943 since the IETF WG has not yet settled on that versus the original
944 "bare" representation.
946 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
947 Previously the millisecond value corrupted the output.
948 Fix also for syslog_pid=no and log_selector +pid, for which the pid
949 corrupted the output.
955 JH/01 Rework error string handling in TLS interface so that the caller in
956 more cases is responsible for logging. This permits library-sourced
957 string to be attached to addresses during delivery, and collapses
958 pairs of long lines into single ones.
960 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
961 during configuration. Wildcards are allowed and expanded.
963 JH/02 Rework error string handling in DKIM to pass more info back to callers.
964 This permits better logging.
966 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
967 do not close it down and have the child transport start it up again on
968 the passed-on TCP connection. Instead, proxy the child (and any
969 subsequent ones) for TLS via a unix-domain socket channel. Logging is
970 affected: the continued delivery log lines do not have any DNSSEC, TLS
971 Certificate or OCSP information. TLS cipher information is still logged.
973 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
974 identical IP addresses on different listening ports. Will also affect
977 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
978 add noisy ifdef guards to special-case this sillyness.
979 Patch from Bernd Kuhls.
981 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
982 than 255 are no longer allowed.
984 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
985 Disable the session-cache too, which might reduce our load. Since we
986 currrectly use a new context for every connection, both as server and
987 client, there is no benefit for these.
988 GnuTLS appears to not support tickets server-side by default (we don't
989 call gnutls_session_ticket_enable_server()) but client side is enabled
990 by default on recent versions (3.1.3 +) unless the PFS priority string
993 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
994 <https://reproducible-builds.org/specs/source-date-epoch/>.
996 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
997 the check for any unsuccessful recipients did not notice the limit, and
998 erroneously found still-pending ones.
1000 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1001 MSG_MORE. Only in-clear (not on TLS connections).
1003 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1004 transport-filter is involved we can buffer the headers in memory for
1005 creating the signature, and read the spool data file once for the
1006 signature and again for transmission.
1008 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1009 4.77 as the kernel support then wasn't solid, having issues in 64bit
1010 mode. Now, it's been long enough. Add support for FreeBSD also.
1012 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1013 case where the routing stage had gathered several addresses to send to
1014 a host before calling the transport for the first, we previously failed
1015 to close down TLS in the old transport process before passing the TCP
1016 connection to the new process. The new one sent a STARTTLS command
1017 which naturally failed, giving a failed delivery and bloating the retry
1018 database. Investigation and fix prototype from Wolfgang Breyha.
1020 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1021 false-negatives in the check that the sender had not preempted a response
1022 or prompt from Exim (running as a server), due to that code's lack of
1023 awareness of the SMTP input buffering.
1025 PP/04 Add commandline_checks_require_admin option.
1026 Exim drops privileges sanely, various checks such as -be aren't a
1027 security problem, as long as you trust local users with access to their
1028 own account. When invoked by services which pass untrusted data to
1029 Exim, this might be an issue. Set this option in main configuration
1030 AND make fixes to the calling application, such as using `--` to stop
1033 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1034 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1035 responses to those, into a single TLS record each way (this usually means
1036 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1039 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1040 affects you only if you're dancing at the edge of the param size limits.
1041 If you are, and this message makes sense to you, then: raise the
1042 configured limit or use OpenSSL 1.1. Nothing we can do for older
1045 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1046 cmdline element to get the documented default one. Previously it was
1049 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1050 are unsuable for non-message-specific reasons - eg. connection timeout,
1051 banner-time rejection.
1053 JH/16 Fix logging of delivery remote port, when specified by router, under
1056 PP/06 Repair manualroute's ability to take options in any order, even if one
1057 is the name of a transport.
1060 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1062 JH/17 Change the list-building routines interface to use the expanding-string
1063 triplet model, for better allocation and copying behaviour.
1065 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1066 Previously it was constructed the first time a possibly-matching string
1067 was met in the configuration file input during startup; now it is done
1070 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1071 compatible one, to avoid the (poorly documented) possibility of a config
1072 file in the working directory redirecting the DB files, possibly correpting
1073 some existing file. CVE-2017-10140 assigned for BDB.
1075 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1076 cache-hot. Previously, although the result was properly cached, the
1077 initial verify call returned a defer.
1079 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1080 the main verify for receipient in uncached-mode.
1082 JH/22 Retire historical build files to an "unsupported" subdir. These are
1083 defined as "ones for which we have no current evidence of testing".
1085 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1086 if present. Previously it was ignored.
1088 JH/24 Start using specified-initialisers in C structure init coding. This is
1089 a C99 feature (it's 2017, so now considered safe).
1091 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1092 if was a fixed-sized field and bitmask ops via macros; it is now more
1095 PP/07 GitHub PR 56: Apply MariaDB build fix.
1096 Patch provided by Jaroslav Škarvada.
1098 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1099 during Coverity cleanups [4.87 JH/47]
1100 Diagnosis and fix provided by Michael Fischer v. Mollard.
1102 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1103 the right size to place the terminating semicolon on its own folded
1104 line, the header hash was calculated to an incorrect value thanks to
1105 the (relaxed) space the fold became.
1107 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1108 and confused the parent.
1110 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1111 which could crash as a result. This could lead to undeliverable messages.
1113 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1116 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1117 matters on fast-turnover and PID-randomising systems, which were getting
1118 out-of-order delivery.
1120 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1121 a possibly-overlapping copy. The symptom was that "Remote host closed
1122 connection in response to HELO" was logged instead of the actual 4xx
1125 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1126 Previously only that bufferd was discarded, resulting in SYMTP command
1129 JH/32 DKIM: when a message has multiple signatures matching an identity given
1130 in dkim_verify_signers, run the dkim acl once for each. Previously only
1131 one run was done. Bug 2189.
1133 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1134 "panic the current process" to "deliberately defer". The panic log is
1135 still written with the problem list name; the mail and reject logs now
1136 get a temp-reject line for the message that was being handled, saying
1137 something like "domains check lookup or other defer". The SMTP 451
1138 message is still "Temporary local problem".
1140 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1141 A crafted sequence of BDAT commands could result in in-use memory beeing
1142 freed. CVE-2017-16943.
1144 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1145 from SMTP input. Previously it was always done; now only done for DATA
1146 and not BDAT commands. CVE-2017-16944.
1148 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1149 to the message (such as an overlong header line). Previously this was
1150 not done and we did not exit BDAT mode. Followon from the previous item
1151 though a different problem.
1157 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1158 than -2003 did; needs libidn2 in addition to libidn.
1160 JH/02 The path option on a pipe transport is now expanded before use.
1162 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1163 Patch provided by "Björn", documentation fix added too.
1165 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1166 missing a wire-to-host endian conversion.
1168 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1169 close after a BDAT command line could be taken as a following command,
1170 giving a synch failure. Fix by only checking for synch immediately
1171 before acknowledging the chunk.
1173 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1174 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1175 macro. Patches provided by Josh Soref.
1177 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1178 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1179 by IANA as a service extension. However, John Klensin suggests that we
1182 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1183 the dkim code may be unix-mode line endings rather than smtp wire-format
1184 CRLF, so prepend a CR to any bare LF.
1186 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1187 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1189 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1190 the first were themselves being wrongly included in the feed into dkim
1191 processing; with most chunk sizes in use this resulted in an incorrect
1192 body hash calculated value.
1194 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1195 DKIM signature block, for verification. Although advised against by
1196 standards it is specifically not ruled illegal.
1198 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1200 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1201 missing a body hash (the bh= tag).
1203 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1204 It seems that HAProxy sends the Proxy Protocol information in clear and
1205 only then does a TLS startup, so do the same.
1207 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1208 TCP connections (such as for Spamd) unless the daemon successfully set
1209 Fast Open mode on its listening sockets. This fixes breakage seen on
1210 too-old kernels or those not configured for Fast Open, at the cost of
1211 requiring both directions being enabled for TFO, and TFO never being used
1212 by non-daemon-related Exim processes.
1214 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1215 endings, at least on the first header line. Try to canonify any that get
1216 past that check, despite the cost.
1218 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1219 now limited to an arbitrary five deep, while parsing addresses with the
1220 strip_excess_angle_brackets option enabled.
1222 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1223 instead leave the unprompted TLS handshake in socket buffer for the
1224 TLS library to consume.
1226 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1228 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1230 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1231 region in one operation, for speed, and this leaves assigned pointers
1232 dangling. Add checks run only under the testsuite which checks all
1233 variables at a store-reset and panics on a dangling pointer; add code
1234 explicitly nulling out all the variables discovered. Fixes one known
1235 bug: a transport crash, where a dangling pointer for $sending_ip_address
1236 originally assigned in a verify callout, is re-used.
1238 PP/06 Drop '.' from @INC in various Perl scripts.
1240 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1242 PP/08 Reduce a number of compilation warnings under clang; building with
1243 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1244 should be warning-free.
1246 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1248 HS/01 Fix portability problems introduced by PP/08 for platforms where
1249 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1251 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1252 chunk. This allows us to accept broken chunked messages. We need a more
1253 general solution here.
1255 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1256 already-broken messages in the queue.
1258 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1260 JH/19 Fix reference counting bug in routing-generated-address tracking.
1266 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1267 supports it and a size is available (ie. the sending peer gave us one).
1269 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1270 years of being deprecated). The replacements are the ACLs
1271 acl_smtp_mime and acl_not_smtp_mime.
1273 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1274 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1275 or even an in-clear connection were permitted. Now, if the host lookup
1276 was dnssec and dane was requested then the host is only used if the
1277 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1278 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1279 if one fails this test.
1280 This means that a poorly-configured remote DNS will make it incommunicado;
1281 but it protects against a DNS-interception attack on it.
1283 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1284 non-noisy when a race steals the message being considered.
1286 JH/05 If main configuration option tls_certificate is unset, generate a
1287 self-signed certificate for inbound TLS connections.
1289 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1290 in rewrites and routers.
1292 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1293 and logged a warning sing 4.83; now they are a configuration file error.
1295 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1296 (lacking @domain). Apply the same qualification processing as RCPT.
1298 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1300 JH/10 Support ${sha256:} applied to a string (as well as the previous
1303 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1304 a cutthrough deliver is pending, as we always want to make a connection.
1305 This also avoids re-routing the message when later placing the cutthrough
1306 connection after a verify cache hit.
1307 Do not update it with the verify result either.
1309 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1310 when routing results in more than one destination address.
1312 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1313 signing (which inhibits the cutthrough capability). Previously only
1314 the presence of an option was tested; now an expansion evaluating as
1315 empty is permissible (obviously it should depend only on data available
1316 when the cutthrough connection is made).
1318 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1319 the relevant preceding SMTP command did not note the pipelining mode.
1321 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1322 Previously they were not counted.
1324 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1325 as one having no matching records. Previously we deferred the message
1326 that needed the lookup.
1328 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1329 distinguished as "(=".
1331 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1332 for missing MX records. Previously it only worked for missing A records.
1334 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1336 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1337 after the data-go-ahead and data-ack. Patch from Jason Betts.
1339 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1340 even for a "none" policy. Patch from Tony Meyer.
1342 JH/22 Fix continued use of a connection for further deliveries. If a port was
1343 specified by a router, it must also match for the delivery to be
1346 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1347 When one of the recipients of a message was unsuitable for the connection
1348 (has no matching addresses), we lost track of needing to mark it
1349 deferred. As a result mail would be lost.
1351 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1353 JH/25 Decoding ACL controls is now done using a binary search; the source code
1354 takes up less space and should be simpler to maintain. Merge the ACL
1355 condition decode tables also, with similar effect.
1357 JH/26 Fix problem with one_time used on a redirect router which returned the
1358 parent address unchanged. A retry would see the parent address marked as
1359 delivered, so not attempt the (identical) child. As a result mail would
1362 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1363 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1364 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1367 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1368 Main config options check_{log,spool}_{inodes,space} are now
1369 100 inodes, 10MB unless set otherwise in the configuration.
1371 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1372 Previously it only applied to the main-section connection policy
1375 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1377 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1378 by me. Added RFC7919 DH primes as an alternative.
1380 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1381 are not in the system include path.
1383 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1384 GnuTLS, when a session startup failed (eg because the client disconnected)
1385 Exim did stdio operations after fclose. This was exposed by a recent
1386 change which nulled out the file handle after the fclose.
1388 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1389 signed directly by the cert-signing cert, rather than an intermediate
1390 OCSP-signing cert. This is the model used by LetsEncrypt.
1392 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1394 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1395 an incoming connection.
1397 HS/02 Bug 1802: Do not half-close the connection after sending a request
1400 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1401 fallback to "prime256v1".
1403 JH/34 SECURITY: Use proper copy of DATA command in error message.
1404 Could leak key material. Remotely exploitable. CVE-2016-9963.
1410 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1411 and 3.4.4 - once the server is enabled to respond to an OCSP request
1412 it does even when not requested, resulting in a stapling non-aware
1413 client dropping the TLS connection.
1415 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1416 support variable-length bit vectors. No functional change.
1418 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1419 The I= interface field on outgoing lines is now after the H= remote
1420 host field, same as incoming lines. There is a separate
1421 outgoing_interface log selector which allows you to disable the
1424 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1425 If not running log_selector +smtp_connection the mainlog would be held
1426 open indefinitely after a "too many connections" event, including to a
1427 deleted file after a log rotate. Leave the per net connection logging
1428 leaving it open for efficiency as that will be quickly detected by the
1429 check on the next write.
1431 HS/01 Bug 1671: Fix post transport crash.
1432 Processing the wait-<transport> messages could crash the delivery
1433 process if the message IDs didn't exist for some reason. When
1434 using 'split_spool_directory=yes' the construction of the spool
1435 file name failed already, exposing the same netto behaviour.
1437 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1438 mime_regex ACL conditions.
1440 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1441 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1442 to HELO, local diagnostic string.
1444 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1445 log line to debug. Even when configured with a tls authenticator many
1446 client connections are expected to not authenticate in this way, so
1447 an authenticate fail is not an error.
1449 HS/02 Add the Exim version string to the process info. This way exiwhat
1450 gives some more detail about the running daemon.
1452 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1453 matter for fast-change records such as DNSBLs.
1455 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1456 retry record, even if constant. There may be multiple transports with
1457 different interface settings and the retry behaviour needs to be kept
1460 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1462 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1464 JH/10 Bug 840: fix log_defer_output option of pipe transport
1466 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1467 hosts_randomize. This matters a lot when combined with mua_wrapper.
1469 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1470 ${quote_pgsql:<string>} operator.
1472 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1474 JH/14 Change header file name for freeradius-client. Relevant if compiling
1475 with Radius support; from the Gentoo tree and checked under Fedora.
1477 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1479 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1480 option transport_filter, meaning no filtering.
1482 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1484 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1485 defaults to "*" (all hosts). The variable is now available when not built
1486 with TLS, default unset, mainly to enable keeping the testsuite sane.
1487 If a server certificate is not supplied (via tls_certificate) an error is
1488 logged, and clients will find TLS connections fail on startup. Presumably
1489 they will retry in-clear.
1490 Packagers of Exim are strongly encouraged to create a server certificate
1491 at installation time.
1493 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1494 with the $config_file variable.
1496 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1497 in transport context, after the attempt, and per-recipient. The latter type
1498 is per host attempted. The event data is the error message, and the errno
1499 information encodes the lookup type (A vs. MX) used for the (first) host,
1500 and the trailing two digits of the smtp 4xx response.
1502 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1503 to write to mainlog (or rejectlog, paniclog) in the window between file
1504 creation and permissions/ownership being changed. Particularly affects
1505 installations where exicyclog is run as root, rather than exim user;
1506 result is that the running daemon panics and dies.
1508 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1510 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1511 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1512 "pri" and "weight". Note that the previous implicit priority given by the
1513 list order is no longer honoured.
1515 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1516 for DKIM processing.
1518 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1519 by defining SUPPORT_SOCKS.
1521 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1522 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1523 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1524 variables are renamed to proxy_{local,external}_{address,port}.
1526 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1527 for a build by defining SUPPORT_I18N
1529 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1530 of the query string, and make ${quote_redis:} do that quoting.
1532 JH/29 Move Events support from Experimental to mainline, enabled by default
1533 and removable for a build by defining DISABLE_EVENT.
1535 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1537 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1538 cached by the daemon.
1540 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1541 by defining LOOKUP_REDIS. The libhiredis library is required.
1543 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1544 keys are given for lookup.
1546 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1547 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1548 only supported when built with TLS support. The PolarSSL SHA routines
1549 are still used when the TLS library is too old for convenient support.
1551 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1552 openssl_options), for security. OpenSSL forces this from version 1.1.0
1553 server-side so match that on older versions.
1555 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1556 allocation for $value could be released as the expansion processing
1557 concluded, but leaving the global pointer active for it.
1559 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1560 and to use the domains and local_parts ACL conditions.
1562 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1563 incorrectly not doubled on cutthrough transmission, hence seen as a
1564 body-termination at the receiving system - resulting in truncated mails.
1565 Commonly the sender saw a TCP-level error, and retransmitted the message
1566 via the normal store-and-forward channel. This could result in duplicates
1567 received - but deduplicating mailstores were liable to retain only the
1568 initial truncated version.
1570 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1572 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1574 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1575 we're in there, support oversigning also; bug 1309.
1577 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1579 HS/04 Add support for keep_environment and add_environment options.
1581 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1582 either intentional arithmetic overflow during PRNG, or testing config-
1585 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1586 delivery resulted in actual delivery. Cancel cutthrough before DATA
1589 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1590 rejects a recipient: pass the reject to the originator.
1592 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1593 Many were false-positives and ignorable, but it's worth fixing the
1596 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1597 for the new environment-manipulation done at startup. Move the routines
1598 from being local to tls.c to being global via the os.c file.
1600 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1601 an extract embedded as result-arg for a map, the first arg for extract
1602 is unavailable so we cannot tell if this is a numbered or keyed
1603 extraction. Accept either.
1609 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1612 JH/02 The smtp transport option "multi_domain" is now expanded.
1614 JH/03 The smtp transport now requests PRDR by default, if the server offers
1617 JH/04 Certificate name checking on server certificates, when exim is a client,
1618 is now done by default. The transport option tls_verify_cert_hostnames
1619 can be used to disable this per-host. The build option
1620 EXPERIMENTAL_CERTNAMES is withdrawn.
1622 JH/05 The value of the tls_verify_certificates smtp transport and main options
1623 default to the word "system" to access the system default CA bundle.
1624 For GnuTLS, only version 3.0.20 or later.
1626 JH/06 Verification of the server certificate for a TLS connection is now tried
1627 (but not required) by default. The verification status is now logged by
1628 default, for both outbound TLS and client-certificate supplying inbound
1631 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1634 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1635 Status Notification (bounce) messages are now MIME format per RFC 3464.
1636 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1637 under the control of the dsn_advertise_hosts option, and routers may
1638 have a dsn_lasthop option.
1640 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1641 default, modifiable by a malware= option. The list separator for
1642 the options can now be changed in the usual way. Bug 68.
1644 JH/10 The smtp_receive_timeout main option is now expanded before use.
1646 JH/11 The incoming_interface log option now also enables logging of the
1647 local interface on delivery outgoing connections.
1649 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1650 if the interface and destination host and port all match.
1652 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1655 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1656 Patch from Andrew Lewis.
1658 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1659 now supports optional time-restrictions, weighting, and priority
1660 modifiers per server. Patch originally by <rommer@active.by>.
1662 JH/16 The spamd_address main option now supports a mixed list of local
1663 and remote servers. Remote servers can be IPv6 addresses, and
1664 specify a port-range.
1666 JH/17 Bug 68: The spamd_address main option now supports an optional
1667 timeout value per server.
1669 JH/18 Bug 1581: Router and transport options headers_add/remove can
1670 now have the list separator specified.
1672 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1675 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1678 JH/21 Support for the A6 type of dns record is withdrawn.
1680 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1681 rather than the verbs used.
1683 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1684 from 255 to 1024 chars.
1686 JH/24 Verification callouts now attempt to use TLS by default.
1688 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1689 are generic router options now. The defaults didn't change.
1691 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1692 Original patch from Alexander Shikoff, worked over by JH.
1694 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1695 files if ZCAT_COMMAND is not executable.
1697 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1699 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1701 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1702 Normally benign, it bites when the pair was led to by a CNAME;
1703 modern usage is to not canonicalize the domain to a CNAME target
1704 (and we were inconsistent anyway for A-only vs AAAA+A).
1706 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1708 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1709 when evaluating $sender_host_dnssec.
1711 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1712 $sender_helo_dnssec variable.
1714 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1716 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1718 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1720 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1721 documented as working, but never had. Support all but $spam_report.
1723 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1724 added for tls authenticator.
1726 HS/03 Add perl_taintmode main config option
1732 TL/01 When running the test suite, the README says that variables such as
1733 no_msglog_check are global and can be placed anywhere in a specific
1734 test's script, however it was observed that placement needed to be near
1735 the beginning for it to behave that way. Changed the runtest perl
1736 script to read through the entire script once to detect and set these
1737 variables, reset to the beginning of the script, and then run through
1738 the script parsing/test process like normal.
1740 TL/02 The BSD's have an arc4random API. One of the functions to induce
1741 adding randomness was arc4random_stir(), but it has been removed in
1742 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1743 function when detected.
1745 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1746 cause callback expansion.
1748 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1749 syntax errors in an expansion can be treated as a string instead of
1750 logging or causing an error, due to the internal use of bool_lax
1751 instead of bool when processing it.
1753 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1754 server certificates when making smtp deliveries.
1756 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1758 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1760 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1762 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1763 Merged patch from Sebastian Wiedenroth.
1765 JH/05 Fix results-pipe from transport process. Several recipients, combined
1766 with certificate use, exposed issues where response data items split
1767 over buffer boundaries were not parsed properly. This eventually
1768 resulted in duplicates being sent. This issue only became common enough
1769 to notice due to the introduction of connection certificate information,
1770 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1772 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1773 size buffer was used, resulting in syntax errors when an expansion
1776 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1777 version 3.3.6 or later.
1779 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1780 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1781 both become "event_action", the variables become $event_name, $event_data
1782 and $event_defer_errno. There is a new variable $verify_mode, usable in
1783 routers, transports and related events. The tls:cert event is now also
1784 raised for inbound connections, if the main configuration event_action
1787 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1788 early OCSP support, but no stapling (appears to be less than 1.0.0).
1790 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1791 server certificate names available under the smtp transport option
1792 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1795 JH/10 Time-related extraction expansions from certificates now use the main
1796 option "timezone" setting for output formatting, and are consistent
1797 between OpenSSL and GnuTLS compilations. Bug 1541.
1799 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1800 encoded parameter in the incoming message. Bug 1558.
1802 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1803 include certificate info, eximon was claiming there were spoolfile
1806 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1808 JH/14 Log delivery-related information more consistently, using the sequence
1809 "H=<name> [<ip>]" wherever possible.
1811 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1812 are problematic for Debian distribution, omit them from the release
1815 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1817 JH/16 Fix string representation of time values on 64bit time_t architectures.
1820 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1821 output list separator was used.
1826 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1827 checkers that were complaining about end of non-void function with no
1830 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1831 This was a regression introduced in 4.83 by another bugfix.
1833 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1835 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1836 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1842 TF/01 Correctly close the server side of TLS when forking for delivery.
1844 When a message was received over SMTP with TLS, Exim failed to clear up
1845 the incoming connection properly after forking off the child process to
1846 deliver the message. In some situations the subsequent outgoing
1847 delivery connection happened to have the same fd number as the incoming
1848 connection previously had. Exim would try to use TLS and fail, logging
1849 a "Bad file descriptor" error.
1851 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1852 utilities have not been installed.
1854 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1855 temporary space as the ACL may create new global variables.
1857 TL/01 LDAP support uses per connection or global context settings, depending
1858 upon the detected version of the libraries at build time.
1860 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1861 to extract and use the src ip:port in logging and expansions as if it
1862 were a direct connection from the outside internet. PPv2 support was
1863 updated based on HAProxy spec change in May 2014.
1865 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1867 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1868 Properly escape header and check for NULL return.
1870 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1873 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1875 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1876 characters in header names, implemented as a verify condition.
1877 Contributed by Michael Fischer v. Mollard.
1879 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1880 results permerror and temperror. Previous values are deprecated but
1881 still accepted. In a future release, err_perm and err_temp will be
1882 completely removed, which will be a backward incompatibility if the
1883 ACL tests for either of these two old results. Patch contributed by
1884 user bes-internal on the mailing list.
1886 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1888 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1889 selectors, in both main and reject logs.
1891 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1894 JH/07 Add malware type "sock" for talking to simple daemon.
1896 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1898 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1899 routers/transports under cutthrough routing.
1901 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1902 numbers. Touch up "bool" conditional to keep the same definition.
1904 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1906 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1908 JH/12 Expand items in router/transport headers_add or headers_remove lists
1909 individually rather than the list as a whole. Bug 1452.
1911 Required for reasonable handling of multiple headers_ options when
1912 they may be empty; requires that headers_remove items with embedded
1913 colons must have them doubled (or the list-separator changed).
1915 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1916 view the policy declared in the DMARC record. Currently, $dmarc_status
1917 is a combined value of both the record presence and the result of the
1920 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1922 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1923 dnslookup router and the smtp transport (applying to the forward
1926 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1927 of ldap servers used for a specific lookup. Patch provided by Heiko
1930 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1931 New variable $lookup_dnssec_authenticated for observability.
1933 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1934 Patch submitted by Lars Timman.
1936 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1938 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1939 Requires trusted mode and valid format message id, aborts otherwise.
1940 Patch contributed by Heiko Schlichting.
1942 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1943 certextract with support for various fields. Bug 1358.
1945 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1946 is requested by default, modifiable by smtp transport option
1949 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1950 operate on certificate variables to give certificate fingerprints
1951 Also new ${sha256:cert_variable}.
1953 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1955 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1956 Christian Aistleitner.
1958 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1960 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1961 file. Patch from Wolfgang Breyha.
1963 JH/25 Expand the coverage of the delivery $host and $host_address to
1964 client authenticators run in verify callout. Bug 1476.
1966 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1967 align with daemon_smtp_ports. Bug 72.
1969 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1970 support and error reporting did not work properly.
1972 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1973 and is readable. Patch from Andrew Colin Kissa.
1975 TL/14 Enhance documentation of ${run expansion and how it parses the
1976 commandline after expansion, particularly in the case when an
1977 unquoted variable expansion results in an empty value.
1979 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1981 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1982 by renaming away from it. Addresses GH issue 15, reported by
1985 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1987 TL/15 SECURITY: prevent double expansion in math comparison functions
1988 (can expand unsanitized data). Not remotely exploitable.
1995 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1997 PP/02 Make -n do something, by making it not do something.
1998 When combined with -bP, the name of an option is not output.
2000 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2003 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2004 $sender_host_name and config options to manage this, and basic check
2007 PP/05 DSCP support for outbound connections and control modifier for inbound.
2009 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2010 (Only plugin which currently uses this is kerberos4, which nobody should
2011 be using, but we should make it available and other future plugins might
2012 conceivably use it, even though it would break NAT; stuff *should* be
2013 using channel bindings instead).
2015 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2016 name; added for Sendmail compatibility; requires admin caller.
2017 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2018 just ignore it); requires trusted caller.
2019 Also parse but ignore: -Ac -Am -X<logfile>
2022 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2024 TL/02 Add +smtp_confirmation as a default logging option.
2026 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2027 Patch by Magnus Holmgren from 2007-02-20.
2029 TL/04 Bugzilla 1281 - Spec typo.
2030 Bugzilla 1283 - Spec typo.
2031 Bugzilla 1290 - Spec grammar fixes.
2033 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2035 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2037 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2038 mailing list by by Dmitry Isaikin.
2040 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2042 JH/02 Support "G" suffix to numbers in ${if comparisons.
2044 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2046 NM/01 Bugzilla 1197 - Spec typo
2047 Bugzilla 1196 - Spec examples corrections
2049 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2051 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2052 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2055 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2056 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2058 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2059 "acl {{name}{arg}...}", and optional args on acl condition
2062 JH/05 Permit multiple router/transport headers_add/remove lines.
2064 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2066 JH/07 Avoid using a waiting database for a single-message-only transport.
2067 Performance patch from Paul Fisher. Bugzilla 1262.
2069 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2072 JH/09 Add $headers_added variable, with content from use of ACL modifier
2073 add_header (but not yet added to the message). Bugzilla 199.
2075 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2076 Pulled from Bugzilla 817 by Wolfgang Breyha.
2078 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2080 (nb: this is the same fix as in Exim 4.80.1)
2082 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2085 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2088 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2090 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2091 advertise SMTP AUTH mechanism to us, instead of a generic
2092 protocol violation error. Also, make Exim more robust to bad
2093 data from the Dovecot auth socket.
2095 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2097 When a queue runner is handling a message, Exim first routes the
2098 recipient addresses, during which it prunes them based on the retry
2099 hints database. After that it attempts to deliver the message to
2100 any remaining recipients. It then updates the hints database using
2103 So if a recipient address works intermittently, it can get repeatedly
2104 deferred at routing time. The retry hints record remains fresh so the
2105 address never reaches the final cutoff time.
2107 This is a fairly common occurrence when a user is bumping up against
2108 their storage quota. Exim had some logic in its local delivery code
2109 to deal with this. However it did not apply to per-recipient defers
2110 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2112 This change adds a proper retry rule check during routing so that the
2113 final cutoff time is checked against the message's age. We only do
2114 this check if there is an address retry record and there is not a
2115 domain retry record; this implies that previous attempts to handle
2116 the address had the retry_use_local_parts option turned on. We use
2117 this as an approximation for the destination being like a local
2118 delivery, as in LMTP.
2120 I suspect this new check makes the old local delivery cutoff check
2121 redundant, but I have not verified this so I left the code in place.
2123 TF/02 Correct gecos expansion when From: is a prefix of the username.
2125 Test 0254 submits a message to Exim with the header
2129 When I ran the test suite under the user fanf2, Exim expanded
2130 the header to contain my full name, whereas it should have added
2131 a Resent-Sender: header. It erroneously treats any prefix of the
2132 username as equal to the username.
2134 This change corrects that bug.
2136 GF/01 DCC debug and logging tidyup
2137 Error conditions log to paniclog rather than rejectlog.
2138 Debug lines prefixed by "DCC: " to remove any ambiguity.
2140 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2142 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2143 Bug spotted by Jeremy Harris; was flawed since initial commit.
2144 Would have resulted in OCSP responses post-SNI triggering an Exim
2145 NULL dereference and crash.
2147 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2149 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2150 Bug detection, analysis and fix by Samuel Thibault.
2151 Bugzilla 1331, Debian bug #698092.
2153 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2155 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2156 Server implementation by Todd Lyons, client by JH.
2157 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2158 config variable "prdr_enable" controls whether the server
2159 advertises the facility. If the client requests PRDR a new
2160 acl_data_smtp_prdr ACL is called once for each recipient, after
2161 the body content is received and before the acl_smtp_data ACL.
2162 The client is controlled by both of: a hosts_try_prdr option
2163 on the smtp transport, and the server advertisement.
2164 Default client logging of deliveries and rejections involving
2165 PRDR are flagged with the string "PRDR".
2167 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2168 fclose(). Diagnosis by Todd Lyons.
2170 PP/17 Update configure.default to handle IPv6 localhost better.
2171 Patch by Alain Williams (plus minor tweaks).
2174 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2175 This is now consistent with GnuTLS, and is now documented: the
2176 previous undocumented portable approach to treating the option as
2177 unset was to force an expansion failure. That still works, and
2178 an empty string is now equivalent.
2180 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2181 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2182 not performing validation itself.
2184 PP/20 Added force_command boolean option to pipe transport.
2185 Patch from Nick Koston, of cPanel Inc.
2187 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2190 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2192 PP/21 Fix eximon continuous updating with timestamped log-files.
2193 Broken in a format-string cleanup in 4.80, missed when I repaired the
2194 other false fix of the same issue.
2195 Report and fix from Heiko Schlichting.
2198 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2199 Report from Prashanth Katuri.
2201 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2202 It's SecureTransport, so affects any MacOS clients which use the
2203 system-integrated TLS libraries, including email clients.
2205 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2206 using a MIME ACL for non-SMTP local injection.
2207 Report and assistance in diagnosis by Warren Baker.
2209 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2211 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2213 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2214 last id that failed so it may be referenced in subsequent ACL's.
2216 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2219 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2220 ldap library initialization, allowing self-signed CA's to be
2221 used. Also properly sets require_cert option later in code by
2222 using NULL (global ldap config) instead of ldap handle (per
2223 session). Bug diagnosis and testing by alxgomz.
2225 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2226 the src/util/ subdirectory.
2228 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2229 renamed to Transport Post Delivery Action by Jeremy Harris, as
2232 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2233 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2234 redis_servers = needs to be configured which will be used by the redis
2235 lookup. Patch from Warren Baker, of The Packet Hub.
2237 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2239 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2240 hostname or reverse DNS when processing a host list. Used suggestions
2241 from multiple comments on this bug.
2243 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2245 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2246 Missed a few lines, added it to make the runtest require no keyboard
2249 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2250 contains upper case chars. Make router use caseful_local_part.
2252 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2253 support when GnuTLS has been built with p11-kit.
2259 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2261 This, or similar/improved, will also be change PP/11 of 4.82.
2267 PP/01 Handle short writes when writing local log-files.
2268 In practice, only affects FreeBSD (8 onwards).
2269 Bugzilla 1053, with thanks to Dmitry Isaikin.
2271 NM/01 Bugzilla 949 - Documentation tweak
2273 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2276 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2278 PP/02 Implemented gsasl authenticator.
2280 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2282 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2283 `pkg-config foo` for cflags/libs.
2285 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2286 with rest of GSASL and with heimdal_gssapi.
2288 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2289 `pkg-config foo` for cflags/libs for the TLS implementation.
2291 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2292 properties get this fed in as external SSF. A number of robustness
2293 and debugging improvements to the cyrus_sasl authenticator.
2295 PP/08 cyrus_sasl server now expands the server_realm option.
2297 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2298 Patch by Jeremy Harris.
2300 PP/10 Added dbmjz lookup type.
2302 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2304 PP/12 MAIL args handles TAB as well as SP, for better interop with
2305 non-compliant senders.
2306 Analysis and variant patch by Todd Lyons.
2308 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2309 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2310 Patch from Dirk Mueller <dmueller@suse.com>
2312 PP/13 tls_peerdn now print-escaped for spool files.
2313 Observed some $tls_peerdn in wild which contained \n, which resulted
2314 in spool file corruption.
2316 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2317 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2318 or write after TLS renegotiation, which otherwise led to messages
2321 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2322 as a tracking header (ie: a signed header comes before the signature).
2323 Patch from Wolfgang Breyha.
2325 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2326 comma-sep list; embedded commas doubled.
2328 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2330 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2332 Report and patch from Dmitry Banschikov.
2334 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2335 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2336 failures. We appear to now support TLS1.1+ with Exim.
2338 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2339 lets Exim select keys and certificates based upon TLS SNI from client.
2340 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2341 before an outbound SMTP session. New log_selector, +tls_sni.
2343 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2344 NULL dereference. Report and patch from Alun Jones.
2346 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2347 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2348 Not seeing resolver debug output on NetBSD, but suspect this is a
2349 resolver implementation change.
2351 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2352 Left warnings. Added "eximon gdb" invocation mode.
2354 PP/21 Defaulting "accept_8bitmime" to true, not false.
2356 PP/22 Added -bw for inetd wait mode support.
2358 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2359 locate the relevant includes and libraries. Made this the default.
2361 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2362 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2364 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2365 This may cause build issues on older platforms.
2367 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2368 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2369 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2370 Added SNI support via GnuTLS too.
2371 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2373 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2375 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2376 Applied second patch from Janne, implementing suggestion to default
2377 multiple-strings-in-record handling to match SPF spec.
2379 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2381 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2382 read-only, out of scope).
2383 Patch from Wolfgang Breyha, report from Stuart Northfield.
2385 PP/29 Fix three issues highlighted by clang analyser static analysis.
2386 Only crash-plausible issue would require the Cambridge-specific
2387 iplookup router and a misconfiguration.
2388 Report from Marcin Mirosław.
2390 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2392 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2393 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2394 As part of this, removing so much warning spew let me fix some minor
2395 real issues in debug logging.
2397 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2398 assignment on my part. Fixed.
2400 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2401 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2402 Janne Snabb (who went above and beyond: thank you).
2404 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2405 string otherwise requires a connection and a bunch more work and it's
2406 relatively easy to get wrong. Should also expose TLS library linkage
2409 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2410 64-bit ${eval} (JH/03).
2412 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2413 GNU libc to support some of the 64-bit stuff, should not lead to
2414 conflicts. Defined before os.h is pulled in, so if a given platform
2415 needs to override this, it can.
2417 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2418 protection layer was required, which is not implemented.
2419 Bugzilla 1254, patch from Wolfgang Breyha.
2421 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2422 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2423 tls_dhparam take prime identifiers. Also unbreak combination of
2424 OpenSSL+DH_params+TLSSNI.
2426 PP/39 Disable SSLv2 by default in OpenSSL support.
2432 PP/01 Solaris build fix for Oracle's LDAP libraries.
2433 Bugzilla 1109, patch from Stephen Usher.
2435 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2437 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2440 TF/02 Fix a couple more cases where we did not log the error message
2441 when unlink() failed. See also change 4.74-TF/03.
2443 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2444 lock up or crash if it happened to be inside a call to libc when it
2445 got a SIGUSR1 from exiwhat.
2447 The SIGUSR1 handler appends the current process status to the process
2448 log which is later printed by exiwhat. It used to use the general
2449 purpose logging code to do this, but several functions it calls are
2450 not safe for signals.
2452 The new output code in the SIGUSR1 handler is specific to the process
2453 log, and simple enough that it's easy to inspect for signal safety.
2454 Removing some special cases also simplifies the general logging code.
2455 Removing the spurious timestamps from the process log simplifies
2458 TF/04 Improved ratelimit ACL condition.
2460 The /noupdate option has been deprecated in favour of /readonly which
2461 has clearer semantics. The /leaky, /strict, and /readonly update modes
2462 are mutually exclusive. The update mode is no longer included in the
2463 database key; it just determines when the database is updated. (This
2464 means that when you upgrade Exim will forget old rate measurements.)
2466 Exim now checks that the per_* options are used with an update mode that
2467 makes sense for the current ACL. For example, when Exim is processing a
2468 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2469 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2470 must specify per_mail/readonly. If you omit the update mode it defaults to
2471 /leaky where that makes sense (as before) or /readonly where required.
2473 The /noupdate option is now undocumented but still supported for
2474 backwards compatibility. It is equivalent to /readonly except that in
2475 ACLs where /readonly is required you may specify /leaky/noupdate or
2476 /strict/noupdate which are treated the same as /readonly.
2478 A useful new feature is the /count= option. This is a generalization
2479 of the per_byte option, so that you can measure the throughput of other
2480 aggregate values. For example, the per_byte option is now equivalent
2481 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2483 The per_rcpt option has been generalized using the /count= mechanism
2484 (though it's more complicated than the per_byte equivalence). When it is
2485 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2486 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2487 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2488 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2489 Note that using per_rcpt with a non-readonly update mode in more than
2490 one ACL will cause the recipients to be double-counted. (The per_mail
2491 and per_byte options don't have this problem.)
2493 The handling of very low rates has changed slightly. If the computed rate
2494 is less than the event's count (usually one) then this event is the first
2495 after a long gap. In this case the rate is set to the same as this event's
2496 count, so that the first message of a spam run is counted properly.
2498 The major new feature is a mechanism for counting the rate of unique
2499 events. The new per_addr option counts the number of different
2500 recipients that someone has sent messages to in the last time period. It
2501 behaves like per_rcpt if all the recipient addresses are different, but
2502 duplicate recipient addresses do not increase the measured rate. Like
2503 the /count= option this is a general mechanism, so the per_addr option
2504 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2505 example, measure the rate that a client uses different sender addresses
2506 with the options per_mail/unique=$sender_address. There are further
2507 details in the main documentation.
2509 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2511 TF/06 Removed a few PCRE remnants.
2513 TF/07 Automatically extract Exim's version number from tags in the git
2514 repository when doing development or release builds.
2516 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2517 Bugzilla 879. Patch from Paul Fisher.
2519 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2520 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2523 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2525 PP/05 Variable $av_failed, true if the AV scanner deferred.
2526 Bugzilla 1078. Patch from John Horne.
2528 PP/06 Stop make process more reliably on build failure.
2529 Bugzilla 1087. Patch from Heiko Schlittermann.
2531 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2532 Bugzilla 1089. Patch from Heiko Schlittermann.
2534 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2535 Bugzilla 1131. Patch from Holger Weiß.
2537 PP/09 Handle IPv6 addresses with SPF.
2538 Bugzilla 860. Patch from Wolfgang Breyha.
2540 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2542 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2545 PP/11 match_* no longer expand right-hand-side by default.
2546 New compile-time build option, EXPAND_LISTMATCH_RHS.
2547 New expansion conditions, "inlist", "inlisti".
2549 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2551 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2553 PP/14 fix log_write() format string regression from TF/03.
2554 Bugzilla 1152. Patch from Dmitry Isaikin.
2560 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2562 PP/02 Harmonised TLS library version reporting; only show if debugging.
2563 Layout now matches that introduced for other libraries in 4.74 PP/03.
2565 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2567 PP/04 New "dns_use_edns0" global option.
2569 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2572 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2573 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2575 TK/01 Updated PolarSSL code to 0.14.2.
2576 Bugzilla 1097. Patch from Andreas Metzler.
2578 PP/07 Catch divide-by-zero in ${eval:...}.
2579 Fixes bugzilla 1102.
2581 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2584 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2585 format-string attack -- SECURITY: remote arbitrary code execution.
2587 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2588 time unintentionally subject to list matching rules, letting the header
2589 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2590 arbitrary string expansion). This allowed for information disclosure.
2592 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2593 INT_MIN/-1 -- value coerced to INT_MAX.
2599 NM/01 Workaround for PCRE version dependency in version reporting
2602 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2603 This fixes portability to compilers other than gcc, notably
2604 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2606 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2607 makefiles for portability to HP-UX and POSIX correctness.
2609 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2610 Also via indented variable definition in the Makefile.
2611 (Debugging by Oliver Heesakkers).
2613 PP/02 Restore caching of spamd results with expanded spamd_address.
2614 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2616 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2617 Improves build reliability. Fix from: Frank Elsner
2619 NM/02 Fix wide character breakage in the rfc2047 coding
2620 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2622 NM/03 Allow underscore in dnslist lookups
2623 Fixes bug 1026. Patch from Graeme Fowler
2625 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2626 Code patches from Adam Ciarcinski of NetBSD.
2628 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2631 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2632 is logged, to avoid truncation. Patch from John Horne.
2634 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2635 Patch from Jakob Hirsch.
2637 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2638 SQL string expansion failure details.
2639 Patch from Andrey Oktyabrski.
2641 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2642 Patch from Simon Arlott.
2644 PP/09 New lookups functionality failed to compile on old gcc which rejects
2645 extern declarations in function scope.
2646 Patch from Oliver Fleischmann
2648 PP/10 Use sig_atomic_t for flags set from signal handlers.
2649 Check getgroups() return and improve debugging.
2650 Fixed developed for diagnosis in bug 927 (which turned out to be
2653 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2654 Patch from Mark Zealey.
2656 PP/12 Bugzilla 1056: Improved spamd server selection.
2657 Patch from Mark Zealey.
2659 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2660 Based on patch from Heiko Schlittermann.
2662 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2663 Patch from Uwe Doering, sign-off by Michael Haardt.
2665 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2666 variable declaration deep within a block. Bug and patch from
2669 PP/15 lookups-Makefile IRIX compatibility coercion.
2671 PP/16 Make DISABLE_DKIM build knob functional.
2673 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2674 Patch by Simon Arlott
2676 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2677 variable argument macros. Our copy now differs from upstream.
2683 TF/01 Failure to get a lock on a hints database can have serious
2684 consequences so log it to the panic log.
2686 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2687 controlled using the smtp_confirmation log selector.
2689 TF/03 Include the error message when we fail to unlink a spool file.
2691 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2692 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2693 for maintaining out-of-tree patches for some time.
2695 PP/01 Bugzilla 139: Documentation and portability issues.
2696 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2697 Handle per-OS dynamic-module compilation flags.
2699 PP/02 Let /dev/null have normal permissions.
2700 The 4.73 fixes were a little too stringent and complained about the
2701 permissions on /dev/null. Exempt it from some checks.
2702 Reported by Andreas M. Kirchwitz.
2704 PP/03 Report version information for many libraries, including
2705 Exim version information for dynamically loaded libraries. Created
2706 version.h, now support a version extension string for distributors
2707 who patch heavily. Dynamic module ABI change.
2709 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2710 privilege escalation vulnerability whereby the Exim run-time user
2711 can cause root to append content of the attacker's choosing to
2714 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2717 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2718 If dropping privileges for untrusted macros, we disabled normal logging
2719 on the basis that it would fail; for the Exim run-time user, this is not
2720 the case, and it resulted in successful deliveries going unlogged.
2721 Fixed. Reported by Andreas Metzler.
2727 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2728 only prepend for the Resent-* case. Fixes regression introduced in
2729 Exim 4.70 by NM/22 for Bugzilla 607.
2731 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2732 increasing numbers of administrators be bitten by this.
2734 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2736 PP/03 Bugzilla 994: added openssl_options main configuration option.
2738 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2740 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2742 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2744 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2745 without a peer certificate, leading to a segfault because of an
2746 assumption that peers always have certificates. Be a little more
2747 paranoid. Problem reported by Martin Tscholak.
2749 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2750 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2751 NB: ClamAV planning to remove STREAM in "middle of 2010".
2752 CL also introduces -bmalware, various -d+acl logging additions and
2753 more caution in buffer sizes.
2755 PP/09 Implemented reverse_ip expansion operator.
2757 PP/10 Bugzilla 937: provide a "debug" ACL control.
2759 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2761 PP/12 Bugzilla 973: Implement --version.
2763 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2765 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2767 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2769 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2770 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2771 ignore trailing whitespace.
2773 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2775 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2778 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2779 Notification from Dr Andrew Aitchison.
2781 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2782 ExtendedDetectionInfo response format.
2783 Notification from John Horne.
2785 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2788 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2789 XSL and documented dependency on system catalogs, with examples of how
2792 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2795 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2796 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2797 configuration file which is writeable by the Exim user or group.
2799 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2800 of configuration files to cover files specified with the -C option if
2801 they are going to be used with root privileges, not just the default
2804 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2805 option (effectively making it always true).
2807 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2808 files to be used while preserving root privileges.
2810 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2811 that rogue child processes cannot use them.
2813 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2814 run-time user, instead of root.
2816 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2817 Exim run-time user without dropping privileges.
2819 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2820 result string, instead of calling string_vformat() twice with the same
2823 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2824 for other users. Others should always drop root privileges if they use
2825 -C on the command line, even for a whitelisted configure file.
2827 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2829 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2835 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2836 $data_path, and $header_path variables; fixed documentation bugs and
2839 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2840 exipick to access non-standard spools, including the "frozen" queue
2843 NM/01 Bugzilla 965: Support mysql stored procedures.
2844 Patch from Alain Williams
2846 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2848 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2849 Patch from Andreas Metzler
2851 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2852 Patch from Kirill Miazine
2854 NM/05 Bugzilla 671: Added umask to procmail example.
2856 JJ/03 installed exipick 20100323.0, fixing doc bug
2858 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2859 directory. Notification and patch from Dan Rosenberg.
2861 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2863 TK/02 Improve log output when DKIM signing operation fails.
2865 MH/01 Treat the transport option dkim_domain as a colon separated
2866 list, not as a single string, and sign the message with each element,
2867 omitting multiple occurences of the same signer.
2869 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2870 Bugzilla 985, 986. Patch by Simon Arlott
2872 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2873 Patch by Simon Arlott
2875 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2876 MBX locking. Notification from Dan Rosenberg.
2882 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2884 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2886 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2888 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2890 NM/04 Bugzilla 915: Flag broken perl installation during build.
2896 TK/01 Added patch by Johannes Berg that expands the main option
2897 "spamd_address" if it starts with a dollar sign.
2899 TK/02 Write list of recipients to X-Envelope-Sender header when building
2900 the mbox-format spool file for content scanning (suggested by Jakob
2903 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2904 (http://www.dcc-servers.net/) support via dccifd. Activated by
2905 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2907 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2908 by Mark Daniel Reidel <mr@df.eu>.
2910 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2911 When building exim an external PCRE library is now needed -
2912 PCRE is a system library on the majority of modern systems.
2913 See entry on PCRE_LIBS in EDITME file.
2915 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2916 conversation. Added nologin parameter to request.
2917 Patch contributed by Kirill Miazine.
2919 TF/01 Do not log submission mode rewrites if they do not change the address.
2921 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2923 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2924 log files in place. Contributed by Roberto Lima.
2926 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2928 TF/03 Bugzilla 615: When checking the local_parts router precondition
2929 after a local_part_suffix or local_part_prefix option, Exim now
2930 does not use the address's named list lookup cache, since this
2931 contains cached lookups for the whole local part.
2933 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2934 Robert Millan. Documentation is in experimental-spec.txt.
2936 TF/04 Bugzilla 668: Fix parallel build (make -j).
2938 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2940 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2941 Patch provided by Jan Srzednicki.
2943 TF/05 Leading white space used to be stripped from $spam_report which
2944 wrecked the formatting. Now it is preserved.
2946 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2947 that they are available at delivery time.
2949 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2951 TF/08 TLS error reporting now respects the incoming_interface and
2952 incoming_port log selectors.
2954 TF/09 Produce a more useful error message if an SMTP transport's hosts
2955 setting expands to an empty string.
2957 NM/06 Bugzilla 744: EXPN did not work under TLS.
2958 Patch provided by Phil Pennock.
2960 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2961 Patch provided by Richard Godbee.
2963 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2964 acl_smtp_notquit, added index entry.
2966 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2967 Patch provided by Eugene Bujak.
2969 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2970 accept(). Patch provided by Maxim Dounin.
2972 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2973 Patch provided by Phil Pennock.
2975 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2977 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2978 Patch provided by Brad "anomie" Jorsch.
2980 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2981 Patch provided by Dean Brooks.
2983 TK/05 Add native DKIM support (does not depend on external libraries).
2985 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2986 Patch provided by Graeme Fowler.
2988 NM/16 Bugzilla 851: Documentation example syntax fix.
2990 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2992 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2995 NM/19 Bugzilla 745: TLS version reporting.
2996 Patch provided by Phil Pennock.
2998 NM/20 Bugzilla 167: bool: condition support.
2999 Patch provided by Phil Pennock.
3001 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3002 clients. Patch provided by Phil Pennock.
3004 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3005 Patch provided by Brad "anomie" Jorsch.
3007 NM/23 Bugzilla 687: Fix misparses in eximstats.
3008 Patch provided by Heiko Schlittermann.
3010 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3011 Patch provided by Heiko Schlittermann.
3013 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3014 plus update to original patch.
3016 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3018 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3019 Patch provided by David Brownlee.
3021 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3023 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3025 NM/30 Bugzilla 888: TLS documentation bugfixes.
3027 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3029 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3030 Unlike the original bugzilla I have changed all shell scripts in src tree.
3032 NM/33 Bugzilla 898: Transport filter timeout fix.
3033 Patch by Todd Rinaldo.
3035 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3036 Patch by Serge Demonchaux.
3038 NM/35 Bugzilla 39: Base64 decode bug fixes.
3039 Patch by Jakob Hirsch.
3041 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3043 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3045 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3047 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3053 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3054 ALT-N's libdkim that I have put here:
3055 http://duncanthrax.net/exim-experimental/
3057 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3058 were called 'true' and it seems that C99 defines that as a reserved
3059 keyword to be used with 'bool' variable types. That means you could
3060 not include C99-style headers which use bools without triggering
3061 build errors in sieve.c.
3063 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3064 as mailq or other aliases. Changed the --help handling significantly
3065 to do whats expected. exim_usage() emits usage/help information.
3067 SC/01 Added the -bylocaldomain option to eximstats.
3069 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3071 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3073 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3079 PH/01 Another patch from the Sieve maintainer.
3081 PH/02 When an IPv6 address is converted to a string for single-key lookup
3082 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3083 dots are used instead of colons so that keys in lsearch files need not
3084 contain colons. This was done some time before quoting was made available
3085 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3086 (notated using the quote facility) so as to distinguish them from IPv4
3087 keys. This meant that lookups for IP addresses in host lists did not work
3088 for iplsearch lookups.
3090 This has been fixed by arranging for IPv6 addresses to be expressed with
3091 colons if the lookup type is iplsearch. This is not incompatible, because
3092 previously such lookups could never work.
3094 The situation is now rather anomalous, since one *can* have colons in
3095 ordinary lsearch keys. However, making the change in all cases is
3096 incompatible and would probably break a number of configurations.
3098 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3101 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3102 conversion specification without a maximum field width, thereby enabling
3103 a rogue spamd server to cause a buffer overflow. While nobody in their
3104 right mind would setup Exim to query an untrusted spamd server, an
3105 attacker that gains access to a server running spamd could potentially
3106 exploit this vulnerability to run arbitrary code as the Exim user.
3108 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3109 $primary_hostname instead of what libspf2 thinks the hosts name is.
3111 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3112 a directory entry by the name of the lookup key. Previously, if a
3113 symlink pointed to a non-existing file or a file in a directory that
3114 Exim lacked permissions to read, a lookup for a key matching that
3115 symlink would fail. Now it is enough that a matching directory entry
3116 exists, symlink or not. (Bugzilla 503.)
3118 PH/03 The body_linecount and body_zerocount variables are now exported in the
3121 PH/04 Added the $dnslist_matched variable.
3123 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3124 This means they are set thereafter only if the connection becomes
3127 PH/06 Added the client_condition to authenticators so that some can be skipped
3128 by clients under certain conditions.
3130 PH/07 The error message for a badly-placed control=no_multiline_responses left
3131 "_responses" off the end of the name.
3133 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3135 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3136 (without spaces) instead of just copying the configuration text.
3138 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3140 PH/11 Added $max_received_linelength.
3142 PH/12 Added +ignore_defer and +include_defer to host lists.
3144 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3145 way in which PCRE > 7.0 is built.
3147 PH/14 Implemented queue_only_load_latch.
3149 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3150 MAIL command. The effect was to mangle the value on 64-bit systems.
3152 PH/16 Another patch from the Sieve maintainer.
3154 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3156 PH/18 If a system quota error occurred while trying to create the file for
3157 a maildir delivery, the message "Mailbox is full" was not appended to the
3158 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3159 only to a quota excession during the actual writing of the file.
3161 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3162 characters?) which causes problems in log lines. The DN values are now
3163 passed through string_printing() before being added to log lines.
3165 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3166 and InterBase are left for another time.)
3168 PH/21 Added message_body_newlines option.
3170 PH/22 Guard against possible overflow in moan_check_errorcopy().
3172 PH/23 POSIX allows open() to be a macro; guard against that.
3174 PH/24 If the recipient of an error message contained an @ in the local part
3175 (suitably quoted, of course), incorrect values were put in $domain and
3176 $local_part during the evaluation of errors_copy.
3182 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3183 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3186 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3187 issue a MAIL command.
3189 PH/02 In an ACL statement such as
3191 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3193 if a client was not listed at all, or was listed with a value other than
3194 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3195 the condition was not true (as it should be), so access was not denied.
3196 The bug was that the ! inversion was incorrectly passed on to the second
3197 item. This has been fixed.
3199 PH/03 Added additional dnslists conditions == and =& which are different from
3200 = and & when the dns lookup returns more than one IP address.
3202 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3203 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3205 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3206 FSYNC, which compiles an option called disable_fsync that allows for
3207 bypassing fsync(). The documentation is heavily laced with warnings.
3209 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3211 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3212 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3213 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3214 including adding "make clean"; (3) Added -fPIC when compiling the test
3215 dynamically loaded module, to get rid of a warning.
3217 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3218 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3219 = 0s. The bug is otherwise harmless.
3221 PH/07 There was a bug in the dovecot authenticator such that the value of
3222 $auth1 could be overwritten, and so not correctly preserved, after a
3223 successful authentication. This usually meant that the value preserved by
3224 the server_setid option was incorrect.
3226 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3228 PH/09 Installed PCRE release 7.0.
3230 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3231 run for batched SMTP input. It is now run at the start of every message
3232 in the batch. While fixing this I discovered that the process information
3233 (output by running exiwhat) was not always getting set for -bs and -bS
3234 input. This is fixed, and it now also says "batched" for BSMTP.
3236 PH/11 Added control=no_pipelining.
3238 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3239 patch, slightly modified), and move the expansion of helo_data till after
3240 the connection is made in the smtp transport (so it can use these
3243 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3245 PH/14 Added log_selector = +pid.
3247 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3249 PH/16 Add ${if forany and ${if forall.
3251 PH/17 Added dsn_from option to vary the From: line in DSNs.
3253 PH/18 Flush SMTP output before performing a callout, unless control =
3254 no_callout_flush is set.
3256 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3257 was true (the default) a successful delivery failed to delete the retry
3258 item, thus causing premature timeout of the address. The bug is now
3261 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3263 PH/21 Long custom messages for fakedefer and fakereject are now split up
3264 into multiline responses in the same way that messages for "deny" and
3265 other ACL rejections are.
3267 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3268 with slight modification.
3270 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3271 draft, changing the syntax and factoring some duplicate code".
3273 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3274 for deliveries of the second and subsequent messages over the same SMTP
3277 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3278 ${reduce, with only minor "tidies".
3280 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3282 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3283 expansion side effects.
3285 PH/27 When a message times out after an over-quota error from an Exim-imposed
3286 quota, the bounce message says "mailbox is full". This message was not
3287 being given when it was a system quota that was exceeded. It now should
3290 MH/03 Made $recipients available in local_scan(). local_scan() already has
3291 better access to the recipient list through recipients_list[], but
3292 $recipients can be useful in postmaster-provided expansion strings.
3294 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3295 in the case of a MAIL command with additional options following the
3296 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3297 were accidentally chopped off.
3299 PH/29 SMTP synchronization checks are implemented when a command is read -
3300 there is a check that no more input is waiting when there shouldn't be
3301 any. However, for some commands, a delay in an ACL can mean that it is
3302 some time before the response is written. In this time, more input might
3303 arrive, invalidly. So now there are extra checks after an ACL has run for
3304 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3305 pipelining has not been advertised.
3307 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3309 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3310 returned in the SMTP response when smtp_return_error_details was false.
3311 This has been fixed.
3313 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3314 instead of the C I/O that was originally supplied, because problems were
3315 reported on Solaris.
3317 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3318 Exim which did not show up earlier: it was assuming that a call to
3319 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3320 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3321 was a macro that became an assignment, so it seemed to work. This has
3322 changed to a proper function call with a void return, hence the compile
3323 error. Exim's code has been fixed.
3325 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3328 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3330 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3332 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3333 because it turns out to be too big or there is a timeout) there is no
3334 "Completed" line in the log. When some messages of this type were
3335 selected by exigrep, they were listed as "not completed". Others were
3336 picked up by some special patterns. I have improved the selection
3337 criteria to be more general.
3339 PH/38 The host_find_failed option in the manualroute router can now be set
3340 to "ignore", to completely ignore a host whose IP address cannot be
3341 found. If all hosts are ignored, the behaviour is controlled by the new
3342 host_all_ignored option.
3344 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3345 homing or because of multiple MX records with /mx) generated more than
3346 one IP address, and the following item turned out to be the local host,
3347 all the secondary addresses of the first item were incorrectly removed
3348 from the list, along with the local host and any following hosts (which
3349 is what is supposed to happen).
3351 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3352 whoever called Exim into the -H file. In the case of the daemon it was
3353 behaving confusingly. When first started, it used values for whoever
3354 started the daemon, but after a SIGHUP it used the Exim user (because it
3355 calls itself on a restart). I have changed the code so that it now always
3358 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3359 message are rejected with the same error (e.g. no authentication or bad
3360 sender address), and a DATA command is nevertheless sent (as can happen
3361 with PIPELINING or a stupid MUA), the error message that was given to the
3362 RCPT commands is included in the rejection of the DATA command. This is
3363 intended to be helpful for MUAs that show only the final error to their
3366 PH/42 Another patch from the Sieve maintainer.
3368 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3369 Eximstats - Fixed some broken HTML links and added missing column headers
3371 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3372 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3374 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3376 PH/43 Yet another patch from the Sieve maintainer.
3378 PH/44 I found a way to check for a TCP/IP connection going away before sending
3379 the response to the final '.' that terminates a message, but only in the
3380 case where the client has not sent further data following the '.'
3381 (unfortunately, this is allowed). However, in many cases there won't be
3382 any further data because there won't be any more messages to send. A call
3383 to select() can be used: if it shows that the input is "ready", there is
3384 either input waiting, or the socket has been closed. An attempt to read
3385 the next input character can distinguish the two cases. Previously, Exim
3386 would have sent an OK response which the client would never have see.
3387 This could lead to message repetition. This fix should cure that, at
3388 least in a lot of common cases.
3390 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3391 advertised in response to EHLO.
3397 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3398 fixed by 4.65/MH/01 (is this a record?) are fixed:
3400 (i) An empty string was always treated as zero by the numeric comparison
3401 operators. This behaviour has been restored.
3403 (ii) It is documented that the numeric comparison operators always treat
3404 their arguments as decimal numbers. This was broken in that numbers
3405 starting with 0 were being interpreted as octal.
3407 While fixing these problems I realized that there was another issue that
3408 hadn't been noticed. Values of message_size_limit (both the global option
3409 and the transport option) were treated as octal if they started with 0.
3410 The documentation was vague. These values are now always treated as
3411 decimal, and I will make that clear in the documentation.
3417 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3418 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3421 MH/01 Don't check that the operands of numeric comparison operators are
3422 integers when their expansion is in "skipping" mode (fixes bug
3423 introduced by 4.64-PH/07).
3425 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3426 child addresses, Exim now panics and dies. Previously, because the count
3427 is held in a short int, deliveries were likely to be lost. As such a
3428 large number of recipients for a single message is ridiculous
3429 (performance will be very, very poor), I have chosen to impose a limit
3430 rather than extend the field.
3436 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3437 leftover -K file (the existence of which was triggered by #402).
3438 While we were at it, introduced process PID as part of the -K
3439 filename. This should rule out race conditions when creating
3442 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3443 processing considerably. Previous code took too long for large mails,
3444 triggering a timeout which in turn triggers #401.
3446 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3447 in the DK code in transports.c. sendfile() is not really portable,
3448 hence the _LINUX specificness.
3450 TF/01 In the add_headers option to the mail command in an Exim filter,
3451 there was a bug that Exim would claim a syntax error in any
3452 header after the first one which had an odd number of characters
3455 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3456 callout verification, Exim cached a "reject" for the entire domain. This
3457 is correct for most verifications, but it is not correct for a recipient
3458 verification with use_sender or use_postmaster set, because in that case
3459 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3460 case of MAIL FROM:<> rejection from other early rejections (e.g.
3461 rejection of HELO). When verifying a recipient using a non-null MAIL
3462 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3463 Whatever the result of the callout, the value of the domain cache is
3464 left unchanged (for any other kind of callout, getting as far as trying
3465 RCPT means that the domain itself is ok).
3467 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3470 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3471 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3472 session, the daemon ignores ECONNECT errors and logs others; it now
3473 ignores EPIPE as well.
3475 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3476 (quoted-printable decoding).
3478 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3479 later the small subsequent patch to fix an introduced bug.
3481 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3483 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3485 PH/08 An error is now given if message_size_limit is specified negative.
3487 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3488 to be given (somewhat) arbitrary names.
3490 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3493 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3494 miscellaneous code fixes
3496 PH/10 Added the log_reject_target ACL modifier to specify where to log
3499 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3500 hostname. This is wrong, because it relates to the incoming message (and
3501 probably the interface on which it is arriving) and not to the outgoing
3502 callout (which could be using a different interface). This has been
3503 changed to use the value of the helo_data option from the smtp transport
3504 instead - this is what is used when a message is actually being sent. If
3505 there is no remote transport (possible with a router that sets up host
3506 addresses), $smtp_active_hostname is used.
3508 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3509 tweaks were necessary in order to get it to work (see also 21 below):
3510 (a) The code assumed that strncpy() returns a negative number on buffer
3511 overflow, which isn't the case. Replaced with Exim's string_format()
3513 (b) There were several signed/unsigned issues. I just did the minimum
3514 hacking in of casts. There is scope for a larger refactoring.
3515 (c) The code used strcasecmp() which is not a standard C function.
3516 Replaced with Exim's strcmpic() function.
3517 (d) The code set only $1; it now sets $auth1 as well.
3518 (e) A simple test gave the error "authentication client didn't specify
3519 service in request". It would seem that Dovecot has changed its
3520 interface. Fortunately there's a specification; I followed it and
3521 changed what the client sends and it appears to be working now.
3523 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3526 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3527 address A is aliased to B and C, where B exists and C does not. Without
3528 -v the output is "A verified" because verification stops after a
3529 successful redirection if more than one address is generated. However,
3530 with -v the child addresses are also verified. Exim was outputting "A
3531 failed to verify" and then showing the successful verification for C,
3532 with its parentage. It now outputs "B failed to verify", showing B's
3533 parentage before showing the successful verification of C.
3535 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3536 look up a TXT record in a specific list after matching in a combined
3539 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3540 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3541 they consult the DNS. I had assumed they would set it the way they
3542 wanted; and indeed my experiments on Linux seem to show that in some
3543 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3544 To be on the safe side, however, I have now made the interface to
3545 host_find_byname() similar to host_find_bydns(), with an argument
3546 containing the DNS resolver options. The host_find_byname() function now
3547 sets these options at its start, just as host_find_bydns() does. The smtp
3548 transport options dns_qualify_single and dns_search_parents are passed to
3549 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3550 of host_find_byname() use the default settings of RES_DEFNAMES
3551 (qualify_single) but not RES_DNSRCH (search_parents).
3553 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3554 spool_read_header() do less string testing, by means of a preliminary
3555 switch on the second character of optional "-foo" lines. (This is
3556 overdue, caused by the large number of possibilities that now exist.
3557 Originally there were few.) While I was there, I also converted the
3558 str(n)cmp tests so they don't re-test the leading "-" and the first
3559 character, in the hope this might squeeze out yet more improvement.
3561 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3562 flag allowing group syntax was set by the header_syntax check but not
3563 turned off, possible causing trouble later; (2) The flag was not being
3564 set at all for the header_verify test, causing "group"-style headers to
3565 be rejected. I have now set it in this case, and also caused header_
3566 verify to ignore an empty address taken from a group. While doing this, I
3567 came across some other cases where the code for allowing group syntax
3568 while scanning a header line wasn't quite right (mostly, not resetting
3569 the flag correctly in the right place). These bugs could have caused
3570 trouble for malformed header lines. I hope it is now all correct.
3572 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3573 with the "reply" argument non-NULL. The code, however (which originally
3574 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3575 but it didn't always do it. This confused somebody who was copying the
3576 code for some other use. I have removed all the tests.
3578 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3579 feature that was used to support insecure browsers during the U.S. crypto
3580 embargo. It requires special client support, and Exim is probably the
3581 only MTA that supported it -- and would never use it because real RSA is
3582 always available. This code has been removed, because it had the bad
3583 effect of slowing Exim down by computing (never used) parameters for the
3584 RSA_EXPORT functionality.
3586 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3587 authenticator to fail if there's a tab character in the incoming data
3588 (there should never be unless someone is messing about, as it's supposed
3589 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3590 if the connection is using TLS or if the remote IP is the same as the
3591 local IP, and the "valid-client-cert option" if a client certificate has
3594 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3595 authenticators. This can be used for authorization after authentication
3596 succeeds. (In the case of plaintext, it servers for both authentication
3599 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3600 if any retry times were supplied.
3602 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3603 connection, where there is no client IP address to check. In this
3604 situation, the verify now always succeeds.
3606 PH/25 Applied John Jetmore's -Mset patch.
3608 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3610 PH/27 In a string expansion for a processed (not raw) header when multiple
3611 headers of the same name were present, leading whitespace was being
3612 removed from all of them, but trailing whitespace was being removed only
3613 from the last one. Now trailing whitespace is removed from each header
3614 before concatenation. Completely empty headers in a concatenation (as
3615 before) are ignored.
3617 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3618 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3620 PH/29 [Removed. This was a change that I later backed out, and forgot to
3621 correct the ChangeLog entry (that I had efficiently created) before
3622 committing the later change.]
3624 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3625 address errors (4xx response to RCPT) over the same connection as other
3626 messages routed to the same hosts. Such deliveries are always "forced",
3627 so retry times are not inspected. This resulted in far too many retries
3628 for the affected addresses. The effect occurred only when there were more
3629 hosts than the hosts_max_try setting in the smtp transport when it had
3630 the 4xx errors. Those hosts that it had tried were not added to the list
3631 of hosts for which the message was waiting, so if all were tried, there
3632 was no problem. Two fixes have been applied:
3634 (i) If there are any address or message errors in an SMTP delivery, none
3635 of the hosts (tried or untried) are now added to the list of hosts
3636 for which the message is waiting, so the message should not be a
3637 candidate for sending over the same connection that was used for a
3638 successful delivery of some other message. This seems entirely
3639 reasonable: after all the message is NOT "waiting for some host".
3640 This is so "obvious" that I'm not sure why it wasn't done
3641 previously. Hope I haven't missed anything, but it can't do any
3642 harm, as the worst effect is to miss an optimization.
3644 (ii) If, despite (i), such a delivery is accidentally attempted, the
3645 routing retry time is respected, so at least it doesn't keep
3646 hammering the server.
3648 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3649 in ${readsocket because some servers need this prod.
3651 PH/32 Added some extra debug output when updating a wait-xxx database.
3653 PH/33 The hint "could be header name not terminated by colon", which has been
3654 given for certain expansion errors for a long time, was not being given
3655 for the ${if def:h_colon_omitted{... case.
3657 PH/34 The spec says: "With one important exception, whenever a domain list is
3658 being scanned, $domain contains the subject domain." There was at least
3659 one case where this was not true.
3661 PH/35 The error "getsockname() failed: connection reset by peer" was being
3662 written to the panic log as well as the main log, but it isn't really
3663 panic-worthy as it just means the connection died rather early on. I have
3664 removed the panic log writing for the ECONNRESET error when getsockname()
3667 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3668 runs only) independently of the message's sender address. This meant
3669 that, if the 4xx error was in fact related to the sender, a different
3670 message to the same recipient with a different sender could confuse
3671 things. In particular, this can happen when sending to a greylisting
3672 server, but other circumstances could also provoke similar problems.
3673 I have changed the default so that the retry time for these errors is now
3674 based a combination of the sender and recipient addresses. This change
3675 can be overridden by setting address_retry_include_sender=false in the
3678 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3679 remote server are returned as part of bounce messages. This was not
3680 happening for LMTP over a pipe (the lmtp transport), but now it is the
3681 same for both kinds of LMTP.
3683 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3684 in header lines that were in fact CNAMEs. This is no longer the case.
3686 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3687 and queue runs started by the daemon processed all messages. This has
3688 been fixed so that -R and -S can now usefully be given with -q<time>.
3690 PH/40 Import PCRE release 6.7 (fixes some bugs).
3692 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3694 PH/42 Give an error if -q is specified more than once.
3696 PH/43 Renamed the variables $interface_address and $interface_port as
3697 $received_ip_address and $received_port, to make it clear that these
3698 values apply to message reception, and not to the outgoing interface when
3699 a message is delivered. (The old names remain recognized, of course.)
3701 PH/44 There was no timeout on the connect() call when using a Unix domain
3702 socket in the ${readsocket expansion. There now is.
3704 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3705 be meaningful with "accept".
3707 SC/01 Eximstats V1.43
3708 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3710 SC/02 Eximstats V1.44
3711 Use a glob alias rather than an array ref in the generated
3712 parser. This improves both readability and performance.
3714 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3715 Collect SpamAssassin and rejection statistics.
3716 Don't display local sender or destination tables unless
3717 there is data to show.
3718 Added average volumes into the top table text output.
3720 SC/04 Eximstats V1.46
3721 Collect data on the number of addresses (recipients)
3722 as well as the number of messages.
3724 SC/05 Eximstats V1.47
3725 Added 'Message too big' to the list of mail rejection
3726 reasons (thanks to Marco Gaiarin).
3728 SC/06 Eximstats V1.48
3729 Mainlog lines which have GMT offsets and are too short to
3730 have a flag are now skipped.
3732 SC/07 Eximstats V1.49 (Alain Williams)
3733 Added the -emptyok flag.
3735 SC/08 Eximstats V1.50
3736 Fixes for obtaining the IP address from reject messages.
3738 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3739 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3740 whitespace changes from 4.64-PH/27
3742 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3745 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3746 are found, allow negative numbers in numeric criteria)
3748 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3750 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3751 to match changes made in 4.64-PH/43
3753 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3755 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3756 "domain" is actually an IP address, and force a failure. This locks out
3757 those revolvers/nameservers that support "A-for-A" lookups, in
3758 contravention of the specifications.
3760 PH/48 When a host name was looked up from an IP address, and the subsequent
3761 forward lookup of the name timed out, the host name was left in
3762 $sender_host_name, contrary to the specification.
3764 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3765 restricted to single-key lookups, Exim was not diagnosing an error if
3766 * or *@ was used with a query-style lookup.
3768 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3770 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3771 long ago, but noone interested enough thought of it. Let's just say that
3772 the "1.1" means that there are some new functions that weren't there at
3773 some point in the past.
3775 PH/51 Error processing for expansion failure of helo_data from an smtp
3776 transport during callout processing was broken.
3778 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3779 tested/used via the -bh/-bhc/-bs options.
3781 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3782 bug, fixed in subsequent PCRE releases).
3784 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3785 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3787 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3793 SC/01 Use a glob alias rather than an array ref in eximstats generated
3794 parser. This improves both readability and performance.
3796 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3797 Don't display local sender or destination tables in eximstats unless
3798 there is data to show.
3799 Added average volumes into the eximstats top table text output.
3801 SC/03 Collect data on the number of addresses (recipients) as well
3802 as the number of messages in eximstats.
3804 TF/01 Correct an error in the documentation for the redirect router. Exim
3805 does (usually) call initgroups() when daemonizing.
3807 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3808 with consistent privilege compared to when running as a daemon.
3810 TF/03 Note in the spec that $authenticated_id is not set for local
3811 submissions from trusted users.
3813 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3814 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3816 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3817 by adding some example configuration directives to the default
3818 configuration file. A little bit of work is required to uncomment the
3819 directives and define how usernames and passwords are checked, but
3820 there is now a framework to start from.
3822 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3823 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3824 without this. I don't know how relevant this is to other LDAP libraries.
3826 PH/02 Add the verb name to the "unknown ACL verb" error.
3828 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3830 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3832 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3833 directory not expanded when it should be if an expanded home directory
3834 was set for the address (which is overridden by the transport).
3836 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3839 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3840 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3841 because it is too late at that time, and has no effect.
3843 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3844 security issue with \' (bugzilla #107). I could not use the
3845 PQescapeStringConn() function, because it needs a PGconn value as one of
3848 PH/08 When testing addresses using -bt, indicate those final addresses that
3849 are duplicates that would not cause an additional delivery. At least one
3850 person was confused, thinking that -bt output corresponded to deliveries.
3851 (Suppressing duplicates isn't a good idea as you lose the information
3852 about possibly different redirections that led to the duplicates.)
3854 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3855 systems where poll() doesn't work, in particular OS X.
3857 PH/10 Added more information to debugging output for retry time not reached.
3859 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3860 operations in malware.c.
3862 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3865 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3866 syslog_duplication set false, log lines that would normally be written
3867 both the the main log and to the reject log were not written to syslog at
3870 PH/14 In the default configuration, change the use of "message" in ACL warn
3871 statements to "add_header".
3873 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3874 not followed by a command (e.g. "seen endif").
3876 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3877 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3880 PH/17 Added extra conditions to the default value of delay_warning_condition
3884 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3885 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3886 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3889 The Auto-Submitted: and various List- headers are standardised, whereas I
3890 don't think Precedence: ever was.
3892 PH/18 Refactored debugging code in route_finduser() to show more information,
3893 in particular, the error code if getpwnam() issues one.
3895 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3896 This is apparently needed in addition to the PH/07 change above to avoid
3897 any possible encoding problems.
3899 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3900 but not after initializing Perl.
3902 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3903 output them only if debugging. By default they are written stderr,
3904 apparently, which is not desirable.
3906 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3909 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3912 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3914 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3915 authenticated or an ident call has been made. Suppress the default
3916 values for $authenticated_id and $authenticated_sender (but permit -oMai
3917 and -oMas) when testing with -bh.
3919 PH/24 Re-jigged the order of the tests in the default configuration so that the
3920 tests for valid domains and recipients precede the DNS black list and CSA
3921 tests, on the grounds that those ones are more expensive.
3923 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3924 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3925 This bug exists in every version of Exim that I still have, right back to
3928 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3929 However, an attempt to turn on case-sensitivity in a regex key by
3930 including (?-i) didn't work because the subject string was already
3931 lowercased, and the effects were non-intuitive. It turns out that a
3932 one-line patch can be used to allow (?-i) to work as expected.
3938 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3939 other effects) broke the use of negated acl sub-conditions.
3941 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3944 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3945 "Deny" causes Exim to reject the incoming connection with a 554 error.
3946 Unfortunately, if there is a major crisis, such as a disk failure,
3947 tcp-wrappers gives "deny", whereas what one would like would be some
3948 kind of temporary error. A kludge has been added to help with this.
3949 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3950 554 error is used if errno is still zero or contains ENOENT (which occurs
3951 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3954 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3956 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3957 errors. Otherwise a message that provokes a temporary error (when other
3958 messages do not) can cause a whole host to time out.
3960 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3961 addresses were routed directly to files or pipes from a redirect router.
3962 File deliveries just didn't batch; pipe deliveries might have suffered
3965 PH/06 A failure to get a lock for a hints database would erroneously always say
3966 "Failed to get write lock", even when it was really a read lock.
3968 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3969 of 0600. This has been changed to use the value of the lockfile_mode
3970 option (which defaults to 0600).
3972 PH/08 Applied small patch from the Sieve maintainer.
3974 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3975 folder from quota calculations, a direct delivery into this folder messed
3976 up the contents of the maildirsize file. This was because the regex was
3977 used only to exclude .Trash (or whatever) when the size of the mailbox
3978 was calculated. There was no check that a delivery was happening into an
3979 excluded directory. This bug has been fixed by ignoring all quota
3980 processing for deliveries into excluded directories.
3982 PH/10 Added the maildirfolder_create_regex option to appendfile.
3988 PH/01 The code for finding all the local interface addresses on a FreeBSD
3989 system running IPv6 was broken. This may well have applied to all BSD
3990 systems, as well as to others that have similar system calls. The broken
3991 code found IPv4 interfaces correctly, but gave incorrect values for the
3992 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3993 that it would not match correctly against @[] and not recognize the IPv6
3996 PH/02 The ipliteral router was not recognizing addresses of the form user@
3997 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3999 PH/03 Added disable_ipv6.
4001 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4002 decoded form, because it is most often used to construct To: headers
4003 lines in autoreplies, and the decoded form may well be syntactically
4004 invalid. However, $reply_address has leading white space removed, and all
4005 newlines turned into spaces so that the autoreply transport does not
4008 PH/05 If group was specified without a user on a router, and no group or user
4009 was specified on a transport, the group from the router was ignored.
4011 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4012 for visible compile-time settings that can be used to change these
4013 numbers, for those that want even more. Backwards compatibility with old
4014 spool files has been maintained. However, going back to a previous Exim
4015 release will lost any variables that are in spool files.
4017 PH/07 Two small changes when running in the test harness: increase delay when
4018 passing a TCP/IP connection to a new process, in case the original
4019 process has to generate a bounce, and remove special handling of
4020 127.0.0.2 (sic), which is no longer necessary.
4022 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4023 be the same on different OS.
4025 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4028 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4029 whether --show-vars was specified or not
4031 JJ/02 exipick: Added support for new ACL variable spool format introduced
4034 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4035 syntactically invalid From: or Reply-to: line, and a filter used this to
4036 generate an autoreply, and therefore failed to obtain an address for the
4037 autoreply, Exim could try to deliver to a non-existent relative file
4038 name, causing unrelated and misleading errors. What now happens is that
4039 it logs this as a hard delivery error, but does not attempt to create a
4042 PH/11 The exinext utility has a -C option for testing purposes, but although
4043 the given file was scanned by exinext itself; it wasn't being passed on
4044 when Exim was called.
4046 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4047 an end-of-file indication when reading a command response.
4049 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4050 compiled. In many other places in Exim, IPv6 addresses are always
4051 recognized, so I have changed this. It also means that IPv4 domain
4052 literals of the form [IPV4:n.n.n.n] are now always recognized.
4054 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4055 used if the router is not running as root, for example, when verifying at
4056 ACL time, or when using -bh. The debugging output from this situation was
4057 non-existent - all you got was a failure to exec. I have made two
4060 (a) Failures to set uid/gid, the current directory, or a process leader
4061 in a subprocess such as that created by queryprogram now generate
4062 suitable debugging output when -d is set.
4064 (b) The queryprogram router detects when it is not running as root,
4065 outputs suitable debugging information if -d is set, and then runs
4066 the subprocess without attempting to change uid/gid.
4068 PH/15 Minor change to Makefile for building test_host (undocumented testing
4071 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4072 additional section of a DNS packet that returns MX or SRV records.
4073 Instead, it always explicitly searches for A/AAAA records. This avoids
4074 major problems that occur when a DNS server includes only records of one
4075 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4076 fixed another bug: if SRV records were looked up and the corresponding
4077 address records were *not* found in the additional section, the port
4078 values from the SRV records were lost.
4080 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4081 using the correct key (the original address) when searching the retry
4082 rules in order to find which one to use for generating the retry hint.
4084 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4085 adding the default one. Similarly, if it contains a Reply-To: header, the
4086 errors_reply_to option, if set, is not used.
4088 PH/19 When calculating a retry time, Exim used to measure the "time since
4089 failure" by looking at the "first failed" field in the retry record. Now
4090 it does not use this if it is later than than the arrival time of the
4091 message. Instead it uses the arrival time. This makes for better
4092 behaviour in cases where some deliveries succeed, thus re-setting the
4093 "first failed" field. An example is a quota failure for a huge message
4094 when small messages continue to be delivered. Without this change, the
4095 "time since failure" will always be short, possible causing more frequent
4096 delivery attempts for the huge message than are intended.
4097 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4099 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4100 $1, $2, $3) because the numerical variables can be reset during some
4101 expansion items (e.g. "match"), thereby losing the authentication data.
4103 PH/21 Make -bV show the size of off_t variables so that the test suite can
4104 decide whether to run tests for quotas > 2G.
4106 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4107 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4108 filecount value is greater than 2G or if a quota value is greater than 2G
4109 on a system where the size of off_t is not greater than 4, a panic error
4112 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4113 never match. The debug and -bh output now contains an explicit error
4114 message indicating a malformed IPv4 address or mask.
4116 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4117 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4118 PH/23 above applies.
4120 PH/25 Do not write to syslog when running in the test harness. The only
4121 occasion when this arises is a failure to open the main or panic logs
4122 (for which there is an explicit test).
4124 PH/26 Added the /no_tell option to "control=freeze".
4126 PH/27 If a host name lookup failed very early in a connection, for example, if
4127 the IP address matched host_lookup and the reverse lookup yielded a name
4128 that did not have a forward lookup, an error message of the form "no IP
4129 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4130 could be logged. Now it outputs the IP address instead of "NULL".
4132 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4133 allows the sender and the authenticated sender to be set when
4134 submitting a message from within Exim. Since child_open_exim() is
4135 documented for local_scan(), the new function should be too.
4137 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4138 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4139 results in an empty string is now treated as unset.
4141 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4143 PH/31 Added sender_verify_fail logging option.
4145 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4146 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4147 tidied the source and removed it altogether.
4149 PH/33 When a queue run was abandoned because the load average was too high, a
4150 log line was always written; now it is written only if the queue_run log
4151 selector is set. In addition, the log line for abandonment now contains
4152 information about the queue run such as the pid. This is always present
4153 in "start" and "stop" lines but was omitted from the "abandon" line.
4155 PH/34 Omit spaces between a header name and the colon in the error message that
4156 is given when verify = headers_syntax fails (if there are lots of them,
4157 the message gets confusing).
4159 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4160 names, as there are now some PTR records that contain slashes. This check
4161 is only to protect against broken name servers that fall over on strange
4162 characters, so the fact that it applies to all lookups doesn't matter.
4164 PH/36 Now that the new test suite is complete, we can remove some of the
4165 special code in Exim that was needed for the old test suite. For example,
4166 sorting DNS records because real resolvers return them in an arbitrary
4167 order. The new test suite's fake resolver always returns records in the
4170 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4171 bounces) except when queue_only is set, to avoid logging races between
4172 the different processes.
4174 PH/38 Panic-die if .include specifies a non-absolute path.
4176 PH/39 A tweak to the "H" retry rule from its user.
4178 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4179 a label. They prevented compilation on older perls.
4181 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4182 a warning to be raised on newish perls.
4184 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4185 on queue. Changes to match documented behaviour of showing count of
4186 messages matching specified criteria.
4188 PH/40 Changed the default ident timeout from 30s to 5s.
4190 PH/41 Added support for the use of login_cap features, on those BSD systems
4191 that have them, for controlling the resources used by pipe deliveries.
4193 PH/42 The content-scanning code uses fopen() to create files in which to put
4194 message data. Previously it was not paying any attention to the mode of
4195 the files. Exim runs with umask(0) because the rest of the code creates
4196 files with open(), and sets the required mode explicitly. Thus, these
4197 files were ending up world-writeable. This was not a big issue, because,
4198 being within the spool directory, they were not world-accessible. I have
4199 created a function called modefopen, which takes an additional mode
4200 argument. It sets umask(777), creates the file, chmods it to the required
4201 mode, then resets the umask. All the relevant calls to fopen() in the
4202 content scanning code have been changed to use this function.
4204 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4205 to 24 hours. This avoids potential overflow problems when processing G
4206 and H retry rules. I suspect nobody ever tinkers with this value.
4208 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4210 PH/45 When the plaintext authenticator is running as a client, the server's
4211 challenges are checked to ensure they are valid base64 strings. By
4212 default, the authentication attempt is cancelled if an invalid string is
4213 received. Setting client_ignore_invalid_base64 true ignores these errors.
4214 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4215 they are received. Thus, the responses can be made to depend on the
4216 challenges. If an invalid string is ignored, an empty string is placed in
4219 PH/46 Messages that are created by the autoreply transport now contains a
4220 References: header, in accordance with RFCs 2822 and 3834.
4222 PH/47 Added authenticated_sender_force to the smtp transport.
4224 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4226 PH/49 Installed latest patch from the Sieve maintainer.
4228 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4229 was also set, the appendfile transport was unnecessarily scanning a
4230 directory of message files (e.g. for maildir delivery) to find the count
4231 of files (along with the size), even though it did not need this
4232 information. It now does the scan only if it needs to find either the
4233 size of the count of files.
4235 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4237 PH/52 Two bugs concerned with error handling when the smtp transport is
4240 (i) Exim was not creating retry information for temporary errors given
4241 for individual recipients after the DATA command when the smtp transport
4242 was used in LMTP mode. This meant that they could be retried too
4243 frequently, and not timed out correctly.
4245 (ii) Exim was setting the flag that allows error details to be returned
4246 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4247 recipients that were returned after the DATA command.
4249 PH/53 This is related to PH/52, but is more general: for any failing address,
4250 when detailed error information was permitted to be returned to the
4251 sender, but the error was temporary, then after the final timeout, only
4252 "retry timeout exceeded" was returned. Now it returns the full error as
4253 well as "retry timeout exceeded".
4255 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4256 do this, and (what is worse) MTAs that accept it.
4258 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4259 will now be deprecated.
4261 PH/56 New os.c-cygwin from the Cygwin maintainer.
4263 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4264 formats (previously only available in exim formats via -bpr, -bpru,
4265 and -bpra. Now also available in native and exiqgrep formats)
4267 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4268 with very large, slow to parse queues
4270 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4272 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4274 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4275 responses to authentication challenges, though it was showing the
4276 challenges; (ii) I've removed the CR characters from the debug output for
4279 PH/58 Allow for the insertion of a newline as well as a space when a string
4280 is turned into more than one encoded-word during RFC 2047 encoding. The
4281 Sieve code now uses this.
4283 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4284 data_4xx, lost_connection, tls_required.
4286 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4287 message was being sent as an SMTP response.
4289 PH/61 Add -l and -k options to exicyclog.
4291 PH/62 When verifying, if an address was redirected to one new address, so that
4292 verification continued, and the new address failed or deferred after
4293 having set something in $address_data, the value of $address_data was not
4294 passed back to the ACL. This was different to the case when no
4295 redirection occurred. The value is now passed back in both cases.
4297 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4298 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4299 use login_cap.h, so on its own it isn't the distinguishing feature. The
4300 new name refers directly to the setclassresources() function.
4302 PH/65 Added configuration files for NetBSD3.
4304 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4306 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4307 is preferred over IPv4.
4309 PH/68 The bounce_return_message and bounce_return_body options were not being
4310 honoured for bounces generated during the reception of non-SMTP messages.
4311 In particular, this applied to messages rejected by the ACL. This bug has
4312 been fixed. However, if bounce_return_message is true and bounce_return_
4313 body is false, the headers that are returned for a non-SMTP message
4314 include only those that have been read before the error was detected.
4315 (In the case of an ACL rejection, they have all been read.)
4317 PH/69 The HTML version of the specification is now built in a directory called
4318 spec_html instead of spec.html, because the latter looks like a path with
4319 a MIME-type, and this confuses some software.
4321 PH/70 Catch two compiler warnings in sieve.c.
4323 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4324 function verify_get_ident() calls ip_connect() to connect a socket, but
4325 if the "connect()" function timed out, ip_connect() used to close the
4326 socket. However, verify_get_ident() also closes the socket later, and in
4327 between Exim writes to the log, which may get opened at this point. When
4328 the socket was closed in ip_connect(), the log could get the same file
4329 descriptor number as the socket. This naturally causes chaos. The fix is
4330 not to close the socket in ip_connect(); the socket should be closed by
4331 the function that creates it. There was only one place in the code where
4332 this was missing, in the iplookup router, which I don't think anybody now
4333 uses, but I've fixed it anyway.
4335 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4336 well as to direct DNS lookups. Otherwise the handling of names in host
4337 lists is inconsistent and therefore confusing.
4343 PH/01 Two changes to the default runtime configuration:
4345 (1) Move the checks for relay_from_hosts and authenticated clients from
4346 after to before the (commented out) DNS black list checks.
4348 (2) Add control=submission to the relay_from_hosts and authenticated
4349 clients checks, on the grounds that messages accepted by these
4350 statements are most likely to be submissions.
4352 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4354 (1) Generate an error if the third argument for the ${prvs expansion is
4357 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4360 (3) Reset the variables that are obtained from the first argument of
4361 ${prvscheck and used in the second argument before leaving the code,
4362 because their memory is reclaimed, so using them afterwards may do
4365 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4366 one (it's much easier than Tom thought :-).
4368 (5) Because of (4), we can now allow for the use of $prvscheck_result
4369 inside the third argument.
4371 PH/03 For some reason, the default setting of PATH when running a command from
4372 a pipe transport was just "/usr/bin". I have changed it to
4375 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4376 anything to be listed in the output from -bV.
4378 PH/05 When a filter generated an autoreply, the entire To: header line was
4379 quoted in the delivery log line, like this:
4381 => >A.N.Other <ano@some.domain> <original@ddress> ...
4383 This has been changed so that it extracts the operative address. There
4384 may be more than one such address. If so, they are comma-separated, like
4387 => >ano@some.domain,ona@other.domain <original@ddress> ...
4389 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4390 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4391 not being looked up in the rDNS to get a host name, Exim was showing the
4392 IP address twice in Received: lines, even though the IP addresses were
4393 identical. For example:
4395 Received: from [1.2.3.4] (helo=[1.2.3.4])
4397 However, if the real host name was known, it was omitting the HELO data
4398 if it matched the actual IP address. This has been tidied up so that it
4399 doesn't show the same IP address twice.
4401 PH/07 When both +timestamp and +memory debugging was on, the value given by
4402 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4403 called by the debug printing, thereby overwriting the timestamp buffer.
4404 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4406 PH/08 When the original message was included in an autoreply transport, it
4407 always said "this is a copy of the message, including all the headers",
4408 even if body_only or headers_only was set. It now gives an appropriate
4411 PH/09 Applied a patch from the Sieve maintainer which:
4413 o fixes some comments
4414 o adds the (disabled) notify extension core
4415 o adds some debug output for the result of if/elsif tests
4416 o points to the current vacation draft in the documentation
4417 and documents the missing references header update
4421 o fixes a bug in processing the envelope test (when testing
4422 multiple envelope elements, the last element determined the
4425 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4426 Electronic Mail") by including:
4428 Auto-submitted: auto-generated
4430 in the messages that it generates (bounce messages and others, such as
4431 warnings). In the case of bounce messages for non-SMTP messages, there was
4432 also a typo: it was using "Auto_submitted" (underscore instead of
4433 hyphen). Since every message generated by Exim is necessarily in response
4434 to another message, thes have all been changed to:
4436 Auto-Submitted: auto-replied
4438 in accordance with these statements in the RFC:
4440 The auto-replied keyword:
4442 - SHOULD be used on messages sent in direct response to another
4443 message by an automatic process,
4445 - MUST NOT be used on manually-generated messages,
4447 - MAY be used on Delivery Status Notifications (DSNs) and Message
4448 Disposition Notifications (MDNs),
4450 - MUST NOT be used on messages generated by automatic or periodic
4451 processes, except for messages which are automatic responses to
4454 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4455 to the default Received: header definition.
4457 PH/12 Added log selector acl_warn_skipped (default on).
4459 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4460 variables because (a) they are in the wrong storage pool and (b) even if
4461 they were copied, it wouldn't work properly because of the caching.
4463 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4464 checking when decoding. Apparently there are clients that generate
4465 overlong encoded strings. Why am I not surprised?
4467 PH/15 If the first argument of "${if match_address" was not empty, but did not
4468 contain an "@" character, Exim crashed. Now it writes a panic log message
4469 and treats the condition as false.
4471 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4473 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4474 extension "envelope-auth". The code is finished and in agreement with
4475 other implementations, but there is no documentation so far and in fact,
4476 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4477 not changing the active code.
4479 Print executed "if" and "elsif" statements when debugging is used. This
4480 helps a great deal to understand what a filter does.
4482 Document more things not specified clearly in RFC3028. I had all this
4483 sorted out, when out of a sudden new issues came to my mind. Oops."
4485 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4488 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4489 canonical form (as documented). However, after a host name lookup from
4490 the IP address, check_host() was doing a simple string comparison with
4491 addresses acquired from the DNS when checking that the found name did
4492 have the original IP as one of its addresses. Since any found IPv6
4493 addresses are likely to be in abbreviated form, the comparison could
4494 fail. Luckily, there already exists a function for doing the comparison
4495 by converting both addresses to binary, so now that is used instead of
4496 the text comparison.
4498 PH/20 There was another similar case to PH/19, when a complete host name was
4499 given in a host list; looking up its IP address could give an abbreviated
4500 form, whereas the current host's name might or might not be abbreviated.
4501 The same fix has been applied.
4507 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4508 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4511 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4512 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4514 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4516 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4517 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4518 Also, for "Auto-Submitted", treat anything other than "no" as
4519 non-personal, in accordance with RFC 3834. (Previously it treated
4520 anything starting "auto-" as non-personal.)
4522 TF/01 The control=submission/name=... option had a problem with syntax
4523 errors if the name included a slash character. The /name= option
4524 now slurps the rest of the string, so it can include any characters
4525 but it must come last in the list of options (after /sender_retain
4528 PH/05 Some modifications to the interface to the fake nameserver for the new
4536 TK/01 Added the "success_on_redirect" address verification option. See
4537 NewStuff for rationale and an example.
4539 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4541 PH/02 Patch to exigrep to allow it to work on syslog lines.
4543 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4544 fread() to skip over the body file's header line, because in Cygwin the
4545 header line is locked and is inaccessible.
4547 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4548 co-exist for some time) to make it clear that it is the Exim ID that is
4549 referenced, not the Message-ID: header line.
4551 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4552 string_format() function, because snprintf() does not exist on all
4555 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4556 ${stat: expansion item.
4558 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4559 protocol synchronization error", to keep the pedants happy.
4561 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4562 well as for IRIX systems, when gcc is being used. See the host.c source
4565 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4567 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4570 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4571 transport and to the smtp transport in LMTP mode.
4573 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4575 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4576 run a filter in a subprocess. This could lead to confusion in subsequent
4577 lookups in the parent process. There should also be a search_tidyup() at
4578 the end of the subprocess.
4580 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4581 only if the host matched helo_try_verify_hosts, which caused the
4582 verification to occur when the EHLO/HELO command was issued. The ACL just
4583 tested the remembered result. Now, if a previous verification attempt has
4584 not happened, "verify = helo" does it there and then.
4586 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4588 TK/03 Fix log output including CR from clamd.
4590 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4591 exist provoked a memory error which could cause a segfault.
4593 PH/15 Installed PCRE 6.2
4595 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4597 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4598 of the problem. Specifically, suggested +O2 rather than +O1 for the
4601 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4603 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4604 record, the port was not correctly logged when the outgoing_port log
4605 selector was set (it logged the transort's default port).
4607 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4608 fallback_hosts, and "hosts" in the smtp transport.
4610 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4611 host errors such as "Connection refused".
4613 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4614 authentication with radiusclient 0.4.9:
4616 - Error returned from rc_read_config was caught wrongly
4617 - Username/password not passed on to radius server due to wrong length.
4619 The presumption is that some radiusclient API changes for 4.51/PH/17
4620 were not taken care of correctly. The code is still untested by me (my
4621 Linux distribution still has 0.3.2 of radiusclient), but it was
4622 contributed by a Radius user.
4624 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4625 expanding the "port" option of the smtp transport.
4627 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4628 while reading a MIME header. Thanks to Tom Hughes for a patch.
4630 PH/24 Include config.h inside local_scan.h so that configuration settings are
4633 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4634 that in an ACL for RCPT (for example), you can examine exactly what was
4637 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4638 commands, but it was not correctly comparing the address with the actual
4639 client host address. Thus, it would show the EHLO address in Received:
4640 header lines when this was not necessary.
4642 PH/27 Added the % operator to ${eval:}.
4644 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4645 it should be ignoring failures (because with -C, for example, it has lost
4646 privilege). It wasn't ignoring creation failures other than "already
4649 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4652 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4653 because an input line was too long, either on its own, or by virtue of
4654 too many continuations, the temporary file was not being removed, and the
4655 return code was incorrect.
4657 PH/31 Missing "BOOL" in function definition in filtertest.c.
4659 PH/32 Applied Sieve patches from the maintainer.
4661 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4663 PH/33 Added "verify = not_blind".
4665 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4666 Local/Makefile (with some defaults set). These are used in built scripts
4667 such as exicyclog, but they have never been used in the exim_install
4668 script (though there are many overriding facilities there). I have
4669 arranged that the exim_install script now takes note of these two
4672 PH/35 Installed configuration files for Dragonfly.
4674 PH/36 When a locally submitted message by a trusted user did not contain a
4675 From: header, and the sender address was obtained from -f or from an SMTP
4676 MAIL command, and the trusted user did not use -F to supply a sender
4677 name, $originator_name was incorrectly used when constructing a From:
4678 header. Furthermore, $originator_name was used for submission mode
4679 messages from external hosts without From: headers in a similar way,
4680 which is clearly wrong.
4682 PH/37 Added control=suppress_local_fixups.
4684 PH/38 When log_selector = +received_sender was set, and the addition of the
4685 sender made the log line's construction buffer exactly full, or one byte
4686 less than full, an overflow happened when the terminating "\n" was
4689 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4690 when the result of a list match is failure because a DNS lookup failed.
4692 PH/40 RM_COMMAND is now used in the building process.
4694 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4695 the "build-* directories that it finds.
4697 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4698 address in a domain literal was a prefix of an interface address.
4700 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4701 when verifying a sender address, unless rewrite_headers is false.
4703 PH/44 Wrote a long comment about why errors_to addresses are verified as
4704 recipients, not senders.
4706 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4707 the ratelimit ACL was added.
4709 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4711 PH/46 Added extra information about PostgreSQL errors to the error string.
4713 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4714 avoiding the need to install special zones in a real server. This is
4715 backwards compatible; if it can't find the fake resolver, it drops back.
4716 Thus, both old and new test suites can be run.
4718 TF/02 Added util/ratelimit.pl
4720 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4721 clock is set back in time.
4723 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4724 Candler <B.Candler@pobox.com>.
4726 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4727 OK for addresses that are the result of redirections.
4729 PH/48 A number of further additions for the benefit of the new test suite,
4730 including a fake gethostbyname() that interfaces to the fake DNS resolver
4733 TF/06 The fix for widen_domains has also been applied to qualify_single and
4734 search_parents which are the other dnslookup options that can cause
4737 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4740 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4742 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4743 service" if it encountered an MX record with an empty target hostname.
4744 The message is now "an MX or SRV record indicated no SMTP service".
4746 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4747 if the DNS of the sending site is misconfigured. This is quite a
4748 common situation. This change restores the behaviour of treating a
4749 helo verification defer as a failure.
4751 PH/51 If self=fail was set on a router, the bounce message did not include the
4752 actual error message.
4758 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4760 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4761 command itself ended in error, the underlying message about the transport
4762 filter timeout was being overwritten with the pipe command error. Now the
4763 underlying error message should be appended to the second error message.
4765 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4767 PH/02 Reduce the amount of output that "make" produces by default. Full output
4768 can still be requested.
4770 PH/03 The warning log line about a condition test deferring for a "warn" verb
4771 was being output only once per connection, rather than after each
4772 occurrence (because it was using the same function as for successful
4773 "warn" verbs). This seems wrong, so I have changed it.
4775 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4776 it should not have, which might have caused a crash in the right
4777 circumstances, but probably never did.
4779 PH/04 Installed a modified version of Tony Finch's patch to make submission
4780 mode fix the return path as well as the Sender: header line, and to
4781 add a /name= option so that you can make the user's friendly name appear
4784 TF/03 Added the control = fakedefer ACL modifier.
4786 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4787 Mark Lowes for thorough testing.
4789 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4791 TK/03 Merged latest SRS patch from Miles Wilton.
4793 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4794 with the definition in sysexits.h (which is #included earlier).
4795 Fortunately, Exim does not actually use EX_OK. The code used to try to
4796 preserve the sysexits.h value, by assuming that macro definitions were
4797 scanned for macro replacements. I have been disabused of this notion,
4798 so now the code just undefines EX_OK before #including unistd.h.
4800 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4801 in the smtp transport. When a block could not be written in a single
4802 write() function, the timeout was being re-applied to each part-write.
4803 This seems wrong - if the receiver was accepting one byte at a time it
4804 would take for ever. The timeout is now adjusted when this happens. It
4805 doesn't have to be particularly precise.
4807 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4808 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4810 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4811 without a domain if the check to <postmaster@domain> fails.
4813 SC/01 Eximstats: added -xls and the ability to specify output files
4814 (patch written by Frank Heydlauf).
4816 SC/02 Eximstats: use FileHandles for outputting results.
4818 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4820 SC/04 Eximstats: fixed display of large numbers with -nvr option
4822 SC/05 Eximstats: fixed merging of reports with empty tables.
4824 SC/06 Eximstats: added the -include_original_destination flag
4826 SC/07 Eximstats: removed tabs and trailing whitespace.
4828 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4830 TK/06 MBOX spool code: Add real "From " MBOX separator line
4831 so the .eml file is really in mbox format (even though
4832 most programs do not really care). Patch from Alex Miller.
4834 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4835 The latter is generated from $received_to and is only set if the
4836 message has one envelope recipient. SA can use these headers,
4837 obviously out-of-the-box. Patch from Alex Miller.
4839 PH/08 The ${def test on a variable was returning false if the variable's
4840 value was "0", contrary to what the specification has always said!
4841 The result should be true unless the variable is empty.
4843 PH/09 The syntax error of a character other than { following "${if
4844 def:variable_name" (after optional whitespace) was not being diagnosed.
4845 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4846 accidental colon was present, for example, could give incorrect results.
4848 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4849 result is used (not including appendfile, where other changes are about
4852 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4853 This involved changing a lot of size variables from int to off_t. It
4854 should work with maildirs and everything.
4856 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4857 spamd dying while we are connected to it.
4859 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4862 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4863 function for BDB 4.3.
4865 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4867 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4868 the sources for PCRE have been re-arranged and more files are now
4871 PH/15 The code I had for printing potentially long long variables in PH/11
4872 above was not the best (it lost precision). The length of off_t variables
4873 is now inspected at build time, and an appropriate printing format (%ld
4874 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4875 to be "long long int" or "long int". This is needed for the internal
4876 formatting function string_vformat().
4878 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4879 the configuration file to be ":syslog", then the script "guesses" where
4880 the logs files are, rather than using the compiled in default. In our
4881 case the guess is not the same as the compiled default, so the script
4882 suddenly stopped working when I started to use syslog. The patch checks
4883 to see if log_file_path is "". If so, it attempts to read it from exim
4884 with no configuration file to get the compiled in version, before it
4885 falls back to the previous guessing code."
4887 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4888 implementing BATV in an Exim configuration. See NewStuff for the gory
4891 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4892 Makefile that are specific to HP-UX.
4894 PH/18 If the "use_postmaster" option was set for a recipient callout together
4895 with the "random" option, the postmaster address was used as the MAIL
4896 FROM address for the random test, but not for the subsequent recipient
4897 test. It is now used for both.
4899 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4900 patch removes a few documentation additions to RFC 3028, because the
4901 latest draft now contains them. It adds the new en;ascii-case comparator
4902 and a new error check for 8bit text in MIME parts. Comparator and
4903 require names are now matched exactly. I enabled the subaddress
4904 extension, but it is not well tested yet (read: it works for me)."
4906 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4907 rework some of the code of TK/09 above to avoid the hardwired use of
4908 "%lld" and "long long". Replaced the call to snprintf() with a call to
4911 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4912 records point to non-existent hosts", "retry timeout exceeded", and
4913 "retry time not reached for any host after a long failure period".
4915 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4916 experimental DomainKeys support:
4918 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4919 (2) On an error such as an illegally used "control", the wrong name for
4920 the control was given.
4922 These problems did NOT occur unless DomainKeys support was compiled.
4924 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4926 PH/24 Added ${if match_ip condition.
4928 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4929 databases so that it will be absolutely obvious if a crash occurs in the
4930 DB library. This is a regular occurrence (often caused by mis-matched
4933 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4934 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4935 picked up on a user's system that detects such things. There doesn't seem
4936 to be a gcc warning option for this - only an attribute that has to be
4937 put on the function's prototype. It seems that in Fedora Core 4 they have
4938 set this on a number of new functions. No doubt there will be more in due
4941 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4942 specify a transport. However, if an address that was verified by such a
4943 router was the subject of a callout, Exim crashed because it tried to
4944 read the rcpt_include_affixes from the non-existent transport. Now it
4945 just assumes that the setting of that option is false. This bug was
4946 introduced by 4.51/PH/31.
4948 PH/28 Changed -d+all to exclude +memory, because that information is very
4949 rarely of interest, but it makes the output a lot bigger. People tend to
4950 do -d+all out of habit.
4952 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4953 code in os-type was giving problems when libc.so lives in lib64, like on
4956 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4957 aren't the modern standard, and it seems that some systems' include files
4958 don't always have them. Exim was already checking for some of the newer
4959 ones like T_AAAA, and defining it itself. I've added checks for all the
4960 record types that Exim uses.
4962 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4963 not automatically generating a new one, as it is supposed to. This
4964 prevented TLS from working. If the file did exist, but contained invalid
4965 data, a new version was generated, as expected. It was only the case of a
4966 non-existent file that was broken.
4968 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4969 with a change in libdomainkeys > 0.64.
4971 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4972 from DNS. If the selector record carries the flag, it now has
4973 precedence over the domain-wide flag.
4975 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4977 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4978 the use of a port name that isn't defined in /etc/services) occurred, the
4979 message was deferred as in a normal delivery, and thus remained on the
4980 spool, instead of being failed because of the mua_wrapper setting. This
4981 is now fixed, and I tidied up some of the mua_wrapper messages at the
4984 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4985 the messages in a hash of arrays rather than using individual hashes.
4986 This is a bit cleaner and results in dramatic memory savings, albeit
4987 at a slight CPU cost.
4989 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4990 as requested by Marc Sherman.
4992 SC/10 Eximstats: added histograms for user specified patterns as requested
4995 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4997 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4998 fopen() in the content-scanning modules that did not already have it.
5004 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5005 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5007 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5009 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5011 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5012 to test Sieve filters that use "vacation".
5014 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5015 that changes the way the GnuTLS parameters are stored in the cache file.
5016 The new format can be generated externally. For backward compatibility,
5017 if the data in the cache doesn't make sense, Exim assumes it has read an
5018 old-format file, and it generates new data and writes a new file. This
5019 means that you can't go back to an older release without removing the
5022 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5023 work if there are any delivery delays because "one_time" forces the
5024 parent to be marked "delivered", so its unseen clone is never tried
5025 again. For this reason, Exim now forbids the simultaneous setting of
5028 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5029 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5030 entry if you want to know the details. The fix, however, neglected to
5031 consider the case when local delivery batching is involved. The test for
5032 "previously delivered" was not happening when checking to see if an
5033 address could be batched with a previous (undelivered) one; under
5034 certain circumstances this could lead to multiple deliveries to the same
5037 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5038 in its include files, and this causes problems building Exim.
5040 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5041 header_syntax) but Exim was just ignoring anything given after a slash.
5042 In particular, this caused confusion with an attempt to use "verify =
5043 reverse_host_lookup/defer_ok". An error is now given when options are
5044 supplied for verify items that do not have them. (Maybe reverse_host_
5045 lookup should have a defer_ok option, but that's a different point.)
5047 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5048 defined by RFC 821) to 2048, because there were problems with some AUTH
5049 commands, and RFC 1869 says the size should be increased for extended
5050 SMTP commands that take arguments.
5052 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5055 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5056 "unknown" error; now it says that the functionality isn't in the binary.
5058 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5059 an address' error message when a string expansion fails (syntax or
5060 whatever). Otherwise the password may appear in the log. Following change
5061 PH/42 below, there is no longer a chance of it appearing in a bounce
5064 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5066 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5067 of its addresses was ever tried. (Bugzilla bug #2.)
5069 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5070 the result incorrectly in the debug output. (It correctly added a newline
5071 to what was transported.)
5073 TF/01 Added $received_time.
5075 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5076 commented out examples of how to interface to a virus scanner and to
5077 SpamAssassin. Also added commented examples of av_scanner and
5078 spamd_address settings.
5080 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5081 and controls are allowed in which ACLs. There were a couple of minor
5082 errors. Some of the entries in the conditions table (which is a table of
5083 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5084 negation of where the condition IS allowed.
5086 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5088 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5089 header file does not have a version number, so I've had to invent a new
5090 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5091 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5092 radiusclient), but it was contributed by a Radius user.
5094 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5095 files or directories, for OpenSSL.
5097 PH/19 When an Exim process that is running as root has to create an Exim log
5098 file, it does so in a subprocess that runs as exim:exim so as to get the
5099 ownership right at creation (otherwise, other Exim processes might see
5100 the file with the wrong ownership). There was no test for failure of this
5101 fork() call, which would lead to the process getting stuck as it waited
5102 for a non-existent subprocess. Forks do occasionally fail when resources
5103 run out. I reviewed all the other calls to fork(); they all seem to check
5106 PH/20 When checking for unexpected SMTP input at connect time (before writing
5107 the banner), Exim was not dealing correctly with a non-positive return
5108 from the read() function. If the client had disconnected by this time,
5109 the result was a log entry for a synchronization error with an empty
5110 string after "input=" when read() returned zero. If read() returned -1
5111 (an event I could not check), uninitialized data bytes were printed.
5112 There were reports of junk text (parts of files, etc) appearing after
5115 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5117 PH/22 Added support for macro redefinition, and (re)definition in between
5118 driver and ACL definitions.
5120 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5121 forgetting to use the resulting value; it was using the unexpanded value.
5123 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5124 hadn't been configured. The fix is from Juergen Kreileder, who
5125 understands it better than I do:
5127 "Here's what I see happening with three configured cyrus_sasl
5128 authenticators configured (plain, login, cram-md5):
5130 On startup auth_cyrus_sasl_init() gets called for each of these.
5131 This means three calls to sasl_listmech() without a specified mech_list.
5132 => SASL tests which mechs of all available mechs actually work
5133 => three warnings about OTP not working
5134 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5136 With the patch, sasl_listmech() also gets called three times. But now
5137 SASL's mech_list option is set to the server_mech specified in the the
5138 authenticator. Or in other words, the answer from sasl_listmech()
5139 gets limited to just the mech you're testing for (which is different
5141 => the return list contains just 'plain' or 'login', 'cram-md5' or
5142 nothing depending on the value of ob->server_mech.
5144 I've just tested the patch: Authentication still works fine,
5145 unavailable mechs specified in the exim configuration are still
5146 caught, and the auth.log warnings about OTP are gone."
5148 PH/25 When debugging is enabled, the contents of the command line are added
5149 to the debugging output, even when log_selector=+arguments is not
5152 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5153 answer is "GNU", and only if the return is "GNU/something" is the answer
5156 PH/27 $acl_verify_message is now set immediately after the failure of a
5157 verification in an ACL, and so is available in subsequent modifiers. In
5158 particular, the message can be preserved by coding like this:
5160 warn !verify = sender
5161 set acl_m0 = $acl_verify_message
5163 Previously, $acl_verify_message was set only while expanding "message"
5164 and "log_message" when a very denied access.
5166 PH/28 Modified OS/os.c-Linux with
5168 -#ifndef OS_LOAD_AVERAGE
5169 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5171 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5172 nomenclature these days.)
5174 PH/29 Installed patch from the Sieve maintainer that adds the options
5175 sieve_useraddress and sieve_subaddress to the redirect router.
5177 PH/30 In these circumstances:
5178 . Two addresses routed to the same list of hosts;
5179 . First host does not offer TLS;
5180 . First host accepts first address;
5181 . First host gives temporary error to second address;
5182 . Second host offers TLS and a TLS session is established;
5183 . Second host accepts second address.
5184 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5185 and peerdn, if requested) that were in fact used only for the second
5188 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5189 attention to any local part prefix or suffix that was matched by the
5190 router that accepted the address. It now behaves in the same way as it
5191 does for delivery: the affixes are removed from the local part unless
5192 rcpt_include_affixes is set on the transport.
5194 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5195 timeout during the DATA phase of an incoming message.
5197 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5198 applied a patch sanctioned by the Sieve maintainer.
5200 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5201 the uid or gid is negative. A case of a negative gid caused this to be
5202 noticed. The fix allows for either to be negative.
5204 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5205 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5208 PH/36 The change PH/12 above was broken. Fixed it.
5210 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5211 the grounds that routing the same address twice would always produce the
5212 same answer. This might have been true once, but it is certainly no
5213 longer true now. Routing a child address may depend on the previous
5214 routing that produced that child. Some complicated redirection strategies
5215 went wrong when messages had multiple recipients, and made Exim's
5216 behaviour dependent on the order in which the addresses were given.
5218 I have moved the duplicate checking until after the routing is complete.
5219 Exim scans the addresses that are assigned to local and remote
5220 transports, and removes any duplicates. This means that more work will be
5221 done, as duplicates will always all be routed, but duplicates are
5222 presumably rare, so I don't expect this is of any significance.
5224 For deliveries to pipes, files, and autoreplies, the duplicate checking
5225 still happens during the routing process, since they are not going to be
5228 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5229 It corrects a timeout issue with spamd. This is Ian's comment: "The
5230 background is that sometimes spamd either never reads data from a
5231 connection it has accepted, or it never writes response data. The exiscan
5232 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5233 blindly assumes that writes won't block so it may never time out."
5235 PH/39 Allow G after quota size as well as K and M.
5237 PH/40 The value set for $authenticated_id in an authenticator may not contain
5238 binary zeroes or newlines because the value is written to log lines and
5239 to spool files. There was no check on this. Now the value is run through
5240 the string_printing() function so that such characters are converted to
5241 printable escape sequences.
5243 PH/41 $message_linecount is a new variable that contains the total number of
5244 lines in the message. Compare $body_linecount, which is the count for the
5247 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5248 bounce and delay warning messages, except in certain special cases, which
5251 (a) An SMTP error message from a remote host;
5252 (b) A message specified in a :fail: redirection;
5253 (c) A message specified in a "fail" command in a system filter;
5254 (d) A message specified in a FAIL return from the queryprogram router;
5255 (e) A message specified by the cannot_route_message router option.
5257 In these cases only, Exim does include the error details in bounce and
5258 warning messages. There are also a few cases where bland messages such
5259 as "unrouteable address" or "local delivery error" are given.
5261 PH/43 $value is now also set for the "else" part of a ${run expansion.
5263 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5264 being worked on, but at least Exim now implements the latest version to
5267 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5268 process to complete was treated as a delivery failure, a timeout while
5269 writing the message to the pipe was logged, but erroneously treated as a
5270 successful delivery. Such timeouts include transport filter timeouts. For
5271 consistency with the overall process timeout, these timeouts are now
5272 treated as errors, giving rise to delivery failures by default. However,
5273 there is now a new Boolean option for the pipe transport called
5274 timeout_defer, which, if set TRUE, converts the failures into defers for
5275 both kinds of timeout. A transport filter timeout is now identified in
5278 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5279 systems where "make" and "gmake" are different, calling "gmake" at top
5280 level broke things. I've arranged for the value of $(MAKE) to be passed
5281 from the Makefile to this script so that it can call the same version of
5285 A note about Exim versions 4.44 and 4.50
5286 ----------------------------------------
5288 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5289 changes of various kinds. As a consequence, a big documentation update was
5290 needed. This delayed the release for rather longer than seemed good, especially
5291 in the light of a couple of (minor) security issues. Therefore, the changes
5292 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5293 release. So 4.44 and 4.50 are in effect two different branches that both start
5296 I have left the 4.50 change log unchanged; it contains all the changes since
5297 4.43. The change log for 4.44 is below; many of its items are identical to
5298 those for 4.50. This seems to be the most sensible way to preserve the
5299 historical information.
5305 1. Minor wording change to the doc/README.SIEVE file.
5307 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5308 computation of the current number of files was incorrect.
5310 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5311 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5314 4. Give more explanation in the error message when the command for a transport
5315 filter fails to execute.
5317 5. There are several places where Exim runs a non-Exim command in a
5318 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5319 was being done only for the command run by the queryprogram router. It is
5320 now done for all such subprocesses. The other cases are: ${run, transport
5321 filters, and the commands run by the lmtp and pipe transports.
5323 6. Added CONFIGURE_GROUP build-time option.
5325 7. Some older OS have a limit of 256 on the maximum number of file
5326 descriptors. Exim was using setrlimit() to set 1000 as a large value
5327 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5328 systems. I've change it so that if it can't get 1000, it tries for 256.
5330 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5331 was an oversight, and furthermore, ever since the addition of extra
5332 controls (e.g. 4.43/32), the checks on when to allow different forms of
5333 "control" were broken. There should now be diagnostics for all cases when a
5334 control that does not make sense is encountered.
5336 9. Added the /retain_sender option to "control=submission".
5338 10. $recipients is now available in the predata ACL (oversight).
5340 11. Tidy the search cache before the fork to do a delivery from a message
5341 received from the command line. Otherwise the child will trigger a lookup
5342 failure and thereby defer the delivery if it tries to use (for example) a
5343 cached ldap connection that the parent has called unbind on.
5345 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5346 of $address_data from the recipient verification was clobbered by the
5347 sender verification.
5349 13. The value of address_data from a sender verification is now available in
5350 $sender_address_data in subsequent conditions in the ACL statement.
5352 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5354 15. Added a new option "connect=<time>" to callout options, to set a different
5357 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5358 was its contents. (It was OK if the option was not defined at all.)
5360 17. A "Completed" log line is now written for messages that are removed from
5361 the spool by the -Mrm option.
5363 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5364 information about exactly what failed.
5366 19. Added -dd to debug only the daemon process.
5368 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5369 handles timeouts, both on the server side and network timeouts. Renamed the
5370 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5372 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5373 It is now set to "smtps".
5375 22. $host_address is now set to the target address during the checking of
5376 ignore_target_hosts.
5378 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5379 being passed; this would have caused $sender_host_name to have been used if
5380 matching the list had actually called for a host name (not very likely,
5381 since this list is usually IP addresses). A host name is now passed as
5384 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5385 code that specifies a non-restarting handler (typically sigaction() in
5386 modern systems) in an attempt to fix a rare and obscure crash bug.
5388 25. Narrowed the window for a race in the daemon that could cause it to ignore
5389 SIGCHLD signals. This is not a major problem, because they are used only to
5390 wake it up if nothing else does.
5392 26. A malformed maildirsize file could cause Exim to calculate negative values
5393 for the mailbox size or file count. Odd effects could occur as a result.
5394 The maildirsize information is now recalculated if the size or filecount
5397 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5398 support for a long time. Removed HAVE_SYS_VFS_H.
5400 28. Installed the latest version of exipick from John Jetmore.
5402 29. In an address list, if the pattern was not a regular expression, an empty
5403 subject address (from a bounce message) matched only if the pattern was an
5404 empty string. Non-empty patterns were not even tested. This was the wrong
5405 because it is perfectly reasonable to use an empty address as part of a
5406 database query. An empty address is now tested by patterns that are
5407 lookups. However, all the other forms of pattern expect the subject to
5408 contain a local part and a domain, and therefore, for them, an empty
5409 address still always fails if the pattern is not itself empty.
5411 30. Exim went into a mad DNS loop when attempting to do a callout where the
5412 host was specified on an smtp transport, and looking it up yielded more
5413 than one IP address.
5415 31. Re-factored the code for checking spool and log partition space into a
5416 function that finds that data and another that does the check. The former
5417 is then used to implement four new variables: $spool_space, $log_space,
5418 $spool_inodes, and $log_inodes.
5420 32. The RFC2047 encoding function was originally intended for short strings
5421 such as real names; it was not keeping to the 75-character limit for
5422 encoded words that the RFC imposes. It now respects the limit, and
5423 generates multiple encoded words if necessary. To be on the safe side, I
5424 have increased the buffer size for the ${rfc2047: expansion operator from
5427 33. It is now permitted to omit both strings after an "if" condition; if the
5428 condition is true, the result is "true". As before, when the second string
5429 is omitted, a false condition yields an empty string. This makes it less
5430 cumbersome to write custom ACL and router conditions.
5432 34. Failure to deliver a bounce message always caused it to be frozen, even if
5433 there was an errors_to setting on the router. The errors_to setting is now
5436 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5437 canonical form (fully expanded) before being placed in
5438 $sender_host_address.
5440 36. The table in the code that translates DNS record types into text (T_A to
5441 "A" for instance) was missing entries for NS and CNAME. It is just possible
5442 that this could have caused confusion if both these types were looked up
5443 for the same domain, because the text type is used as part of Exim's
5444 per-process caching. But the chance of anyone hitting this buglet seems
5447 37. The dnsdb lookup has been extended in a number of ways.
5449 (1) There is a new type, "zns", which walks up the domain tree until it
5450 finds some nameserver records. It should be used with care.
5452 (2) There is a new type, "mxh", which is like "mx" except that it returns
5453 just the host names, not the priorities.
5455 (3) It is now possible to give a list of domains (or IP addresses) to be
5456 looked up. The behaviour when one of the lookups defers can be
5457 controlled by a keyword.
5459 (4) It is now possible to specify the separator character for use when
5460 multiple records are returned.
5462 38. The dnslists ACL condition has been extended: it is now possible to supply
5463 a list of IP addresses and/or domains to be looked up in a particular DNS
5466 39. Added log_selector=+queue_time_overall.
5468 40. When running the queue in the test harness, wait just a tad after forking a
5469 delivery process, to get repeatability of debugging output.
5471 41. Include certificate and key file names in error message when GnuTLS fails
5472 to set them up, because the GnuTLS error message doesn't include the name
5473 of the failing file when there is a problem reading it.
5475 42. Allow both -bf and -bF in the same test run.
5477 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5479 44. The "Exiscan patch" is now merged into the mainline Exim source.
5481 45. Sometimes the final signoff response after QUIT could fail to get
5482 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5483 < 0 before doing a fflush(). This bug looks as though it goes back to the
5484 introduction of TLS in release 3.20, but "sometimes" must have been rare
5485 because the tests only now provoked it.
5487 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5488 (this can affect the format of dates).
5490 47. exim_tidydb, when checking for the continued existence of a message for
5491 which it has found a message-specific retry record, was not finding
5492 messages that were in split spool directories. Consequently, it was
5493 deleting retry records that should have stayed in existence.
5495 48. Steve fixed some bugs in eximstats.
5497 49. The SPA authentication driver was not abandoning authentication and moving
5498 on to the next authenticator when an expansion was forced to fail,
5499 contradicting the general specification for all authenticators. Instead it
5500 was generating a temporary error. It now behaves as specified.
5502 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5503 (the order specifies the preference for clients). The order is now AES256,
5504 AES128, 3DES, ARCFOUR128.
5506 51. Small patch to Sieve code - explicitly set From: when generating an
5509 52. Exim crashed if a remote delivery caused a very long error message to be
5510 recorded - for instance if somebody sent an entire SpamAssassin report back
5511 as a large number of 550 error lines. This bug was coincidentally fixed by
5512 increasing the size of one of Exim's internal buffers (big_buffer) that
5513 happened as part of the Exiscan merge. However, to be on the safe side, I
5514 have made the code more robust (and fixed the comments that describe what
5517 53. Now that there can be additional text after "Completed" in log lines (if
5518 the queue_time_overall log selector is set), a one-byte patch to exigrep
5519 was needed to allow it to recognize "Completed" as not the last thing in
5522 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5523 patch that reportedly fixes this has been added. I am not expert enough to
5524 create a test for it. This is what the patch creator wrote:
5526 "I found a little strange behaviour of ldap code when working with
5527 Windows 2003 AD Domain, where users was placed in more than one
5528 Organization Units. When I tried to give exim partial DN, the exit code
5529 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5530 But simultaneously result of request was absolutely normal ldap result,
5531 so I produce this patch..."
5533 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5534 REFERENCE, so I have modified the code to exclude the patch when that macro
5537 55. Some experimental protocols are using DNS PTR records for new purposes. The
5538 keys for these records are domain names, not reversed IP addresses. The
5539 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5540 leaves it alone. Component reversal etc. now happens only for IP addresses.
5543 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5545 57. Double the size of the debug message buffer (to 2048) so that more of very
5546 long debug lines gets shown.
5548 58. The exicyclog utility now does better if the number of log files to keep
5549 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5551 59. Two changes related to the smtp_active_hostname option:
5553 (1) $smtp_active_hostname is now available as a variable.
5554 (2) The default for smtp_banner uses $smtp_active_hostname instead
5555 of $primary_hostname.
5557 60. The host_aton() function is supposed to be passed a string that is known
5558 to be a valid IP address. However, in the case of IPv6 addresses, it was
5559 not checking this. This is a hostage to fortune. Exim now panics and dies
5560 if the condition is not met. A case was found where this could be provoked
5561 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5562 components; fortuitously, this particular loophole had already been fixed
5563 by change 4.50/55 above.
5565 If there are any other similar loopholes, the new check in host_aton()
5566 itself should stop them being exploited. The report I received stated that
5567 data on the command line could provoke the exploit when Exim was running as
5568 exim, but did not say which command line option was involved. All I could
5569 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5570 running as the user.
5573 61. There was a buffer overflow vulnerability in the SPA authentication code
5574 (which came originally from the Samba project). I have added a test to the
5575 spa_base64_to_bits() function which I hope fixes it.
5578 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5579 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5581 63. The daemon start-up calls getloadavg() while still root for those OS that
5582 need the first call to be done as root, but it missed one case: when
5583 deliver_queue_load_max is set with deliver_drop_privilege. This is
5584 necessary for the benefit of the queue runner, because there is no re-exec
5585 when deliver_drop_privilege is set.
5587 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5588 This has been fixed.
5590 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5591 was in use, was not putting the data itself into the right store pool;
5592 consequently, it could be overwritten for a subsequent message in the same
5593 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5596 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5598 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5599 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5600 the function were treating the return as a boolean value, which happened to
5601 work because 0=false and not-0=true, but is not correct code.
5603 68. The host_aton() function was not handling scoped IPv6 addresses (those
5604 with, for example, "%eth0" on the end) correctly.
5606 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5607 negated items (that is, ~something) in unsigned ints. Some compilers
5608 apparently mutter when there is no cast.
5610 70. If an address verification called from an ACL failed, and did not produce a
5611 user-specific message (i.e. there was only a "system" message), nothing was
5612 put in $acl_verify_message. In this situation, it now puts the system
5615 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5616 session; change 4.31/43 added the unwanted input to the log line - except
5617 that it did not do this in the start of session case. It now does.
5619 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5620 This is wrong and can cause the other end to generate a synchronization
5621 error if it is another Exim or anything else that does the synchronization
5622 check. A QUIT command is no longer sent after a timeout.
5624 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5625 during host lookups.
5627 74. The defer_ok option of callout verification was not working if it was used
5628 when verifying addresses in header lines, that is, for this case:
5630 verify = header_sender/callout=defer_ok
5632 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5633 those file descriptors could be used for SMTP connections. If anything
5634 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5635 could be sent to the SMTP client, causing chaos. The daemon now opens
5636 stdin, stdout, and stderr to /dev/null when it puts itself into the
5639 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5640 log by default. The user can override this with suitable Perl magic.
5642 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5643 the log message when discard triggers, was not working for the DATA ACL or
5644 for the non-SMTP ACL.
5646 78. Error message wording change in sieve.c.
5648 79. If smtp_accept_max_per_host was set, the number of connections could be
5649 restricted to fewer than expected, because the daemon was trying to set up
5650 a new connection before checking whether the processes handling previous
5651 connections had finished. The check for completed processes is now done
5652 earlier. On busy systems, this bug wouldn't be noticed because something
5653 else would have woken the daemon, and it would have reaped the completed
5656 80. If a message was submitted locally by a user whose login name contained one
5657 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5658 It caused a spool format error. I have fixed the spool reading code. A
5659 related problem was that the "from" clause in the Received: line became
5660 illegal because of the space(s). It is now covered by ${quote_local_part.
5662 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5665 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5667 83. Patch from Sieve maintainer to fix unterminated string problem in
5668 "vacation" handling.
5670 84. Some minor changes to the Linux configuration files to help with other
5671 OS variants using glibc.
5673 85. One more patch for Sieve to update vacation handling to latest spec.
5676 ----------------------------------------------------
5677 See the note above about the 4.44 and 4.50 releases.
5678 ----------------------------------------------------
5684 1. Change 4.43/35 introduced a bug that caused file counts to be
5685 incorrectly computed when quota_filecount was set in an appendfile
5688 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5689 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5692 3. Give more explanation in the error message when the command for a transport
5693 filter fails to execute.
5695 4. There are several places where Exim runs a non-Exim command in a
5696 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5697 was being done only for the command run by the queryprogram router. It is
5698 now done for all such subprocesses. The other cases are: ${run, transport
5699 filters, and the commands run by the lmtp and pipe transports.
5701 5. Some older OS have a limit of 256 on the maximum number of file
5702 descriptors. Exim was using setrlimit() to set 1000 as a large value
5703 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5704 systems. I've change it so that if it can't get 1000, it tries for 256.
5706 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5707 was an oversight, and furthermore, ever since the addition of extra
5708 controls (e.g. 4.43/32), the checks on when to allow different forms of
5709 "control" were broken. There should now be diagnostics for all cases when a
5710 control that does not make sense is encountered.
5712 7. $recipients is now available in the predata ACL (oversight).
5714 8. Tidy the search cache before the fork to do a delivery from a message
5715 received from the command line. Otherwise the child will trigger a lookup
5716 failure and thereby defer the delivery if it tries to use (for example) a
5717 cached ldap connection that the parent has called unbind on.
5719 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5720 of $address_data from the recipient verification was clobbered by the
5721 sender verification.
5723 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5724 was its contents. (It was OK if the option was not defined at all.)
5726 11. A "Completed" log line is now written for messages that are removed from
5727 the spool by the -Mrm option.
5729 12. $host_address is now set to the target address during the checking of
5730 ignore_target_hosts.
5732 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5733 being passed; this would have caused $sender_host_name to have been used if
5734 matching the list had actually called for a host name (not very likely,
5735 since this list is usually IP addresses). A host name is now passed as
5738 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5739 code that specifies a non-restarting handler (typically sigaction() in
5740 modern systems) in an attempt to fix a rare and obscure crash bug.
5742 15. Narrowed the window for a race in the daemon that could cause it to ignore
5743 SIGCHLD signals. This is not a major problem, because they are used only to
5744 wake it up if nothing else does.
5746 16. A malformed maildirsize file could cause Exim to calculate negative values
5747 for the mailbox size or file count. Odd effects could occur as a result.
5748 The maildirsize information is now recalculated if the size or filecount
5751 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5752 support for a long time. Removed HAVE_SYS_VFS_H.
5754 18. Updated exipick to current release from John Jetmore.
5756 19. Allow an empty sender to be matched against a lookup in an address list.
5757 Previously the only cases considered were a regular expression, or an
5760 20. Exim went into a mad DNS lookup loop when doing a callout where the
5761 host was specified on the transport, if the DNS lookup yielded more than
5764 21. The RFC2047 encoding function was originally intended for short strings
5765 such as real names; it was not keeping to the 75-character limit for
5766 encoded words that the RFC imposes. It now respects the limit, and
5767 generates multiple encoded words if necessary. To be on the safe side, I
5768 have increased the buffer size for the ${rfc2047: expansion operator from
5771 22. Failure to deliver a bounce message always caused it to be frozen, even if
5772 there was an errors_to setting on the router. The errors_to setting is now
5775 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5776 canonical form (fully expanded) before being placed in
5777 $sender_host_address.
5779 24. Updated eximstats to version 1.33
5781 25. Include certificate and key file names in error message when GnuTLS fails
5782 to set them up, because the GnuTLS error message doesn't include the name
5783 of the failing file when there is a problem reading it.
5785 26. Expand error message when OpenSSL has problems setting up cert/key files.
5788 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5789 (this can affect the format of dates).
5791 28. exim_tidydb, when checking for the continued existence of a message for
5792 which it has found a message-specific retry record, was not finding
5793 messages that were in split spool directories. Consequently, it was
5794 deleting retry records that should have stayed in existence.
5796 29. eximstats updated to version 1.35
5797 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5798 1.35 - bugfix such that pie charts by volume are generated correctly
5800 30. The SPA authentication driver was not abandoning authentication and moving
5801 on to the next authenticator when an expansion was forced to fail,
5802 contradicting the general specification for all authenticators. Instead it
5803 was generating a temporary error. It now behaves as specified.
5805 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5806 (the order specifies the preference for clients). The order is now AES256,
5807 AES128, 3DES, ARCFOUR128.
5809 31. Small patch to Sieve code - explicitly set From: when generating an
5812 32. Exim crashed if a remote delivery caused a very long error message to be
5813 recorded - for instance if somebody sent an entire SpamAssassin report back
5814 as a large number of 550 error lines. This bug was coincidentally fixed by
5815 increasing the size of one of Exim's internal buffers (big_buffer) that
5816 happened as part of the Exiscan merge. However, to be on the safe side, I
5817 have made the code more robust (and fixed the comments that describe what
5820 33. Some experimental protocols are using DNS PTR records for new purposes. The
5821 keys for these records are domain names, not reversed IP addresses. The
5822 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5823 leaves it alone. Component reversal etc. now happens only for IP addresses.
5826 34. The host_aton() function is supposed to be passed a string that is known
5827 to be a valid IP address. However, in the case of IPv6 addresses, it was
5828 not checking this. This is a hostage to fortune. Exim now panics and dies
5829 if the condition is not met. A case was found where this could be provoked
5830 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5831 components; fortuitously, this particular loophole had already been fixed
5832 by change 4.50/55 or 4.44/33 above.
5834 If there are any other similar loopholes, the new check in host_aton()
5835 itself should stop them being exploited. The report I received stated that
5836 data on the command line could provoke the exploit when Exim was running as
5837 exim, but did not say which command line option was involved. All I could
5838 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5839 running as the user.
5842 35. There was a buffer overflow vulnerability in the SPA authentication code
5843 (which came originally from the Samba project). I have added a test to the
5844 spa_base64_to_bits() function which I hope fixes it.
5847 36. The daemon start-up calls getloadavg() while still root for those OS that
5848 need the first call to be done as root, but it missed one case: when
5849 deliver_queue_load_max is set with deliver_drop_privilege. This is
5850 necessary for the benefit of the queue runner, because there is no re-exec
5851 when deliver_drop_privilege is set.
5853 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5854 was in use, was not putting the data itself into the right store pool;
5855 consequently, it could be overwritten for a subsequent message in the same
5856 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5859 38. Sometimes the final signoff response after QUIT could fail to get
5860 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5861 < 0 before doing a fflush(). This bug looks as though it goes back to the
5862 introduction of TLS in release 3.20, but "sometimes" must have been rare
5863 because the tests only now provoked it.
5869 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5870 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5871 Now it returns an integer. A number of calls were still expecting a T/F
5872 return. Fortuitously, in all cases, the tests worked in OK situations,
5873 which is the norm. However, things would have gone wrong on any write
5874 failures on the smtp file descriptor. This function is used when sending
5875 messages over SMTP and also when doing verify callouts.
5877 2. When Exim is called to do synchronous delivery of a locally submitted
5878 message (the -odf or -odi options), it no longer closes stderr before doing
5881 3. Implemented the mua_wrapper option.
5883 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5885 5. Implemented the functions header_remove(), header_testname(),
5886 header_add_at_position(), and receive_remove_recipient(), and exported them
5889 6. If an ACL "warn" statement specified the addition of headers, Exim already
5890 inserted X-ACL-Warn: at the start if there was no header name. However, it
5891 was not making this test for the second and subsequent header lines if
5892 there were newlines in the string. This meant that an invalid header could
5893 be inserted if Exim was badly configured.
5895 7. Allow an ACL "warn" statement to add header lines at the start or after all
5896 the Received: headers, as well as at the end.
5898 8. Added the rcpt_4xx retry error code.
5900 9. Added postmaster_mailfrom=xxx to callout verification option.
5902 10. Added mailfrom=xxxx to the callout verification option, for verify=
5905 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5906 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5908 12. Inserted some casts to stop certain compilers warning when using pointer
5909 differences as field lengths or precisions in printf-type calls (mostly
5910 affecting debugging statements).
5912 13. Added optional readline() support for -be (dynamically loaded).
5914 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5915 same clock tick as a message's arrival, so that its received time was the
5916 same as the "first fail" time on the retry record, and that message
5917 remained on the queue past the ultimate address timeout, every queue runner
5918 would try a delivery (because it was past the ultimate address timeout) but
5919 after another failure, the ultimate address timeout, which should have then
5920 bounced the address, did not kick in. This was a "< instead of <=" error;
5921 in most cases the first failure would have been in the next clock tick
5922 after the received time, and all would be well.
5924 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5925 being recognized when the domain list was tested by the match_domain
5926 condition in an expansion string.
5928 16. Added the ${str2b64: operator.
5930 17. Exim was always calling setrlimit() to set a large limit for the number of
5931 processes, without checking whether the existing limit was already
5932 adequate. (It did check for the limit on file descriptors.) Furthermore,
5933 errors from getrlimit() and setrlimit() were being ignored. Now they are
5934 logged to the main and panic logs, but Exim does carry on, to try to do its
5935 job under whatever limits there are.
5937 18. Imported PCRE 5.0.
5939 19. Trivial typo in log message " temporarily refused connection" (the leading
5942 20. If the log selector return_path_on_delivery was set and an address was
5943 redirected to /dev/null, the delivery process crashed because it assumed
5944 that a return path would always be set for a "successful" delivery. In this
5945 case, the whole delivery is bypassed as an optimization, and therefore no
5948 21. Internal re-arrangement: the function for sending a challenge and reading
5949 a response while authentication was assuming a zero-terminated challenge
5950 string. It's now changed to take a pointer and a length, to allow for
5951 binary data in such strings.
5953 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5955 23. Exim was not respecting finduser_retries when seeking the login of the
5956 uid under which it was called; it was always trying 10 times. (The default
5957 setting of finduser_retries is zero.) Also, it was sleeping after the final
5958 failure, which is pointless.
5960 24. Implemented tls_on_connect_ports.
5962 25. Implemented acl_smtp_predata.
5964 26. If the domain in control=submission is set empty, Exim assumes that the
5965 authenticated id is a complete email address when it generates From: or
5966 Sender: header lines.
5968 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5969 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5970 chown and chgrp in /bin and hostname in /usr/bin.
5972 28. Exim was keeping the "process log" file open after each use, just as it
5973 does for the main log. This opens the possibility of it remaining open for
5974 long periods when the USR1 signal hits a daemon. Occasional processlog
5975 errors were reported, that could have been caused by this. Anyway, it seems
5976 much more sensible not to leave this file open at all, so that is what now
5979 29. The long-running daemon process does not normally write to the log once it
5980 has entered its main loop, and it closes the log before doing so. This is
5981 so that log files can straightforwardly be renamed and moved. However,
5982 there are a couple of unusual error situations where the daemon does write
5983 log entries, and I had neglected to close the log afterwards.
5985 30. The text of an SMTP error response that was received during a remote
5986 delivery was being truncated at 512 bytes. This is too short for some of
5987 the long messages that one sometimes sees. I've increased the limit to
5990 31. It is now possible to make retry rules that apply only when a message has a
5991 specific sender, in particular, an empty sender.
5993 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5994 it possible to be selective about when SMTP synchronization is enforced.
5996 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5998 32. Implemented hosts_connection_nolog.
6000 33. Added an ACL for QUIT.
6002 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6005 35. Added mailbox_size and mailbox_filecount to appendfile.
6007 36. Added control = no_multiline_responses to ACLs.
6009 37. There was a bug in the logic of the code that waits for the clock to tick
6010 in the case where the clock went backwards by a substantial amount such
6011 that the microsecond fraction of "now" was more than the microsecond
6012 fraction of "then" (but the whole seconds number was less).
6014 38. Added support for the libradius Radius client library this is found on
6015 FreeBSD (previously only the radiusclient library was supported).
6021 1. When certain lookups returned multiple values in the form name=value, the
6022 quoting of the values was not always being done properly. Specifically:
6023 (a) If the value started with a double quote, but contained no whitespace,
6025 (b) If the value contained whitespace other than a space character (i.e.
6026 tabs or newlines or carriage returns) it was not quoted.
6027 This fix has been applied to the mysql and pgsql lookups by writing a
6028 separate quoting function and calling it from the lookup code. The fix
6029 should probably also be applied to nisplus, ibase and oracle lookups, but
6030 since I cannot test any of those, I have not disturbed their existing code.
6032 2. A hit in the callout cache for a specific address caused a log line with no
6033 reason for rejecting RCPT. Now it says "Previous (cached) callout
6034 verification failure".
6036 3. There was an off-by-one bug in the queryprogram router. An over-long
6037 return line was truncated at 256 instead of 255 characters, thereby
6038 overflowing its buffer with the terminating zero. As well as fixing this, I
6039 have increased the buffer size to 1024 (and made a note to document this).
6041 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6042 when Exim is waiting for an SMTP response from a remote server, Exim
6043 restarts its select() call on the socket, thereby resetting its timeout.
6044 This is not a problem when such interrupts are rare. Somebody set up a cron
6045 job to run exiwhat every 2 minutes, which is less than the normal select()
6046 timeout (5 or 10 minutes). This meant that the select() timeout never
6047 kicked in because it was always reset. I have fixed this by comparing the
6048 time when an interrupt arrives with the time at the start of the first call
6049 to select(). If more time than the timeout has elapsed, the interrupt is
6050 treated as a timeout.
6052 5. Some internal re-factoring in preparation for the addition of Sieve
6053 extensions (by MH). In particular, the "personal" test is moved to a
6054 separate function, and given an option for scanning Cc: and Bcc: (which is
6055 not set for Exim filters).
6057 6. When Exim created an email address using the login of the caller as the
6058 local part (e.g. when creating a From: or Sender: header line), it was not
6059 quoting the local part when it contained special characters such as @.
6061 7. Installed new OpenBSD configuration files.
6063 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6064 try to make them clearer.
6066 9. Callout options, other than the timeout value, were being ignored when
6067 verifying sender addresses in header lines. For example, when using
6069 verify = header_sender/callout=no_cache
6071 the cache was (incorrectly) being used.
6073 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6074 only the Cygwin environment.
6076 11. When return_path_on_delivery was set as a log selector, if different remote
6077 addresses in the same message used different return paths and parallel
6078 remote delivery occurred, the wrong values would sometimes be logged.
6079 (Whenever a remote delivery process finished, the return path value from
6080 the most recently started remote delivery process was logged.)
6082 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6083 lines when AUTH and/or TLS are in use. This is the "received protocol"
6084 field. Exim used to use "asmtp" for authenticated SMTP, without any
6085 indication (in the protocol name) for TLS use. Now it follows the RFC and
6086 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6087 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6088 names appear in log lines as well as in Received: header lines.
6090 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6091 extensions, and comparison tests, and to fix some bugs.
6093 14. Changes to the "personal" filter test:
6095 (1) The test was buggy in that it was just doing the equivalent of
6096 "contains" tests on header lines. For example, if a user's address was
6097 anne@some.where, the "personal" test would incorrectly be true for
6099 To: susanne@some.where
6101 This test is now done by extracting each address from the header in turn,
6102 and checking the entire address. Other tests that are part of "personal"
6103 are now done using regular expressions (for example, to check local parts
6104 of addresses in From: header lines).
6106 (2) The list of non-personal local parts in From: addresses has been
6107 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6108 taken from the Sieve specification recommendations.
6110 (3) If the message contains any header line starting with "List-" it is
6111 treated as non-personal.
6113 (4) The test for "circular" in the Subject: header line has been removed
6114 because it now seems ill-conceived.
6116 15. Minor typos in src/EDITME comments corrected.
6118 16. Installed latest exipick from John Jetmore.
6120 17. If headers_add on a router specified a text string that was too long for
6121 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6122 of string_sprintf() is now avoided.
6124 18. $message_body_size was not set (it was always zero) when running the DATA
6125 ACL and the local_scan() function.
6127 19. For the "mail" command in an Exim filter, no default was being set for
6128 the once_repeat time, causing a random time value to be used if "once" was
6129 specified. (If the value happened to be <= 0, no repeat happened.) The
6130 default is now 0s, meaning "never repeat". The "vacation" command was OK
6131 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6132 (I found it when inspecting the code).
6134 20. There is now an overall timeout for performing a callout verification. It
6135 defaults to 4 times the callout timeout, which applies to individual SMTP
6136 commands during the callout. The overall timeout applies when there is more
6137 than one host that can be tried. The timeout is checked before trying the
6138 next host. This prevents very long delays if there are a large number of
6139 hosts and all are timing out (e.g. when the network connections are timing
6140 out). The value of the overall timeout can be changed by specifying an
6141 additional sub-option for "callout", called "maxwait". For example:
6143 verify = sender/callout=5s,maxwait=20s
6145 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6146 to the end before writing, but this should make it even safer).
6148 22. Exim was forgetting that it had advertised PIPELINING for the second and
6149 subsequent messages on an SMTP connection. It was also not resetting its
6150 memory on STARTTLS and an internal HELO.
6152 23. When Exim logs an SMTP synchronization error within a session, it now
6153 records whether PIPELINING has been advertised or not.
6155 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6156 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6157 rather than long int.
6159 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6161 26. Added the never_mail option to autoreply.
6167 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6168 crash if the getsockname() call failed; this can happen if a connection is
6169 closed very soon after it is established. The problem was simply in the
6170 order in which certain operations were done, causing Exim to try to write
6171 to the SMTP stream before it had set up the file descriptor. The bug has
6172 been fixed by making things happen in the correct order.
6178 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6179 before the connection was closed, thus losing the rejection response.
6181 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6182 some early Solaris releases, but causes trouble in current releases where
6183 socklen_t is defined.
6185 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6188 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6191 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6192 that releasing the top of it at the end releases what was used for sub-
6193 expansions (unless the block got too big). However, discard this block if
6194 the first thing is a variable or header, so that we can use its block when
6195 it is dynamic (useful for very large $message_headers, for example).
6197 6. Lookups now cache *every* query, not just the most recent. A new, separate
6198 store pool is used for this. It can be recovered when all lookup caches are
6199 flushed. Lookups now release memory at the end of their result strings.
6200 This has involved some general refactoring of the lookup sources.
6202 7. Some code has been added to the store_xxx() functions to reduce the amount
6203 of flapping under certain conditions.
6205 8. log_incoming_interface used to affect only the <= reception log lines. Now
6206 it causes the local interface and port to be added to several more SMTP log
6207 lines, for example "SMTP connection from", and rejection lines.
6209 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6211 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6213 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6214 could be overwritten at the end of the current message (or the start of a
6215 new message if it was set in a HELO ACL). The value is now preserved for
6216 the duration of the SMTP connection.
6218 12. If a transport had a headers_rewrite setting, and a matching header line
6219 contained an unqualified address, that address was qualified, even if it
6220 did not match any rewriting rules. The underlying bug was that the values
6221 of the flags that permit the existence of unqualified sender and recipient
6222 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6223 non-local messages, and by -bnq for local messages) were not being
6224 preserved with the message after it was received.
6226 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6227 "next input=" as part of the text comprising the host identity instead of
6228 the correct text. The code was using the same buffer for two different
6229 strings. However, depending on which order the printing function evaluated
6230 its arguments, the bug did not always show up. Under Linux, for example, my
6231 test suite worked just fine.
6233 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6234 doesn't work with some older versions of Perl. It has been changed to "my",
6235 which in any case is probably the better facility to use.
6237 15. A really picky compiler found some instances of statements for creating
6238 error messages that either had too many or two few arguments for the format
6241 16. The size of the buffer for calls to the DNS resolver has been increased
6242 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6243 for addresses that have a lot of PTR records. This alleviates a problem; it
6244 does not fully solve it.
6246 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6247 buffer now truncates the list and logs the incident, which is the same
6248 action as happens when Exim is looking up a host name and its aliases.
6249 Previously in this situation something unpredictable would happen;
6250 sometimes it was "internal error: store_reset failed".
6252 18. If a server dropped the connection unexpectedly when an Exim client was
6253 using GnuTLS and trying to read a response, the client delivery process
6254 crashed while trying to generate an error log message.
6256 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6257 string, for example:
6259 warn message = H1: something\nH2: something
6261 the text was added as a single header line from Exim's point of view
6262 though it ended up OK in the delivered message. However, searching for the
6263 second and subsequent header lines using $h_h2: did not work. This has been
6264 fixed. Similarly, if a system filter added multiple headers in this way,
6265 the routers could not see them.
6267 20. Expanded the error message when iplsearch is called with an invalid key to
6268 suggest using net-iplsearch in a host list.
6270 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6271 ACLs are no longer actually imposed (and a message to that effect is
6274 22. If a "gecos" field in a passwd entry contained escaped characters, in
6275 particular, if it contained a \" sequence, Exim got it wrong when building
6276 a From: or a Sender: header from that name. A second bug also caused
6277 incorrect handling when an unquoted " was present following a character
6278 that needed quoting.
6280 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6281 was not being matched caselessly.
6283 24. Arranged for all hyphens in the exim.8 source to be escaped with
6286 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6287 properly. Recipient callout cache records were still being keyed to include
6288 the sender, even when use_sender was set false. This led to far more
6289 callouts that were necessary. The sender is no longer included in the key
6290 when use_sender is false.
6292 26. Added "control = submission" modifier to ACLs.
6294 27. Added the ${base62d: operator to decode base 62 numbers.
6296 28. dnsdb lookups can now access SRV records.
6298 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6299 the configuration file.
6301 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6302 (-v) mode. This makes the output for a verbose queue run more intelligible.
6304 31. Added a use_postmaster feature to recipient callouts.
6306 32. Added the $body_zerocount variable, containing the number of binary zero
6307 bytes in the message body.
6309 33. The time of last modification of the "new" subdirectory is now used as the
6310 "mailbox time last read" when there is a quota error for a maildir
6313 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6315 35. Added +ignore_unknown as a special item in host lists.
6317 36. Code for decoding IPv6 addresses in host lists is now included, even if
6318 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6319 address was recognized as an IP address, but was then not correctly decoded
6320 into binary, causing unexpected and incorrect effects when compared with
6327 1. Very minor rewording of debugging text in manualroute to say "list of
6328 hosts" instead of "hostlist".
6330 2. If verify=header_syntax was set, and a header line with an unqualified
6331 address (no domain) and a large number of spaces between the end of the
6332 name and the colon was received, the reception process suffered a buffer
6333 overflow, and (when I tested it) crashed. This was caused by some obsolete
6334 code that should have been removed. The fix is to remove it!
6336 3. When running in the test harness, delay a bit after writing a bounce
6337 message to get a bit more predictability in the log output.
6339 4. Added a call to search_tidyup() just before forking a reception process. In
6340 theory, someone could use a lookup in the expansion of smtp_accept_max_
6341 per_host which, without the tidyup, could leave open a database connection.
6343 5. Added the variables $recipient_data and $sender_data which get set from a
6344 lookup success in an ACL "recipients" or "senders" condition, or a router
6345 "senders" option, similar to $domain_data and $local_part_data.
6347 6. Moved the writing of debug_print from before to after the "senders" test
6350 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6351 problems for message scanning, either using a data ACL, or using
6352 local_scan() because the Received: header was not generated till after they
6353 were called (in order to set the time as the time of reception completion).
6354 I have revised the way this works. The header is now generated after the
6355 body is received, but before the ACL or local_scan() are called. After they
6356 are run, the timestamp in the header is updated.
6362 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6363 before starting a queue runner without re-exec. This happened only when
6364 deliver_drop_privilege was set or when the Exim user was set to root. The
6365 effect of the bug was that timeouts during subsequent deliveries caused
6366 crashes instead of being properly handled. The handler is now left at its
6367 default (and expected) setting.
6369 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6370 message, again when deliver_drop_privilege is set or Exim is run as root.
6371 The bug described in (1) was not present in this case, but the tidying up
6372 of the other signals was missing. I have made the two cases consistent.
6374 3. The ignore_target_hosts setting on a manualroute router was being ignored
6375 for hosts that were looked up using the /MX notation.
6377 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6380 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6381 operated on the sender address. After changing the $sender_address to <>
6382 for the sender address verify, Exim was re-instated it as the original
6383 (before rewriting) address, but remembering that it had rewritten it, so it
6384 wasn't rewriting it again. This bug also had the effect of breaking the
6385 sender address verification caching when the sender address was rewritten.
6387 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6388 This has been changed so that if the ip literal address matches
6389 ignore_target_hosts, the router declines.
6391 7. Added expansion conditions match_domain, match_address, and match_local_
6392 part (NOT match_host).
6394 8. The placeholder for the Received: header didn't have a length field set.
6396 9. Added code to Exim itself and to exim_lock to test for a specific race
6397 condition that could lead to file corruption when using MBX delivery. The
6398 issue is with the lockfile that is created in /tmp. If this file is removed
6399 after a process has opened it but before that process has acquired a lock,
6400 there is the potential for a second process to recreate the file and also
6401 acquire a lock. This could lead to two Exim processes writing to the file
6402 at the same time. The added code performs the same test as UW imapd; it
6403 checks after acquiring the lock that its file descriptor still refers to
6404 the same named file.
6406 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6407 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6410 11. Added the smtp_active_hostname option. If used, this will typically be made
6411 to depend on the incoming interface address. Because $interface_address is
6412 not set up until the daemon has forked a reception process, error responses
6413 that can happen earlier (such as "too many connections") no longer contain
6416 12. If an expansion in a condition on a "warn" statement fails because a lookup
6417 defers, the "warn" statement is abandoned, and the next ACL statement is
6418 processed. Previously this caused the whole ACL to be aborted.
6420 13. Added the iplsearch lookup type.
6422 14. Added ident_timeout as a log selector.
6424 15. Added tls_certificate_verified as a log selector.
6426 16. Added a global option tls_require_ciphers (compare the smtp transport
6427 option of the same name). This controls incoming TLS connections.
6429 17. I finally figured out how to make tls_require_ciphers do a similar thing
6430 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6431 before starting the TLS session.
6433 18. Tabs are now shown as \t in -bP output.
6435 19. If the log selector return_path_on_delivery was set, Exim crashed when
6436 bouncing a message because it had too many Received: header lines.
6438 20. If two routers both had headers_remove settings, and the first one included
6439 a superfluous trailing colon, the final name in the first list and the
6440 first name in the second list were incorrectly joined into one item (with a
6441 colon in the middle).
6447 1. Added -C and -D options to the exinext utility, mainly to make it easier
6448 to include in the automated testing, but these could be helpful when
6449 multiple configurations are in use.
6451 2. The exinext utility was not formatting the output nicely when there was
6452 an alternate port involved in the retry record key, nor when there was a
6453 message id as well (for retries that were specific to a specific message
6454 and a specific host). It was also confused by IPv6 addresses, because of
6455 the additional colons they contain. I have fixed the IPv4 problem, and
6456 patched it up to do a reasonable job for IPv6.
6458 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6459 delivery, the log line now contains "pipelined" if PIPELINING was used.
6461 4. An SMTP transport process used to panic and die if the bind() call to set
6462 an explicit outgoing interface failed. This has been changed; it is now
6463 treated in the same way as a connect() failure.
6465 5. A reference to $sender_host_name in the part of a conditional expansion
6466 that was being skipped was still causing a DNS lookup. This no longer
6469 6. The def: expansion condition was not recognizing references to header lines
6470 that used bh_ and bheader_.
6472 7. Added the _cache feature to named lists.
6474 8. The code for checking quota_filecount in the appendfile transport was
6475 allowing one more file than it should have been.
6477 9. For compatibility with Sendmail, the command line option
6485 and sets the incoming protocol and host name (for trusted callers). The
6486 host name and its colon can be omitted when only the protocol is to be set.
6487 Note the Exim already has two private options, -pd and -ps, that refer to
6488 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6489 "s", but I don't think that's a major issue.
6491 10. A number of refactoring changes to the code, none of which should affect
6494 (a) The number of logging options was getting close to filling up the
6495 32-bit word that was used as a bit map. I have split them into two classes:
6496 those that are passed in the argument to log_write(), and those that are
6497 only ever tested independently outside of that function. These are now in
6498 separate 32-bit words, so there is plenty of room for expansion again.
6499 There is no change in the user interface or the logging behaviour.
6501 (b) When building, for example, log lines, the code previously used a
6502 macro that called string_cat() twice, in order to add two strings. This is
6503 not really sufficiently general. Furthermore, there was one instance where
6504 it was actually wrong because one of the argument was used twice, and in
6505 one call a function was used. (As it happened, calling the function twice
6506 did not affect the overall behaviour.) The macro has been replaced by a
6507 function that can join an arbitrary number of extra strings onto a growing
6510 (c) The code for expansion conditions now uses a table and a binary chop
6511 instead of a serial search (which was left over from when there were very
6512 few conditions). Also, it now recognizes conditions like "pam" even when
6513 the relevant support is not compiled in: a suitably worded error message is
6514 given if an attempt is made to use such a condition.
6516 11. Added ${time_interval:xxxxx}.
6518 12. A bug was causing one of the ddress fields not to be passed back correctly
6519 from remote delivery subprocesses. The field in question was not being
6520 subsequently used, so this caused to problems in practice.
6522 13. Added new log selectors queue_time and deliver_time.
6524 14. Might have fixed a bug in maildirsizefile handling that threw up
6525 "unexpected character" debug warnings, and recalculated the data
6526 unnecessarily. In any case, I expanded the warning message to give more
6529 15. Added the message "Restricted characters in address" to the statements in
6530 the default ACL that block characters like @ and % in local parts.
6532 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6533 Three changes have been made:
6535 (a) There was a serious bug; a negative response to MAIL caused the whole
6536 recipient domain to be cached as invalid, thereby blocking all messages
6537 to all local parts at the same domain, from all senders. This bug has
6538 been fixed. The domain is no longer cached after a negative response to
6539 MAIL if the sender used is not empty.
6541 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6544 (c) A new callout option, "use_sender" has been added for people who want
6545 the modified behaviour.
6551 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6554 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6555 indeed breaks things for older releases.
6557 3. Added additional logging to the case where there is a problem reading data
6558 from a filter that is running in a subprocess using a pipe, in order to
6559 try to track down a specific problem.
6561 4. Testing facility fudge: when running in the test harness and attempting
6562 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6563 getting "No route to host". Convert this to a timeout.
6565 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6568 6. Some OS don't have socklen_t but use size_t instead. This affects the
6569 fifth argument of getsockopt() amongst other things. This is now
6570 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6571 can be set for individual OS. I have set it for SunOS5, OSF1, and
6572 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6573 some earlier ones do not.
6575 7. Change 4.30/15 was not doing the test caselessly.
6577 8. The standard form for an IPv6 address literal was being rejected by address
6578 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6579 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6580 this, as well as the form without the "IPv6" on the front (but only when
6581 address literals are enabled, of course).
6583 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6585 10. Exim crashed if a message with an empty sender address specified by -f
6586 encountered a router with an errors_to setting. This could be provoked only
6587 by a command such as
6591 where an empty string was supplied; "<>" did not hit this bug.
6593 11. Installed PCRE release 4.5.
6595 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6596 remained set. It is now erased.
6598 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6599 times from message ids (which are base 36 rather than the normal 62).
6601 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6602 were being counted as actual protocol errors, and logged if the log
6603 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6604 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6605 and DATA following a set of rejected RCPTs do not count as protocol errors.
6606 In other words, Exim assumes they were pipelined, though this may not
6607 actually be the case. Of course, in all cases the client gets an
6608 appropriate error code.
6610 15. If a lookup fails in an ACL condition, a message about the failure may
6611 be available; it is used if testing the ACL cannot continue, because most
6612 such messages specify what the cause of the deferral is. However, some
6613 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6614 that caused an old message to be retained and used if a later statement
6615 caused a defer, replacing the real cause of the deferral.
6617 16. If an IP address had so many PTR records that the DNS lookup buffer
6618 was not large enough to hold them, Exim could crash while trying to process
6619 the truncated data. It now detects and logs this case.
6621 17. Further to 4.21/58, another change has been made: if (and only if) the
6622 first line of a message (the first header line) ends with CRLF, a bare LF
6623 in a subsequent header line has a space inserted after it, so as not to
6624 terminate the header.
6626 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6627 unnecessarily, used atoi() instead of strtol(), and didn't check the
6628 termination when getting file sizes from file names by regex.
6630 19. Completely re-implemented the support for maildirsize files, in the light
6631 of a number of problems with the previous contributed implementation
6632 (4.30/29). In particular:
6634 . If the quota is zero, the maildirsize file is maintained, but no quota is
6637 . If the maildir directory does not exist, it is created before any attempt
6638 to write a maildirsize file.
6640 . The quota value in the file is just a cache; if the quota is changed in
6641 the transport, the new value overrides.
6643 . A regular expression is available for excluding directories from the
6646 20. The autoreply transport checks the characters in options that define the
6647 message's headers; it allows continued headers, but it was checking with
6648 isspace() after an embedded newline instead of explicitly looking for a
6651 21. If all the "regular" hosts to which an address was routed had passed their
6652 expiry times, and had not reached their retry times, the address was
6653 bounced, even if fallback hosts were defined. Now Exim should go on to try
6656 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6657 equivalent code in the SMTP transport. Some hosts send humungous responses
6658 to HELO/EHLO, more than 1024 it seems.
6660 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6661 gives compiler warnings in some environments. I've now done it "properly",
6664 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6665 (because of problems with the built-in one) was declared to return uschar *
6666 instead of char *, causing compiler failure.
6668 25. Fixed a file descriptor leak when processing alias/forward files.
6670 26. Fixed a minor format string issue in dbfn.c.
6672 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6674 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6675 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6678 29. When a sender address is verified, it is cached, to save repeating the test
6679 when there is more than one recipient in a message. However, when the
6680 verification involves a callout, it is possible for different callout
6681 options to be set for different recipients. It is too complicated to keep
6682 track of this in the cache, so now Exim always runs a verification when a
6683 callout is required, relying on the callout cache for the optimization.
6684 The overhead is duplication of the address routing, but this should not be
6687 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6688 to be verified with callout=postmaster, and the main callout worked but the
6689 postmaster check failed, the verification correctly failed. However, if a
6690 subsequent RCPT command asked for sender verification *without* the
6691 postmaster check, incorrect caching caused this verification also to fail,
6694 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6695 it was not caching the DNS options (qualify_single, search_parents) that
6696 were used when the lookup failed. A subsequent lookup with different
6697 options therefore always gave the same answer, though there were cases
6698 where it should not have. (Example: a "domains = !$mx_any" option on a
6699 dnslookup router: the "domains" option is always processed without any
6700 widening, but the router might have qualify_single set.) Now Exim uses the
6701 cached value only when the same options are set.
6703 32. Added John Jetmore's "exipick" utility to the distribution.
6705 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6706 than a timeout (e.g. a certificate is required, and is not provided), an
6707 Exim server now closes the connection immediately. Previously it waited for
6708 the client to close - but if the client is SSL, it seems that they each
6709 wait for each other, leading to a delay before one of them times out.
6711 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6712 maintained 0.8.x compatibility because I don't think many are using it, and
6713 it is clearly obsolete.
6715 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6718 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6719 client certificate was expired. A simple patch fixes this, though I don't
6720 understand the full logic of why the verify callback is called multiple
6723 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6724 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6725 which causes problems with some clients (such as the Certicom SSL Plus
6726 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6727 disables the coutermeasure allowing Eudora to connect."
6729 38. Exim was not checking that a write() to a log file succeeded. This could
6730 lead to Bad Things if a log got too big, in particular if it hit a file
6731 size limit. Exim now panics and dies if it cannot write to a log file, just
6732 as it does if it cannot open a log file.
6734 39. Modified OS/Makefile-Linux so that it now contains
6736 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6738 The two -D definitions ensure that Exim is compiled with large file
6739 support, which makes it possible to handle log files that are bigger than
6742 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6743 instance) a domain was checked against a named list that involved a lookup,
6744 causing $domain_data to be set, then another domain was checked against the
6745 same list, then the first domain was re-checked, the value of $domain_data
6746 after the final check could be wrong. In particular, if the second check
6747 failed, it could be set empty. This bug probably also applied to
6750 41. The strip_trailing_dot option was not being applied to the address given
6751 with the -f command-line option.
6753 42. The code for reading a message's header from the spool was incrementing
6754 $received_count, but never initializing it. This meant that the value was
6755 incorrect (doubled) while delivering a message in the same process in which
6756 it was received. In the most common configuration of Exim, this never
6757 happens - a fresh exec is done - but it can happen when
6758 deliver_drop_privilege is set.
6760 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6761 it now includes up to 150 characters of the unexpected data in the log
6764 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6765 and building data strings. The size of both of these buffers was 10 000
6766 bytes - far larger than anybody would *ever* want, thought I. Needless to
6767 say, somebody hit the limit. I have increased the maximum line length to
6768 20 000 and the maximum data length of concatenated lines to 100 000. I have
6769 also fixed two bugs, because there was no checking on these buffers. Tsk,
6770 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6771 buffer is too small.
6773 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6774 lsearch lookups. Now it does.
6776 46. When parsing a route_list item in a manualroute router, a fixed-length
6777 buffer was used for the list of hosts. I made this 1024 bytes long,
6778 thinking that nobody would ever have a list of hosts that long. Wrong.
6779 Somebody had a whole pile of complicated expansion conditions, and the
6780 string was silently truncated, leading to an expansion error. It turns out
6781 that it is easier to change to an unlimited length (owing to other changes
6782 that have happened since this code was originally written) than to build
6783 structure for giving a limitation error. The length of the item that
6784 expands into the list of hosts is now unlimited.
6786 47. The lsearch lookup could not handle data where the length of text line was
6787 more than 4095 characters. Such lines were truncated, leading to shortened
6788 data being returned. It should now handle lines of any length.
6790 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6791 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6794 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6795 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6796 no longer happens in comment lines. A list of replacements is now placed
6797 at the head of all of the source files, except those whose only change is
6798 to replace PERL_COMMAND in the very first #! line.
6800 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6801 messages on the queue, with a bottom-up merge sort, using code contributed
6802 by Michael Haardt. This should make operations like -bp somewhat faster on
6803 large queues. It won't affect queue runners, except when queue_run_in_order
6806 51. Installed eximstats 1.31 in the distribution.
6808 52. Added support for SRV lookups to the dnslookup router.
6810 53. If an ACL referred to $message_body or $message_body_end, the value was not
6811 reset for any messages that followed in the same SMTP session.
6813 54. The store-handling optimization for building very long strings was not
6814 differentiating between the different store pools. I don't think this
6815 actually made any difference in practice, but I've tidied it.
6817 55. While running the routers to verify a sender address, $sender_address
6818 was still set to the sender address. This is wrong, because when routing to
6819 send a bounce to the sender, it would be empty. Therefore, I have changed
6820 it so that, while verifying a sender address, $sender_address is set to <>.
6821 (There is no change to what happens when verifying a recipient address.)
6823 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6824 target A or AAAA records (if not already returned) without resetting the
6825 qualify_single or search_parents options of the DNS resolver. These are
6826 inappropriate in this case because the targets of MX and SRV records must
6827 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6828 target that, when qualified, matched something in the local domain. These
6829 two options are now turned off when doing these lookups.
6831 57. It seems that at least some releases of Reiserfs (which does not have the
6832 concept of a fixed number of inodes) returns zero and not -1 for the
6833 number of available inodes. This interacted badly with check_spool_inodes,
6834 which assumed that -1 was the "no such thing" setting. What I have done is
6835 to check that the total number of inodes is greater than zero before doing
6836 the test of how many are available.
6838 58. When a "warn" ACL statement has a log_message modifier, the message is
6839 remembered, and not repeated. This is to avoid a lot of repetition when a
6840 message has many recipients that cause the same warning to be written.
6841 However, Exim was preserving the list of already written lines for an
6842 entire SMTP session, which doesn't seem right. The memory is now reset if a
6843 new message is started.
6845 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6846 debugging output unless log_rewrite was also set.
6848 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6849 of (int)(handle) when we know that handle contains (void *)(-1).
6851 61. The Exim daemon panic-logs an error return when it closes the incoming
6852 connection. However "connection reset by peer" seems to be common, and
6853 isn't really an error worthy of noting specially, so that particular error
6856 62. When Exim is trying to find all the local interfaces, it used to panic and
6857 die if the ioctl to get the interface flags failed. However, it seems that
6858 on at least one OS (Solaris 9) it is possible to have an interface that is
6859 included in the list of interfaces, but for which you get a failure error
6860 for this call. This happens when the interface is not "plumbed" into a
6861 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6862 failure of the "get flags" call assumes that the interface is down.
6864 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6865 makes life easier for people who are doing arithmetic on fields extracted
6866 from dates, where you often get leading zeros that should not be
6867 interpreted as octal.
6869 64. Added qualify_domain to the redirect router, to override the global
6872 65. If a pathologically long header line contained very many addresses (the
6873 report of this problem mentioned 10 000) and each of them was rewritten,
6874 Exim could use up a very large amount of memory. (It kept on making new
6875 copies of the header line as it rewrote, and never released the old ones.)
6876 At the expense of a bit more processing, the header rewriting function has
6877 been changed so that it no longer eats memory in this way.
6879 66. The generation of the Received: header has been moved from the time that a
6880 message starts to be received, to the time that it finishes. The timestamp
6881 in the Received: header should now be very close to that of the <= log
6882 line. There are two side-effects of this change:
6884 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6885 logged header lines no longer include the local Received: line, because
6886 it has not yet been created. The same applies to a copy of the message
6887 that is returned to a non-SMTP sender when a message is rejected.
6889 (b) When a filter file is tested using -bf, no additional Received: header
6890 is added to the test message. After some thought, I decided that this
6893 This change does not affect the value of $received_for. It is still set
6894 after address rewriting, but before local_scan() is called.
6896 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6898 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6899 gave an unhelpful panic error message, and a defer error. I have managed to
6900 change this behaviour so that it now rejects any supplied certificate,
6901 which seems right, as the list of acceptable certificates is empty.
6903 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6904 gave an unhelpful defer error. I have not managed to make this reject any
6905 supplied certificates, but the error message it gives is "no certificate
6906 supplied", which is not helpful.
6908 70. exigrep's output now also includes lines that are not associated with any
6909 message, but which match the given pattern. Implemented by a patch from
6910 Martin Sluka, which also tidied up the Perl a bit.
6912 71. Recipient callout verification, like sender verification, was using <> in
6913 the MAIL FROM command. This isn't really the right thing, since the actual
6914 sender may affect whether the remote host accepts the recipient or not. I
6915 have changed it to use the actual sender in the callout; this means that
6916 the cache record is now keyed on a recipient/sender pair, not just the
6917 recipient address. There doesn't seem to be a real danger of callout loops,
6918 since a callout by the remote host to check the sender would use <>.
6919 [SEE ABOVE: changed after hitting problems.]
6921 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6922 temporary errors. However, in the case of such a code being given after
6923 the end of a data transmission (i.e. after ".") Exim was failing to write
6924 a retry record for the message. (Yes, there was some broken host that was
6925 actually sending 8xx at this point.)
6927 73. An unknown lookup type in a host list could cause Exim to panic-die when
6928 the list was checked. (An example that provoked this was putting <; in the
6929 middle of a list instead of at the start.) If this happened during a DATA
6930 ACL check, a -D file could be left lying around. This kind of configuration
6931 error no longer causes Exim to die; instead it causes a defer error. The
6932 incident is still logged to the main and panic logs.
6934 74. Buglet left over from Exim 3 conversion. The message "too many messages
6935 in one connection" was written to the rejectlog but not the mainlog, except
6936 when address rewriting (yes!) was being logged.
6938 75. Added write_rejectlog option.
6940 76. When a system filter was run not as root (that is, when system_filter_user
6941 was set), the values of the $n variables were not being returned to the
6942 main process; thus, they were not subsequently available in the $sn
6945 77. Added +return_path_on_delivery log selector.
6947 78. A connection timeout was being treated differently from recipients deferred
6948 when testing hosts_max_try with a message that was older than the host's
6949 retry timeout. (The host should not be counted, thus allowing all hosts to
6950 be tried at least once before bouncing.) This may have been the cause of an
6951 occasionally reported bug whereby a message would remain on the queue
6952 longer than the retry timeout, but would be bounced if a delivery was
6953 forced. I say "may" because I never totally pinned down the problem;
6954 setting up timeout/retry tests is difficult. See also the next item.
6956 79. The ultimate address timeout was not being applied to errors that involved
6957 a combination of host plus message (for example, a timeout on a MAIL
6958 command). When an address resolved to a number of possible hosts, and they
6959 were not all tried for each delivery (e.g. because of hosts_max_try), a
6960 message could remain on the queue longer than the retry timeout.
6962 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6965 81. Fixed an obscure SMTP outgoing bug which required at least the following
6966 conditions: (a) there was another message waiting for the same server;
6967 (b) the server returned 5xx to all RCPT commands in the first message so
6968 that the message was not completed; (c) the server dropped the connection
6969 or gave a negative response to the RSET that Exim sends to abort the
6970 transaction. The observed case was a dropped connection after DATA that had
6971 been sent in pipelining mode. That is, the server had advertised PIPELINING
6972 but was not implementing it correctly. The effect of the bug was incorrect
6973 behaviour, such as trying another host, and this could lead to a crash.
6979 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6980 and daemon.c were passed as pointers to ints; they should have been
6981 pointers to socklen_t variables (which are typically unsigned ints).
6983 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6986 3. Fixed a really odd bug that affected only the testing scheme; patching a
6987 certain fixed string in the binary changed the value of another string that
6988 happened to be identical to the end of the original first string.
6990 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6991 name", it returns that address as the IP address. On some operating
6992 systems (e.g. Solaris), it also passes back the IP address string as the
6993 "host name". However, on others (e.g. Linux), it passes back an empty
6994 string. Exim wasn't checking for this, and was changing the host name to an
6995 empty string, assuming it had been canonicalized.
6997 5. Although rare, it is permitted to have more than one PTR record for a given
6998 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6999 all the names associated with an address, because they do in Solaris.
7000 However, it seems that they do not in Linux for data that comes from the
7001 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7002 I found this out when I moved to a new Linux workstation and tried to run
7003 the Exim test suite.
7005 To get round this problem I have changed the code so that it now does its
7006 own call to the DNS to look up PTR records when searching for a host name.
7007 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7008 addresses that are only in /etc/hosts are still found.
7010 This behaviour is, however, controlled by an option called host_lookup_
7011 order, which defaults to "bydns:byaddr". If people want to use the other
7012 order, or indeed, just use one or the other means of lookup, they can
7013 specify it in this variable.
7015 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7016 some operating systems, this comes back from gethostbyaddr() as an empty
7017 string, and this is what Exim used to test for. However, it seems that in
7018 other systems, "." is yielded. Exim now tests for this case too.
7020 7. The values of check_spool_space and check_log_space are now held internally
7021 as a number of kilobytes instead of an absolute number of bytes. If a
7022 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7023 kilobyte. This means that much larger values can be stored.
7025 8. Exim monitor: an attempt to get the action menu when not actually pointing
7026 at a message produces an empty menu entitled "No message selected". This
7027 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7028 no entries in it ("Shell widget menu has zero width and/or height"). So I
7029 have added a single, blank menu entry in this case.
7031 9. Added ${quote_local_part.
7033 10. MIME decoding is now applied to the contents of Subject: header lines when
7036 11. Now that a reference to $sender_host_address automatically causes a reverse
7037 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7038 host lookup before query-style lookups in lists that might use this
7039 variable. This has therefore been abolished, and the "net-" prefix is no
7040 longer necessary for query-style lookups.
7042 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7043 have been a typo for LFLAGS, so it has been changed.
7045 13. The install script calls Exim with "-C /dev/null" in order to find the
7046 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7047 to be output. However, since Exim outputs its version number before the
7048 error, it didn't break the script. It just looked ugly. I fixed this by
7049 always allowing "-C /dev/null" if the caller is root.
7051 14. Ignore overlarge ACL variable number when reading spool file - insurance
7052 against a later release with more variables having written the file.
7054 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7055 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7056 this, as well as the form without the "IPv6" on the front.
7058 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7059 OS/Makefile-Darwin file.
7061 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7062 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7064 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7065 variable. However, some OS use ints for inodes. I've added cast to long int
7066 to get rid of the compiler warning.
7068 19. I had forgotten to lock out "/../" in configuration file names when
7069 ALT_CONFIG_PREFIX was set.
7071 20. Routers used for verification do not need to specify transports. However,
7072 if such a router generated a host list, and callout was configured, Exim
7073 crashed, because it could not find a port number from the (non-existent)
7074 transport. It now assumes port 25 in this circumstance.
7076 21. Added the -t option to exigrep.
7078 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7079 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7080 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7081 them is set, LOOKUP_LSEARCH is forced.
7083 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7085 24. Added sender and host information to the "rejected by local_scan()" log
7086 line; previously there was no indication of these.
7088 25. Added .include_if_exists.
7090 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7091 Linux. It turns out that not all file systems support this. Apparently some
7092 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7093 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7094 not supported on the file descriptor, is now ignored when Exim is trying to
7095 sync a directory. This applies only to Linux.
7097 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7099 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7100 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7103 29. Added support for maildirsize files from supplied patch (modified a bit).
7105 30. The use of :fail: followed by an empty string could lead Exim to respond to
7106 sender verification failures with (e.g.):
7108 550 Verification failed for <xxx>
7109 550 Sender verify failed
7111 where the first response line was missing the '-' that indicates it is not
7112 the final line of the response.
7114 31. The loop for finding the name of the user that called Exim had a hardwired
7115 limit of 10; it now uses the value of finduser_retries, which is used for
7116 all other user lookups.
7118 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7121 33. Exim was neglecting to zero errno before one call of strtol() when
7122 expanding a string and expecting an integer value. On some systems this
7123 resulted in spurious "integer overflow" errors. Also, it was casting the
7124 result into an int without checking.
7126 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7127 did not work. The code looks as if it has *never* worked, though it appears
7128 to have been documented since at least release 1.62. I have made it work.
7130 35. The "timeout_DNS" error in retry rules, also documented since at least
7131 1.62, also never worked. As it isn't clear exactly what this means, and
7132 clearly it isn't a major issue, I have abolished the feature by treating it
7133 as "timeout", and writing a warning to the main and panic logs.
7135 36. The display of retry rules for -brt wasn't always showing the error code
7138 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7139 timeout_connect_A, timeout_connect_MX.
7141 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7142 to the empty sender.
7144 39. The daemon was not analysing the content of -oX till after it had closed
7145 stderr and disconnected from the controlling terminal. This meant that any
7146 syntax errors were only noted on the panic log, and the return code from
7147 the command was 0. By re-arranging the code a little, I've made the
7148 decoding happen first, so such errors now appear on stderr, and the return
7149 code is 1. However, the actual setting up of the sockets still happens in
7150 the disconnected process, so errors there are still only recorded on the
7153 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7154 connections (as happens on some IP stacks) was logged at start up time as
7155 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7156 it from "IPv6 and IPv4", which means that two separate sockets are being
7159 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7160 says whether AF_INET or AF_INET6 was passed as an argument.
7162 42. Exiwhat output was messed up when time zones were included in log
7165 43. Exiwhat now gives more information about the daemon's listening ports,
7166 and whether -tls-on-connect was used.
7168 44. The "port" option of the smtp transport is now expanded.
7170 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7171 silently ignored. Now an error message is written to the main and panic
7174 46. There's a new ACL modifier called "logwrite" which writes to a log file
7175 as soon as it is encountered.
7177 47. Added $local_user_uid and $local_user_gid at routing time.
7179 48. Exim crashed when trying to verify a sender address that was being
7182 49. Exim was recognizing only a space character after ".include". It now also
7183 recognizes a tab character.
7185 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7186 extracting the relevant information from the specification. The man page no
7187 longer contains scrambled data for the -d option, and I've added a section
7188 at the front about calling Exim under different names.
7190 51. Added "extra_headers" argument to the "mail" command in filter files.
7192 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7195 53. Installed eximstats 1.29.
7197 54. Added transport_filter_timeout as a generic transport option.
7199 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7200 Cc: header lines. This was required by RFC 822, but it not required by RFC
7203 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7204 incoming messages that did not have them. Now it does so only if the
7205 message originates locally, that is, if there is no associated remote host
7206 address. When Resent- header lines are present, this applies to the Resent-
7207 lines rather than the non-Resent- lines.
7209 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7210 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7212 58. Messages for configuration errors now include the name of the main
7213 configuration file - useful now that there may be more than one file in a
7214 list (.included file names were always shown).
7216 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7217 for those rare installations that do not start the daemon as root or run it
7218 setuid root. I've cut out the call to initgroups() if the daemon is not
7221 60. The Exim user and group can now be bound into the binary as text strings
7222 that are looked up at the start of Exim's processing.
7224 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7226 62. Added $mailstore_basename variable.
7228 63. Installed patch to sieve.c from Michael Haardt.
7230 64. When Exim failed to open the panic log after failing to open the main log,
7231 the original message it was trying to log was written to stderr and debug
7232 output, but if they were not available (the usual case in production), it
7233 was lost. Now it is written to syslog before the two lines that record the
7234 failures to open the logs.
7236 65. Users' Exim filters run in subprocesses under the user's uid. It is
7237 possible for a "deliver" command or an alias in a "personal" command to
7238 provoke an address rewrite. If logging of address rewriting is configured,
7239 this fails because the process is not running as root or exim. There may be
7240 a better way of dealing with this, but for the moment (because 4.30 needs
7241 to be released), I have disabled address rewrite logging when running a
7242 filter in a non-root, non-exim process.
7248 1. The buildconfig auxiliary program wasn't quoting the value set for
7249 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7250 not defined. This bug was masked in 4.22 by the effect that was fixed in
7253 2. Some messages that were rejected after a message id was allocated were
7254 shown as "incomplete" by exigrep. It no longer does this for messages that
7255 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7257 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7258 have allow_domain_literals set, the ID did not get logged in the <= line.
7259 Domain literals are now always recognized in Message-ID: header lines.
7261 4. The first argument for a ${extract expansion item is the key name or field
7262 number. Leading and trailing spaces in this item were not being ignored,
7263 causing some misleading effects.
7265 5. When deliver_drop_privilege was set, single queue runner processes started
7266 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7267 same command in the process it spins off) were not dropping privilege.
7269 6. When the daemon running as "exim" started a queue runner, it always
7270 re-executed Exim in the spun-off process. This is a waste of effort when
7271 deliver_drop_privilege is set. The new process now just calls the
7272 queue-runner function directly.
7278 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7281 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7282 ignored. Though the use of -oP was forcing the writing of a pid file, it
7283 was always written to the default place.
7285 3. If the message "no IP address found for host xxxx" is generated during
7286 incoming verification, it is now followed by identification of the incoming
7287 connection (so you can more easily find what provoked it).
7289 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7291 5. Added some features to "harden" Exim a bit more against certain attacks:
7293 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7294 be put in Local/Makefile. This is like the never_users runtime option,
7295 but it cannot be overridden. The default setting is "root".
7297 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7298 prefix string with which any file named in a -C command line option
7301 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7302 is retained for -C and -D only if the caller of Exim is root. Without
7303 it, the exim user may also use -C and -D and retain privilege.
7305 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7306 command line option is disabled.
7308 6. Macro names set by the -D option must start with an upper case letter, just
7309 like macro names defined in the configuration file.
7311 7. Added "dereference=" facility to LDAP.
7313 8. Two instances of the typo "uknown" in the source files are fixed.
7315 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7316 the Configure-Makefile script screwed up while processing it.
7318 10. Incorporated PCRE 4.4.
7320 11. The SMTP synchronization check was not operating right at the start of an
7321 SMTP session. For example, it could not catch a HELO sent before the client
7322 waited for the greeting. There is now a check for outstanding input at the
7323 point when the greeting is written. Because of the duplex, asynchronous
7324 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7325 way, but not yet received, when the check is performed.
7327 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7328 on TCP/IP sockets, because this apparently causes some broken clients to
7331 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7332 unchanged) from the Cygwin maintainer.
7334 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7335 was supported instead of testing for mbx. Effectively a typo.
7337 15. The spa authenticator server code was not checking that the input it
7338 received was valid base64.
7340 16. The debug output line for the "set" modifier in ACLs was not showing the
7341 name of the variable that was being set.
7343 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7345 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7346 lookup on its own. Something else was needed to trigger the lookup. For
7347 example, a match in host_lookup or the need for a host name in a host list.
7348 Now, if $sender_host_name is referenced and the host name has not yet been
7349 looked up, a lookup is performed. If the lookup fails, the variable remains
7350 empty, and $host_lookup_failed is set to "1".
7352 19. Added "eqi" as a case-independent comparison operator.
7354 20. The saslauthd authentication condition could segfault if neither service
7355 nor realm was specified.
7357 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7358 error message that was logged was misleading, and incoming SMTP
7359 connections were dropped. The message is now more accurate, and temporary
7360 errors are given to SMTP connections.
7362 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7363 (except RSET) with a 421 error, until QUIT is received. However, it was
7364 failing to send a response to QUIT.
7366 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7367 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7368 ACL is now run after the helo_try_verify_hosts code.
7370 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7371 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7372 case-independent, so other case variants are also recognized). Apparently
7373 some systems use these upper case variants.
7375 25. If more than two messages were waiting for the same host, and a transport
7376 filter was specified for the transport, Exim sent two messages over the
7377 same TCP/IP connection, and then failed with "socket operation on non-
7378 socket" when it tried to send the third.
7380 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7382 27. The extern definition of crypt16() in expand.c was not being excluded when
7383 the OS had its own crypt16() function.
7385 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7386 as a preferred synonym for return_size_limit, both as an option and as an
7389 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7391 30. Changed the default configuration ACL to relax the local part checking rule
7392 for addresses that are not in any local domains. For these addresses,
7393 slashes and pipe symbols are allowed within local parts, but the sequence
7394 /../ is explicitly forbidden.
7396 31. SPA server authentication was not clearing the challenge buffer before
7399 32. log_message in a "warn" ACL statement was writing to the reject log as
7400 well as to the main log, which contradicts the documentation and doesn't
7401 seem right (because no rejection is happening). So I have stopped it.
7403 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7404 However, I am unable to do any testing of this.
7406 34. Fixed an infelicity in the appendfile transport. When checking directories
7407 for a mailbox, to see if any needed to be created, it was accidentally
7408 using path names with one or more superfluous leading slashes; tracing
7409 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7411 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7412 contents are added to the log line that is written for every discarded
7413 recipient. (Previously a log_message setting was ignored.)
7415 36. The ${quote: operator now quotes the string if it is empty.
7417 37. The install script runs exim in order to find its version number. If for
7418 some reason other than non-existence or emptiness, which it checks, it
7419 could not run './exim', it was installing it with an empty version number,
7420 i.e. as "exim-". This error state is now caught, and the installation is
7423 38. An argument was missing from the function that creates an error message
7424 when Exim fails to connect to the socket for saslauthd authentication.
7425 This could cause Exim to crash, or give a corrupted message.
7427 39. Added isip, isip4, and isip6 to ${if conditions.
7429 40. The ACL variables $acl_xx are now saved with the message, and can be
7430 accessed later in routers, transports, and filters.
7432 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7433 strings in the file are not string-expanded.
7435 42. If a MAIL command specified a SIZE value that was too large to fit into an
7436 int variable, the check against message_size_limit failed. Such values are
7437 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7438 day this will have to be increased, but I don't think I want to be around
7439 when emails are that large.
7446 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7447 iconv() is not standard in FreeBSD.
7449 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7450 IPv6 enabled. The observed symptom was a segmentation fault on return from
7451 the function os_common_find_running_interfaces() in src/os.c.
7453 3. In the check_special_case() function in daemon.c I had used "errno" as an
7454 argument name, which causes warnings on some systems. This was basically a
7455 typo, since it was named "eno" in the comments!
7457 4. The code that waits for the clock to tick (at a resolution of some fraction
7458 of a second) so as to ensure message-id uniqueness was always waiting for
7459 at least one whole tick, when it could have waited for less. [This is
7460 almost certainly not relevant at current processor speeds, where it is
7461 unlikely to ever wait at all. But we try to future-proof.]
7463 5. The function that sleeps for a time interval that includes fractions of a
7464 second contained a race. It did not block SIGALRM between setting the
7465 timer, and suspending (a couple of lines later). If the interval was short
7466 and the sigsuspend() was delayed until after it had expired, the suspension
7467 never ended. On busy systems this could lead to processes getting stuck for
7470 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7471 process, before it forks any delivery processes. The open lookup caching
7472 mechanism meant that the open file or database connection was passed into
7473 the delivery process. The problem was that delivery processes always tidy
7474 up cached lookup data. This could cause a problem for the next delivery
7475 process started by the queue runner, because the external queue runner
7476 process does not know about the closure. So the next delivery process
7477 still has data in the lookup cache. In the case of a file lookup, there was
7478 no problem because closing a file descriptor in a subprocess doesn't affect
7479 the parent. However, if the lookup was caching a connection to a database,
7480 the connection was closed, and the second delivery process was likely to
7481 see errors such as "PGSQL: query failed: server closed the connection
7482 unexpectedly". The problem has been fixed by closing all cached lookups
7483 in a queue runner before running a delivery process.
7485 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7486 seem to have the "const" qualifier which it has on other OS. I've
7489 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7490 *of the same type* (client or server) with the same public name that an
7491 error should be diagnosed.
7493 9. When Exim looked up a host name for an IP address, but failed to find the
7494 original IP address when looking up the host name (a safety check), it
7495 output the message "<ip address> does not match any IP for NULL", which was
7496 confusing, to say the least. The bug was that the host name should have
7497 appeared instead of "NULL".
7499 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7500 user that is built into the binary, and the -C or -D options is used, root
7501 privilege is dropped before the configuration file is read. In addition,
7502 logging is switched to stderr instead of the normal log files. If the
7503 configuration then re-defines the Exim user, the unprivileged environment
7504 is probably not what is expected, so Exim logs a panic warning message (but
7507 However, if deliver_drop_privilege is set, the unprivileged state may well
7508 be exactly what is intended, so the warning has been cut out in that case,
7509 and Exim is allowed to try to write to its normal log files.
7515 1. smtp_return_error_details was not giving details for temporary sender
7516 or receiver verification errors.
7518 2. Diagnose a configuration error if two authenticators have the same public
7521 3. Exim used not to create the message log file for a message until the first
7522 delivery attempt. This could be confusing when incoming messages were held
7523 for policy or load reasons. The message log file is now created at the time
7524 the message is received, and an initial "Received" line is written to it.
7526 4. The automatically generated man page for command line options had a minor
7527 bug that caused no ill effects; however, a more serious problem was that
7528 the procedure for building the man page automatically didn't always
7529 operate. Consequently, release 4.20 contains an out-of-date version. This
7530 shouldn't happen again.
7532 5. When building Exim with embedded Perl support, the script that builds the
7533 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7534 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7536 6. The freeze_tell option was not being used for messages that were frozen on
7537 arrival, either by an ACL or by local_scan().
7539 7. Added the smtp_incomplete_transaction log selector.
7541 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7542 was accepting AUTH without a new EHLO.
7544 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7545 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7548 10. Logging of TCP/IP connections (when configured) now happens in the main
7549 daemon process instead of the child process, so that the TCP/IP connection
7550 count is more accurate (but it can never be perfect).
7552 11. The use of "drop" in a nested ACL was not being handled correctly in the
7553 outer ACL. Now, if condition failure induced by the nested "drop" causes
7554 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7555 or "require"), the connection is dropped.
7557 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7558 that yield "discard" can now be used with an "accept" or a "discard" verb,
7559 but an error is generated for any others (because I can't see a useful way
7560 to define what should happen).
7562 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7563 are now processed in the same way as lines in the Exim configuration file.
7564 In particular, continuation lines are supported.
7566 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7568 15. Added -ti meaning -t -i.
7570 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7571 domains, and warn by logging if others are found.
7573 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7574 returned from GIFCONF when Exim is trying to find the list of interfaces on
7575 a host. The code in os.c has been modified to copy each ifreq to an aligned
7576 structure in all cases.
7578 Also, in some cases, the returned ifreq's were being copied to a 'struct
7579 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7580 means the last couple of bytes of an IPv6 address could be chopped if the
7581 ifreq contained only a normal sockaddr (14 bytes storage).
7583 18. Named domain lists were not supported in the hosts_treat_as_local option.
7584 An entry such as +xxxx was not recognized, and was treated as a literal
7587 19. Ensure that header lines added by a DATA ACL are included in the reject log
7588 if the ACL subsequently rejects the message.
7590 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7591 MD5 (which is deprecated).
7593 21. When testing a filter file using -bf, Exim was writing a message when it
7594 took the sender from a "From " line in the message, but it was not doing so
7595 when it took $return_path from a Return-Path: header line. It now does.
7597 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7598 with a valid header line field name (a series of printing characters
7599 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7601 23. Changed "disc" in the source to "disk" to conform to the documentation and
7602 the book and for uniformity.
7604 24. Ignore Sendmail's -Ooption=value command line item.
7606 25. When execve() failed while trying to run a command in a pipe transport,
7607 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7608 could be confused with a return value of 69 from the command itself. This
7609 has been changed to 127, the value the shell returns if it is asked to run
7610 a non-existent command. The wording for the related log line suggests a
7611 non-existent command as the problem.
7613 26. If received_header_text expands to an empty string, do not add a Received:
7614 header line to the message. (Well, it adds a token one on the spool, but
7615 marks it "old" so that it doesn't get used or transmitted.)
7617 27. Installed eximstats 1.28 (addition of -nt option).
7619 28. There was no check for failure on the call to getsockname() in the daemon
7620 code. This can fail if there is a shortage of resources on the system, with
7621 ENOMEM, for example. A temporary error is now given on failure.
7623 29. Contrary to the C standard, it seems that in some environments, the
7624 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7625 program. Exim now does this explicitly; it affects the formatting of
7626 timestamps using strftime().
7628 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7629 complaints. I've now initialized all the variables, to avoid this.
7631 32. Header lines added by a system filter were not being "seen" during
7632 transport-time rewrites.
7634 33. The info_callback() function passed to OpenSSL is set up with type void
7635 (*)(SSL *, int, int), as described somewhere. However, when calling the
7636 function (actually a macro) that sets it up, the type void(*)() is
7637 expected. I've put in a cast to prevent warnings from picky compilers.
7639 34. If a DNS black list lookup found a CNAME record, but there were no A
7640 records associated with the domain it pointed at, Exim crashed.
7642 35. If a DNS black list lookup returned more than one A record, Exim ignored
7643 all but the first. It now scans all returned addresses if a particular IP
7644 value is being sought. In this situation, the contents of the
7645 $dnslist_value variable are a list of all the addresses, separated by a
7648 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7649 to accept a host name and all its aliases if the forward lookup for any of
7650 them yielded the IP address of the incoming connection. Now it accepts only
7651 those names whose forward lookup yields the correct IP address. Any other
7652 names are discarded. This closes a loophole whereby a rogue DNS
7653 administrator could create reverse DNS records to break through a
7654 wildcarded host restriction in an ACL.
7656 37. If a user filter or a system filter that ran in a subprocess used any of
7657 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7658 the wrong values were passed to the pipe command ($thisaddress had the
7659 value of $0, $0 had the value of $1, etc). This bug was introduced by
7660 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7662 38. Improved the line breaking for long SMTP error messages from ACLs.
7663 Previously, if there was no break point between 40 and 75 characters, Exim
7664 left the rest of the message alone. Two changes have been made: (a) I've
7665 reduced the minimum length to 35 characters; (b) if it can't find a break
7666 point between 35 and 75 characters, it looks ahead and uses the first one
7667 that it finds. This may give the occasional overlong line, but at least the
7668 remaining text gets split now.
7670 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7671 file descriptors might be low, and that setting 1000 would always raise it.
7672 It turns out that in some environments, the limit is already over 1000 and
7673 that lowering it causes trouble. So now Exim takes care not to decrease it.
7675 40. When delivering a message, the value of $return_path is set to $sender_
7676 address at the start of routing (routers may change the value). By an
7677 oversight, this default was not being set up when an address was tested by
7678 -bt or -bv, which affected the outcome if any router or filter referred to
7681 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7682 log only when "message" or "log_message" are set. However, if one of the
7683 conditions was an address verification, or a call to a nested ACL, the
7684 messages generated by the underlying test were being passed through. This
7685 no longer happens. The underlying message is available in $acl_verify_
7686 message for both "message" and "log_message" expansions, so it can be
7687 passed through if needed.
7689 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7690 new expansion $bh_ to give the encoded byte string without charset
7691 translation. Translation happens only if iconv() is available; HAVE_ICONV
7692 indicates this at build time. HEADERS_CHARSET gives the charset to
7693 translate to; headers_charset can change it in the configuration, and
7694 "headers charset" can change it in an individual filter file.
7696 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7697 that creates RFC 2047 encoded "words" labels them as that charset instead
7698 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7699 expansion operator; (ii) when Exim creates a From: line for a local
7700 message; (iii) when a header line is rewritten to include a "phrase" part.
7702 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7703 buggy, causing it to skip the first lines of messages whose message ID
7704 ended in 'D'. This would not have bitten before Exim release 4.14, because
7705 message IDs were unlikely to end in 'D' before then. The effect was to have
7706 incorrect size information for certain domains.
7708 45. #include "config.h" was missing at the start of the crypt16.c module. This
7709 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7712 46. If there was a timeout during a "random" callout check, Exim treated it as
7713 a failure of the random address, and carried on sending RSET and the real
7714 address. If the delay was just some slowness somewhere, the response to the
7715 original RCPT would be taken as a response to RSET and so on, causing
7716 mayhem of various kinds.
7718 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7719 when I implemented it. It didn't allow for the fact that some option values
7720 may legitimately be negative (e.g. size_addition), and it didn't even do
7721 the right test for positive values.
7723 48. Domain names in DNS records are case-independent. Exim always looks them up
7724 in lower case. Some resolvers return domain names in exactly the case they
7725 appear in the zone file, that is, they may contain uppercase letters. Not
7726 all resolvers do this - some return always lower case. Exim was treating a
7727 change of case by a resolver as a change of domain, similar to a widening
7728 of a domain abbreviation. This triggered its re-routing code and so it was
7729 trying to route what was effectively the same domain again. This normally
7730 caused routing to fail (because the router wouldn't handle the domain
7731 twice). Now Exim checks for this case specially, and just changes the
7732 casing of the domain that it ultimately uses when it transmits the message
7735 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7738 50. If a filter generated a file delivery with a non-absolute name (possible if
7739 no home directory exists for the router), the forbid_file option was not
7742 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7743 the existing equality matching.
7745 52. Exim was using ints instead of ino_t variables in some places where it was
7746 dealing with inode numbers.
7748 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7749 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7750 called TMPDIR, and if it finds it is different, it changes its value.
7752 54. The smtp_printf() function is now made available to local_scan() so
7753 additional output lines can be written before returning. There is also an
7754 smtp_fflush() function to enable the detection of a dropped connection.
7755 The variables smtp_input and smtp_batched_input are exported to
7758 55. Changed the default runtime configuration: the message "Unknown user"
7759 has been removed from the ACL, and instead placed on the localuser router,
7760 using the cannot_route_message feature. This means that any verification
7761 failures that generate their own messages won't get overridden. Similarly,
7762 the "Unrouteable address" message that was in the ACL for unverifiable
7763 relay addresses has also been removed.
7765 56. Added hosts_avoid_esmtp to the smtp transport.
7767 57. The exicyclog script was not checking for the esoteric option
7768 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7769 will work only if exicyclog is run under the appropriate euid.
7771 58. Following a discussion on the list, the rules by which Exim recognises line
7772 endings on incoming messages have been changed. The -dropcr and drop_cr
7773 options are now no-ops, retained only for backwards compatibility. The
7774 following line terminators are recognized: LF CRLF CR. However, special
7775 processing applies to CR:
7777 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7778 nor a local message in the state where . is a terminator.
7780 (ii) If a bare CR is encountered in a header line, an extra space is added
7781 after the line terminator so as not to end the header. The reasoning
7782 behind this is that bare CRs in header lines are most likely either
7783 to be mistakes, or people trying to play silly games.
7785 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7786 was being incorrectly given as 18 bytes larger than it should have been.
7787 This is a VOB (very old bug).
7789 60. This may never have affected anything current, but just in case it has:
7790 When the local host is found other than at the start of a list of hosts,
7791 the local host, those with the same MX, and any that follow, are discarded.
7792 When the list in question was part of a longer list of hosts, the following
7793 hosts (not currently being processed) were also being discarded. This no
7794 longer happens. I'm not sure if this situation could ever has previously
7797 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7800 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7801 header. This is something that is recommended in a new Internet Draft, and
7802 is something that is documented as being done by Sendmail. There are two
7803 possible values. For messages generated by the autoreply transport, Exim
7806 Auto-Submitted: auto-replied
7808 whereas for all other generated messages (e.g. bounces) it adds
7810 Auto-Submitted: auto-generated
7812 63. The "personal" condition in filters now includes a test for the
7813 Auto-Submitted: header. If it contains the string "auto-" the message it
7814 not considered personal.
7816 64. Added rcpt_include_affixes as a generic transport option.
7818 65. Added queue_only_override (default true).
7820 66. Added the syslog_duplication option.
7822 67. If what should have been the first header line of a message consisted of
7823 a space followed by a colon, Exim was mis-interpreting it as a header line.
7824 It isn't of course - it is syntactically invalid and should therefore be
7825 treated as the start of the message body. The misbehaviour could have
7826 caused a number of strange effects, including loss of data in subsequent
7827 header lines, and spool format errors.
7829 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7830 client host had authenticated. This control can now be exercised by an ACL
7831 for more flexibility.
7833 69. By default, callouts do not happen when testing with -bh. There is now a
7834 variant, -bhc, which does actually run the callout code, including
7835 consulting and updating the callout cache.
7837 70. Added support for saslauthd authentication, courtesy of Alexander
7840 71. If statvfs() failed on the spool or log directories while checking their
7841 size for availability, Exim confusingly gave the error "space shortage".
7842 Furthermore, in debugging mode it crashed with a floating point exception.
7843 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7844 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7845 serious problem, Exim now writes to the main and panic logs when this
7846 happens, with details of the failure. It then refuses to accept the
7847 incoming message, giving the message "spool directory problem" or "log
7848 directory problem" with a 421 code for SMTP messages.
7850 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7851 0, 1, and 2 exist, because some OS complain for execs without them (see
7852 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7853 descriptors. However, the code omitted to check that the open succeeded,
7854 causing mysterious errors if for some reason the permissions on /dev/null
7855 got screwed. Now Exim writes a message to the main and panic logs, and
7856 bombs out if it can't open /dev/null.
7858 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7859 interact so that it is all more flexible. It is supposed to remain
7860 backwards compatible. Also added extra_local_interfaces.
7862 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7863 to bomb out with an assertion failure - to the client this appears as a
7864 connection drop. This problem occurs in the part of the code that was taken
7865 from the Samba project. Fortunately, the assertion is in a very simple
7866 function, so I have fixed this by reproducing the function inline in the
7867 one place where it is called, and arranging for authentication to fail
7868 instead of killing the process with assert().
7870 75. The SPA client code was not working when the server requested OEM rather
7871 than Unicode encoding.
7873 76. Added code to make require_files with a specific uid setting more usable in
7874 the case where statting the file as root fails - usually a non-root-mounted
7875 NFS file system. When this happens and the failure is EACCES, Exim now
7876 forks a subprocess and does the per-uid checking as the relevant uid.
7878 77. Added process_log_path.
7880 78. If log_file_path was not explicitly set, a setting of check_log_space or
7881 check_log_inodes was ignored.
7883 79. If a space check for the spool or log partitions fails, the incident is now
7884 logged. Of course, in the latter case the data may get lost...
7886 80. Added the %p formatting code to string_format() so that it can be used to
7887 print addresses in debug_print(). Adjusted all the address printing in the
7888 debugging in store.c to use %p rather than %d.
7890 81. There was a concern that a line of code in smtp_in.c could overflow a
7891 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7892 initially expressed, the concern was not well-founded, because trailing
7893 spaces are removed early. However, if the trailing spaces were followed by
7894 a NULL, they did not get removed, so the overflow was possible. Two fixes
7897 (a) I re-wrote the offending code in a cleaner fashion.
7898 (b) If an incoming SMTP command contains a NULL character, it is rejected
7901 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7902 runs initgroups(), so that if the Exim user is in any additional groups,
7903 they will be used during message reception.
7909 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob