1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
9 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
10 related applications. Previously an "H" was used where available info
11 says that "M" should be, so change to match.
13 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
14 as arguments, so an implementation trying to copy these into a local
15 buffer was taking a taint-enforcement trap. Fix by using dynamically
16 created buffers. Similar fix for radius expansion condition.
18 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
19 reasonable, eg. to count headers. Fix by using dynamically created
20 buffers rather than a local. Do similar fixes for ACL actions "dcc",
21 "log_reject_target", "malware" and "spam"; the arguments are expanded
22 so could be handling tainted values.
24 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
25 broken the (no-op) support for this sendmail command. Restore it
26 to doing nothing, silently, and returning good status.
28 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
29 record path was given (or the default used) without a leading directory
30 path, an error occurred on trying to open it. Use the transport's working
33 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
34 transport. Previously it was the name on the DNS A-record; use instead
35 the head of the CNAME chain leading there (if there is one). This seems
36 to align better with RFC 6125.
38 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
39 smtp_accept_max_per_host allocated resources which were not released
40 when the limit was exceeded. This eventually crashed the daemon. Fix
41 by adding a relase action in that path.
43 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
44 expanded; previously using tainted values was rejected. Fix by using
45 dynamically-created buffers.
47 JH/09 Relax restrictions on ACL verify condition needing access to message
48 headers. Previously they were only permitted in data and non-smtp ACLs;
49 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
50 not_blind, header_sender and header_names_ascii verification.
52 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
53 Previously a macro used one argument twice; when called with the
54 argument as an expression having side-effects, incorrect operation
55 resulted. Use an inlineable function.
57 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
58 held open for a verify callout. Previously this wan not accounted for
59 and a corrupt onward SMTP conversation resulted.
61 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
62 Previously they were embedded naively in the constructed address; when
63 needed, strip the quoting and quote the entire local_part.
64 Also make the inbound_srs expansion condition handle quoting.
66 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
67 excluded, not matching the documentation.
69 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
70 was given for the sqlite_dbfile a trap resulted.
72 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
73 "name" argument resulted in a trap. There is no reason to disallow such;
74 this was a coding error.
76 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
77 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
78 spent suspended, ignoring the Posix definition. Previously we assumed
79 it did and a constant offset from real time could be used as a correction.
80 Change to using the same clock source for the start-of-message and the
81 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
82 exists, just to get a clock slightly more aligned to reality.
84 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
85 RFC says it is optional some validators care. The missing char was not
86 intended but triggered by a line-wrap alignement. Discovery and fix by
87 Guillaume Outters, hacked on by JH.
89 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
90 name being quoted was tainted a trap would be taken. Fix by using
91 dynamicaly created buffers. The routine could have been called by a
92 rewrite with the "h" flag, by using the "-F" command-line option, or
93 by using a "name=" option on a control=submission ACL modifier.
95 JH/19 SPF: change the Authentication-Results expansion component to give
96 smtp.helo when the sender domain is empty. Previously it gave
99 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
100 not in 127.0.0.0/8 to help in spotting list domains taken over by a
101 domain-parking registrar.
103 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
104 Previously when a whitespace character was specified it was not inserted
105 after removing the newline.
107 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
108 the domain part of the recipient address. This overrides any tls_sni
109 option set, which was previously used.
111 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
114 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
115 is_tainted() had an off-by-one error in the overenthusiastic direction.
116 Find and fix by Gavan. Although NetBSD is not a supported platform for
117 4.94 this bug could affect other platforms.
119 PP/01 Fix default prime selection to be consistent.
120 One path used ike23 still, instead of exim.dev.20160529.3; now both
121 execution flows will use the same DH primes (currently
122 exim.dev.20160529.3).
124 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
125 option in smtp transport, to match the documentation. Previously
126 verification was not being done in some cases where it should have been.
128 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
129 than one server was defined and depending on the platform memory layout
130 details, an internal consistency trap could be hit while walking the list
133 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
134 files. The value is used by the authresults expansion item. Previously
135 if this was used in a router or transport, a crash could result.
137 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
138 have a sufficient incidence of EINTR returns from select that an
139 interaction with other operations done by the main daemon loop exposed
140 a bug in the error-handling. This was benign apart from the log
143 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
144 for consistency with delivered "=>" and failed "**" lines. While we're
145 there, handle PRX and TFO.
151 JH/01 Avoid costly startup code when not strictly needed. This reduces time
152 for some exim process initialisations. It does mean that the logging
153 of TLS configuration problems is only done for the daemon startup.
155 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
157 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
158 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
161 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
162 DKIM signing is being done. Previously a transport_filter would always
163 disable CHUNKING, falling back to traditional DATA.
165 JH/05 Regard command-line recipients as tainted.
167 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
169 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
170 PAM library frees one of the arguments given to it, despite the
171 documentation. Therefore a plain malloc must be used.
173 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
174 on-stack buffers were used, resulting in a taint trap when DSN information
175 copied from a received message was written into the buffer.
177 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
178 the ordering of its ARC headers. This caused a crash.
180 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
181 a new record was being constructed with information from the peer, a trap
184 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
185 installation would get error messages from DMARC verify, when it hit the
186 nonexistent file indicated by the default. Distros wanting DMARC enabled
187 should both provide the file and set the option.
188 Also enforce no DMARC verification for command-line sourced messages.
190 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
191 could, depending on the platform, hang at the STARTTLS response.
193 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
194 message on a connection. Previously if one message had ARC headers and
195 the following one did not, a crash could result when adding an
196 Authentication-Results: header.
198 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
199 Exim main code and Exim-related utities. The introduction of taint
200 tracking also did many adjustments to string handling. Since then, eximon
201 frequently terminated with an assert failure.
203 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
204 check for 452 responses. This slightly helps the inefficieny of doing
205 a large alias-expansion into a recipient-limited target. The max_rcpt
206 transport option still applies (and at the current default, will override
207 the new feature). The check is done for either cause of synch, and forces
208 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
209 connection. The new facility is not tunable at this time.
211 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
212 library live data was being used, so the results became garbage. Make
213 copies while it is still usable.
215 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
216 on delivery deferred (==) and failed (**) lines (if a delivery was
217 attemtped). Previously it was only on completion (=>) lines.
219 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
220 for the expansion of the server_scram_iter and server_scram_salt options.
222 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
223 are now specifically given a NO_DATA response without hitting the system
224 resolver. The library goes on to do the now-standard TXT lookup.
225 Use of dnsdb lookups is not affected.
227 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
228 only retrieve the errormessage once. Previously two calls to dlerror()
229 were used, and the second one (for mainlog/paniclog) retrieved null
232 JH/20 Taint checking: disallow use of tainted data for
233 - the appendfile transport file and directory options
234 - the pipe transport command
235 - the autoreply transport file, log and once options
236 - file names used by the redirect router (including filter files)
238 - paths used by single-key lookups
239 Previously this was permitted.
241 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
242 adjusted the size of a major service buffer; this failed because the
243 buffer was in use at the time. Change to a compile-time increase in the
244 buffer size, when this authenticator is compiled into exim.
246 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
247 previous fast-mode was untenable in the face of glibs using mmap to
248 support larger malloc requests.
250 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
251 New values supported, if defined on system where compiled:
252 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
253 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
255 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
256 running a limited number of proceses in parallel, a benefit is gained. The
257 amount varies with the platform hardware and load. The use of the option
258 queue_run_in_order means we cannot do this, as ordering becomes
261 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
262 had introduced a string-copy (for ensuring NUL-termination) which was not
263 appropriate for that case, which can include embedded NUL bytes in the
264 block of data. Investigation showed the copy to actually be needless, the
265 data being length-specified.
267 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
268 done during a receiving connection, and both used TLS, global info was
269 used rather than per-connection info for tracking the state of data
270 queued for transmission. This could result in a connection hang.
272 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
273 Previously, when delivering serveral messages down a single connection
274 only the first would provide a SIZE. This was due to the size information
275 not being properly tracked.
277 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
278 TAI (at 37 seconds currently), pretend to be in UTC for time-related
279 expansion and logging. Previously, spurious values such as a future
280 minute could be seen.
282 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
283 it could crash from a null-deref. This could also affect the
284 ${addresses: } operator and ${readsock } item.
286 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
287 message following a mime one, the variable was not reset.
289 JH/30 When an pipelined-connect fails at the first response, assume incorrect
290 cached capability (perhaps the peer reneged?) and immediately retry in
293 JH/31 Fix spurious detection of timeout while writing to transport filter.
295 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
296 an attempt to copy the string was made before checking it.
298 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
299 taint of the lookup key was maintained; we now regard the presence in the
300 filesystem as sufficient validation.
302 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
303 argument is supplied.
305 JH/35 The dsearch lookup now requires that the directory is an absolute path.
306 Previously this was not checked, and nonempty relative paths made an
307 access under Exim's current working directory.
309 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
310 Previously no event was raised.
312 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
313 parameter supplied by the sender MAIL FROM command. Previously it was
314 ignored, and only the check_spool_space option value for the required
317 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
318 the size of the signing public-key. Previously it was instead giving
319 the size of the signature hash.
321 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
322 the default. See the (new) dkim_verify_min_keysizes option.
324 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
325 and an ACL use a lookup for checking either the local_part or domain,
326 stale data could be accessed. Ensure that variable references are
327 dropped between messages.
329 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
330 by the client was not checked as pointing within response data before
331 being used. A malicious client could thus cause an out-of-bounds read and
332 possibly gain authentication. Fix by adding the check.
334 JH/42 Internationalisation: change the default for downconversion in the smtp
335 transport to be "if needed". Previously it was "as previously set" for
336 the message, which usually meant "if needed" for message-submission but
337 "no" for everything else. However, MTAs have been seen using SMTPUTF8
338 even when the envelope addresses did not need it, resulting in forwarding
339 failures to non-supporting MTAs. A downconvert in such cases will be
340 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
341 transport. The change does mean that addresses needing conversion will
342 be converted when previously a delivery failure would occur.
344 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
345 response was received it would be used unchecked in a fail-DSN, violating
346 standards on line-length limits. Truncate if needed.
348 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
349 doesn't work. (Additionally add a new main config option to configure the
356 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
357 side, to decode a TLS 1.3 packet capture.
359 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
360 Previously the default library behaviour applied, sending two, each in
363 JH/03 Debug output for ACL now gives the config file name and line number for
366 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
368 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
370 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
371 buffer overrun for (non-chunking) other transports.
373 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
374 TLS1.3, means that a server rejecting a client certificate is not visible
375 to the client until the first read of encrypted data (typically the
376 response to EHLO). Add detection for that case and treat it as a failed
377 TLS connection attempt, so that the normal retry-in-clear can work (if
378 suitably configured).
380 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
381 and/or domain. Found and fixed by Jason Betts.
383 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
384 configuration). If a CNAME target was not a wellformed name pattern, a
387 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
388 the OS reports them interleaved with other addresses.
390 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
391 used both for input and for a verify callout, both encrypted, SMTP
392 responses being sent by the server could be lost. This resulted in
393 dropped connections and sometimes bounces generated by a peer sending
396 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
397 string. Previously it was possible to cause undefined behaviour in a
398 library routine (usually a crash). Found by "zerons".
400 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
403 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
404 API was removed, so update to use the newer ones.
406 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
407 any timeout set, is taking a long time. Previously we would hang on to a
408 rotated logfile "forever" if the input was arriving with long gaps
409 (a previous attempt to fix addressed lack, for a long time, of initial
412 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
413 shared (NFS) environment. The length of the tempfile name is now
414 4 + 16 ("hdr.$message_exim_id") which might break on file
415 systems which restrict the file name length to lower values.
418 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
419 shared (NFS) environment.
421 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
422 did for all versions <4.90). Notably -M, -m, --invert, -I may be
425 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
426 on some platforms for bit 31.
428 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
429 to changes apparently associated with TLS1.3 handling some of the APIs
430 previously used were either nonfunctional or inappropriate. Strings
431 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
432 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
433 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
434 This affects log line X= elements, the $tls_{in,out}_cipher variables,
435 and the use of specific cipher names in the encrypted= ACL condition.
437 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
439 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
440 verification result was not updated unless hosts_require_ocsp applied.
442 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
443 queue_list_requires_admin set to false, non-admin users were denied the
446 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
447 directory-of-certs mode. Previously they were advertised despite the
450 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
451 A single TCP connection by a client will now hold a TLS connection open
452 for multiple message deliveries, by default. Previously the default was to
455 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
456 default. If built with the facility, DANE will be used. The facility
457 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
459 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
460 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
461 must be defined and you must still, unless you define DISABLE_TLS, manage
462 the the include-dir and library-file requirements that go with that
463 choice. Non-TLS builds are still supported.
465 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
468 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
469 default. If the platform supports and has the facility enabled, it will
470 be requested on all coneections.
472 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
473 controlled by the build-time option SUPPORT_PIPE_CONNECT.
475 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
477 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
478 success-DSN messages. Previously the From: header was always the default
479 one for these; the option was ignored.
481 JH/28 Fix the timeout on smtp response to apply to the whole response.
482 Previously it was reset for every read, so a teergrubing peer sending
483 single bytes within the time limit could extend the connection for a
484 long time. Credit to Qualsys Security Advisory Team for the discovery.
486 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
487 delivery address, which leaked information of the results of local
488 forwarding. Change to the original envelope recipient address, per
491 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
492 requested. Previously not bounce was generated and a log entry of
493 error ignored was made.
495 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
497 JH/32 Introduce a general tainting mechanism for values read from the input
498 channel, and values derived from them. Refuse to expand any tainted
499 values, to catch one form of exploit.
501 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
502 was unused and the unexpanded text used for the test. Found and
503 fixed by Ruben Jenster.
505 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
506 an attempt to use a TLS library read routine dereffed a nul pointer,
509 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
510 them in our smtp response.
512 JH/36 Have the general router option retry_use_local_part default to true when
513 any of the restrictive preconditions are set (to anything). Previously it
514 was only for check_local user. The change removes one item of manual
515 configuration which is required for proper retries when a remote router
516 handles a subset of addresses for a domain.
518 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
519 link count into consideration.
521 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
522 caused the extension of big_buffer, the following lines were ignored.
524 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
525 accordance with RFC 2308. Previously there was no expiry, so a longlived
526 receive process (eg. due to ACL delays) versus a short SOA value could
529 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
531 JH/39 Promote DMARC support to mainline.
533 JH/40 Bug 2452: Add a References: header to DSNs.
535 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
536 parameters. The relevant library call is documented as "Deprecated: This
537 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
538 3.6.0, DH parameters are negotiated following RFC7919."
540 HS/06 Change the default of dnssec_request_domains to "*"
542 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
543 carried on and emitted a BDAT command, even when PIPELINING was not
546 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
547 buffer was used for the filename, resulting in a trap when tainted
548 arguments (eg. $domain) were used.
550 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
551 recommended to avoid a possible server-load attack. The feature can be
552 re-enabled via the openssl_options main cofiguration option.
554 JH/45 local_scan API: documented the current smtp_printf() call. This changed
555 for version 4.90 - adding a "more data" boolean to the arguments.
556 Bumped the ABI version number also, this having been missed previously;
557 release versions 4.90 to 4.92.3 inclusive were effectively broken in
558 respect of usage of smtp_printf() by either local_scan code or libraries
559 accessed via the ${dlfunc } expansion item. Both will need coding
560 adjustment for any calls to smtp_printf() to match the new function
561 signature; a FALSE value for the new argument is always safe.
563 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
564 the file-offset (which the Linux syscall does, and exim expects); this
565 resulted in an indefinite loop.
567 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
568 to do ARC verification. The Authentication-Results: header line added
569 by the configuration then had no ARC item.
575 JH/01 Remove code calling the customisable local_scan function, unless a new
576 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
578 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
579 non-signal-safe functions being used.
581 JH/03 Bug 2269: When presented with a received message having a stupidly large
582 number of DKIM-Signature headers, disable DKIM verification to avoid
583 a resource-consumption attack. The limit is set at twenty.
585 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
586 report of oldest_pass in ${authres } in consequence, and separate out
587 some descriptions of reasons for verification fail.
589 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
590 files in the spool were present and unlocked. A queue-runner could spot
591 them, resulting in a duplicate delivery. Fix that by doing the unlock
592 after the unlink. Investigation by Tim Stewart. Take the opportunity to
593 add more error-checking on spoolfile handling while that code is being
596 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
597 No known attacks, no CVE, this is defensive hardening.
599 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
600 a queue-runner could start a delivery while other operations were ongoing.
601 Cutthrough delivery was a common victim, resulting in duplicate delivery.
602 Found and investigated by Tim Stewart. Fix by using the open message data
603 file handle rather than opening another, and not locally closing it (which
604 releases a lock) for that case, while creating the temporary .eml format
605 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
607 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
608 $sender_verify_failure/$recipient_verify_failure to "random".
610 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
613 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
614 Previously this would segfault.
616 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
619 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
620 like zero, since the resolver should be doing this for us, But we need one
621 as a CNAME but no MX presence gets the CNAME returned; we need to check
622 that doesn't point to an MX to declare it "no MX returned" rather than
623 "error, loop". A new main option is added so the older capability of
624 following some limited number of chain links is maintained.
626 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
628 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
629 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
630 now-deprecated earlier definitions used only specified the range up to TLS
631 1.2 (in the older-version library docs).
633 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
635 JH/15 Rework TLS client-side context management. Stop using a global, and
636 explicitly pass a context around. This enables future use of TLS for
637 connections to service-daemons (eg. malware scanning) while a client smtp
638 connection is using TLS; with cutthrough connections this is quite likely.
640 JH/16 Fix ARC verification to do AS checks in reverse order.
642 JH/17 Support a "tls" option on the ${readsocket } expansion item.
644 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
645 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
646 Previously the "utf8" would be re-prepended for every additional message.
648 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
649 Previously thery were accepted, resulting in issues when attempting to
650 forward messages to a non-supporting MTA.
652 PP/02 Let -n work with printing macros too, not just options.
654 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
655 one parent address was copied, and bogus data was used at delivery-logging
656 time. Either a crash (after delivery) or bogus log data could result.
657 Discovery and analysis by Tim Stewart.
659 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
660 Previously if the string ended mid-character, we did not insert the
661 promised '?' replacement.
663 PP/04 Documentation: current string operators work on bytes, not codepoints.
665 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
666 should pack well giving a smaller memory footprint so better caching and
667 therefore performance. Group the declarations where this can't be done so
668 that the byte-sized flag variables are not interspersed among pointer
669 variables, giving a better chance of good packing by the compiler.
671 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
672 non-null, to avoid issues with sites running BATV. Previously reports were
673 sent with an empty envelope sender so looked like bounces.
675 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
676 The ignore_error flag wasn't being returned from the filter subprocess so
677 was not set for later routers. Investigation and fix by Matthias Kurz.
679 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
680 and a msg:complete for the whole, when a message is manually removed using
681 -Mrm. Developement by Matthias Kurz, hacked on by JH.
683 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
684 a "Gnu special" function, asprintf() in the DB utility binary builds; I
685 hope that is portable enough.
687 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
688 requiring a known-CA anchor certificate; make it now rely entirely on the
689 TLSA as an anchor. Checking the name on the leaf cert against the name
690 on the A-record for the host is still done for TA (but not for EE mode).
692 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
693 included in delivery lines for non-proxied connections, when compiled with
694 SUPPORT_SOCKS and running with proxy logging enabled.
696 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
697 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
698 move the existing event to fire before the normal logging of message
699 failure so that custom logging is bracketed by normal logging.
701 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
702 msg:fail:internal event. Developement by Matthias Kurz.
704 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
705 far too small for todays use of crypto signatures stored there. Go all
706 the way to the max DNS message size of 64kB, even though this might be
707 overmuch for IOT constrained device use.
709 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
710 copy a string over itself. The library routine is documented as not
711 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
713 JH/32 For main options check_spool_space and check_inode_space, where the
714 platform supports 64b integers, support more than the previous 2^31 kB
715 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
716 the previous G, M, k.
718 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
719 $authenticated_fail_id variable on authentication failure. Previously
722 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
723 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
724 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
725 GNUTLS_SEC_PARAM_MEDIUM.
727 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
728 side. Previously we would continue as if no SNI had been received.
730 JH/36 Harden the handling of string-lists. When a list consisted of a sole
731 "<" character, which should be a list-separator specification, we walked
732 off past the nul-terimation.
734 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
735 causes) even when the retry time is not yet met. Previously they were
736 not, meaning that when (say) an account was over-quota and temp-rejecting,
737 and multiple senders' messages were queued, only one sender would get
738 notified on each configured delay_warning cycle.
740 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
742 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
743 cipher-suites, an error can be left on the stack even for a succeeding
744 accept; this results in impossible error messages when a later operation
747 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
748 return error codes indicating retry. Under TLS1.3 this becomes required.
750 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
751 it only wrote the new authenticators, resulting in a lack of tracking of
752 peer changes of ESMTP extensions until the next cache flush.
754 JH/41 Fix the loop reading a message header line to check for integer overflow,
755 and more-often against header_maxsize. Previously a crafted message could
756 induce a crash of the recive process; now the message is cleanly rejected.
758 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
759 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
765 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
766 When redis_servers is set to a list of > 1 element, and the Redis servers
767 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
768 case of MOVED into a DEFER case instead, thus moving the query onto the
769 next server in the list. For a cluster of N elements, all N servers must
770 be defined in redis_servers.
772 GF/02 Catch and remove uninitialized value warning in exiqsumm
773 Check for existence of @ARGV before looking at $ARGV[0]
775 JH/01 Replace the store_release() internal interface with store_newblock(),
776 which internalises the check required to safely use the old one, plus
777 the allocate and data copy operations duplicated in both (!) of the
778 extant use locations.
780 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
781 modifier. This matches the restriction on the commandline.
783 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
784 Previously only the last row was returned.
786 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
787 we assumed that tags in the header were well-formed, and parsed the
788 element content after inspecting only the first char of the tag.
789 Assumptions at that stage could crash the receive process on malformed
792 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
793 While running the DKIM ACL we operate on the Permanent memory pool so that
794 variables created with "set" persist to the DATA ACL. Also (at any time)
795 DNS lookups that fail create cache records using the Permanent pool. But
796 expansions release any allocations made on the current pool - so a dnsdb
797 lookup expansion done in the DKIM ACL releases the memory used for the
798 DNS negative-cache, and bad things result. Solution is to switch to the
799 Main pool for expansions.
800 While we're in that code, add checks on the DNS cache during store_reset,
801 active in the testsuite.
802 Problem spotted, and debugging aided, by Wolfgang Breyha.
804 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
805 When none of the hosts presented to a transport match an already-open
806 connection, close it and proceed with the list. Previously we would
807 queue the message. Spotted by Lena with Yahoo, probably involving
810 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
811 Previously a spurious "250 OK id=" response was appended to the proper
814 JH/08 The "support for" informational output now, which built with Content
815 Scanning support, has a line for the malware scanner interfaces compiled
816 in. Interface can be individually included or not at build time.
818 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
819 by the template makefile "src/EDITME". The "STREAM" support for an older
820 ClamAV interface method is removed.
822 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
823 rows affected is given instead).
825 JH/11 The runtime Berkeley DB library version is now additionally output by
826 "exim -d -bV". Previously only the compile-time version was shown.
828 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
829 SMTP connection. Previously, when one had more recipients than the
830 first, an abortive onward connection was made. Move to full support for
831 multiple onward connections in sequence, handling cutthrough connection
832 for all multi-message initiating connections.
834 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
835 routers. Previously, a multi-recipient message would fail to match the
836 onward-connection opened for the first recipient, and cause its closure.
838 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
839 a timeout on read on a GnuTLS initiating connection, resulting in the
840 initiating connection being dropped. This mattered most when the callout
841 was marked defer_ok. Fix to keep the two timeout-detection methods
844 JH/15 Relax results from ACL control request to enable cutthrough, in
845 unsupported situations, from error to silently (except under debug)
846 ignoring. This covers use with PRDR, frozen messages, queue-only and
849 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
851 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
852 metadata, resulting in a crash in free().
854 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
855 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
856 Broken also in d185889f4, with init system revamp.
858 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
859 Previously we abruptly closed the connection after reading a malware-
860 found indication; now we go on to read the "scan ok" response line,
863 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
864 ACL. Previously, a crash would result.
866 JH/19 Speed up macro lookups during configuration file read, by skipping non-
867 macro text after a replacement (previously it was only once per line) and
868 by skipping builtin macros when searching for an uppercase lead character.
870 JH/20 DANE support moved from Experimental to mainline. The Makefile control
871 for the build is renamed.
873 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
874 was allocated for every new TLS startup, meaning one per message. Fix
875 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
877 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
878 reported the original. Fix to report (as far as possible) the ACL
879 result replacing the original.
881 JH/23 Fix memory leak during multi-message connections using STARTTLS under
882 OpenSSL. Certificate information is loaded for every new TLS startup,
883 and the resources needed to be freed.
885 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
887 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
890 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
891 DATA response info to the (existing) per-recipient response info for
892 the "C=" log element. It can have useful tracking info from the
893 destination system. Patch from Simon Arlott.
895 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
896 length value. Previously this would segfault.
898 HS/02 Support Avast multiline protoocol, this allows passing flags to
899 newer versions of the scanner.
901 JH/28 Ensure that variables possibly set during message acceptance are marked
902 dead before release of memory in the daemon loop. This stops complaints
903 about them when the debug_store option is enabled. Discovered specifically
904 for sender_rate_period, but applies to a whole set of variables.
905 Do the same for the queue-runner and queue-list loops, for variables set
906 from spool message files. Do the same for the SMTP per-message loop, for
907 certain variables indirectly set in ACL operations.
909 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
910 as a multi-recipient message from a mailinglist manager). The coding had
911 an arbitrary cutoff number of characters while checking for more input;
912 enforced by writing a NUL into the buffer. This corrupted long / fast
913 input. The problem was exposed more widely when more pipelineing of SMTP
914 responses was introduced, and one Exim system was feeding another.
915 The symptom is log complaints of SMTP syntax error (NUL chars) on the
916 receiving system, and refused recipients seen by the sending system
917 (propating to people being dropped from mailing lists).
918 Discovered and pinpointed by David Carter.
920 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
921 replaced by the ${authresults } expansion.
923 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
925 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
926 allows proper process termination in container environments.
928 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
929 Previously the "final dot" had a newline after it; ensure it is CR,LF.
931 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
932 and "err_perm", deprecated since 4.83 when the RFC-defined words
933 "temperror" and "permerror" were introduced.
935 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
936 transport-filters or DKIM-signing. The restriction was lost in the
937 consolidation of verify-callout and delivery SMTP handling.
938 Extend the restriction to also cover ARC-signing.
940 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
941 in defer=pass mode supply a 450 to the initiator. Previously the message
944 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
945 tls_require_ciphers is used as before.
947 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
948 "pass_unscanned". Only tmpfails from the scanner are written to
949 the paniclog, as they may require admin intervention (permission
950 denied, license issues). Other scanner errors (like decompression
951 bombs) do not cause a paniclog entry.
953 JH/36 Fix reinitialisation of DKIM logging variable between messages.
954 Previously it was possible to log spurious information in receive log
957 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
958 triggered odd behaviour from Outlook Express clients.
960 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
963 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
964 since the IETF WG has not yet settled on that versus the original
965 "bare" representation.
967 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
968 Previously the millisecond value corrupted the output.
969 Fix also for syslog_pid=no and log_selector +pid, for which the pid
970 corrupted the output.
976 JH/01 Rework error string handling in TLS interface so that the caller in
977 more cases is responsible for logging. This permits library-sourced
978 string to be attached to addresses during delivery, and collapses
979 pairs of long lines into single ones.
981 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
982 during configuration. Wildcards are allowed and expanded.
984 JH/02 Rework error string handling in DKIM to pass more info back to callers.
985 This permits better logging.
987 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
988 do not close it down and have the child transport start it up again on
989 the passed-on TCP connection. Instead, proxy the child (and any
990 subsequent ones) for TLS via a unix-domain socket channel. Logging is
991 affected: the continued delivery log lines do not have any DNSSEC, TLS
992 Certificate or OCSP information. TLS cipher information is still logged.
994 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
995 identical IP addresses on different listening ports. Will also affect
998 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
999 add noisy ifdef guards to special-case this sillyness.
1000 Patch from Bernd Kuhls.
1002 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1003 than 255 are no longer allowed.
1005 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1006 Disable the session-cache too, which might reduce our load. Since we
1007 currrectly use a new context for every connection, both as server and
1008 client, there is no benefit for these.
1009 GnuTLS appears to not support tickets server-side by default (we don't
1010 call gnutls_session_ticket_enable_server()) but client side is enabled
1011 by default on recent versions (3.1.3 +) unless the PFS priority string
1014 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1015 <https://reproducible-builds.org/specs/source-date-epoch/>.
1017 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1018 the check for any unsuccessful recipients did not notice the limit, and
1019 erroneously found still-pending ones.
1021 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1022 MSG_MORE. Only in-clear (not on TLS connections).
1024 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1025 transport-filter is involved we can buffer the headers in memory for
1026 creating the signature, and read the spool data file once for the
1027 signature and again for transmission.
1029 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1030 4.77 as the kernel support then wasn't solid, having issues in 64bit
1031 mode. Now, it's been long enough. Add support for FreeBSD also.
1033 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1034 case where the routing stage had gathered several addresses to send to
1035 a host before calling the transport for the first, we previously failed
1036 to close down TLS in the old transport process before passing the TCP
1037 connection to the new process. The new one sent a STARTTLS command
1038 which naturally failed, giving a failed delivery and bloating the retry
1039 database. Investigation and fix prototype from Wolfgang Breyha.
1041 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1042 false-negatives in the check that the sender had not preempted a response
1043 or prompt from Exim (running as a server), due to that code's lack of
1044 awareness of the SMTP input buffering.
1046 PP/04 Add commandline_checks_require_admin option.
1047 Exim drops privileges sanely, various checks such as -be aren't a
1048 security problem, as long as you trust local users with access to their
1049 own account. When invoked by services which pass untrusted data to
1050 Exim, this might be an issue. Set this option in main configuration
1051 AND make fixes to the calling application, such as using `--` to stop
1054 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1055 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1056 responses to those, into a single TLS record each way (this usually means
1057 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1060 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1061 affects you only if you're dancing at the edge of the param size limits.
1062 If you are, and this message makes sense to you, then: raise the
1063 configured limit or use OpenSSL 1.1. Nothing we can do for older
1066 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1067 cmdline element to get the documented default one. Previously it was
1070 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1071 are unsuable for non-message-specific reasons - eg. connection timeout,
1072 banner-time rejection.
1074 JH/16 Fix logging of delivery remote port, when specified by router, under
1077 PP/06 Repair manualroute's ability to take options in any order, even if one
1078 is the name of a transport.
1081 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1083 JH/17 Change the list-building routines interface to use the expanding-string
1084 triplet model, for better allocation and copying behaviour.
1086 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1087 Previously it was constructed the first time a possibly-matching string
1088 was met in the configuration file input during startup; now it is done
1091 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1092 compatible one, to avoid the (poorly documented) possibility of a config
1093 file in the working directory redirecting the DB files, possibly correpting
1094 some existing file. CVE-2017-10140 assigned for BDB.
1096 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1097 cache-hot. Previously, although the result was properly cached, the
1098 initial verify call returned a defer.
1100 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1101 the main verify for receipient in uncached-mode.
1103 JH/22 Retire historical build files to an "unsupported" subdir. These are
1104 defined as "ones for which we have no current evidence of testing".
1106 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1107 if present. Previously it was ignored.
1109 JH/24 Start using specified-initialisers in C structure init coding. This is
1110 a C99 feature (it's 2017, so now considered safe).
1112 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1113 if was a fixed-sized field and bitmask ops via macros; it is now more
1116 PP/07 GitHub PR 56: Apply MariaDB build fix.
1117 Patch provided by Jaroslav Škarvada.
1119 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1120 during Coverity cleanups [4.87 JH/47]
1121 Diagnosis and fix provided by Michael Fischer v. Mollard.
1123 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1124 the right size to place the terminating semicolon on its own folded
1125 line, the header hash was calculated to an incorrect value thanks to
1126 the (relaxed) space the fold became.
1128 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1129 and confused the parent.
1131 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1132 which could crash as a result. This could lead to undeliverable messages.
1134 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1137 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1138 matters on fast-turnover and PID-randomising systems, which were getting
1139 out-of-order delivery.
1141 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1142 a possibly-overlapping copy. The symptom was that "Remote host closed
1143 connection in response to HELO" was logged instead of the actual 4xx
1146 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1147 Previously only that bufferd was discarded, resulting in SYMTP command
1150 JH/32 DKIM: when a message has multiple signatures matching an identity given
1151 in dkim_verify_signers, run the dkim acl once for each. Previously only
1152 one run was done. Bug 2189.
1154 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1155 "panic the current process" to "deliberately defer". The panic log is
1156 still written with the problem list name; the mail and reject logs now
1157 get a temp-reject line for the message that was being handled, saying
1158 something like "domains check lookup or other defer". The SMTP 451
1159 message is still "Temporary local problem".
1161 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1162 A crafted sequence of BDAT commands could result in in-use memory beeing
1163 freed. CVE-2017-16943.
1165 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1166 from SMTP input. Previously it was always done; now only done for DATA
1167 and not BDAT commands. CVE-2017-16944.
1169 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1170 to the message (such as an overlong header line). Previously this was
1171 not done and we did not exit BDAT mode. Followon from the previous item
1172 though a different problem.
1178 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1179 than -2003 did; needs libidn2 in addition to libidn.
1181 JH/02 The path option on a pipe transport is now expanded before use.
1183 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1184 Patch provided by "Björn", documentation fix added too.
1186 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1187 missing a wire-to-host endian conversion.
1189 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1190 close after a BDAT command line could be taken as a following command,
1191 giving a synch failure. Fix by only checking for synch immediately
1192 before acknowledging the chunk.
1194 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1195 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1196 macro. Patches provided by Josh Soref.
1198 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1199 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1200 by IANA as a service extension. However, John Klensin suggests that we
1203 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1204 the dkim code may be unix-mode line endings rather than smtp wire-format
1205 CRLF, so prepend a CR to any bare LF.
1207 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1208 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1210 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1211 the first were themselves being wrongly included in the feed into dkim
1212 processing; with most chunk sizes in use this resulted in an incorrect
1213 body hash calculated value.
1215 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1216 DKIM signature block, for verification. Although advised against by
1217 standards it is specifically not ruled illegal.
1219 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1221 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1222 missing a body hash (the bh= tag).
1224 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1225 It seems that HAProxy sends the Proxy Protocol information in clear and
1226 only then does a TLS startup, so do the same.
1228 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1229 TCP connections (such as for Spamd) unless the daemon successfully set
1230 Fast Open mode on its listening sockets. This fixes breakage seen on
1231 too-old kernels or those not configured for Fast Open, at the cost of
1232 requiring both directions being enabled for TFO, and TFO never being used
1233 by non-daemon-related Exim processes.
1235 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1236 endings, at least on the first header line. Try to canonify any that get
1237 past that check, despite the cost.
1239 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1240 now limited to an arbitrary five deep, while parsing addresses with the
1241 strip_excess_angle_brackets option enabled.
1243 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1244 instead leave the unprompted TLS handshake in socket buffer for the
1245 TLS library to consume.
1247 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1249 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1251 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1252 region in one operation, for speed, and this leaves assigned pointers
1253 dangling. Add checks run only under the testsuite which checks all
1254 variables at a store-reset and panics on a dangling pointer; add code
1255 explicitly nulling out all the variables discovered. Fixes one known
1256 bug: a transport crash, where a dangling pointer for $sending_ip_address
1257 originally assigned in a verify callout, is re-used.
1259 PP/06 Drop '.' from @INC in various Perl scripts.
1261 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1263 PP/08 Reduce a number of compilation warnings under clang; building with
1264 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1265 should be warning-free.
1267 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1269 HS/01 Fix portability problems introduced by PP/08 for platforms where
1270 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1272 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1273 chunk. This allows us to accept broken chunked messages. We need a more
1274 general solution here.
1276 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1277 already-broken messages in the queue.
1279 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1281 JH/19 Fix reference counting bug in routing-generated-address tracking.
1287 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1288 supports it and a size is available (ie. the sending peer gave us one).
1290 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1291 years of being deprecated). The replacements are the ACLs
1292 acl_smtp_mime and acl_not_smtp_mime.
1294 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1295 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1296 or even an in-clear connection were permitted. Now, if the host lookup
1297 was dnssec and dane was requested then the host is only used if the
1298 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1299 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1300 if one fails this test.
1301 This means that a poorly-configured remote DNS will make it incommunicado;
1302 but it protects against a DNS-interception attack on it.
1304 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1305 non-noisy when a race steals the message being considered.
1307 JH/05 If main configuration option tls_certificate is unset, generate a
1308 self-signed certificate for inbound TLS connections.
1310 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1311 in rewrites and routers.
1313 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1314 and logged a warning sing 4.83; now they are a configuration file error.
1316 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1317 (lacking @domain). Apply the same qualification processing as RCPT.
1319 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1321 JH/10 Support ${sha256:} applied to a string (as well as the previous
1324 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1325 a cutthrough deliver is pending, as we always want to make a connection.
1326 This also avoids re-routing the message when later placing the cutthrough
1327 connection after a verify cache hit.
1328 Do not update it with the verify result either.
1330 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1331 when routing results in more than one destination address.
1333 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1334 signing (which inhibits the cutthrough capability). Previously only
1335 the presence of an option was tested; now an expansion evaluating as
1336 empty is permissible (obviously it should depend only on data available
1337 when the cutthrough connection is made).
1339 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1340 the relevant preceding SMTP command did not note the pipelining mode.
1342 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1343 Previously they were not counted.
1345 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1346 as one having no matching records. Previously we deferred the message
1347 that needed the lookup.
1349 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1350 distinguished as "(=".
1352 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1353 for missing MX records. Previously it only worked for missing A records.
1355 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1357 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1358 after the data-go-ahead and data-ack. Patch from Jason Betts.
1360 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1361 even for a "none" policy. Patch from Tony Meyer.
1363 JH/22 Fix continued use of a connection for further deliveries. If a port was
1364 specified by a router, it must also match for the delivery to be
1367 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1368 When one of the recipients of a message was unsuitable for the connection
1369 (has no matching addresses), we lost track of needing to mark it
1370 deferred. As a result mail would be lost.
1372 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1374 JH/25 Decoding ACL controls is now done using a binary search; the source code
1375 takes up less space and should be simpler to maintain. Merge the ACL
1376 condition decode tables also, with similar effect.
1378 JH/26 Fix problem with one_time used on a redirect router which returned the
1379 parent address unchanged. A retry would see the parent address marked as
1380 delivered, so not attempt the (identical) child. As a result mail would
1383 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1384 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1385 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1388 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1389 Main config options check_{log,spool}_{inodes,space} are now
1390 100 inodes, 10MB unless set otherwise in the configuration.
1392 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1393 Previously it only applied to the main-section connection policy
1396 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1398 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1399 by me. Added RFC7919 DH primes as an alternative.
1401 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1402 are not in the system include path.
1404 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1405 GnuTLS, when a session startup failed (eg because the client disconnected)
1406 Exim did stdio operations after fclose. This was exposed by a recent
1407 change which nulled out the file handle after the fclose.
1409 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1410 signed directly by the cert-signing cert, rather than an intermediate
1411 OCSP-signing cert. This is the model used by LetsEncrypt.
1413 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1415 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1416 an incoming connection.
1418 HS/02 Bug 1802: Do not half-close the connection after sending a request
1421 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1422 fallback to "prime256v1".
1424 JH/34 SECURITY: Use proper copy of DATA command in error message.
1425 Could leak key material. Remotely exploitable. CVE-2016-9963.
1431 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1432 and 3.4.4 - once the server is enabled to respond to an OCSP request
1433 it does even when not requested, resulting in a stapling non-aware
1434 client dropping the TLS connection.
1436 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1437 support variable-length bit vectors. No functional change.
1439 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1440 The I= interface field on outgoing lines is now after the H= remote
1441 host field, same as incoming lines. There is a separate
1442 outgoing_interface log selector which allows you to disable the
1445 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1446 If not running log_selector +smtp_connection the mainlog would be held
1447 open indefinitely after a "too many connections" event, including to a
1448 deleted file after a log rotate. Leave the per net connection logging
1449 leaving it open for efficiency as that will be quickly detected by the
1450 check on the next write.
1452 HS/01 Bug 1671: Fix post transport crash.
1453 Processing the wait-<transport> messages could crash the delivery
1454 process if the message IDs didn't exist for some reason. When
1455 using 'split_spool_directory=yes' the construction of the spool
1456 file name failed already, exposing the same netto behaviour.
1458 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1459 mime_regex ACL conditions.
1461 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1462 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1463 to HELO, local diagnostic string.
1465 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1466 log line to debug. Even when configured with a tls authenticator many
1467 client connections are expected to not authenticate in this way, so
1468 an authenticate fail is not an error.
1470 HS/02 Add the Exim version string to the process info. This way exiwhat
1471 gives some more detail about the running daemon.
1473 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1474 matter for fast-change records such as DNSBLs.
1476 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1477 retry record, even if constant. There may be multiple transports with
1478 different interface settings and the retry behaviour needs to be kept
1481 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1483 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1485 JH/10 Bug 840: fix log_defer_output option of pipe transport
1487 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1488 hosts_randomize. This matters a lot when combined with mua_wrapper.
1490 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1491 ${quote_pgsql:<string>} operator.
1493 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1495 JH/14 Change header file name for freeradius-client. Relevant if compiling
1496 with Radius support; from the Gentoo tree and checked under Fedora.
1498 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1500 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1501 option transport_filter, meaning no filtering.
1503 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1505 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1506 defaults to "*" (all hosts). The variable is now available when not built
1507 with TLS, default unset, mainly to enable keeping the testsuite sane.
1508 If a server certificate is not supplied (via tls_certificate) an error is
1509 logged, and clients will find TLS connections fail on startup. Presumably
1510 they will retry in-clear.
1511 Packagers of Exim are strongly encouraged to create a server certificate
1512 at installation time.
1514 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1515 with the $config_file variable.
1517 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1518 in transport context, after the attempt, and per-recipient. The latter type
1519 is per host attempted. The event data is the error message, and the errno
1520 information encodes the lookup type (A vs. MX) used for the (first) host,
1521 and the trailing two digits of the smtp 4xx response.
1523 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1524 to write to mainlog (or rejectlog, paniclog) in the window between file
1525 creation and permissions/ownership being changed. Particularly affects
1526 installations where exicyclog is run as root, rather than exim user;
1527 result is that the running daemon panics and dies.
1529 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1531 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1532 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1533 "pri" and "weight". Note that the previous implicit priority given by the
1534 list order is no longer honoured.
1536 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1537 for DKIM processing.
1539 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1540 by defining SUPPORT_SOCKS.
1542 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1543 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1544 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1545 variables are renamed to proxy_{local,external}_{address,port}.
1547 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1548 for a build by defining SUPPORT_I18N
1550 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1551 of the query string, and make ${quote_redis:} do that quoting.
1553 JH/29 Move Events support from Experimental to mainline, enabled by default
1554 and removable for a build by defining DISABLE_EVENT.
1556 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1558 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1559 cached by the daemon.
1561 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1562 by defining LOOKUP_REDIS. The libhiredis library is required.
1564 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1565 keys are given for lookup.
1567 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1568 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1569 only supported when built with TLS support. The PolarSSL SHA routines
1570 are still used when the TLS library is too old for convenient support.
1572 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1573 openssl_options), for security. OpenSSL forces this from version 1.1.0
1574 server-side so match that on older versions.
1576 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1577 allocation for $value could be released as the expansion processing
1578 concluded, but leaving the global pointer active for it.
1580 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1581 and to use the domains and local_parts ACL conditions.
1583 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1584 incorrectly not doubled on cutthrough transmission, hence seen as a
1585 body-termination at the receiving system - resulting in truncated mails.
1586 Commonly the sender saw a TCP-level error, and retransmitted the message
1587 via the normal store-and-forward channel. This could result in duplicates
1588 received - but deduplicating mailstores were liable to retain only the
1589 initial truncated version.
1591 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1593 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1595 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1596 we're in there, support oversigning also; bug 1309.
1598 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1600 HS/04 Add support for keep_environment and add_environment options.
1602 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1603 either intentional arithmetic overflow during PRNG, or testing config-
1606 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1607 delivery resulted in actual delivery. Cancel cutthrough before DATA
1610 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1611 rejects a recipient: pass the reject to the originator.
1613 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1614 Many were false-positives and ignorable, but it's worth fixing the
1617 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1618 for the new environment-manipulation done at startup. Move the routines
1619 from being local to tls.c to being global via the os.c file.
1621 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1622 an extract embedded as result-arg for a map, the first arg for extract
1623 is unavailable so we cannot tell if this is a numbered or keyed
1624 extraction. Accept either.
1630 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1633 JH/02 The smtp transport option "multi_domain" is now expanded.
1635 JH/03 The smtp transport now requests PRDR by default, if the server offers
1638 JH/04 Certificate name checking on server certificates, when exim is a client,
1639 is now done by default. The transport option tls_verify_cert_hostnames
1640 can be used to disable this per-host. The build option
1641 EXPERIMENTAL_CERTNAMES is withdrawn.
1643 JH/05 The value of the tls_verify_certificates smtp transport and main options
1644 default to the word "system" to access the system default CA bundle.
1645 For GnuTLS, only version 3.0.20 or later.
1647 JH/06 Verification of the server certificate for a TLS connection is now tried
1648 (but not required) by default. The verification status is now logged by
1649 default, for both outbound TLS and client-certificate supplying inbound
1652 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1655 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1656 Status Notification (bounce) messages are now MIME format per RFC 3464.
1657 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1658 under the control of the dsn_advertise_hosts option, and routers may
1659 have a dsn_lasthop option.
1661 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1662 default, modifiable by a malware= option. The list separator for
1663 the options can now be changed in the usual way. Bug 68.
1665 JH/10 The smtp_receive_timeout main option is now expanded before use.
1667 JH/11 The incoming_interface log option now also enables logging of the
1668 local interface on delivery outgoing connections.
1670 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1671 if the interface and destination host and port all match.
1673 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1676 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1677 Patch from Andrew Lewis.
1679 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1680 now supports optional time-restrictions, weighting, and priority
1681 modifiers per server. Patch originally by <rommer@active.by>.
1683 JH/16 The spamd_address main option now supports a mixed list of local
1684 and remote servers. Remote servers can be IPv6 addresses, and
1685 specify a port-range.
1687 JH/17 Bug 68: The spamd_address main option now supports an optional
1688 timeout value per server.
1690 JH/18 Bug 1581: Router and transport options headers_add/remove can
1691 now have the list separator specified.
1693 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1696 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1699 JH/21 Support for the A6 type of dns record is withdrawn.
1701 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1702 rather than the verbs used.
1704 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1705 from 255 to 1024 chars.
1707 JH/24 Verification callouts now attempt to use TLS by default.
1709 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1710 are generic router options now. The defaults didn't change.
1712 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1713 Original patch from Alexander Shikoff, worked over by JH.
1715 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1716 files if ZCAT_COMMAND is not executable.
1718 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1720 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1722 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1723 Normally benign, it bites when the pair was led to by a CNAME;
1724 modern usage is to not canonicalize the domain to a CNAME target
1725 (and we were inconsistent anyway for A-only vs AAAA+A).
1727 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1729 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1730 when evaluating $sender_host_dnssec.
1732 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1733 $sender_helo_dnssec variable.
1735 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1737 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1739 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1741 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1742 documented as working, but never had. Support all but $spam_report.
1744 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1745 added for tls authenticator.
1747 HS/03 Add perl_taintmode main config option
1753 TL/01 When running the test suite, the README says that variables such as
1754 no_msglog_check are global and can be placed anywhere in a specific
1755 test's script, however it was observed that placement needed to be near
1756 the beginning for it to behave that way. Changed the runtest perl
1757 script to read through the entire script once to detect and set these
1758 variables, reset to the beginning of the script, and then run through
1759 the script parsing/test process like normal.
1761 TL/02 The BSD's have an arc4random API. One of the functions to induce
1762 adding randomness was arc4random_stir(), but it has been removed in
1763 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1764 function when detected.
1766 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1767 cause callback expansion.
1769 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1770 syntax errors in an expansion can be treated as a string instead of
1771 logging or causing an error, due to the internal use of bool_lax
1772 instead of bool when processing it.
1774 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1775 server certificates when making smtp deliveries.
1777 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1779 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1781 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1783 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1784 Merged patch from Sebastian Wiedenroth.
1786 JH/05 Fix results-pipe from transport process. Several recipients, combined
1787 with certificate use, exposed issues where response data items split
1788 over buffer boundaries were not parsed properly. This eventually
1789 resulted in duplicates being sent. This issue only became common enough
1790 to notice due to the introduction of connection certificate information,
1791 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1793 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1794 size buffer was used, resulting in syntax errors when an expansion
1797 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1798 version 3.3.6 or later.
1800 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1801 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1802 both become "event_action", the variables become $event_name, $event_data
1803 and $event_defer_errno. There is a new variable $verify_mode, usable in
1804 routers, transports and related events. The tls:cert event is now also
1805 raised for inbound connections, if the main configuration event_action
1808 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1809 early OCSP support, but no stapling (appears to be less than 1.0.0).
1811 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1812 server certificate names available under the smtp transport option
1813 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1816 JH/10 Time-related extraction expansions from certificates now use the main
1817 option "timezone" setting for output formatting, and are consistent
1818 between OpenSSL and GnuTLS compilations. Bug 1541.
1820 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1821 encoded parameter in the incoming message. Bug 1558.
1823 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1824 include certificate info, eximon was claiming there were spoolfile
1827 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1829 JH/14 Log delivery-related information more consistently, using the sequence
1830 "H=<name> [<ip>]" wherever possible.
1832 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1833 are problematic for Debian distribution, omit them from the release
1836 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1838 JH/16 Fix string representation of time values on 64bit time_t architectures.
1841 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1842 output list separator was used.
1847 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1848 checkers that were complaining about end of non-void function with no
1851 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1852 This was a regression introduced in 4.83 by another bugfix.
1854 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1856 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1857 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1863 TF/01 Correctly close the server side of TLS when forking for delivery.
1865 When a message was received over SMTP with TLS, Exim failed to clear up
1866 the incoming connection properly after forking off the child process to
1867 deliver the message. In some situations the subsequent outgoing
1868 delivery connection happened to have the same fd number as the incoming
1869 connection previously had. Exim would try to use TLS and fail, logging
1870 a "Bad file descriptor" error.
1872 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1873 utilities have not been installed.
1875 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1876 temporary space as the ACL may create new global variables.
1878 TL/01 LDAP support uses per connection or global context settings, depending
1879 upon the detected version of the libraries at build time.
1881 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1882 to extract and use the src ip:port in logging and expansions as if it
1883 were a direct connection from the outside internet. PPv2 support was
1884 updated based on HAProxy spec change in May 2014.
1886 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1888 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1889 Properly escape header and check for NULL return.
1891 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1894 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1896 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1897 characters in header names, implemented as a verify condition.
1898 Contributed by Michael Fischer v. Mollard.
1900 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1901 results permerror and temperror. Previous values are deprecated but
1902 still accepted. In a future release, err_perm and err_temp will be
1903 completely removed, which will be a backward incompatibility if the
1904 ACL tests for either of these two old results. Patch contributed by
1905 user bes-internal on the mailing list.
1907 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1909 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1910 selectors, in both main and reject logs.
1912 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1915 JH/07 Add malware type "sock" for talking to simple daemon.
1917 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1919 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1920 routers/transports under cutthrough routing.
1922 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1923 numbers. Touch up "bool" conditional to keep the same definition.
1925 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1927 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1929 JH/12 Expand items in router/transport headers_add or headers_remove lists
1930 individually rather than the list as a whole. Bug 1452.
1932 Required for reasonable handling of multiple headers_ options when
1933 they may be empty; requires that headers_remove items with embedded
1934 colons must have them doubled (or the list-separator changed).
1936 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1937 view the policy declared in the DMARC record. Currently, $dmarc_status
1938 is a combined value of both the record presence and the result of the
1941 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1943 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1944 dnslookup router and the smtp transport (applying to the forward
1947 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1948 of ldap servers used for a specific lookup. Patch provided by Heiko
1951 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1952 New variable $lookup_dnssec_authenticated for observability.
1954 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1955 Patch submitted by Lars Timman.
1957 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1959 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1960 Requires trusted mode and valid format message id, aborts otherwise.
1961 Patch contributed by Heiko Schlichting.
1963 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1964 certextract with support for various fields. Bug 1358.
1966 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1967 is requested by default, modifiable by smtp transport option
1970 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1971 operate on certificate variables to give certificate fingerprints
1972 Also new ${sha256:cert_variable}.
1974 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1976 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1977 Christian Aistleitner.
1979 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1981 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1982 file. Patch from Wolfgang Breyha.
1984 JH/25 Expand the coverage of the delivery $host and $host_address to
1985 client authenticators run in verify callout. Bug 1476.
1987 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1988 align with daemon_smtp_ports. Bug 72.
1990 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1991 support and error reporting did not work properly.
1993 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1994 and is readable. Patch from Andrew Colin Kissa.
1996 TL/14 Enhance documentation of ${run expansion and how it parses the
1997 commandline after expansion, particularly in the case when an
1998 unquoted variable expansion results in an empty value.
2000 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2002 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2003 by renaming away from it. Addresses GH issue 15, reported by
2006 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2008 TL/15 SECURITY: prevent double expansion in math comparison functions
2009 (can expand unsanitized data). Not remotely exploitable.
2016 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2018 PP/02 Make -n do something, by making it not do something.
2019 When combined with -bP, the name of an option is not output.
2021 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2024 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2025 $sender_host_name and config options to manage this, and basic check
2028 PP/05 DSCP support for outbound connections and control modifier for inbound.
2030 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2031 (Only plugin which currently uses this is kerberos4, which nobody should
2032 be using, but we should make it available and other future plugins might
2033 conceivably use it, even though it would break NAT; stuff *should* be
2034 using channel bindings instead).
2036 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2037 name; added for Sendmail compatibility; requires admin caller.
2038 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2039 just ignore it); requires trusted caller.
2040 Also parse but ignore: -Ac -Am -X<logfile>
2043 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2045 TL/02 Add +smtp_confirmation as a default logging option.
2047 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2048 Patch by Magnus Holmgren from 2007-02-20.
2050 TL/04 Bugzilla 1281 - Spec typo.
2051 Bugzilla 1283 - Spec typo.
2052 Bugzilla 1290 - Spec grammar fixes.
2054 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2056 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2058 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2059 mailing list by by Dmitry Isaikin.
2061 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2063 JH/02 Support "G" suffix to numbers in ${if comparisons.
2065 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2067 NM/01 Bugzilla 1197 - Spec typo
2068 Bugzilla 1196 - Spec examples corrections
2070 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2072 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2073 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2076 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2077 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2079 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2080 "acl {{name}{arg}...}", and optional args on acl condition
2083 JH/05 Permit multiple router/transport headers_add/remove lines.
2085 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2087 JH/07 Avoid using a waiting database for a single-message-only transport.
2088 Performance patch from Paul Fisher. Bugzilla 1262.
2090 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2093 JH/09 Add $headers_added variable, with content from use of ACL modifier
2094 add_header (but not yet added to the message). Bugzilla 199.
2096 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2097 Pulled from Bugzilla 817 by Wolfgang Breyha.
2099 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2101 (nb: this is the same fix as in Exim 4.80.1)
2103 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2106 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2109 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2111 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2112 advertise SMTP AUTH mechanism to us, instead of a generic
2113 protocol violation error. Also, make Exim more robust to bad
2114 data from the Dovecot auth socket.
2116 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2118 When a queue runner is handling a message, Exim first routes the
2119 recipient addresses, during which it prunes them based on the retry
2120 hints database. After that it attempts to deliver the message to
2121 any remaining recipients. It then updates the hints database using
2124 So if a recipient address works intermittently, it can get repeatedly
2125 deferred at routing time. The retry hints record remains fresh so the
2126 address never reaches the final cutoff time.
2128 This is a fairly common occurrence when a user is bumping up against
2129 their storage quota. Exim had some logic in its local delivery code
2130 to deal with this. However it did not apply to per-recipient defers
2131 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2133 This change adds a proper retry rule check during routing so that the
2134 final cutoff time is checked against the message's age. We only do
2135 this check if there is an address retry record and there is not a
2136 domain retry record; this implies that previous attempts to handle
2137 the address had the retry_use_local_parts option turned on. We use
2138 this as an approximation for the destination being like a local
2139 delivery, as in LMTP.
2141 I suspect this new check makes the old local delivery cutoff check
2142 redundant, but I have not verified this so I left the code in place.
2144 TF/02 Correct gecos expansion when From: is a prefix of the username.
2146 Test 0254 submits a message to Exim with the header
2150 When I ran the test suite under the user fanf2, Exim expanded
2151 the header to contain my full name, whereas it should have added
2152 a Resent-Sender: header. It erroneously treats any prefix of the
2153 username as equal to the username.
2155 This change corrects that bug.
2157 GF/01 DCC debug and logging tidyup
2158 Error conditions log to paniclog rather than rejectlog.
2159 Debug lines prefixed by "DCC: " to remove any ambiguity.
2161 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2163 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2164 Bug spotted by Jeremy Harris; was flawed since initial commit.
2165 Would have resulted in OCSP responses post-SNI triggering an Exim
2166 NULL dereference and crash.
2168 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2170 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2171 Bug detection, analysis and fix by Samuel Thibault.
2172 Bugzilla 1331, Debian bug #698092.
2174 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2176 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2177 Server implementation by Todd Lyons, client by JH.
2178 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2179 config variable "prdr_enable" controls whether the server
2180 advertises the facility. If the client requests PRDR a new
2181 acl_data_smtp_prdr ACL is called once for each recipient, after
2182 the body content is received and before the acl_smtp_data ACL.
2183 The client is controlled by both of: a hosts_try_prdr option
2184 on the smtp transport, and the server advertisement.
2185 Default client logging of deliveries and rejections involving
2186 PRDR are flagged with the string "PRDR".
2188 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2189 fclose(). Diagnosis by Todd Lyons.
2191 PP/17 Update configure.default to handle IPv6 localhost better.
2192 Patch by Alain Williams (plus minor tweaks).
2195 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2196 This is now consistent with GnuTLS, and is now documented: the
2197 previous undocumented portable approach to treating the option as
2198 unset was to force an expansion failure. That still works, and
2199 an empty string is now equivalent.
2201 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2202 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2203 not performing validation itself.
2205 PP/20 Added force_command boolean option to pipe transport.
2206 Patch from Nick Koston, of cPanel Inc.
2208 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2211 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2213 PP/21 Fix eximon continuous updating with timestamped log-files.
2214 Broken in a format-string cleanup in 4.80, missed when I repaired the
2215 other false fix of the same issue.
2216 Report and fix from Heiko Schlichting.
2219 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2220 Report from Prashanth Katuri.
2222 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2223 It's SecureTransport, so affects any MacOS clients which use the
2224 system-integrated TLS libraries, including email clients.
2226 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2227 using a MIME ACL for non-SMTP local injection.
2228 Report and assistance in diagnosis by Warren Baker.
2230 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2232 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2234 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2235 last id that failed so it may be referenced in subsequent ACL's.
2237 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2240 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2241 ldap library initialization, allowing self-signed CA's to be
2242 used. Also properly sets require_cert option later in code by
2243 using NULL (global ldap config) instead of ldap handle (per
2244 session). Bug diagnosis and testing by alxgomz.
2246 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2247 the src/util/ subdirectory.
2249 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2250 renamed to Transport Post Delivery Action by Jeremy Harris, as
2253 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2254 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2255 redis_servers = needs to be configured which will be used by the redis
2256 lookup. Patch from Warren Baker, of The Packet Hub.
2258 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2260 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2261 hostname or reverse DNS when processing a host list. Used suggestions
2262 from multiple comments on this bug.
2264 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2266 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2267 Missed a few lines, added it to make the runtest require no keyboard
2270 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2271 contains upper case chars. Make router use caseful_local_part.
2273 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2274 support when GnuTLS has been built with p11-kit.
2280 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2282 This, or similar/improved, will also be change PP/11 of 4.82.
2288 PP/01 Handle short writes when writing local log-files.
2289 In practice, only affects FreeBSD (8 onwards).
2290 Bugzilla 1053, with thanks to Dmitry Isaikin.
2292 NM/01 Bugzilla 949 - Documentation tweak
2294 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2297 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2299 PP/02 Implemented gsasl authenticator.
2301 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2303 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2304 `pkg-config foo` for cflags/libs.
2306 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2307 with rest of GSASL and with heimdal_gssapi.
2309 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2310 `pkg-config foo` for cflags/libs for the TLS implementation.
2312 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2313 properties get this fed in as external SSF. A number of robustness
2314 and debugging improvements to the cyrus_sasl authenticator.
2316 PP/08 cyrus_sasl server now expands the server_realm option.
2318 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2319 Patch by Jeremy Harris.
2321 PP/10 Added dbmjz lookup type.
2323 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2325 PP/12 MAIL args handles TAB as well as SP, for better interop with
2326 non-compliant senders.
2327 Analysis and variant patch by Todd Lyons.
2329 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2330 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2331 Patch from Dirk Mueller <dmueller@suse.com>
2333 PP/13 tls_peerdn now print-escaped for spool files.
2334 Observed some $tls_peerdn in wild which contained \n, which resulted
2335 in spool file corruption.
2337 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2338 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2339 or write after TLS renegotiation, which otherwise led to messages
2342 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2343 as a tracking header (ie: a signed header comes before the signature).
2344 Patch from Wolfgang Breyha.
2346 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2347 comma-sep list; embedded commas doubled.
2349 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2351 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2353 Report and patch from Dmitry Banschikov.
2355 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2356 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2357 failures. We appear to now support TLS1.1+ with Exim.
2359 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2360 lets Exim select keys and certificates based upon TLS SNI from client.
2361 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2362 before an outbound SMTP session. New log_selector, +tls_sni.
2364 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2365 NULL dereference. Report and patch from Alun Jones.
2367 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2368 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2369 Not seeing resolver debug output on NetBSD, but suspect this is a
2370 resolver implementation change.
2372 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2373 Left warnings. Added "eximon gdb" invocation mode.
2375 PP/21 Defaulting "accept_8bitmime" to true, not false.
2377 PP/22 Added -bw for inetd wait mode support.
2379 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2380 locate the relevant includes and libraries. Made this the default.
2382 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2383 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2385 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2386 This may cause build issues on older platforms.
2388 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2389 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2390 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2391 Added SNI support via GnuTLS too.
2392 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2394 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2396 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2397 Applied second patch from Janne, implementing suggestion to default
2398 multiple-strings-in-record handling to match SPF spec.
2400 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2402 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2403 read-only, out of scope).
2404 Patch from Wolfgang Breyha, report from Stuart Northfield.
2406 PP/29 Fix three issues highlighted by clang analyser static analysis.
2407 Only crash-plausible issue would require the Cambridge-specific
2408 iplookup router and a misconfiguration.
2409 Report from Marcin Mirosław.
2411 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2413 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2414 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2415 As part of this, removing so much warning spew let me fix some minor
2416 real issues in debug logging.
2418 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2419 assignment on my part. Fixed.
2421 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2422 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2423 Janne Snabb (who went above and beyond: thank you).
2425 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2426 string otherwise requires a connection and a bunch more work and it's
2427 relatively easy to get wrong. Should also expose TLS library linkage
2430 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2431 64-bit ${eval} (JH/03).
2433 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2434 GNU libc to support some of the 64-bit stuff, should not lead to
2435 conflicts. Defined before os.h is pulled in, so if a given platform
2436 needs to override this, it can.
2438 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2439 protection layer was required, which is not implemented.
2440 Bugzilla 1254, patch from Wolfgang Breyha.
2442 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2443 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2444 tls_dhparam take prime identifiers. Also unbreak combination of
2445 OpenSSL+DH_params+TLSSNI.
2447 PP/39 Disable SSLv2 by default in OpenSSL support.
2453 PP/01 Solaris build fix for Oracle's LDAP libraries.
2454 Bugzilla 1109, patch from Stephen Usher.
2456 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2458 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2461 TF/02 Fix a couple more cases where we did not log the error message
2462 when unlink() failed. See also change 4.74-TF/03.
2464 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2465 lock up or crash if it happened to be inside a call to libc when it
2466 got a SIGUSR1 from exiwhat.
2468 The SIGUSR1 handler appends the current process status to the process
2469 log which is later printed by exiwhat. It used to use the general
2470 purpose logging code to do this, but several functions it calls are
2471 not safe for signals.
2473 The new output code in the SIGUSR1 handler is specific to the process
2474 log, and simple enough that it's easy to inspect for signal safety.
2475 Removing some special cases also simplifies the general logging code.
2476 Removing the spurious timestamps from the process log simplifies
2479 TF/04 Improved ratelimit ACL condition.
2481 The /noupdate option has been deprecated in favour of /readonly which
2482 has clearer semantics. The /leaky, /strict, and /readonly update modes
2483 are mutually exclusive. The update mode is no longer included in the
2484 database key; it just determines when the database is updated. (This
2485 means that when you upgrade Exim will forget old rate measurements.)
2487 Exim now checks that the per_* options are used with an update mode that
2488 makes sense for the current ACL. For example, when Exim is processing a
2489 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2490 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2491 must specify per_mail/readonly. If you omit the update mode it defaults to
2492 /leaky where that makes sense (as before) or /readonly where required.
2494 The /noupdate option is now undocumented but still supported for
2495 backwards compatibility. It is equivalent to /readonly except that in
2496 ACLs where /readonly is required you may specify /leaky/noupdate or
2497 /strict/noupdate which are treated the same as /readonly.
2499 A useful new feature is the /count= option. This is a generalization
2500 of the per_byte option, so that you can measure the throughput of other
2501 aggregate values. For example, the per_byte option is now equivalent
2502 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2504 The per_rcpt option has been generalized using the /count= mechanism
2505 (though it's more complicated than the per_byte equivalence). When it is
2506 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2507 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2508 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2509 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2510 Note that using per_rcpt with a non-readonly update mode in more than
2511 one ACL will cause the recipients to be double-counted. (The per_mail
2512 and per_byte options don't have this problem.)
2514 The handling of very low rates has changed slightly. If the computed rate
2515 is less than the event's count (usually one) then this event is the first
2516 after a long gap. In this case the rate is set to the same as this event's
2517 count, so that the first message of a spam run is counted properly.
2519 The major new feature is a mechanism for counting the rate of unique
2520 events. The new per_addr option counts the number of different
2521 recipients that someone has sent messages to in the last time period. It
2522 behaves like per_rcpt if all the recipient addresses are different, but
2523 duplicate recipient addresses do not increase the measured rate. Like
2524 the /count= option this is a general mechanism, so the per_addr option
2525 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2526 example, measure the rate that a client uses different sender addresses
2527 with the options per_mail/unique=$sender_address. There are further
2528 details in the main documentation.
2530 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2532 TF/06 Removed a few PCRE remnants.
2534 TF/07 Automatically extract Exim's version number from tags in the git
2535 repository when doing development or release builds.
2537 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2538 Bugzilla 879. Patch from Paul Fisher.
2540 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2541 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2544 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2546 PP/05 Variable $av_failed, true if the AV scanner deferred.
2547 Bugzilla 1078. Patch from John Horne.
2549 PP/06 Stop make process more reliably on build failure.
2550 Bugzilla 1087. Patch from Heiko Schlittermann.
2552 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2553 Bugzilla 1089. Patch from Heiko Schlittermann.
2555 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2556 Bugzilla 1131. Patch from Holger Weiß.
2558 PP/09 Handle IPv6 addresses with SPF.
2559 Bugzilla 860. Patch from Wolfgang Breyha.
2561 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2563 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2566 PP/11 match_* no longer expand right-hand-side by default.
2567 New compile-time build option, EXPAND_LISTMATCH_RHS.
2568 New expansion conditions, "inlist", "inlisti".
2570 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2572 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2574 PP/14 fix log_write() format string regression from TF/03.
2575 Bugzilla 1152. Patch from Dmitry Isaikin.
2581 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2583 PP/02 Harmonised TLS library version reporting; only show if debugging.
2584 Layout now matches that introduced for other libraries in 4.74 PP/03.
2586 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2588 PP/04 New "dns_use_edns0" global option.
2590 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2593 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2594 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2596 TK/01 Updated PolarSSL code to 0.14.2.
2597 Bugzilla 1097. Patch from Andreas Metzler.
2599 PP/07 Catch divide-by-zero in ${eval:...}.
2600 Fixes bugzilla 1102.
2602 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2605 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2606 format-string attack -- SECURITY: remote arbitrary code execution.
2608 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2609 time unintentionally subject to list matching rules, letting the header
2610 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2611 arbitrary string expansion). This allowed for information disclosure.
2613 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2614 INT_MIN/-1 -- value coerced to INT_MAX.
2620 NM/01 Workaround for PCRE version dependency in version reporting
2623 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2624 This fixes portability to compilers other than gcc, notably
2625 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2627 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2628 makefiles for portability to HP-UX and POSIX correctness.
2630 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2631 Also via indented variable definition in the Makefile.
2632 (Debugging by Oliver Heesakkers).
2634 PP/02 Restore caching of spamd results with expanded spamd_address.
2635 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2637 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2638 Improves build reliability. Fix from: Frank Elsner
2640 NM/02 Fix wide character breakage in the rfc2047 coding
2641 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2643 NM/03 Allow underscore in dnslist lookups
2644 Fixes bug 1026. Patch from Graeme Fowler
2646 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2647 Code patches from Adam Ciarcinski of NetBSD.
2649 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2652 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2653 is logged, to avoid truncation. Patch from John Horne.
2655 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2656 Patch from Jakob Hirsch.
2658 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2659 SQL string expansion failure details.
2660 Patch from Andrey Oktyabrski.
2662 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2663 Patch from Simon Arlott.
2665 PP/09 New lookups functionality failed to compile on old gcc which rejects
2666 extern declarations in function scope.
2667 Patch from Oliver Fleischmann
2669 PP/10 Use sig_atomic_t for flags set from signal handlers.
2670 Check getgroups() return and improve debugging.
2671 Fixed developed for diagnosis in bug 927 (which turned out to be
2674 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2675 Patch from Mark Zealey.
2677 PP/12 Bugzilla 1056: Improved spamd server selection.
2678 Patch from Mark Zealey.
2680 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2681 Based on patch from Heiko Schlittermann.
2683 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2684 Patch from Uwe Doering, sign-off by Michael Haardt.
2686 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2687 variable declaration deep within a block. Bug and patch from
2690 PP/15 lookups-Makefile IRIX compatibility coercion.
2692 PP/16 Make DISABLE_DKIM build knob functional.
2694 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2695 Patch by Simon Arlott
2697 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2698 variable argument macros. Our copy now differs from upstream.
2704 TF/01 Failure to get a lock on a hints database can have serious
2705 consequences so log it to the panic log.
2707 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2708 controlled using the smtp_confirmation log selector.
2710 TF/03 Include the error message when we fail to unlink a spool file.
2712 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2713 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2714 for maintaining out-of-tree patches for some time.
2716 PP/01 Bugzilla 139: Documentation and portability issues.
2717 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2718 Handle per-OS dynamic-module compilation flags.
2720 PP/02 Let /dev/null have normal permissions.
2721 The 4.73 fixes were a little too stringent and complained about the
2722 permissions on /dev/null. Exempt it from some checks.
2723 Reported by Andreas M. Kirchwitz.
2725 PP/03 Report version information for many libraries, including
2726 Exim version information for dynamically loaded libraries. Created
2727 version.h, now support a version extension string for distributors
2728 who patch heavily. Dynamic module ABI change.
2730 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2731 privilege escalation vulnerability whereby the Exim run-time user
2732 can cause root to append content of the attacker's choosing to
2735 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2738 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2739 If dropping privileges for untrusted macros, we disabled normal logging
2740 on the basis that it would fail; for the Exim run-time user, this is not
2741 the case, and it resulted in successful deliveries going unlogged.
2742 Fixed. Reported by Andreas Metzler.
2748 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2749 only prepend for the Resent-* case. Fixes regression introduced in
2750 Exim 4.70 by NM/22 for Bugzilla 607.
2752 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2753 increasing numbers of administrators be bitten by this.
2755 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2757 PP/03 Bugzilla 994: added openssl_options main configuration option.
2759 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2761 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2763 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2765 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2766 without a peer certificate, leading to a segfault because of an
2767 assumption that peers always have certificates. Be a little more
2768 paranoid. Problem reported by Martin Tscholak.
2770 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2771 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2772 NB: ClamAV planning to remove STREAM in "middle of 2010".
2773 CL also introduces -bmalware, various -d+acl logging additions and
2774 more caution in buffer sizes.
2776 PP/09 Implemented reverse_ip expansion operator.
2778 PP/10 Bugzilla 937: provide a "debug" ACL control.
2780 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2782 PP/12 Bugzilla 973: Implement --version.
2784 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2786 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2788 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2790 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2791 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2792 ignore trailing whitespace.
2794 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2796 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2799 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2800 Notification from Dr Andrew Aitchison.
2802 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2803 ExtendedDetectionInfo response format.
2804 Notification from John Horne.
2806 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2809 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2810 XSL and documented dependency on system catalogs, with examples of how
2813 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2816 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2817 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2818 configuration file which is writeable by the Exim user or group.
2820 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2821 of configuration files to cover files specified with the -C option if
2822 they are going to be used with root privileges, not just the default
2825 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2826 option (effectively making it always true).
2828 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2829 files to be used while preserving root privileges.
2831 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2832 that rogue child processes cannot use them.
2834 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2835 run-time user, instead of root.
2837 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2838 Exim run-time user without dropping privileges.
2840 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2841 result string, instead of calling string_vformat() twice with the same
2844 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2845 for other users. Others should always drop root privileges if they use
2846 -C on the command line, even for a whitelisted configure file.
2848 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2850 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2856 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2857 $data_path, and $header_path variables; fixed documentation bugs and
2860 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2861 exipick to access non-standard spools, including the "frozen" queue
2864 NM/01 Bugzilla 965: Support mysql stored procedures.
2865 Patch from Alain Williams
2867 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2869 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2870 Patch from Andreas Metzler
2872 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2873 Patch from Kirill Miazine
2875 NM/05 Bugzilla 671: Added umask to procmail example.
2877 JJ/03 installed exipick 20100323.0, fixing doc bug
2879 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2880 directory. Notification and patch from Dan Rosenberg.
2882 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2884 TK/02 Improve log output when DKIM signing operation fails.
2886 MH/01 Treat the transport option dkim_domain as a colon separated
2887 list, not as a single string, and sign the message with each element,
2888 omitting multiple occurences of the same signer.
2890 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2891 Bugzilla 985, 986. Patch by Simon Arlott
2893 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2894 Patch by Simon Arlott
2896 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2897 MBX locking. Notification from Dan Rosenberg.
2903 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2905 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2907 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2909 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2911 NM/04 Bugzilla 915: Flag broken perl installation during build.
2917 TK/01 Added patch by Johannes Berg that expands the main option
2918 "spamd_address" if it starts with a dollar sign.
2920 TK/02 Write list of recipients to X-Envelope-Sender header when building
2921 the mbox-format spool file for content scanning (suggested by Jakob
2924 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2925 (http://www.dcc-servers.net/) support via dccifd. Activated by
2926 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2928 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2929 by Mark Daniel Reidel <mr@df.eu>.
2931 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2932 When building exim an external PCRE library is now needed -
2933 PCRE is a system library on the majority of modern systems.
2934 See entry on PCRE_LIBS in EDITME file.
2936 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2937 conversation. Added nologin parameter to request.
2938 Patch contributed by Kirill Miazine.
2940 TF/01 Do not log submission mode rewrites if they do not change the address.
2942 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2944 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2945 log files in place. Contributed by Roberto Lima.
2947 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2949 TF/03 Bugzilla 615: When checking the local_parts router precondition
2950 after a local_part_suffix or local_part_prefix option, Exim now
2951 does not use the address's named list lookup cache, since this
2952 contains cached lookups for the whole local part.
2954 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2955 Robert Millan. Documentation is in experimental-spec.txt.
2957 TF/04 Bugzilla 668: Fix parallel build (make -j).
2959 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2961 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2962 Patch provided by Jan Srzednicki.
2964 TF/05 Leading white space used to be stripped from $spam_report which
2965 wrecked the formatting. Now it is preserved.
2967 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2968 that they are available at delivery time.
2970 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2972 TF/08 TLS error reporting now respects the incoming_interface and
2973 incoming_port log selectors.
2975 TF/09 Produce a more useful error message if an SMTP transport's hosts
2976 setting expands to an empty string.
2978 NM/06 Bugzilla 744: EXPN did not work under TLS.
2979 Patch provided by Phil Pennock.
2981 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2982 Patch provided by Richard Godbee.
2984 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2985 acl_smtp_notquit, added index entry.
2987 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2988 Patch provided by Eugene Bujak.
2990 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2991 accept(). Patch provided by Maxim Dounin.
2993 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2994 Patch provided by Phil Pennock.
2996 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2998 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2999 Patch provided by Brad "anomie" Jorsch.
3001 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3002 Patch provided by Dean Brooks.
3004 TK/05 Add native DKIM support (does not depend on external libraries).
3006 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3007 Patch provided by Graeme Fowler.
3009 NM/16 Bugzilla 851: Documentation example syntax fix.
3011 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3013 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3016 NM/19 Bugzilla 745: TLS version reporting.
3017 Patch provided by Phil Pennock.
3019 NM/20 Bugzilla 167: bool: condition support.
3020 Patch provided by Phil Pennock.
3022 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3023 clients. Patch provided by Phil Pennock.
3025 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3026 Patch provided by Brad "anomie" Jorsch.
3028 NM/23 Bugzilla 687: Fix misparses in eximstats.
3029 Patch provided by Heiko Schlittermann.
3031 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3032 Patch provided by Heiko Schlittermann.
3034 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3035 plus update to original patch.
3037 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3039 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3040 Patch provided by David Brownlee.
3042 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3044 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3046 NM/30 Bugzilla 888: TLS documentation bugfixes.
3048 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3050 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3051 Unlike the original bugzilla I have changed all shell scripts in src tree.
3053 NM/33 Bugzilla 898: Transport filter timeout fix.
3054 Patch by Todd Rinaldo.
3056 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3057 Patch by Serge Demonchaux.
3059 NM/35 Bugzilla 39: Base64 decode bug fixes.
3060 Patch by Jakob Hirsch.
3062 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3064 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3066 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3068 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3074 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3075 ALT-N's libdkim that I have put here:
3076 http://duncanthrax.net/exim-experimental/
3078 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3079 were called 'true' and it seems that C99 defines that as a reserved
3080 keyword to be used with 'bool' variable types. That means you could
3081 not include C99-style headers which use bools without triggering
3082 build errors in sieve.c.
3084 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3085 as mailq or other aliases. Changed the --help handling significantly
3086 to do whats expected. exim_usage() emits usage/help information.
3088 SC/01 Added the -bylocaldomain option to eximstats.
3090 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3092 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3094 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3100 PH/01 Another patch from the Sieve maintainer.
3102 PH/02 When an IPv6 address is converted to a string for single-key lookup
3103 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3104 dots are used instead of colons so that keys in lsearch files need not
3105 contain colons. This was done some time before quoting was made available
3106 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3107 (notated using the quote facility) so as to distinguish them from IPv4
3108 keys. This meant that lookups for IP addresses in host lists did not work
3109 for iplsearch lookups.
3111 This has been fixed by arranging for IPv6 addresses to be expressed with
3112 colons if the lookup type is iplsearch. This is not incompatible, because
3113 previously such lookups could never work.
3115 The situation is now rather anomalous, since one *can* have colons in
3116 ordinary lsearch keys. However, making the change in all cases is
3117 incompatible and would probably break a number of configurations.
3119 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3122 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3123 conversion specification without a maximum field width, thereby enabling
3124 a rogue spamd server to cause a buffer overflow. While nobody in their
3125 right mind would setup Exim to query an untrusted spamd server, an
3126 attacker that gains access to a server running spamd could potentially
3127 exploit this vulnerability to run arbitrary code as the Exim user.
3129 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3130 $primary_hostname instead of what libspf2 thinks the hosts name is.
3132 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3133 a directory entry by the name of the lookup key. Previously, if a
3134 symlink pointed to a non-existing file or a file in a directory that
3135 Exim lacked permissions to read, a lookup for a key matching that
3136 symlink would fail. Now it is enough that a matching directory entry
3137 exists, symlink or not. (Bugzilla 503.)
3139 PH/03 The body_linecount and body_zerocount variables are now exported in the
3142 PH/04 Added the $dnslist_matched variable.
3144 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3145 This means they are set thereafter only if the connection becomes
3148 PH/06 Added the client_condition to authenticators so that some can be skipped
3149 by clients under certain conditions.
3151 PH/07 The error message for a badly-placed control=no_multiline_responses left
3152 "_responses" off the end of the name.
3154 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3156 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3157 (without spaces) instead of just copying the configuration text.
3159 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3161 PH/11 Added $max_received_linelength.
3163 PH/12 Added +ignore_defer and +include_defer to host lists.
3165 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3166 way in which PCRE > 7.0 is built.
3168 PH/14 Implemented queue_only_load_latch.
3170 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3171 MAIL command. The effect was to mangle the value on 64-bit systems.
3173 PH/16 Another patch from the Sieve maintainer.
3175 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3177 PH/18 If a system quota error occurred while trying to create the file for
3178 a maildir delivery, the message "Mailbox is full" was not appended to the
3179 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3180 only to a quota excession during the actual writing of the file.
3182 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3183 characters?) which causes problems in log lines. The DN values are now
3184 passed through string_printing() before being added to log lines.
3186 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3187 and InterBase are left for another time.)
3189 PH/21 Added message_body_newlines option.
3191 PH/22 Guard against possible overflow in moan_check_errorcopy().
3193 PH/23 POSIX allows open() to be a macro; guard against that.
3195 PH/24 If the recipient of an error message contained an @ in the local part
3196 (suitably quoted, of course), incorrect values were put in $domain and
3197 $local_part during the evaluation of errors_copy.
3203 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3204 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3207 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3208 issue a MAIL command.
3210 PH/02 In an ACL statement such as
3212 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3214 if a client was not listed at all, or was listed with a value other than
3215 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3216 the condition was not true (as it should be), so access was not denied.
3217 The bug was that the ! inversion was incorrectly passed on to the second
3218 item. This has been fixed.
3220 PH/03 Added additional dnslists conditions == and =& which are different from
3221 = and & when the dns lookup returns more than one IP address.
3223 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3224 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3226 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3227 FSYNC, which compiles an option called disable_fsync that allows for
3228 bypassing fsync(). The documentation is heavily laced with warnings.
3230 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3232 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3233 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3234 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3235 including adding "make clean"; (3) Added -fPIC when compiling the test
3236 dynamically loaded module, to get rid of a warning.
3238 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3239 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3240 = 0s. The bug is otherwise harmless.
3242 PH/07 There was a bug in the dovecot authenticator such that the value of
3243 $auth1 could be overwritten, and so not correctly preserved, after a
3244 successful authentication. This usually meant that the value preserved by
3245 the server_setid option was incorrect.
3247 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3249 PH/09 Installed PCRE release 7.0.
3251 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3252 run for batched SMTP input. It is now run at the start of every message
3253 in the batch. While fixing this I discovered that the process information
3254 (output by running exiwhat) was not always getting set for -bs and -bS
3255 input. This is fixed, and it now also says "batched" for BSMTP.
3257 PH/11 Added control=no_pipelining.
3259 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3260 patch, slightly modified), and move the expansion of helo_data till after
3261 the connection is made in the smtp transport (so it can use these
3264 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3266 PH/14 Added log_selector = +pid.
3268 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3270 PH/16 Add ${if forany and ${if forall.
3272 PH/17 Added dsn_from option to vary the From: line in DSNs.
3274 PH/18 Flush SMTP output before performing a callout, unless control =
3275 no_callout_flush is set.
3277 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3278 was true (the default) a successful delivery failed to delete the retry
3279 item, thus causing premature timeout of the address. The bug is now
3282 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3284 PH/21 Long custom messages for fakedefer and fakereject are now split up
3285 into multiline responses in the same way that messages for "deny" and
3286 other ACL rejections are.
3288 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3289 with slight modification.
3291 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3292 draft, changing the syntax and factoring some duplicate code".
3294 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3295 for deliveries of the second and subsequent messages over the same SMTP
3298 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3299 ${reduce, with only minor "tidies".
3301 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3303 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3304 expansion side effects.
3306 PH/27 When a message times out after an over-quota error from an Exim-imposed
3307 quota, the bounce message says "mailbox is full". This message was not
3308 being given when it was a system quota that was exceeded. It now should
3311 MH/03 Made $recipients available in local_scan(). local_scan() already has
3312 better access to the recipient list through recipients_list[], but
3313 $recipients can be useful in postmaster-provided expansion strings.
3315 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3316 in the case of a MAIL command with additional options following the
3317 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3318 were accidentally chopped off.
3320 PH/29 SMTP synchronization checks are implemented when a command is read -
3321 there is a check that no more input is waiting when there shouldn't be
3322 any. However, for some commands, a delay in an ACL can mean that it is
3323 some time before the response is written. In this time, more input might
3324 arrive, invalidly. So now there are extra checks after an ACL has run for
3325 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3326 pipelining has not been advertised.
3328 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3330 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3331 returned in the SMTP response when smtp_return_error_details was false.
3332 This has been fixed.
3334 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3335 instead of the C I/O that was originally supplied, because problems were
3336 reported on Solaris.
3338 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3339 Exim which did not show up earlier: it was assuming that a call to
3340 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3341 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3342 was a macro that became an assignment, so it seemed to work. This has
3343 changed to a proper function call with a void return, hence the compile
3344 error. Exim's code has been fixed.
3346 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3349 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3351 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3353 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3354 because it turns out to be too big or there is a timeout) there is no
3355 "Completed" line in the log. When some messages of this type were
3356 selected by exigrep, they were listed as "not completed". Others were
3357 picked up by some special patterns. I have improved the selection
3358 criteria to be more general.
3360 PH/38 The host_find_failed option in the manualroute router can now be set
3361 to "ignore", to completely ignore a host whose IP address cannot be
3362 found. If all hosts are ignored, the behaviour is controlled by the new
3363 host_all_ignored option.
3365 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3366 homing or because of multiple MX records with /mx) generated more than
3367 one IP address, and the following item turned out to be the local host,
3368 all the secondary addresses of the first item were incorrectly removed
3369 from the list, along with the local host and any following hosts (which
3370 is what is supposed to happen).
3372 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3373 whoever called Exim into the -H file. In the case of the daemon it was
3374 behaving confusingly. When first started, it used values for whoever
3375 started the daemon, but after a SIGHUP it used the Exim user (because it
3376 calls itself on a restart). I have changed the code so that it now always
3379 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3380 message are rejected with the same error (e.g. no authentication or bad
3381 sender address), and a DATA command is nevertheless sent (as can happen
3382 with PIPELINING or a stupid MUA), the error message that was given to the
3383 RCPT commands is included in the rejection of the DATA command. This is
3384 intended to be helpful for MUAs that show only the final error to their
3387 PH/42 Another patch from the Sieve maintainer.
3389 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3390 Eximstats - Fixed some broken HTML links and added missing column headers
3392 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3393 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3395 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3397 PH/43 Yet another patch from the Sieve maintainer.
3399 PH/44 I found a way to check for a TCP/IP connection going away before sending
3400 the response to the final '.' that terminates a message, but only in the
3401 case where the client has not sent further data following the '.'
3402 (unfortunately, this is allowed). However, in many cases there won't be
3403 any further data because there won't be any more messages to send. A call
3404 to select() can be used: if it shows that the input is "ready", there is
3405 either input waiting, or the socket has been closed. An attempt to read
3406 the next input character can distinguish the two cases. Previously, Exim
3407 would have sent an OK response which the client would never have see.
3408 This could lead to message repetition. This fix should cure that, at
3409 least in a lot of common cases.
3411 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3412 advertised in response to EHLO.
3418 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3419 fixed by 4.65/MH/01 (is this a record?) are fixed:
3421 (i) An empty string was always treated as zero by the numeric comparison
3422 operators. This behaviour has been restored.
3424 (ii) It is documented that the numeric comparison operators always treat
3425 their arguments as decimal numbers. This was broken in that numbers
3426 starting with 0 were being interpreted as octal.
3428 While fixing these problems I realized that there was another issue that
3429 hadn't been noticed. Values of message_size_limit (both the global option
3430 and the transport option) were treated as octal if they started with 0.
3431 The documentation was vague. These values are now always treated as
3432 decimal, and I will make that clear in the documentation.
3438 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3439 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3442 MH/01 Don't check that the operands of numeric comparison operators are
3443 integers when their expansion is in "skipping" mode (fixes bug
3444 introduced by 4.64-PH/07).
3446 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3447 child addresses, Exim now panics and dies. Previously, because the count
3448 is held in a short int, deliveries were likely to be lost. As such a
3449 large number of recipients for a single message is ridiculous
3450 (performance will be very, very poor), I have chosen to impose a limit
3451 rather than extend the field.
3457 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3458 leftover -K file (the existence of which was triggered by #402).
3459 While we were at it, introduced process PID as part of the -K
3460 filename. This should rule out race conditions when creating
3463 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3464 processing considerably. Previous code took too long for large mails,
3465 triggering a timeout which in turn triggers #401.
3467 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3468 in the DK code in transports.c. sendfile() is not really portable,
3469 hence the _LINUX specificness.
3471 TF/01 In the add_headers option to the mail command in an Exim filter,
3472 there was a bug that Exim would claim a syntax error in any
3473 header after the first one which had an odd number of characters
3476 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3477 callout verification, Exim cached a "reject" for the entire domain. This
3478 is correct for most verifications, but it is not correct for a recipient
3479 verification with use_sender or use_postmaster set, because in that case
3480 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3481 case of MAIL FROM:<> rejection from other early rejections (e.g.
3482 rejection of HELO). When verifying a recipient using a non-null MAIL
3483 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3484 Whatever the result of the callout, the value of the domain cache is
3485 left unchanged (for any other kind of callout, getting as far as trying
3486 RCPT means that the domain itself is ok).
3488 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3491 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3492 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3493 session, the daemon ignores ECONNECT errors and logs others; it now
3494 ignores EPIPE as well.
3496 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3497 (quoted-printable decoding).
3499 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3500 later the small subsequent patch to fix an introduced bug.
3502 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3504 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3506 PH/08 An error is now given if message_size_limit is specified negative.
3508 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3509 to be given (somewhat) arbitrary names.
3511 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3514 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3515 miscellaneous code fixes
3517 PH/10 Added the log_reject_target ACL modifier to specify where to log
3520 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3521 hostname. This is wrong, because it relates to the incoming message (and
3522 probably the interface on which it is arriving) and not to the outgoing
3523 callout (which could be using a different interface). This has been
3524 changed to use the value of the helo_data option from the smtp transport
3525 instead - this is what is used when a message is actually being sent. If
3526 there is no remote transport (possible with a router that sets up host
3527 addresses), $smtp_active_hostname is used.
3529 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3530 tweaks were necessary in order to get it to work (see also 21 below):
3531 (a) The code assumed that strncpy() returns a negative number on buffer
3532 overflow, which isn't the case. Replaced with Exim's string_format()
3534 (b) There were several signed/unsigned issues. I just did the minimum
3535 hacking in of casts. There is scope for a larger refactoring.
3536 (c) The code used strcasecmp() which is not a standard C function.
3537 Replaced with Exim's strcmpic() function.
3538 (d) The code set only $1; it now sets $auth1 as well.
3539 (e) A simple test gave the error "authentication client didn't specify
3540 service in request". It would seem that Dovecot has changed its
3541 interface. Fortunately there's a specification; I followed it and
3542 changed what the client sends and it appears to be working now.
3544 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3547 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3548 address A is aliased to B and C, where B exists and C does not. Without
3549 -v the output is "A verified" because verification stops after a
3550 successful redirection if more than one address is generated. However,
3551 with -v the child addresses are also verified. Exim was outputting "A
3552 failed to verify" and then showing the successful verification for C,
3553 with its parentage. It now outputs "B failed to verify", showing B's
3554 parentage before showing the successful verification of C.
3556 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3557 look up a TXT record in a specific list after matching in a combined
3560 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3561 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3562 they consult the DNS. I had assumed they would set it the way they
3563 wanted; and indeed my experiments on Linux seem to show that in some
3564 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3565 To be on the safe side, however, I have now made the interface to
3566 host_find_byname() similar to host_find_bydns(), with an argument
3567 containing the DNS resolver options. The host_find_byname() function now
3568 sets these options at its start, just as host_find_bydns() does. The smtp
3569 transport options dns_qualify_single and dns_search_parents are passed to
3570 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3571 of host_find_byname() use the default settings of RES_DEFNAMES
3572 (qualify_single) but not RES_DNSRCH (search_parents).
3574 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3575 spool_read_header() do less string testing, by means of a preliminary
3576 switch on the second character of optional "-foo" lines. (This is
3577 overdue, caused by the large number of possibilities that now exist.
3578 Originally there were few.) While I was there, I also converted the
3579 str(n)cmp tests so they don't re-test the leading "-" and the first
3580 character, in the hope this might squeeze out yet more improvement.
3582 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3583 flag allowing group syntax was set by the header_syntax check but not
3584 turned off, possible causing trouble later; (2) The flag was not being
3585 set at all for the header_verify test, causing "group"-style headers to
3586 be rejected. I have now set it in this case, and also caused header_
3587 verify to ignore an empty address taken from a group. While doing this, I
3588 came across some other cases where the code for allowing group syntax
3589 while scanning a header line wasn't quite right (mostly, not resetting
3590 the flag correctly in the right place). These bugs could have caused
3591 trouble for malformed header lines. I hope it is now all correct.
3593 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3594 with the "reply" argument non-NULL. The code, however (which originally
3595 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3596 but it didn't always do it. This confused somebody who was copying the
3597 code for some other use. I have removed all the tests.
3599 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3600 feature that was used to support insecure browsers during the U.S. crypto
3601 embargo. It requires special client support, and Exim is probably the
3602 only MTA that supported it -- and would never use it because real RSA is
3603 always available. This code has been removed, because it had the bad
3604 effect of slowing Exim down by computing (never used) parameters for the
3605 RSA_EXPORT functionality.
3607 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3608 authenticator to fail if there's a tab character in the incoming data
3609 (there should never be unless someone is messing about, as it's supposed
3610 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3611 if the connection is using TLS or if the remote IP is the same as the
3612 local IP, and the "valid-client-cert option" if a client certificate has
3615 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3616 authenticators. This can be used for authorization after authentication
3617 succeeds. (In the case of plaintext, it servers for both authentication
3620 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3621 if any retry times were supplied.
3623 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3624 connection, where there is no client IP address to check. In this
3625 situation, the verify now always succeeds.
3627 PH/25 Applied John Jetmore's -Mset patch.
3629 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3631 PH/27 In a string expansion for a processed (not raw) header when multiple
3632 headers of the same name were present, leading whitespace was being
3633 removed from all of them, but trailing whitespace was being removed only
3634 from the last one. Now trailing whitespace is removed from each header
3635 before concatenation. Completely empty headers in a concatenation (as
3636 before) are ignored.
3638 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3639 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3641 PH/29 [Removed. This was a change that I later backed out, and forgot to
3642 correct the ChangeLog entry (that I had efficiently created) before
3643 committing the later change.]
3645 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3646 address errors (4xx response to RCPT) over the same connection as other
3647 messages routed to the same hosts. Such deliveries are always "forced",
3648 so retry times are not inspected. This resulted in far too many retries
3649 for the affected addresses. The effect occurred only when there were more
3650 hosts than the hosts_max_try setting in the smtp transport when it had
3651 the 4xx errors. Those hosts that it had tried were not added to the list
3652 of hosts for which the message was waiting, so if all were tried, there
3653 was no problem. Two fixes have been applied:
3655 (i) If there are any address or message errors in an SMTP delivery, none
3656 of the hosts (tried or untried) are now added to the list of hosts
3657 for which the message is waiting, so the message should not be a
3658 candidate for sending over the same connection that was used for a
3659 successful delivery of some other message. This seems entirely
3660 reasonable: after all the message is NOT "waiting for some host".
3661 This is so "obvious" that I'm not sure why it wasn't done
3662 previously. Hope I haven't missed anything, but it can't do any
3663 harm, as the worst effect is to miss an optimization.
3665 (ii) If, despite (i), such a delivery is accidentally attempted, the
3666 routing retry time is respected, so at least it doesn't keep
3667 hammering the server.
3669 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3670 in ${readsocket because some servers need this prod.
3672 PH/32 Added some extra debug output when updating a wait-xxx database.
3674 PH/33 The hint "could be header name not terminated by colon", which has been
3675 given for certain expansion errors for a long time, was not being given
3676 for the ${if def:h_colon_omitted{... case.
3678 PH/34 The spec says: "With one important exception, whenever a domain list is
3679 being scanned, $domain contains the subject domain." There was at least
3680 one case where this was not true.
3682 PH/35 The error "getsockname() failed: connection reset by peer" was being
3683 written to the panic log as well as the main log, but it isn't really
3684 panic-worthy as it just means the connection died rather early on. I have
3685 removed the panic log writing for the ECONNRESET error when getsockname()
3688 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3689 runs only) independently of the message's sender address. This meant
3690 that, if the 4xx error was in fact related to the sender, a different
3691 message to the same recipient with a different sender could confuse
3692 things. In particular, this can happen when sending to a greylisting
3693 server, but other circumstances could also provoke similar problems.
3694 I have changed the default so that the retry time for these errors is now
3695 based a combination of the sender and recipient addresses. This change
3696 can be overridden by setting address_retry_include_sender=false in the
3699 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3700 remote server are returned as part of bounce messages. This was not
3701 happening for LMTP over a pipe (the lmtp transport), but now it is the
3702 same for both kinds of LMTP.
3704 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3705 in header lines that were in fact CNAMEs. This is no longer the case.
3707 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3708 and queue runs started by the daemon processed all messages. This has
3709 been fixed so that -R and -S can now usefully be given with -q<time>.
3711 PH/40 Import PCRE release 6.7 (fixes some bugs).
3713 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3715 PH/42 Give an error if -q is specified more than once.
3717 PH/43 Renamed the variables $interface_address and $interface_port as
3718 $received_ip_address and $received_port, to make it clear that these
3719 values apply to message reception, and not to the outgoing interface when
3720 a message is delivered. (The old names remain recognized, of course.)
3722 PH/44 There was no timeout on the connect() call when using a Unix domain
3723 socket in the ${readsocket expansion. There now is.
3725 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3726 be meaningful with "accept".
3728 SC/01 Eximstats V1.43
3729 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3731 SC/02 Eximstats V1.44
3732 Use a glob alias rather than an array ref in the generated
3733 parser. This improves both readability and performance.
3735 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3736 Collect SpamAssassin and rejection statistics.
3737 Don't display local sender or destination tables unless
3738 there is data to show.
3739 Added average volumes into the top table text output.
3741 SC/04 Eximstats V1.46
3742 Collect data on the number of addresses (recipients)
3743 as well as the number of messages.
3745 SC/05 Eximstats V1.47
3746 Added 'Message too big' to the list of mail rejection
3747 reasons (thanks to Marco Gaiarin).
3749 SC/06 Eximstats V1.48
3750 Mainlog lines which have GMT offsets and are too short to
3751 have a flag are now skipped.
3753 SC/07 Eximstats V1.49 (Alain Williams)
3754 Added the -emptyok flag.
3756 SC/08 Eximstats V1.50
3757 Fixes for obtaining the IP address from reject messages.
3759 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3760 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3761 whitespace changes from 4.64-PH/27
3763 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3766 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3767 are found, allow negative numbers in numeric criteria)
3769 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3771 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3772 to match changes made in 4.64-PH/43
3774 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3776 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3777 "domain" is actually an IP address, and force a failure. This locks out
3778 those revolvers/nameservers that support "A-for-A" lookups, in
3779 contravention of the specifications.
3781 PH/48 When a host name was looked up from an IP address, and the subsequent
3782 forward lookup of the name timed out, the host name was left in
3783 $sender_host_name, contrary to the specification.
3785 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3786 restricted to single-key lookups, Exim was not diagnosing an error if
3787 * or *@ was used with a query-style lookup.
3789 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3791 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3792 long ago, but noone interested enough thought of it. Let's just say that
3793 the "1.1" means that there are some new functions that weren't there at
3794 some point in the past.
3796 PH/51 Error processing for expansion failure of helo_data from an smtp
3797 transport during callout processing was broken.
3799 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3800 tested/used via the -bh/-bhc/-bs options.
3802 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3803 bug, fixed in subsequent PCRE releases).
3805 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3806 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3808 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3814 SC/01 Use a glob alias rather than an array ref in eximstats generated
3815 parser. This improves both readability and performance.
3817 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3818 Don't display local sender or destination tables in eximstats unless
3819 there is data to show.
3820 Added average volumes into the eximstats top table text output.
3822 SC/03 Collect data on the number of addresses (recipients) as well
3823 as the number of messages in eximstats.
3825 TF/01 Correct an error in the documentation for the redirect router. Exim
3826 does (usually) call initgroups() when daemonizing.
3828 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3829 with consistent privilege compared to when running as a daemon.
3831 TF/03 Note in the spec that $authenticated_id is not set for local
3832 submissions from trusted users.
3834 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3835 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3837 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3838 by adding some example configuration directives to the default
3839 configuration file. A little bit of work is required to uncomment the
3840 directives and define how usernames and passwords are checked, but
3841 there is now a framework to start from.
3843 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3844 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3845 without this. I don't know how relevant this is to other LDAP libraries.
3847 PH/02 Add the verb name to the "unknown ACL verb" error.
3849 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3851 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3853 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3854 directory not expanded when it should be if an expanded home directory
3855 was set for the address (which is overridden by the transport).
3857 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3860 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3861 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3862 because it is too late at that time, and has no effect.
3864 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3865 security issue with \' (bugzilla #107). I could not use the
3866 PQescapeStringConn() function, because it needs a PGconn value as one of
3869 PH/08 When testing addresses using -bt, indicate those final addresses that
3870 are duplicates that would not cause an additional delivery. At least one
3871 person was confused, thinking that -bt output corresponded to deliveries.
3872 (Suppressing duplicates isn't a good idea as you lose the information
3873 about possibly different redirections that led to the duplicates.)
3875 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3876 systems where poll() doesn't work, in particular OS X.
3878 PH/10 Added more information to debugging output for retry time not reached.
3880 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3881 operations in malware.c.
3883 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3886 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3887 syslog_duplication set false, log lines that would normally be written
3888 both the the main log and to the reject log were not written to syslog at
3891 PH/14 In the default configuration, change the use of "message" in ACL warn
3892 statements to "add_header".
3894 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3895 not followed by a command (e.g. "seen endif").
3897 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3898 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3901 PH/17 Added extra conditions to the default value of delay_warning_condition
3905 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3906 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3907 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3910 The Auto-Submitted: and various List- headers are standardised, whereas I
3911 don't think Precedence: ever was.
3913 PH/18 Refactored debugging code in route_finduser() to show more information,
3914 in particular, the error code if getpwnam() issues one.
3916 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3917 This is apparently needed in addition to the PH/07 change above to avoid
3918 any possible encoding problems.
3920 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3921 but not after initializing Perl.
3923 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3924 output them only if debugging. By default they are written stderr,
3925 apparently, which is not desirable.
3927 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3930 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3933 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3935 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3936 authenticated or an ident call has been made. Suppress the default
3937 values for $authenticated_id and $authenticated_sender (but permit -oMai
3938 and -oMas) when testing with -bh.
3940 PH/24 Re-jigged the order of the tests in the default configuration so that the
3941 tests for valid domains and recipients precede the DNS black list and CSA
3942 tests, on the grounds that those ones are more expensive.
3944 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3945 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3946 This bug exists in every version of Exim that I still have, right back to
3949 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3950 However, an attempt to turn on case-sensitivity in a regex key by
3951 including (?-i) didn't work because the subject string was already
3952 lowercased, and the effects were non-intuitive. It turns out that a
3953 one-line patch can be used to allow (?-i) to work as expected.
3959 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3960 other effects) broke the use of negated acl sub-conditions.
3962 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3965 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3966 "Deny" causes Exim to reject the incoming connection with a 554 error.
3967 Unfortunately, if there is a major crisis, such as a disk failure,
3968 tcp-wrappers gives "deny", whereas what one would like would be some
3969 kind of temporary error. A kludge has been added to help with this.
3970 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3971 554 error is used if errno is still zero or contains ENOENT (which occurs
3972 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3975 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3977 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3978 errors. Otherwise a message that provokes a temporary error (when other
3979 messages do not) can cause a whole host to time out.
3981 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3982 addresses were routed directly to files or pipes from a redirect router.
3983 File deliveries just didn't batch; pipe deliveries might have suffered
3986 PH/06 A failure to get a lock for a hints database would erroneously always say
3987 "Failed to get write lock", even when it was really a read lock.
3989 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3990 of 0600. This has been changed to use the value of the lockfile_mode
3991 option (which defaults to 0600).
3993 PH/08 Applied small patch from the Sieve maintainer.
3995 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3996 folder from quota calculations, a direct delivery into this folder messed
3997 up the contents of the maildirsize file. This was because the regex was
3998 used only to exclude .Trash (or whatever) when the size of the mailbox
3999 was calculated. There was no check that a delivery was happening into an
4000 excluded directory. This bug has been fixed by ignoring all quota
4001 processing for deliveries into excluded directories.
4003 PH/10 Added the maildirfolder_create_regex option to appendfile.
4009 PH/01 The code for finding all the local interface addresses on a FreeBSD
4010 system running IPv6 was broken. This may well have applied to all BSD
4011 systems, as well as to others that have similar system calls. The broken
4012 code found IPv4 interfaces correctly, but gave incorrect values for the
4013 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4014 that it would not match correctly against @[] and not recognize the IPv6
4017 PH/02 The ipliteral router was not recognizing addresses of the form user@
4018 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4020 PH/03 Added disable_ipv6.
4022 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4023 decoded form, because it is most often used to construct To: headers
4024 lines in autoreplies, and the decoded form may well be syntactically
4025 invalid. However, $reply_address has leading white space removed, and all
4026 newlines turned into spaces so that the autoreply transport does not
4029 PH/05 If group was specified without a user on a router, and no group or user
4030 was specified on a transport, the group from the router was ignored.
4032 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4033 for visible compile-time settings that can be used to change these
4034 numbers, for those that want even more. Backwards compatibility with old
4035 spool files has been maintained. However, going back to a previous Exim
4036 release will lost any variables that are in spool files.
4038 PH/07 Two small changes when running in the test harness: increase delay when
4039 passing a TCP/IP connection to a new process, in case the original
4040 process has to generate a bounce, and remove special handling of
4041 127.0.0.2 (sic), which is no longer necessary.
4043 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4044 be the same on different OS.
4046 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4049 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4050 whether --show-vars was specified or not
4052 JJ/02 exipick: Added support for new ACL variable spool format introduced
4055 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4056 syntactically invalid From: or Reply-to: line, and a filter used this to
4057 generate an autoreply, and therefore failed to obtain an address for the
4058 autoreply, Exim could try to deliver to a non-existent relative file
4059 name, causing unrelated and misleading errors. What now happens is that
4060 it logs this as a hard delivery error, but does not attempt to create a
4063 PH/11 The exinext utility has a -C option for testing purposes, but although
4064 the given file was scanned by exinext itself; it wasn't being passed on
4065 when Exim was called.
4067 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4068 an end-of-file indication when reading a command response.
4070 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4071 compiled. In many other places in Exim, IPv6 addresses are always
4072 recognized, so I have changed this. It also means that IPv4 domain
4073 literals of the form [IPV4:n.n.n.n] are now always recognized.
4075 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4076 used if the router is not running as root, for example, when verifying at
4077 ACL time, or when using -bh. The debugging output from this situation was
4078 non-existent - all you got was a failure to exec. I have made two
4081 (a) Failures to set uid/gid, the current directory, or a process leader
4082 in a subprocess such as that created by queryprogram now generate
4083 suitable debugging output when -d is set.
4085 (b) The queryprogram router detects when it is not running as root,
4086 outputs suitable debugging information if -d is set, and then runs
4087 the subprocess without attempting to change uid/gid.
4089 PH/15 Minor change to Makefile for building test_host (undocumented testing
4092 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4093 additional section of a DNS packet that returns MX or SRV records.
4094 Instead, it always explicitly searches for A/AAAA records. This avoids
4095 major problems that occur when a DNS server includes only records of one
4096 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4097 fixed another bug: if SRV records were looked up and the corresponding
4098 address records were *not* found in the additional section, the port
4099 values from the SRV records were lost.
4101 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4102 using the correct key (the original address) when searching the retry
4103 rules in order to find which one to use for generating the retry hint.
4105 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4106 adding the default one. Similarly, if it contains a Reply-To: header, the
4107 errors_reply_to option, if set, is not used.
4109 PH/19 When calculating a retry time, Exim used to measure the "time since
4110 failure" by looking at the "first failed" field in the retry record. Now
4111 it does not use this if it is later than than the arrival time of the
4112 message. Instead it uses the arrival time. This makes for better
4113 behaviour in cases where some deliveries succeed, thus re-setting the
4114 "first failed" field. An example is a quota failure for a huge message
4115 when small messages continue to be delivered. Without this change, the
4116 "time since failure" will always be short, possible causing more frequent
4117 delivery attempts for the huge message than are intended.
4118 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4120 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4121 $1, $2, $3) because the numerical variables can be reset during some
4122 expansion items (e.g. "match"), thereby losing the authentication data.
4124 PH/21 Make -bV show the size of off_t variables so that the test suite can
4125 decide whether to run tests for quotas > 2G.
4127 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4128 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4129 filecount value is greater than 2G or if a quota value is greater than 2G
4130 on a system where the size of off_t is not greater than 4, a panic error
4133 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4134 never match. The debug and -bh output now contains an explicit error
4135 message indicating a malformed IPv4 address or mask.
4137 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4138 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4139 PH/23 above applies.
4141 PH/25 Do not write to syslog when running in the test harness. The only
4142 occasion when this arises is a failure to open the main or panic logs
4143 (for which there is an explicit test).
4145 PH/26 Added the /no_tell option to "control=freeze".
4147 PH/27 If a host name lookup failed very early in a connection, for example, if
4148 the IP address matched host_lookup and the reverse lookup yielded a name
4149 that did not have a forward lookup, an error message of the form "no IP
4150 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4151 could be logged. Now it outputs the IP address instead of "NULL".
4153 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4154 allows the sender and the authenticated sender to be set when
4155 submitting a message from within Exim. Since child_open_exim() is
4156 documented for local_scan(), the new function should be too.
4158 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4159 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4160 results in an empty string is now treated as unset.
4162 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4164 PH/31 Added sender_verify_fail logging option.
4166 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4167 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4168 tidied the source and removed it altogether.
4170 PH/33 When a queue run was abandoned because the load average was too high, a
4171 log line was always written; now it is written only if the queue_run log
4172 selector is set. In addition, the log line for abandonment now contains
4173 information about the queue run such as the pid. This is always present
4174 in "start" and "stop" lines but was omitted from the "abandon" line.
4176 PH/34 Omit spaces between a header name and the colon in the error message that
4177 is given when verify = headers_syntax fails (if there are lots of them,
4178 the message gets confusing).
4180 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4181 names, as there are now some PTR records that contain slashes. This check
4182 is only to protect against broken name servers that fall over on strange
4183 characters, so the fact that it applies to all lookups doesn't matter.
4185 PH/36 Now that the new test suite is complete, we can remove some of the
4186 special code in Exim that was needed for the old test suite. For example,
4187 sorting DNS records because real resolvers return them in an arbitrary
4188 order. The new test suite's fake resolver always returns records in the
4191 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4192 bounces) except when queue_only is set, to avoid logging races between
4193 the different processes.
4195 PH/38 Panic-die if .include specifies a non-absolute path.
4197 PH/39 A tweak to the "H" retry rule from its user.
4199 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4200 a label. They prevented compilation on older perls.
4202 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4203 a warning to be raised on newish perls.
4205 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4206 on queue. Changes to match documented behaviour of showing count of
4207 messages matching specified criteria.
4209 PH/40 Changed the default ident timeout from 30s to 5s.
4211 PH/41 Added support for the use of login_cap features, on those BSD systems
4212 that have them, for controlling the resources used by pipe deliveries.
4214 PH/42 The content-scanning code uses fopen() to create files in which to put
4215 message data. Previously it was not paying any attention to the mode of
4216 the files. Exim runs with umask(0) because the rest of the code creates
4217 files with open(), and sets the required mode explicitly. Thus, these
4218 files were ending up world-writeable. This was not a big issue, because,
4219 being within the spool directory, they were not world-accessible. I have
4220 created a function called modefopen, which takes an additional mode
4221 argument. It sets umask(777), creates the file, chmods it to the required
4222 mode, then resets the umask. All the relevant calls to fopen() in the
4223 content scanning code have been changed to use this function.
4225 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4226 to 24 hours. This avoids potential overflow problems when processing G
4227 and H retry rules. I suspect nobody ever tinkers with this value.
4229 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4231 PH/45 When the plaintext authenticator is running as a client, the server's
4232 challenges are checked to ensure they are valid base64 strings. By
4233 default, the authentication attempt is cancelled if an invalid string is
4234 received. Setting client_ignore_invalid_base64 true ignores these errors.
4235 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4236 they are received. Thus, the responses can be made to depend on the
4237 challenges. If an invalid string is ignored, an empty string is placed in
4240 PH/46 Messages that are created by the autoreply transport now contains a
4241 References: header, in accordance with RFCs 2822 and 3834.
4243 PH/47 Added authenticated_sender_force to the smtp transport.
4245 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4247 PH/49 Installed latest patch from the Sieve maintainer.
4249 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4250 was also set, the appendfile transport was unnecessarily scanning a
4251 directory of message files (e.g. for maildir delivery) to find the count
4252 of files (along with the size), even though it did not need this
4253 information. It now does the scan only if it needs to find either the
4254 size of the count of files.
4256 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4258 PH/52 Two bugs concerned with error handling when the smtp transport is
4261 (i) Exim was not creating retry information for temporary errors given
4262 for individual recipients after the DATA command when the smtp transport
4263 was used in LMTP mode. This meant that they could be retried too
4264 frequently, and not timed out correctly.
4266 (ii) Exim was setting the flag that allows error details to be returned
4267 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4268 recipients that were returned after the DATA command.
4270 PH/53 This is related to PH/52, but is more general: for any failing address,
4271 when detailed error information was permitted to be returned to the
4272 sender, but the error was temporary, then after the final timeout, only
4273 "retry timeout exceeded" was returned. Now it returns the full error as
4274 well as "retry timeout exceeded".
4276 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4277 do this, and (what is worse) MTAs that accept it.
4279 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4280 will now be deprecated.
4282 PH/56 New os.c-cygwin from the Cygwin maintainer.
4284 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4285 formats (previously only available in exim formats via -bpr, -bpru,
4286 and -bpra. Now also available in native and exiqgrep formats)
4288 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4289 with very large, slow to parse queues
4291 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4293 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4295 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4296 responses to authentication challenges, though it was showing the
4297 challenges; (ii) I've removed the CR characters from the debug output for
4300 PH/58 Allow for the insertion of a newline as well as a space when a string
4301 is turned into more than one encoded-word during RFC 2047 encoding. The
4302 Sieve code now uses this.
4304 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4305 data_4xx, lost_connection, tls_required.
4307 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4308 message was being sent as an SMTP response.
4310 PH/61 Add -l and -k options to exicyclog.
4312 PH/62 When verifying, if an address was redirected to one new address, so that
4313 verification continued, and the new address failed or deferred after
4314 having set something in $address_data, the value of $address_data was not
4315 passed back to the ACL. This was different to the case when no
4316 redirection occurred. The value is now passed back in both cases.
4318 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4319 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4320 use login_cap.h, so on its own it isn't the distinguishing feature. The
4321 new name refers directly to the setclassresources() function.
4323 PH/65 Added configuration files for NetBSD3.
4325 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4327 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4328 is preferred over IPv4.
4330 PH/68 The bounce_return_message and bounce_return_body options were not being
4331 honoured for bounces generated during the reception of non-SMTP messages.
4332 In particular, this applied to messages rejected by the ACL. This bug has
4333 been fixed. However, if bounce_return_message is true and bounce_return_
4334 body is false, the headers that are returned for a non-SMTP message
4335 include only those that have been read before the error was detected.
4336 (In the case of an ACL rejection, they have all been read.)
4338 PH/69 The HTML version of the specification is now built in a directory called
4339 spec_html instead of spec.html, because the latter looks like a path with
4340 a MIME-type, and this confuses some software.
4342 PH/70 Catch two compiler warnings in sieve.c.
4344 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4345 function verify_get_ident() calls ip_connect() to connect a socket, but
4346 if the "connect()" function timed out, ip_connect() used to close the
4347 socket. However, verify_get_ident() also closes the socket later, and in
4348 between Exim writes to the log, which may get opened at this point. When
4349 the socket was closed in ip_connect(), the log could get the same file
4350 descriptor number as the socket. This naturally causes chaos. The fix is
4351 not to close the socket in ip_connect(); the socket should be closed by
4352 the function that creates it. There was only one place in the code where
4353 this was missing, in the iplookup router, which I don't think anybody now
4354 uses, but I've fixed it anyway.
4356 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4357 well as to direct DNS lookups. Otherwise the handling of names in host
4358 lists is inconsistent and therefore confusing.
4364 PH/01 Two changes to the default runtime configuration:
4366 (1) Move the checks for relay_from_hosts and authenticated clients from
4367 after to before the (commented out) DNS black list checks.
4369 (2) Add control=submission to the relay_from_hosts and authenticated
4370 clients checks, on the grounds that messages accepted by these
4371 statements are most likely to be submissions.
4373 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4375 (1) Generate an error if the third argument for the ${prvs expansion is
4378 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4381 (3) Reset the variables that are obtained from the first argument of
4382 ${prvscheck and used in the second argument before leaving the code,
4383 because their memory is reclaimed, so using them afterwards may do
4386 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4387 one (it's much easier than Tom thought :-).
4389 (5) Because of (4), we can now allow for the use of $prvscheck_result
4390 inside the third argument.
4392 PH/03 For some reason, the default setting of PATH when running a command from
4393 a pipe transport was just "/usr/bin". I have changed it to
4396 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4397 anything to be listed in the output from -bV.
4399 PH/05 When a filter generated an autoreply, the entire To: header line was
4400 quoted in the delivery log line, like this:
4402 => >A.N.Other <ano@some.domain> <original@ddress> ...
4404 This has been changed so that it extracts the operative address. There
4405 may be more than one such address. If so, they are comma-separated, like
4408 => >ano@some.domain,ona@other.domain <original@ddress> ...
4410 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4411 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4412 not being looked up in the rDNS to get a host name, Exim was showing the
4413 IP address twice in Received: lines, even though the IP addresses were
4414 identical. For example:
4416 Received: from [1.2.3.4] (helo=[1.2.3.4])
4418 However, if the real host name was known, it was omitting the HELO data
4419 if it matched the actual IP address. This has been tidied up so that it
4420 doesn't show the same IP address twice.
4422 PH/07 When both +timestamp and +memory debugging was on, the value given by
4423 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4424 called by the debug printing, thereby overwriting the timestamp buffer.
4425 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4427 PH/08 When the original message was included in an autoreply transport, it
4428 always said "this is a copy of the message, including all the headers",
4429 even if body_only or headers_only was set. It now gives an appropriate
4432 PH/09 Applied a patch from the Sieve maintainer which:
4434 o fixes some comments
4435 o adds the (disabled) notify extension core
4436 o adds some debug output for the result of if/elsif tests
4437 o points to the current vacation draft in the documentation
4438 and documents the missing references header update
4442 o fixes a bug in processing the envelope test (when testing
4443 multiple envelope elements, the last element determined the
4446 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4447 Electronic Mail") by including:
4449 Auto-submitted: auto-generated
4451 in the messages that it generates (bounce messages and others, such as
4452 warnings). In the case of bounce messages for non-SMTP messages, there was
4453 also a typo: it was using "Auto_submitted" (underscore instead of
4454 hyphen). Since every message generated by Exim is necessarily in response
4455 to another message, thes have all been changed to:
4457 Auto-Submitted: auto-replied
4459 in accordance with these statements in the RFC:
4461 The auto-replied keyword:
4463 - SHOULD be used on messages sent in direct response to another
4464 message by an automatic process,
4466 - MUST NOT be used on manually-generated messages,
4468 - MAY be used on Delivery Status Notifications (DSNs) and Message
4469 Disposition Notifications (MDNs),
4471 - MUST NOT be used on messages generated by automatic or periodic
4472 processes, except for messages which are automatic responses to
4475 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4476 to the default Received: header definition.
4478 PH/12 Added log selector acl_warn_skipped (default on).
4480 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4481 variables because (a) they are in the wrong storage pool and (b) even if
4482 they were copied, it wouldn't work properly because of the caching.
4484 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4485 checking when decoding. Apparently there are clients that generate
4486 overlong encoded strings. Why am I not surprised?
4488 PH/15 If the first argument of "${if match_address" was not empty, but did not
4489 contain an "@" character, Exim crashed. Now it writes a panic log message
4490 and treats the condition as false.
4492 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4494 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4495 extension "envelope-auth". The code is finished and in agreement with
4496 other implementations, but there is no documentation so far and in fact,
4497 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4498 not changing the active code.
4500 Print executed "if" and "elsif" statements when debugging is used. This
4501 helps a great deal to understand what a filter does.
4503 Document more things not specified clearly in RFC3028. I had all this
4504 sorted out, when out of a sudden new issues came to my mind. Oops."
4506 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4509 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4510 canonical form (as documented). However, after a host name lookup from
4511 the IP address, check_host() was doing a simple string comparison with
4512 addresses acquired from the DNS when checking that the found name did
4513 have the original IP as one of its addresses. Since any found IPv6
4514 addresses are likely to be in abbreviated form, the comparison could
4515 fail. Luckily, there already exists a function for doing the comparison
4516 by converting both addresses to binary, so now that is used instead of
4517 the text comparison.
4519 PH/20 There was another similar case to PH/19, when a complete host name was
4520 given in a host list; looking up its IP address could give an abbreviated
4521 form, whereas the current host's name might or might not be abbreviated.
4522 The same fix has been applied.
4528 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4529 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4532 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4533 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4535 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4537 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4538 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4539 Also, for "Auto-Submitted", treat anything other than "no" as
4540 non-personal, in accordance with RFC 3834. (Previously it treated
4541 anything starting "auto-" as non-personal.)
4543 TF/01 The control=submission/name=... option had a problem with syntax
4544 errors if the name included a slash character. The /name= option
4545 now slurps the rest of the string, so it can include any characters
4546 but it must come last in the list of options (after /sender_retain
4549 PH/05 Some modifications to the interface to the fake nameserver for the new
4557 TK/01 Added the "success_on_redirect" address verification option. See
4558 NewStuff for rationale and an example.
4560 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4562 PH/02 Patch to exigrep to allow it to work on syslog lines.
4564 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4565 fread() to skip over the body file's header line, because in Cygwin the
4566 header line is locked and is inaccessible.
4568 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4569 co-exist for some time) to make it clear that it is the Exim ID that is
4570 referenced, not the Message-ID: header line.
4572 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4573 string_format() function, because snprintf() does not exist on all
4576 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4577 ${stat: expansion item.
4579 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4580 protocol synchronization error", to keep the pedants happy.
4582 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4583 well as for IRIX systems, when gcc is being used. See the host.c source
4586 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4588 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4591 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4592 transport and to the smtp transport in LMTP mode.
4594 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4596 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4597 run a filter in a subprocess. This could lead to confusion in subsequent
4598 lookups in the parent process. There should also be a search_tidyup() at
4599 the end of the subprocess.
4601 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4602 only if the host matched helo_try_verify_hosts, which caused the
4603 verification to occur when the EHLO/HELO command was issued. The ACL just
4604 tested the remembered result. Now, if a previous verification attempt has
4605 not happened, "verify = helo" does it there and then.
4607 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4609 TK/03 Fix log output including CR from clamd.
4611 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4612 exist provoked a memory error which could cause a segfault.
4614 PH/15 Installed PCRE 6.2
4616 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4618 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4619 of the problem. Specifically, suggested +O2 rather than +O1 for the
4622 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4624 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4625 record, the port was not correctly logged when the outgoing_port log
4626 selector was set (it logged the transort's default port).
4628 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4629 fallback_hosts, and "hosts" in the smtp transport.
4631 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4632 host errors such as "Connection refused".
4634 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4635 authentication with radiusclient 0.4.9:
4637 - Error returned from rc_read_config was caught wrongly
4638 - Username/password not passed on to radius server due to wrong length.
4640 The presumption is that some radiusclient API changes for 4.51/PH/17
4641 were not taken care of correctly. The code is still untested by me (my
4642 Linux distribution still has 0.3.2 of radiusclient), but it was
4643 contributed by a Radius user.
4645 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4646 expanding the "port" option of the smtp transport.
4648 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4649 while reading a MIME header. Thanks to Tom Hughes for a patch.
4651 PH/24 Include config.h inside local_scan.h so that configuration settings are
4654 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4655 that in an ACL for RCPT (for example), you can examine exactly what was
4658 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4659 commands, but it was not correctly comparing the address with the actual
4660 client host address. Thus, it would show the EHLO address in Received:
4661 header lines when this was not necessary.
4663 PH/27 Added the % operator to ${eval:}.
4665 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4666 it should be ignoring failures (because with -C, for example, it has lost
4667 privilege). It wasn't ignoring creation failures other than "already
4670 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4673 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4674 because an input line was too long, either on its own, or by virtue of
4675 too many continuations, the temporary file was not being removed, and the
4676 return code was incorrect.
4678 PH/31 Missing "BOOL" in function definition in filtertest.c.
4680 PH/32 Applied Sieve patches from the maintainer.
4682 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4684 PH/33 Added "verify = not_blind".
4686 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4687 Local/Makefile (with some defaults set). These are used in built scripts
4688 such as exicyclog, but they have never been used in the exim_install
4689 script (though there are many overriding facilities there). I have
4690 arranged that the exim_install script now takes note of these two
4693 PH/35 Installed configuration files for Dragonfly.
4695 PH/36 When a locally submitted message by a trusted user did not contain a
4696 From: header, and the sender address was obtained from -f or from an SMTP
4697 MAIL command, and the trusted user did not use -F to supply a sender
4698 name, $originator_name was incorrectly used when constructing a From:
4699 header. Furthermore, $originator_name was used for submission mode
4700 messages from external hosts without From: headers in a similar way,
4701 which is clearly wrong.
4703 PH/37 Added control=suppress_local_fixups.
4705 PH/38 When log_selector = +received_sender was set, and the addition of the
4706 sender made the log line's construction buffer exactly full, or one byte
4707 less than full, an overflow happened when the terminating "\n" was
4710 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4711 when the result of a list match is failure because a DNS lookup failed.
4713 PH/40 RM_COMMAND is now used in the building process.
4715 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4716 the "build-* directories that it finds.
4718 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4719 address in a domain literal was a prefix of an interface address.
4721 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4722 when verifying a sender address, unless rewrite_headers is false.
4724 PH/44 Wrote a long comment about why errors_to addresses are verified as
4725 recipients, not senders.
4727 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4728 the ratelimit ACL was added.
4730 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4732 PH/46 Added extra information about PostgreSQL errors to the error string.
4734 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4735 avoiding the need to install special zones in a real server. This is
4736 backwards compatible; if it can't find the fake resolver, it drops back.
4737 Thus, both old and new test suites can be run.
4739 TF/02 Added util/ratelimit.pl
4741 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4742 clock is set back in time.
4744 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4745 Candler <B.Candler@pobox.com>.
4747 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4748 OK for addresses that are the result of redirections.
4750 PH/48 A number of further additions for the benefit of the new test suite,
4751 including a fake gethostbyname() that interfaces to the fake DNS resolver
4754 TF/06 The fix for widen_domains has also been applied to qualify_single and
4755 search_parents which are the other dnslookup options that can cause
4758 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4761 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4763 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4764 service" if it encountered an MX record with an empty target hostname.
4765 The message is now "an MX or SRV record indicated no SMTP service".
4767 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4768 if the DNS of the sending site is misconfigured. This is quite a
4769 common situation. This change restores the behaviour of treating a
4770 helo verification defer as a failure.
4772 PH/51 If self=fail was set on a router, the bounce message did not include the
4773 actual error message.
4779 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4781 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4782 command itself ended in error, the underlying message about the transport
4783 filter timeout was being overwritten with the pipe command error. Now the
4784 underlying error message should be appended to the second error message.
4786 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4788 PH/02 Reduce the amount of output that "make" produces by default. Full output
4789 can still be requested.
4791 PH/03 The warning log line about a condition test deferring for a "warn" verb
4792 was being output only once per connection, rather than after each
4793 occurrence (because it was using the same function as for successful
4794 "warn" verbs). This seems wrong, so I have changed it.
4796 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4797 it should not have, which might have caused a crash in the right
4798 circumstances, but probably never did.
4800 PH/04 Installed a modified version of Tony Finch's patch to make submission
4801 mode fix the return path as well as the Sender: header line, and to
4802 add a /name= option so that you can make the user's friendly name appear
4805 TF/03 Added the control = fakedefer ACL modifier.
4807 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4808 Mark Lowes for thorough testing.
4810 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4812 TK/03 Merged latest SRS patch from Miles Wilton.
4814 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4815 with the definition in sysexits.h (which is #included earlier).
4816 Fortunately, Exim does not actually use EX_OK. The code used to try to
4817 preserve the sysexits.h value, by assuming that macro definitions were
4818 scanned for macro replacements. I have been disabused of this notion,
4819 so now the code just undefines EX_OK before #including unistd.h.
4821 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4822 in the smtp transport. When a block could not be written in a single
4823 write() function, the timeout was being re-applied to each part-write.
4824 This seems wrong - if the receiver was accepting one byte at a time it
4825 would take for ever. The timeout is now adjusted when this happens. It
4826 doesn't have to be particularly precise.
4828 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4829 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4831 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4832 without a domain if the check to <postmaster@domain> fails.
4834 SC/01 Eximstats: added -xls and the ability to specify output files
4835 (patch written by Frank Heydlauf).
4837 SC/02 Eximstats: use FileHandles for outputting results.
4839 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4841 SC/04 Eximstats: fixed display of large numbers with -nvr option
4843 SC/05 Eximstats: fixed merging of reports with empty tables.
4845 SC/06 Eximstats: added the -include_original_destination flag
4847 SC/07 Eximstats: removed tabs and trailing whitespace.
4849 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4851 TK/06 MBOX spool code: Add real "From " MBOX separator line
4852 so the .eml file is really in mbox format (even though
4853 most programs do not really care). Patch from Alex Miller.
4855 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4856 The latter is generated from $received_to and is only set if the
4857 message has one envelope recipient. SA can use these headers,
4858 obviously out-of-the-box. Patch from Alex Miller.
4860 PH/08 The ${def test on a variable was returning false if the variable's
4861 value was "0", contrary to what the specification has always said!
4862 The result should be true unless the variable is empty.
4864 PH/09 The syntax error of a character other than { following "${if
4865 def:variable_name" (after optional whitespace) was not being diagnosed.
4866 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4867 accidental colon was present, for example, could give incorrect results.
4869 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4870 result is used (not including appendfile, where other changes are about
4873 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4874 This involved changing a lot of size variables from int to off_t. It
4875 should work with maildirs and everything.
4877 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4878 spamd dying while we are connected to it.
4880 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4883 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4884 function for BDB 4.3.
4886 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4888 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4889 the sources for PCRE have been re-arranged and more files are now
4892 PH/15 The code I had for printing potentially long long variables in PH/11
4893 above was not the best (it lost precision). The length of off_t variables
4894 is now inspected at build time, and an appropriate printing format (%ld
4895 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4896 to be "long long int" or "long int". This is needed for the internal
4897 formatting function string_vformat().
4899 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4900 the configuration file to be ":syslog", then the script "guesses" where
4901 the logs files are, rather than using the compiled in default. In our
4902 case the guess is not the same as the compiled default, so the script
4903 suddenly stopped working when I started to use syslog. The patch checks
4904 to see if log_file_path is "". If so, it attempts to read it from exim
4905 with no configuration file to get the compiled in version, before it
4906 falls back to the previous guessing code."
4908 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4909 implementing BATV in an Exim configuration. See NewStuff for the gory
4912 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4913 Makefile that are specific to HP-UX.
4915 PH/18 If the "use_postmaster" option was set for a recipient callout together
4916 with the "random" option, the postmaster address was used as the MAIL
4917 FROM address for the random test, but not for the subsequent recipient
4918 test. It is now used for both.
4920 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4921 patch removes a few documentation additions to RFC 3028, because the
4922 latest draft now contains them. It adds the new en;ascii-case comparator
4923 and a new error check for 8bit text in MIME parts. Comparator and
4924 require names are now matched exactly. I enabled the subaddress
4925 extension, but it is not well tested yet (read: it works for me)."
4927 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4928 rework some of the code of TK/09 above to avoid the hardwired use of
4929 "%lld" and "long long". Replaced the call to snprintf() with a call to
4932 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4933 records point to non-existent hosts", "retry timeout exceeded", and
4934 "retry time not reached for any host after a long failure period".
4936 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4937 experimental DomainKeys support:
4939 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4940 (2) On an error such as an illegally used "control", the wrong name for
4941 the control was given.
4943 These problems did NOT occur unless DomainKeys support was compiled.
4945 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4947 PH/24 Added ${if match_ip condition.
4949 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4950 databases so that it will be absolutely obvious if a crash occurs in the
4951 DB library. This is a regular occurrence (often caused by mis-matched
4954 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4955 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4956 picked up on a user's system that detects such things. There doesn't seem
4957 to be a gcc warning option for this - only an attribute that has to be
4958 put on the function's prototype. It seems that in Fedora Core 4 they have
4959 set this on a number of new functions. No doubt there will be more in due
4962 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4963 specify a transport. However, if an address that was verified by such a
4964 router was the subject of a callout, Exim crashed because it tried to
4965 read the rcpt_include_affixes from the non-existent transport. Now it
4966 just assumes that the setting of that option is false. This bug was
4967 introduced by 4.51/PH/31.
4969 PH/28 Changed -d+all to exclude +memory, because that information is very
4970 rarely of interest, but it makes the output a lot bigger. People tend to
4971 do -d+all out of habit.
4973 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4974 code in os-type was giving problems when libc.so lives in lib64, like on
4977 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4978 aren't the modern standard, and it seems that some systems' include files
4979 don't always have them. Exim was already checking for some of the newer
4980 ones like T_AAAA, and defining it itself. I've added checks for all the
4981 record types that Exim uses.
4983 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4984 not automatically generating a new one, as it is supposed to. This
4985 prevented TLS from working. If the file did exist, but contained invalid
4986 data, a new version was generated, as expected. It was only the case of a
4987 non-existent file that was broken.
4989 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4990 with a change in libdomainkeys > 0.64.
4992 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4993 from DNS. If the selector record carries the flag, it now has
4994 precedence over the domain-wide flag.
4996 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4998 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4999 the use of a port name that isn't defined in /etc/services) occurred, the
5000 message was deferred as in a normal delivery, and thus remained on the
5001 spool, instead of being failed because of the mua_wrapper setting. This
5002 is now fixed, and I tidied up some of the mua_wrapper messages at the
5005 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5006 the messages in a hash of arrays rather than using individual hashes.
5007 This is a bit cleaner and results in dramatic memory savings, albeit
5008 at a slight CPU cost.
5010 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5011 as requested by Marc Sherman.
5013 SC/10 Eximstats: added histograms for user specified patterns as requested
5016 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5018 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5019 fopen() in the content-scanning modules that did not already have it.
5025 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5026 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5028 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5030 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5032 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5033 to test Sieve filters that use "vacation".
5035 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5036 that changes the way the GnuTLS parameters are stored in the cache file.
5037 The new format can be generated externally. For backward compatibility,
5038 if the data in the cache doesn't make sense, Exim assumes it has read an
5039 old-format file, and it generates new data and writes a new file. This
5040 means that you can't go back to an older release without removing the
5043 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5044 work if there are any delivery delays because "one_time" forces the
5045 parent to be marked "delivered", so its unseen clone is never tried
5046 again. For this reason, Exim now forbids the simultaneous setting of
5049 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5050 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5051 entry if you want to know the details. The fix, however, neglected to
5052 consider the case when local delivery batching is involved. The test for
5053 "previously delivered" was not happening when checking to see if an
5054 address could be batched with a previous (undelivered) one; under
5055 certain circumstances this could lead to multiple deliveries to the same
5058 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5059 in its include files, and this causes problems building Exim.
5061 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5062 header_syntax) but Exim was just ignoring anything given after a slash.
5063 In particular, this caused confusion with an attempt to use "verify =
5064 reverse_host_lookup/defer_ok". An error is now given when options are
5065 supplied for verify items that do not have them. (Maybe reverse_host_
5066 lookup should have a defer_ok option, but that's a different point.)
5068 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5069 defined by RFC 821) to 2048, because there were problems with some AUTH
5070 commands, and RFC 1869 says the size should be increased for extended
5071 SMTP commands that take arguments.
5073 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5076 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5077 "unknown" error; now it says that the functionality isn't in the binary.
5079 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5080 an address' error message when a string expansion fails (syntax or
5081 whatever). Otherwise the password may appear in the log. Following change
5082 PH/42 below, there is no longer a chance of it appearing in a bounce
5085 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5087 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5088 of its addresses was ever tried. (Bugzilla bug #2.)
5090 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5091 the result incorrectly in the debug output. (It correctly added a newline
5092 to what was transported.)
5094 TF/01 Added $received_time.
5096 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5097 commented out examples of how to interface to a virus scanner and to
5098 SpamAssassin. Also added commented examples of av_scanner and
5099 spamd_address settings.
5101 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5102 and controls are allowed in which ACLs. There were a couple of minor
5103 errors. Some of the entries in the conditions table (which is a table of
5104 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5105 negation of where the condition IS allowed.
5107 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5109 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5110 header file does not have a version number, so I've had to invent a new
5111 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5112 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5113 radiusclient), but it was contributed by a Radius user.
5115 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5116 files or directories, for OpenSSL.
5118 PH/19 When an Exim process that is running as root has to create an Exim log
5119 file, it does so in a subprocess that runs as exim:exim so as to get the
5120 ownership right at creation (otherwise, other Exim processes might see
5121 the file with the wrong ownership). There was no test for failure of this
5122 fork() call, which would lead to the process getting stuck as it waited
5123 for a non-existent subprocess. Forks do occasionally fail when resources
5124 run out. I reviewed all the other calls to fork(); they all seem to check
5127 PH/20 When checking for unexpected SMTP input at connect time (before writing
5128 the banner), Exim was not dealing correctly with a non-positive return
5129 from the read() function. If the client had disconnected by this time,
5130 the result was a log entry for a synchronization error with an empty
5131 string after "input=" when read() returned zero. If read() returned -1
5132 (an event I could not check), uninitialized data bytes were printed.
5133 There were reports of junk text (parts of files, etc) appearing after
5136 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5138 PH/22 Added support for macro redefinition, and (re)definition in between
5139 driver and ACL definitions.
5141 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5142 forgetting to use the resulting value; it was using the unexpanded value.
5144 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5145 hadn't been configured. The fix is from Juergen Kreileder, who
5146 understands it better than I do:
5148 "Here's what I see happening with three configured cyrus_sasl
5149 authenticators configured (plain, login, cram-md5):
5151 On startup auth_cyrus_sasl_init() gets called for each of these.
5152 This means three calls to sasl_listmech() without a specified mech_list.
5153 => SASL tests which mechs of all available mechs actually work
5154 => three warnings about OTP not working
5155 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5157 With the patch, sasl_listmech() also gets called three times. But now
5158 SASL's mech_list option is set to the server_mech specified in the the
5159 authenticator. Or in other words, the answer from sasl_listmech()
5160 gets limited to just the mech you're testing for (which is different
5162 => the return list contains just 'plain' or 'login', 'cram-md5' or
5163 nothing depending on the value of ob->server_mech.
5165 I've just tested the patch: Authentication still works fine,
5166 unavailable mechs specified in the exim configuration are still
5167 caught, and the auth.log warnings about OTP are gone."
5169 PH/25 When debugging is enabled, the contents of the command line are added
5170 to the debugging output, even when log_selector=+arguments is not
5173 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5174 answer is "GNU", and only if the return is "GNU/something" is the answer
5177 PH/27 $acl_verify_message is now set immediately after the failure of a
5178 verification in an ACL, and so is available in subsequent modifiers. In
5179 particular, the message can be preserved by coding like this:
5181 warn !verify = sender
5182 set acl_m0 = $acl_verify_message
5184 Previously, $acl_verify_message was set only while expanding "message"
5185 and "log_message" when a very denied access.
5187 PH/28 Modified OS/os.c-Linux with
5189 -#ifndef OS_LOAD_AVERAGE
5190 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5192 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5193 nomenclature these days.)
5195 PH/29 Installed patch from the Sieve maintainer that adds the options
5196 sieve_useraddress and sieve_subaddress to the redirect router.
5198 PH/30 In these circumstances:
5199 . Two addresses routed to the same list of hosts;
5200 . First host does not offer TLS;
5201 . First host accepts first address;
5202 . First host gives temporary error to second address;
5203 . Second host offers TLS and a TLS session is established;
5204 . Second host accepts second address.
5205 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5206 and peerdn, if requested) that were in fact used only for the second
5209 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5210 attention to any local part prefix or suffix that was matched by the
5211 router that accepted the address. It now behaves in the same way as it
5212 does for delivery: the affixes are removed from the local part unless
5213 rcpt_include_affixes is set on the transport.
5215 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5216 timeout during the DATA phase of an incoming message.
5218 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5219 applied a patch sanctioned by the Sieve maintainer.
5221 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5222 the uid or gid is negative. A case of a negative gid caused this to be
5223 noticed. The fix allows for either to be negative.
5225 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5226 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5229 PH/36 The change PH/12 above was broken. Fixed it.
5231 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5232 the grounds that routing the same address twice would always produce the
5233 same answer. This might have been true once, but it is certainly no
5234 longer true now. Routing a child address may depend on the previous
5235 routing that produced that child. Some complicated redirection strategies
5236 went wrong when messages had multiple recipients, and made Exim's
5237 behaviour dependent on the order in which the addresses were given.
5239 I have moved the duplicate checking until after the routing is complete.
5240 Exim scans the addresses that are assigned to local and remote
5241 transports, and removes any duplicates. This means that more work will be
5242 done, as duplicates will always all be routed, but duplicates are
5243 presumably rare, so I don't expect this is of any significance.
5245 For deliveries to pipes, files, and autoreplies, the duplicate checking
5246 still happens during the routing process, since they are not going to be
5249 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5250 It corrects a timeout issue with spamd. This is Ian's comment: "The
5251 background is that sometimes spamd either never reads data from a
5252 connection it has accepted, or it never writes response data. The exiscan
5253 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5254 blindly assumes that writes won't block so it may never time out."
5256 PH/39 Allow G after quota size as well as K and M.
5258 PH/40 The value set for $authenticated_id in an authenticator may not contain
5259 binary zeroes or newlines because the value is written to log lines and
5260 to spool files. There was no check on this. Now the value is run through
5261 the string_printing() function so that such characters are converted to
5262 printable escape sequences.
5264 PH/41 $message_linecount is a new variable that contains the total number of
5265 lines in the message. Compare $body_linecount, which is the count for the
5268 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5269 bounce and delay warning messages, except in certain special cases, which
5272 (a) An SMTP error message from a remote host;
5273 (b) A message specified in a :fail: redirection;
5274 (c) A message specified in a "fail" command in a system filter;
5275 (d) A message specified in a FAIL return from the queryprogram router;
5276 (e) A message specified by the cannot_route_message router option.
5278 In these cases only, Exim does include the error details in bounce and
5279 warning messages. There are also a few cases where bland messages such
5280 as "unrouteable address" or "local delivery error" are given.
5282 PH/43 $value is now also set for the "else" part of a ${run expansion.
5284 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5285 being worked on, but at least Exim now implements the latest version to
5288 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5289 process to complete was treated as a delivery failure, a timeout while
5290 writing the message to the pipe was logged, but erroneously treated as a
5291 successful delivery. Such timeouts include transport filter timeouts. For
5292 consistency with the overall process timeout, these timeouts are now
5293 treated as errors, giving rise to delivery failures by default. However,
5294 there is now a new Boolean option for the pipe transport called
5295 timeout_defer, which, if set TRUE, converts the failures into defers for
5296 both kinds of timeout. A transport filter timeout is now identified in
5299 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5300 systems where "make" and "gmake" are different, calling "gmake" at top
5301 level broke things. I've arranged for the value of $(MAKE) to be passed
5302 from the Makefile to this script so that it can call the same version of
5306 A note about Exim versions 4.44 and 4.50
5307 ----------------------------------------
5309 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5310 changes of various kinds. As a consequence, a big documentation update was
5311 needed. This delayed the release for rather longer than seemed good, especially
5312 in the light of a couple of (minor) security issues. Therefore, the changes
5313 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5314 release. So 4.44 and 4.50 are in effect two different branches that both start
5317 I have left the 4.50 change log unchanged; it contains all the changes since
5318 4.43. The change log for 4.44 is below; many of its items are identical to
5319 those for 4.50. This seems to be the most sensible way to preserve the
5320 historical information.
5326 1. Minor wording change to the doc/README.SIEVE file.
5328 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5329 computation of the current number of files was incorrect.
5331 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5332 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5335 4. Give more explanation in the error message when the command for a transport
5336 filter fails to execute.
5338 5. There are several places where Exim runs a non-Exim command in a
5339 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5340 was being done only for the command run by the queryprogram router. It is
5341 now done for all such subprocesses. The other cases are: ${run, transport
5342 filters, and the commands run by the lmtp and pipe transports.
5344 6. Added CONFIGURE_GROUP build-time option.
5346 7. Some older OS have a limit of 256 on the maximum number of file
5347 descriptors. Exim was using setrlimit() to set 1000 as a large value
5348 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5349 systems. I've change it so that if it can't get 1000, it tries for 256.
5351 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5352 was an oversight, and furthermore, ever since the addition of extra
5353 controls (e.g. 4.43/32), the checks on when to allow different forms of
5354 "control" were broken. There should now be diagnostics for all cases when a
5355 control that does not make sense is encountered.
5357 9. Added the /retain_sender option to "control=submission".
5359 10. $recipients is now available in the predata ACL (oversight).
5361 11. Tidy the search cache before the fork to do a delivery from a message
5362 received from the command line. Otherwise the child will trigger a lookup
5363 failure and thereby defer the delivery if it tries to use (for example) a
5364 cached ldap connection that the parent has called unbind on.
5366 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5367 of $address_data from the recipient verification was clobbered by the
5368 sender verification.
5370 13. The value of address_data from a sender verification is now available in
5371 $sender_address_data in subsequent conditions in the ACL statement.
5373 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5375 15. Added a new option "connect=<time>" to callout options, to set a different
5378 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5379 was its contents. (It was OK if the option was not defined at all.)
5381 17. A "Completed" log line is now written for messages that are removed from
5382 the spool by the -Mrm option.
5384 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5385 information about exactly what failed.
5387 19. Added -dd to debug only the daemon process.
5389 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5390 handles timeouts, both on the server side and network timeouts. Renamed the
5391 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5393 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5394 It is now set to "smtps".
5396 22. $host_address is now set to the target address during the checking of
5397 ignore_target_hosts.
5399 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5400 being passed; this would have caused $sender_host_name to have been used if
5401 matching the list had actually called for a host name (not very likely,
5402 since this list is usually IP addresses). A host name is now passed as
5405 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5406 code that specifies a non-restarting handler (typically sigaction() in
5407 modern systems) in an attempt to fix a rare and obscure crash bug.
5409 25. Narrowed the window for a race in the daemon that could cause it to ignore
5410 SIGCHLD signals. This is not a major problem, because they are used only to
5411 wake it up if nothing else does.
5413 26. A malformed maildirsize file could cause Exim to calculate negative values
5414 for the mailbox size or file count. Odd effects could occur as a result.
5415 The maildirsize information is now recalculated if the size or filecount
5418 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5419 support for a long time. Removed HAVE_SYS_VFS_H.
5421 28. Installed the latest version of exipick from John Jetmore.
5423 29. In an address list, if the pattern was not a regular expression, an empty
5424 subject address (from a bounce message) matched only if the pattern was an
5425 empty string. Non-empty patterns were not even tested. This was the wrong
5426 because it is perfectly reasonable to use an empty address as part of a
5427 database query. An empty address is now tested by patterns that are
5428 lookups. However, all the other forms of pattern expect the subject to
5429 contain a local part and a domain, and therefore, for them, an empty
5430 address still always fails if the pattern is not itself empty.
5432 30. Exim went into a mad DNS loop when attempting to do a callout where the
5433 host was specified on an smtp transport, and looking it up yielded more
5434 than one IP address.
5436 31. Re-factored the code for checking spool and log partition space into a
5437 function that finds that data and another that does the check. The former
5438 is then used to implement four new variables: $spool_space, $log_space,
5439 $spool_inodes, and $log_inodes.
5441 32. The RFC2047 encoding function was originally intended for short strings
5442 such as real names; it was not keeping to the 75-character limit for
5443 encoded words that the RFC imposes. It now respects the limit, and
5444 generates multiple encoded words if necessary. To be on the safe side, I
5445 have increased the buffer size for the ${rfc2047: expansion operator from
5448 33. It is now permitted to omit both strings after an "if" condition; if the
5449 condition is true, the result is "true". As before, when the second string
5450 is omitted, a false condition yields an empty string. This makes it less
5451 cumbersome to write custom ACL and router conditions.
5453 34. Failure to deliver a bounce message always caused it to be frozen, even if
5454 there was an errors_to setting on the router. The errors_to setting is now
5457 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5458 canonical form (fully expanded) before being placed in
5459 $sender_host_address.
5461 36. The table in the code that translates DNS record types into text (T_A to
5462 "A" for instance) was missing entries for NS and CNAME. It is just possible
5463 that this could have caused confusion if both these types were looked up
5464 for the same domain, because the text type is used as part of Exim's
5465 per-process caching. But the chance of anyone hitting this buglet seems
5468 37. The dnsdb lookup has been extended in a number of ways.
5470 (1) There is a new type, "zns", which walks up the domain tree until it
5471 finds some nameserver records. It should be used with care.
5473 (2) There is a new type, "mxh", which is like "mx" except that it returns
5474 just the host names, not the priorities.
5476 (3) It is now possible to give a list of domains (or IP addresses) to be
5477 looked up. The behaviour when one of the lookups defers can be
5478 controlled by a keyword.
5480 (4) It is now possible to specify the separator character for use when
5481 multiple records are returned.
5483 38. The dnslists ACL condition has been extended: it is now possible to supply
5484 a list of IP addresses and/or domains to be looked up in a particular DNS
5487 39. Added log_selector=+queue_time_overall.
5489 40. When running the queue in the test harness, wait just a tad after forking a
5490 delivery process, to get repeatability of debugging output.
5492 41. Include certificate and key file names in error message when GnuTLS fails
5493 to set them up, because the GnuTLS error message doesn't include the name
5494 of the failing file when there is a problem reading it.
5496 42. Allow both -bf and -bF in the same test run.
5498 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5500 44. The "Exiscan patch" is now merged into the mainline Exim source.
5502 45. Sometimes the final signoff response after QUIT could fail to get
5503 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5504 < 0 before doing a fflush(). This bug looks as though it goes back to the
5505 introduction of TLS in release 3.20, but "sometimes" must have been rare
5506 because the tests only now provoked it.
5508 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5509 (this can affect the format of dates).
5511 47. exim_tidydb, when checking for the continued existence of a message for
5512 which it has found a message-specific retry record, was not finding
5513 messages that were in split spool directories. Consequently, it was
5514 deleting retry records that should have stayed in existence.
5516 48. Steve fixed some bugs in eximstats.
5518 49. The SPA authentication driver was not abandoning authentication and moving
5519 on to the next authenticator when an expansion was forced to fail,
5520 contradicting the general specification for all authenticators. Instead it
5521 was generating a temporary error. It now behaves as specified.
5523 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5524 (the order specifies the preference for clients). The order is now AES256,
5525 AES128, 3DES, ARCFOUR128.
5527 51. Small patch to Sieve code - explicitly set From: when generating an
5530 52. Exim crashed if a remote delivery caused a very long error message to be
5531 recorded - for instance if somebody sent an entire SpamAssassin report back
5532 as a large number of 550 error lines. This bug was coincidentally fixed by
5533 increasing the size of one of Exim's internal buffers (big_buffer) that
5534 happened as part of the Exiscan merge. However, to be on the safe side, I
5535 have made the code more robust (and fixed the comments that describe what
5538 53. Now that there can be additional text after "Completed" in log lines (if
5539 the queue_time_overall log selector is set), a one-byte patch to exigrep
5540 was needed to allow it to recognize "Completed" as not the last thing in
5543 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5544 patch that reportedly fixes this has been added. I am not expert enough to
5545 create a test for it. This is what the patch creator wrote:
5547 "I found a little strange behaviour of ldap code when working with
5548 Windows 2003 AD Domain, where users was placed in more than one
5549 Organization Units. When I tried to give exim partial DN, the exit code
5550 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5551 But simultaneously result of request was absolutely normal ldap result,
5552 so I produce this patch..."
5554 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5555 REFERENCE, so I have modified the code to exclude the patch when that macro
5558 55. Some experimental protocols are using DNS PTR records for new purposes. The
5559 keys for these records are domain names, not reversed IP addresses. The
5560 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5561 leaves it alone. Component reversal etc. now happens only for IP addresses.
5564 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5566 57. Double the size of the debug message buffer (to 2048) so that more of very
5567 long debug lines gets shown.
5569 58. The exicyclog utility now does better if the number of log files to keep
5570 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5572 59. Two changes related to the smtp_active_hostname option:
5574 (1) $smtp_active_hostname is now available as a variable.
5575 (2) The default for smtp_banner uses $smtp_active_hostname instead
5576 of $primary_hostname.
5578 60. The host_aton() function is supposed to be passed a string that is known
5579 to be a valid IP address. However, in the case of IPv6 addresses, it was
5580 not checking this. This is a hostage to fortune. Exim now panics and dies
5581 if the condition is not met. A case was found where this could be provoked
5582 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5583 components; fortuitously, this particular loophole had already been fixed
5584 by change 4.50/55 above.
5586 If there are any other similar loopholes, the new check in host_aton()
5587 itself should stop them being exploited. The report I received stated that
5588 data on the command line could provoke the exploit when Exim was running as
5589 exim, but did not say which command line option was involved. All I could
5590 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5591 running as the user.
5594 61. There was a buffer overflow vulnerability in the SPA authentication code
5595 (which came originally from the Samba project). I have added a test to the
5596 spa_base64_to_bits() function which I hope fixes it.
5599 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5600 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5602 63. The daemon start-up calls getloadavg() while still root for those OS that
5603 need the first call to be done as root, but it missed one case: when
5604 deliver_queue_load_max is set with deliver_drop_privilege. This is
5605 necessary for the benefit of the queue runner, because there is no re-exec
5606 when deliver_drop_privilege is set.
5608 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5609 This has been fixed.
5611 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5612 was in use, was not putting the data itself into the right store pool;
5613 consequently, it could be overwritten for a subsequent message in the same
5614 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5617 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5619 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5620 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5621 the function were treating the return as a boolean value, which happened to
5622 work because 0=false and not-0=true, but is not correct code.
5624 68. The host_aton() function was not handling scoped IPv6 addresses (those
5625 with, for example, "%eth0" on the end) correctly.
5627 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5628 negated items (that is, ~something) in unsigned ints. Some compilers
5629 apparently mutter when there is no cast.
5631 70. If an address verification called from an ACL failed, and did not produce a
5632 user-specific message (i.e. there was only a "system" message), nothing was
5633 put in $acl_verify_message. In this situation, it now puts the system
5636 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5637 session; change 4.31/43 added the unwanted input to the log line - except
5638 that it did not do this in the start of session case. It now does.
5640 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5641 This is wrong and can cause the other end to generate a synchronization
5642 error if it is another Exim or anything else that does the synchronization
5643 check. A QUIT command is no longer sent after a timeout.
5645 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5646 during host lookups.
5648 74. The defer_ok option of callout verification was not working if it was used
5649 when verifying addresses in header lines, that is, for this case:
5651 verify = header_sender/callout=defer_ok
5653 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5654 those file descriptors could be used for SMTP connections. If anything
5655 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5656 could be sent to the SMTP client, causing chaos. The daemon now opens
5657 stdin, stdout, and stderr to /dev/null when it puts itself into the
5660 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5661 log by default. The user can override this with suitable Perl magic.
5663 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5664 the log message when discard triggers, was not working for the DATA ACL or
5665 for the non-SMTP ACL.
5667 78. Error message wording change in sieve.c.
5669 79. If smtp_accept_max_per_host was set, the number of connections could be
5670 restricted to fewer than expected, because the daemon was trying to set up
5671 a new connection before checking whether the processes handling previous
5672 connections had finished. The check for completed processes is now done
5673 earlier. On busy systems, this bug wouldn't be noticed because something
5674 else would have woken the daemon, and it would have reaped the completed
5677 80. If a message was submitted locally by a user whose login name contained one
5678 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5679 It caused a spool format error. I have fixed the spool reading code. A
5680 related problem was that the "from" clause in the Received: line became
5681 illegal because of the space(s). It is now covered by ${quote_local_part.
5683 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5686 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5688 83. Patch from Sieve maintainer to fix unterminated string problem in
5689 "vacation" handling.
5691 84. Some minor changes to the Linux configuration files to help with other
5692 OS variants using glibc.
5694 85. One more patch for Sieve to update vacation handling to latest spec.
5697 ----------------------------------------------------
5698 See the note above about the 4.44 and 4.50 releases.
5699 ----------------------------------------------------
5705 1. Change 4.43/35 introduced a bug that caused file counts to be
5706 incorrectly computed when quota_filecount was set in an appendfile
5709 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5710 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5713 3. Give more explanation in the error message when the command for a transport
5714 filter fails to execute.
5716 4. There are several places where Exim runs a non-Exim command in a
5717 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5718 was being done only for the command run by the queryprogram router. It is
5719 now done for all such subprocesses. The other cases are: ${run, transport
5720 filters, and the commands run by the lmtp and pipe transports.
5722 5. Some older OS have a limit of 256 on the maximum number of file
5723 descriptors. Exim was using setrlimit() to set 1000 as a large value
5724 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5725 systems. I've change it so that if it can't get 1000, it tries for 256.
5727 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5728 was an oversight, and furthermore, ever since the addition of extra
5729 controls (e.g. 4.43/32), the checks on when to allow different forms of
5730 "control" were broken. There should now be diagnostics for all cases when a
5731 control that does not make sense is encountered.
5733 7. $recipients is now available in the predata ACL (oversight).
5735 8. Tidy the search cache before the fork to do a delivery from a message
5736 received from the command line. Otherwise the child will trigger a lookup
5737 failure and thereby defer the delivery if it tries to use (for example) a
5738 cached ldap connection that the parent has called unbind on.
5740 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5741 of $address_data from the recipient verification was clobbered by the
5742 sender verification.
5744 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5745 was its contents. (It was OK if the option was not defined at all.)
5747 11. A "Completed" log line is now written for messages that are removed from
5748 the spool by the -Mrm option.
5750 12. $host_address is now set to the target address during the checking of
5751 ignore_target_hosts.
5753 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5754 being passed; this would have caused $sender_host_name to have been used if
5755 matching the list had actually called for a host name (not very likely,
5756 since this list is usually IP addresses). A host name is now passed as
5759 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5760 code that specifies a non-restarting handler (typically sigaction() in
5761 modern systems) in an attempt to fix a rare and obscure crash bug.
5763 15. Narrowed the window for a race in the daemon that could cause it to ignore
5764 SIGCHLD signals. This is not a major problem, because they are used only to
5765 wake it up if nothing else does.
5767 16. A malformed maildirsize file could cause Exim to calculate negative values
5768 for the mailbox size or file count. Odd effects could occur as a result.
5769 The maildirsize information is now recalculated if the size or filecount
5772 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5773 support for a long time. Removed HAVE_SYS_VFS_H.
5775 18. Updated exipick to current release from John Jetmore.
5777 19. Allow an empty sender to be matched against a lookup in an address list.
5778 Previously the only cases considered were a regular expression, or an
5781 20. Exim went into a mad DNS lookup loop when doing a callout where the
5782 host was specified on the transport, if the DNS lookup yielded more than
5785 21. The RFC2047 encoding function was originally intended for short strings
5786 such as real names; it was not keeping to the 75-character limit for
5787 encoded words that the RFC imposes. It now respects the limit, and
5788 generates multiple encoded words if necessary. To be on the safe side, I
5789 have increased the buffer size for the ${rfc2047: expansion operator from
5792 22. Failure to deliver a bounce message always caused it to be frozen, even if
5793 there was an errors_to setting on the router. The errors_to setting is now
5796 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5797 canonical form (fully expanded) before being placed in
5798 $sender_host_address.
5800 24. Updated eximstats to version 1.33
5802 25. Include certificate and key file names in error message when GnuTLS fails
5803 to set them up, because the GnuTLS error message doesn't include the name
5804 of the failing file when there is a problem reading it.
5806 26. Expand error message when OpenSSL has problems setting up cert/key files.
5809 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5810 (this can affect the format of dates).
5812 28. exim_tidydb, when checking for the continued existence of a message for
5813 which it has found a message-specific retry record, was not finding
5814 messages that were in split spool directories. Consequently, it was
5815 deleting retry records that should have stayed in existence.
5817 29. eximstats updated to version 1.35
5818 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5819 1.35 - bugfix such that pie charts by volume are generated correctly
5821 30. The SPA authentication driver was not abandoning authentication and moving
5822 on to the next authenticator when an expansion was forced to fail,
5823 contradicting the general specification for all authenticators. Instead it
5824 was generating a temporary error. It now behaves as specified.
5826 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5827 (the order specifies the preference for clients). The order is now AES256,
5828 AES128, 3DES, ARCFOUR128.
5830 31. Small patch to Sieve code - explicitly set From: when generating an
5833 32. Exim crashed if a remote delivery caused a very long error message to be
5834 recorded - for instance if somebody sent an entire SpamAssassin report back
5835 as a large number of 550 error lines. This bug was coincidentally fixed by
5836 increasing the size of one of Exim's internal buffers (big_buffer) that
5837 happened as part of the Exiscan merge. However, to be on the safe side, I
5838 have made the code more robust (and fixed the comments that describe what
5841 33. Some experimental protocols are using DNS PTR records for new purposes. The
5842 keys for these records are domain names, not reversed IP addresses. The
5843 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5844 leaves it alone. Component reversal etc. now happens only for IP addresses.
5847 34. The host_aton() function is supposed to be passed a string that is known
5848 to be a valid IP address. However, in the case of IPv6 addresses, it was
5849 not checking this. This is a hostage to fortune. Exim now panics and dies
5850 if the condition is not met. A case was found where this could be provoked
5851 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5852 components; fortuitously, this particular loophole had already been fixed
5853 by change 4.50/55 or 4.44/33 above.
5855 If there are any other similar loopholes, the new check in host_aton()
5856 itself should stop them being exploited. The report I received stated that
5857 data on the command line could provoke the exploit when Exim was running as
5858 exim, but did not say which command line option was involved. All I could
5859 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5860 running as the user.
5863 35. There was a buffer overflow vulnerability in the SPA authentication code
5864 (which came originally from the Samba project). I have added a test to the
5865 spa_base64_to_bits() function which I hope fixes it.
5868 36. The daemon start-up calls getloadavg() while still root for those OS that
5869 need the first call to be done as root, but it missed one case: when
5870 deliver_queue_load_max is set with deliver_drop_privilege. This is
5871 necessary for the benefit of the queue runner, because there is no re-exec
5872 when deliver_drop_privilege is set.
5874 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5875 was in use, was not putting the data itself into the right store pool;
5876 consequently, it could be overwritten for a subsequent message in the same
5877 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5880 38. Sometimes the final signoff response after QUIT could fail to get
5881 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5882 < 0 before doing a fflush(). This bug looks as though it goes back to the
5883 introduction of TLS in release 3.20, but "sometimes" must have been rare
5884 because the tests only now provoked it.
5890 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5891 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5892 Now it returns an integer. A number of calls were still expecting a T/F
5893 return. Fortuitously, in all cases, the tests worked in OK situations,
5894 which is the norm. However, things would have gone wrong on any write
5895 failures on the smtp file descriptor. This function is used when sending
5896 messages over SMTP and also when doing verify callouts.
5898 2. When Exim is called to do synchronous delivery of a locally submitted
5899 message (the -odf or -odi options), it no longer closes stderr before doing
5902 3. Implemented the mua_wrapper option.
5904 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5906 5. Implemented the functions header_remove(), header_testname(),
5907 header_add_at_position(), and receive_remove_recipient(), and exported them
5910 6. If an ACL "warn" statement specified the addition of headers, Exim already
5911 inserted X-ACL-Warn: at the start if there was no header name. However, it
5912 was not making this test for the second and subsequent header lines if
5913 there were newlines in the string. This meant that an invalid header could
5914 be inserted if Exim was badly configured.
5916 7. Allow an ACL "warn" statement to add header lines at the start or after all
5917 the Received: headers, as well as at the end.
5919 8. Added the rcpt_4xx retry error code.
5921 9. Added postmaster_mailfrom=xxx to callout verification option.
5923 10. Added mailfrom=xxxx to the callout verification option, for verify=
5926 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5927 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5929 12. Inserted some casts to stop certain compilers warning when using pointer
5930 differences as field lengths or precisions in printf-type calls (mostly
5931 affecting debugging statements).
5933 13. Added optional readline() support for -be (dynamically loaded).
5935 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5936 same clock tick as a message's arrival, so that its received time was the
5937 same as the "first fail" time on the retry record, and that message
5938 remained on the queue past the ultimate address timeout, every queue runner
5939 would try a delivery (because it was past the ultimate address timeout) but
5940 after another failure, the ultimate address timeout, which should have then
5941 bounced the address, did not kick in. This was a "< instead of <=" error;
5942 in most cases the first failure would have been in the next clock tick
5943 after the received time, and all would be well.
5945 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5946 being recognized when the domain list was tested by the match_domain
5947 condition in an expansion string.
5949 16. Added the ${str2b64: operator.
5951 17. Exim was always calling setrlimit() to set a large limit for the number of
5952 processes, without checking whether the existing limit was already
5953 adequate. (It did check for the limit on file descriptors.) Furthermore,
5954 errors from getrlimit() and setrlimit() were being ignored. Now they are
5955 logged to the main and panic logs, but Exim does carry on, to try to do its
5956 job under whatever limits there are.
5958 18. Imported PCRE 5.0.
5960 19. Trivial typo in log message " temporarily refused connection" (the leading
5963 20. If the log selector return_path_on_delivery was set and an address was
5964 redirected to /dev/null, the delivery process crashed because it assumed
5965 that a return path would always be set for a "successful" delivery. In this
5966 case, the whole delivery is bypassed as an optimization, and therefore no
5969 21. Internal re-arrangement: the function for sending a challenge and reading
5970 a response while authentication was assuming a zero-terminated challenge
5971 string. It's now changed to take a pointer and a length, to allow for
5972 binary data in such strings.
5974 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5976 23. Exim was not respecting finduser_retries when seeking the login of the
5977 uid under which it was called; it was always trying 10 times. (The default
5978 setting of finduser_retries is zero.) Also, it was sleeping after the final
5979 failure, which is pointless.
5981 24. Implemented tls_on_connect_ports.
5983 25. Implemented acl_smtp_predata.
5985 26. If the domain in control=submission is set empty, Exim assumes that the
5986 authenticated id is a complete email address when it generates From: or
5987 Sender: header lines.
5989 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5990 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5991 chown and chgrp in /bin and hostname in /usr/bin.
5993 28. Exim was keeping the "process log" file open after each use, just as it
5994 does for the main log. This opens the possibility of it remaining open for
5995 long periods when the USR1 signal hits a daemon. Occasional processlog
5996 errors were reported, that could have been caused by this. Anyway, it seems
5997 much more sensible not to leave this file open at all, so that is what now
6000 29. The long-running daemon process does not normally write to the log once it
6001 has entered its main loop, and it closes the log before doing so. This is
6002 so that log files can straightforwardly be renamed and moved. However,
6003 there are a couple of unusual error situations where the daemon does write
6004 log entries, and I had neglected to close the log afterwards.
6006 30. The text of an SMTP error response that was received during a remote
6007 delivery was being truncated at 512 bytes. This is too short for some of
6008 the long messages that one sometimes sees. I've increased the limit to
6011 31. It is now possible to make retry rules that apply only when a message has a
6012 specific sender, in particular, an empty sender.
6014 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6015 it possible to be selective about when SMTP synchronization is enforced.
6017 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6019 32. Implemented hosts_connection_nolog.
6021 33. Added an ACL for QUIT.
6023 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6026 35. Added mailbox_size and mailbox_filecount to appendfile.
6028 36. Added control = no_multiline_responses to ACLs.
6030 37. There was a bug in the logic of the code that waits for the clock to tick
6031 in the case where the clock went backwards by a substantial amount such
6032 that the microsecond fraction of "now" was more than the microsecond
6033 fraction of "then" (but the whole seconds number was less).
6035 38. Added support for the libradius Radius client library this is found on
6036 FreeBSD (previously only the radiusclient library was supported).
6042 1. When certain lookups returned multiple values in the form name=value, the
6043 quoting of the values was not always being done properly. Specifically:
6044 (a) If the value started with a double quote, but contained no whitespace,
6046 (b) If the value contained whitespace other than a space character (i.e.
6047 tabs or newlines or carriage returns) it was not quoted.
6048 This fix has been applied to the mysql and pgsql lookups by writing a
6049 separate quoting function and calling it from the lookup code. The fix
6050 should probably also be applied to nisplus, ibase and oracle lookups, but
6051 since I cannot test any of those, I have not disturbed their existing code.
6053 2. A hit in the callout cache for a specific address caused a log line with no
6054 reason for rejecting RCPT. Now it says "Previous (cached) callout
6055 verification failure".
6057 3. There was an off-by-one bug in the queryprogram router. An over-long
6058 return line was truncated at 256 instead of 255 characters, thereby
6059 overflowing its buffer with the terminating zero. As well as fixing this, I
6060 have increased the buffer size to 1024 (and made a note to document this).
6062 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6063 when Exim is waiting for an SMTP response from a remote server, Exim
6064 restarts its select() call on the socket, thereby resetting its timeout.
6065 This is not a problem when such interrupts are rare. Somebody set up a cron
6066 job to run exiwhat every 2 minutes, which is less than the normal select()
6067 timeout (5 or 10 minutes). This meant that the select() timeout never
6068 kicked in because it was always reset. I have fixed this by comparing the
6069 time when an interrupt arrives with the time at the start of the first call
6070 to select(). If more time than the timeout has elapsed, the interrupt is
6071 treated as a timeout.
6073 5. Some internal re-factoring in preparation for the addition of Sieve
6074 extensions (by MH). In particular, the "personal" test is moved to a
6075 separate function, and given an option for scanning Cc: and Bcc: (which is
6076 not set for Exim filters).
6078 6. When Exim created an email address using the login of the caller as the
6079 local part (e.g. when creating a From: or Sender: header line), it was not
6080 quoting the local part when it contained special characters such as @.
6082 7. Installed new OpenBSD configuration files.
6084 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6085 try to make them clearer.
6087 9. Callout options, other than the timeout value, were being ignored when
6088 verifying sender addresses in header lines. For example, when using
6090 verify = header_sender/callout=no_cache
6092 the cache was (incorrectly) being used.
6094 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6095 only the Cygwin environment.
6097 11. When return_path_on_delivery was set as a log selector, if different remote
6098 addresses in the same message used different return paths and parallel
6099 remote delivery occurred, the wrong values would sometimes be logged.
6100 (Whenever a remote delivery process finished, the return path value from
6101 the most recently started remote delivery process was logged.)
6103 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6104 lines when AUTH and/or TLS are in use. This is the "received protocol"
6105 field. Exim used to use "asmtp" for authenticated SMTP, without any
6106 indication (in the protocol name) for TLS use. Now it follows the RFC and
6107 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6108 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6109 names appear in log lines as well as in Received: header lines.
6111 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6112 extensions, and comparison tests, and to fix some bugs.
6114 14. Changes to the "personal" filter test:
6116 (1) The test was buggy in that it was just doing the equivalent of
6117 "contains" tests on header lines. For example, if a user's address was
6118 anne@some.where, the "personal" test would incorrectly be true for
6120 To: susanne@some.where
6122 This test is now done by extracting each address from the header in turn,
6123 and checking the entire address. Other tests that are part of "personal"
6124 are now done using regular expressions (for example, to check local parts
6125 of addresses in From: header lines).
6127 (2) The list of non-personal local parts in From: addresses has been
6128 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6129 taken from the Sieve specification recommendations.
6131 (3) If the message contains any header line starting with "List-" it is
6132 treated as non-personal.
6134 (4) The test for "circular" in the Subject: header line has been removed
6135 because it now seems ill-conceived.
6137 15. Minor typos in src/EDITME comments corrected.
6139 16. Installed latest exipick from John Jetmore.
6141 17. If headers_add on a router specified a text string that was too long for
6142 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6143 of string_sprintf() is now avoided.
6145 18. $message_body_size was not set (it was always zero) when running the DATA
6146 ACL and the local_scan() function.
6148 19. For the "mail" command in an Exim filter, no default was being set for
6149 the once_repeat time, causing a random time value to be used if "once" was
6150 specified. (If the value happened to be <= 0, no repeat happened.) The
6151 default is now 0s, meaning "never repeat". The "vacation" command was OK
6152 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6153 (I found it when inspecting the code).
6155 20. There is now an overall timeout for performing a callout verification. It
6156 defaults to 4 times the callout timeout, which applies to individual SMTP
6157 commands during the callout. The overall timeout applies when there is more
6158 than one host that can be tried. The timeout is checked before trying the
6159 next host. This prevents very long delays if there are a large number of
6160 hosts and all are timing out (e.g. when the network connections are timing
6161 out). The value of the overall timeout can be changed by specifying an
6162 additional sub-option for "callout", called "maxwait". For example:
6164 verify = sender/callout=5s,maxwait=20s
6166 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6167 to the end before writing, but this should make it even safer).
6169 22. Exim was forgetting that it had advertised PIPELINING for the second and
6170 subsequent messages on an SMTP connection. It was also not resetting its
6171 memory on STARTTLS and an internal HELO.
6173 23. When Exim logs an SMTP synchronization error within a session, it now
6174 records whether PIPELINING has been advertised or not.
6176 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6177 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6178 rather than long int.
6180 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6182 26. Added the never_mail option to autoreply.
6188 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6189 crash if the getsockname() call failed; this can happen if a connection is
6190 closed very soon after it is established. The problem was simply in the
6191 order in which certain operations were done, causing Exim to try to write
6192 to the SMTP stream before it had set up the file descriptor. The bug has
6193 been fixed by making things happen in the correct order.
6199 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6200 before the connection was closed, thus losing the rejection response.
6202 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6203 some early Solaris releases, but causes trouble in current releases where
6204 socklen_t is defined.
6206 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6209 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6212 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6213 that releasing the top of it at the end releases what was used for sub-
6214 expansions (unless the block got too big). However, discard this block if
6215 the first thing is a variable or header, so that we can use its block when
6216 it is dynamic (useful for very large $message_headers, for example).
6218 6. Lookups now cache *every* query, not just the most recent. A new, separate
6219 store pool is used for this. It can be recovered when all lookup caches are
6220 flushed. Lookups now release memory at the end of their result strings.
6221 This has involved some general refactoring of the lookup sources.
6223 7. Some code has been added to the store_xxx() functions to reduce the amount
6224 of flapping under certain conditions.
6226 8. log_incoming_interface used to affect only the <= reception log lines. Now
6227 it causes the local interface and port to be added to several more SMTP log
6228 lines, for example "SMTP connection from", and rejection lines.
6230 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6232 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6234 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6235 could be overwritten at the end of the current message (or the start of a
6236 new message if it was set in a HELO ACL). The value is now preserved for
6237 the duration of the SMTP connection.
6239 12. If a transport had a headers_rewrite setting, and a matching header line
6240 contained an unqualified address, that address was qualified, even if it
6241 did not match any rewriting rules. The underlying bug was that the values
6242 of the flags that permit the existence of unqualified sender and recipient
6243 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6244 non-local messages, and by -bnq for local messages) were not being
6245 preserved with the message after it was received.
6247 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6248 "next input=" as part of the text comprising the host identity instead of
6249 the correct text. The code was using the same buffer for two different
6250 strings. However, depending on which order the printing function evaluated
6251 its arguments, the bug did not always show up. Under Linux, for example, my
6252 test suite worked just fine.
6254 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6255 doesn't work with some older versions of Perl. It has been changed to "my",
6256 which in any case is probably the better facility to use.
6258 15. A really picky compiler found some instances of statements for creating
6259 error messages that either had too many or two few arguments for the format
6262 16. The size of the buffer for calls to the DNS resolver has been increased
6263 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6264 for addresses that have a lot of PTR records. This alleviates a problem; it
6265 does not fully solve it.
6267 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6268 buffer now truncates the list and logs the incident, which is the same
6269 action as happens when Exim is looking up a host name and its aliases.
6270 Previously in this situation something unpredictable would happen;
6271 sometimes it was "internal error: store_reset failed".
6273 18. If a server dropped the connection unexpectedly when an Exim client was
6274 using GnuTLS and trying to read a response, the client delivery process
6275 crashed while trying to generate an error log message.
6277 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6278 string, for example:
6280 warn message = H1: something\nH2: something
6282 the text was added as a single header line from Exim's point of view
6283 though it ended up OK in the delivered message. However, searching for the
6284 second and subsequent header lines using $h_h2: did not work. This has been
6285 fixed. Similarly, if a system filter added multiple headers in this way,
6286 the routers could not see them.
6288 20. Expanded the error message when iplsearch is called with an invalid key to
6289 suggest using net-iplsearch in a host list.
6291 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6292 ACLs are no longer actually imposed (and a message to that effect is
6295 22. If a "gecos" field in a passwd entry contained escaped characters, in
6296 particular, if it contained a \" sequence, Exim got it wrong when building
6297 a From: or a Sender: header from that name. A second bug also caused
6298 incorrect handling when an unquoted " was present following a character
6299 that needed quoting.
6301 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6302 was not being matched caselessly.
6304 24. Arranged for all hyphens in the exim.8 source to be escaped with
6307 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6308 properly. Recipient callout cache records were still being keyed to include
6309 the sender, even when use_sender was set false. This led to far more
6310 callouts that were necessary. The sender is no longer included in the key
6311 when use_sender is false.
6313 26. Added "control = submission" modifier to ACLs.
6315 27. Added the ${base62d: operator to decode base 62 numbers.
6317 28. dnsdb lookups can now access SRV records.
6319 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6320 the configuration file.
6322 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6323 (-v) mode. This makes the output for a verbose queue run more intelligible.
6325 31. Added a use_postmaster feature to recipient callouts.
6327 32. Added the $body_zerocount variable, containing the number of binary zero
6328 bytes in the message body.
6330 33. The time of last modification of the "new" subdirectory is now used as the
6331 "mailbox time last read" when there is a quota error for a maildir
6334 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6336 35. Added +ignore_unknown as a special item in host lists.
6338 36. Code for decoding IPv6 addresses in host lists is now included, even if
6339 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6340 address was recognized as an IP address, but was then not correctly decoded
6341 into binary, causing unexpected and incorrect effects when compared with
6348 1. Very minor rewording of debugging text in manualroute to say "list of
6349 hosts" instead of "hostlist".
6351 2. If verify=header_syntax was set, and a header line with an unqualified
6352 address (no domain) and a large number of spaces between the end of the
6353 name and the colon was received, the reception process suffered a buffer
6354 overflow, and (when I tested it) crashed. This was caused by some obsolete
6355 code that should have been removed. The fix is to remove it!
6357 3. When running in the test harness, delay a bit after writing a bounce
6358 message to get a bit more predictability in the log output.
6360 4. Added a call to search_tidyup() just before forking a reception process. In
6361 theory, someone could use a lookup in the expansion of smtp_accept_max_
6362 per_host which, without the tidyup, could leave open a database connection.
6364 5. Added the variables $recipient_data and $sender_data which get set from a
6365 lookup success in an ACL "recipients" or "senders" condition, or a router
6366 "senders" option, similar to $domain_data and $local_part_data.
6368 6. Moved the writing of debug_print from before to after the "senders" test
6371 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6372 problems for message scanning, either using a data ACL, or using
6373 local_scan() because the Received: header was not generated till after they
6374 were called (in order to set the time as the time of reception completion).
6375 I have revised the way this works. The header is now generated after the
6376 body is received, but before the ACL or local_scan() are called. After they
6377 are run, the timestamp in the header is updated.
6383 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6384 before starting a queue runner without re-exec. This happened only when
6385 deliver_drop_privilege was set or when the Exim user was set to root. The
6386 effect of the bug was that timeouts during subsequent deliveries caused
6387 crashes instead of being properly handled. The handler is now left at its
6388 default (and expected) setting.
6390 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6391 message, again when deliver_drop_privilege is set or Exim is run as root.
6392 The bug described in (1) was not present in this case, but the tidying up
6393 of the other signals was missing. I have made the two cases consistent.
6395 3. The ignore_target_hosts setting on a manualroute router was being ignored
6396 for hosts that were looked up using the /MX notation.
6398 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6401 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6402 operated on the sender address. After changing the $sender_address to <>
6403 for the sender address verify, Exim was re-instated it as the original
6404 (before rewriting) address, but remembering that it had rewritten it, so it
6405 wasn't rewriting it again. This bug also had the effect of breaking the
6406 sender address verification caching when the sender address was rewritten.
6408 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6409 This has been changed so that if the ip literal address matches
6410 ignore_target_hosts, the router declines.
6412 7. Added expansion conditions match_domain, match_address, and match_local_
6413 part (NOT match_host).
6415 8. The placeholder for the Received: header didn't have a length field set.
6417 9. Added code to Exim itself and to exim_lock to test for a specific race
6418 condition that could lead to file corruption when using MBX delivery. The
6419 issue is with the lockfile that is created in /tmp. If this file is removed
6420 after a process has opened it but before that process has acquired a lock,
6421 there is the potential for a second process to recreate the file and also
6422 acquire a lock. This could lead to two Exim processes writing to the file
6423 at the same time. The added code performs the same test as UW imapd; it
6424 checks after acquiring the lock that its file descriptor still refers to
6425 the same named file.
6427 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6428 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6431 11. Added the smtp_active_hostname option. If used, this will typically be made
6432 to depend on the incoming interface address. Because $interface_address is
6433 not set up until the daemon has forked a reception process, error responses
6434 that can happen earlier (such as "too many connections") no longer contain
6437 12. If an expansion in a condition on a "warn" statement fails because a lookup
6438 defers, the "warn" statement is abandoned, and the next ACL statement is
6439 processed. Previously this caused the whole ACL to be aborted.
6441 13. Added the iplsearch lookup type.
6443 14. Added ident_timeout as a log selector.
6445 15. Added tls_certificate_verified as a log selector.
6447 16. Added a global option tls_require_ciphers (compare the smtp transport
6448 option of the same name). This controls incoming TLS connections.
6450 17. I finally figured out how to make tls_require_ciphers do a similar thing
6451 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6452 before starting the TLS session.
6454 18. Tabs are now shown as \t in -bP output.
6456 19. If the log selector return_path_on_delivery was set, Exim crashed when
6457 bouncing a message because it had too many Received: header lines.
6459 20. If two routers both had headers_remove settings, and the first one included
6460 a superfluous trailing colon, the final name in the first list and the
6461 first name in the second list were incorrectly joined into one item (with a
6462 colon in the middle).
6468 1. Added -C and -D options to the exinext utility, mainly to make it easier
6469 to include in the automated testing, but these could be helpful when
6470 multiple configurations are in use.
6472 2. The exinext utility was not formatting the output nicely when there was
6473 an alternate port involved in the retry record key, nor when there was a
6474 message id as well (for retries that were specific to a specific message
6475 and a specific host). It was also confused by IPv6 addresses, because of
6476 the additional colons they contain. I have fixed the IPv4 problem, and
6477 patched it up to do a reasonable job for IPv6.
6479 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6480 delivery, the log line now contains "pipelined" if PIPELINING was used.
6482 4. An SMTP transport process used to panic and die if the bind() call to set
6483 an explicit outgoing interface failed. This has been changed; it is now
6484 treated in the same way as a connect() failure.
6486 5. A reference to $sender_host_name in the part of a conditional expansion
6487 that was being skipped was still causing a DNS lookup. This no longer
6490 6. The def: expansion condition was not recognizing references to header lines
6491 that used bh_ and bheader_.
6493 7. Added the _cache feature to named lists.
6495 8. The code for checking quota_filecount in the appendfile transport was
6496 allowing one more file than it should have been.
6498 9. For compatibility with Sendmail, the command line option
6506 and sets the incoming protocol and host name (for trusted callers). The
6507 host name and its colon can be omitted when only the protocol is to be set.
6508 Note the Exim already has two private options, -pd and -ps, that refer to
6509 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6510 "s", but I don't think that's a major issue.
6512 10. A number of refactoring changes to the code, none of which should affect
6515 (a) The number of logging options was getting close to filling up the
6516 32-bit word that was used as a bit map. I have split them into two classes:
6517 those that are passed in the argument to log_write(), and those that are
6518 only ever tested independently outside of that function. These are now in
6519 separate 32-bit words, so there is plenty of room for expansion again.
6520 There is no change in the user interface or the logging behaviour.
6522 (b) When building, for example, log lines, the code previously used a
6523 macro that called string_cat() twice, in order to add two strings. This is
6524 not really sufficiently general. Furthermore, there was one instance where
6525 it was actually wrong because one of the argument was used twice, and in
6526 one call a function was used. (As it happened, calling the function twice
6527 did not affect the overall behaviour.) The macro has been replaced by a
6528 function that can join an arbitrary number of extra strings onto a growing
6531 (c) The code for expansion conditions now uses a table and a binary chop
6532 instead of a serial search (which was left over from when there were very
6533 few conditions). Also, it now recognizes conditions like "pam" even when
6534 the relevant support is not compiled in: a suitably worded error message is
6535 given if an attempt is made to use such a condition.
6537 11. Added ${time_interval:xxxxx}.
6539 12. A bug was causing one of the ddress fields not to be passed back correctly
6540 from remote delivery subprocesses. The field in question was not being
6541 subsequently used, so this caused to problems in practice.
6543 13. Added new log selectors queue_time and deliver_time.
6545 14. Might have fixed a bug in maildirsizefile handling that threw up
6546 "unexpected character" debug warnings, and recalculated the data
6547 unnecessarily. In any case, I expanded the warning message to give more
6550 15. Added the message "Restricted characters in address" to the statements in
6551 the default ACL that block characters like @ and % in local parts.
6553 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6554 Three changes have been made:
6556 (a) There was a serious bug; a negative response to MAIL caused the whole
6557 recipient domain to be cached as invalid, thereby blocking all messages
6558 to all local parts at the same domain, from all senders. This bug has
6559 been fixed. The domain is no longer cached after a negative response to
6560 MAIL if the sender used is not empty.
6562 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6565 (c) A new callout option, "use_sender" has been added for people who want
6566 the modified behaviour.
6572 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6575 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6576 indeed breaks things for older releases.
6578 3. Added additional logging to the case where there is a problem reading data
6579 from a filter that is running in a subprocess using a pipe, in order to
6580 try to track down a specific problem.
6582 4. Testing facility fudge: when running in the test harness and attempting
6583 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6584 getting "No route to host". Convert this to a timeout.
6586 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6589 6. Some OS don't have socklen_t but use size_t instead. This affects the
6590 fifth argument of getsockopt() amongst other things. This is now
6591 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6592 can be set for individual OS. I have set it for SunOS5, OSF1, and
6593 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6594 some earlier ones do not.
6596 7. Change 4.30/15 was not doing the test caselessly.
6598 8. The standard form for an IPv6 address literal was being rejected by address
6599 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6600 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6601 this, as well as the form without the "IPv6" on the front (but only when
6602 address literals are enabled, of course).
6604 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6606 10. Exim crashed if a message with an empty sender address specified by -f
6607 encountered a router with an errors_to setting. This could be provoked only
6608 by a command such as
6612 where an empty string was supplied; "<>" did not hit this bug.
6614 11. Installed PCRE release 4.5.
6616 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6617 remained set. It is now erased.
6619 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6620 times from message ids (which are base 36 rather than the normal 62).
6622 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6623 were being counted as actual protocol errors, and logged if the log
6624 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6625 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6626 and DATA following a set of rejected RCPTs do not count as protocol errors.
6627 In other words, Exim assumes they were pipelined, though this may not
6628 actually be the case. Of course, in all cases the client gets an
6629 appropriate error code.
6631 15. If a lookup fails in an ACL condition, a message about the failure may
6632 be available; it is used if testing the ACL cannot continue, because most
6633 such messages specify what the cause of the deferral is. However, some
6634 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6635 that caused an old message to be retained and used if a later statement
6636 caused a defer, replacing the real cause of the deferral.
6638 16. If an IP address had so many PTR records that the DNS lookup buffer
6639 was not large enough to hold them, Exim could crash while trying to process
6640 the truncated data. It now detects and logs this case.
6642 17. Further to 4.21/58, another change has been made: if (and only if) the
6643 first line of a message (the first header line) ends with CRLF, a bare LF
6644 in a subsequent header line has a space inserted after it, so as not to
6645 terminate the header.
6647 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6648 unnecessarily, used atoi() instead of strtol(), and didn't check the
6649 termination when getting file sizes from file names by regex.
6651 19. Completely re-implemented the support for maildirsize files, in the light
6652 of a number of problems with the previous contributed implementation
6653 (4.30/29). In particular:
6655 . If the quota is zero, the maildirsize file is maintained, but no quota is
6658 . If the maildir directory does not exist, it is created before any attempt
6659 to write a maildirsize file.
6661 . The quota value in the file is just a cache; if the quota is changed in
6662 the transport, the new value overrides.
6664 . A regular expression is available for excluding directories from the
6667 20. The autoreply transport checks the characters in options that define the
6668 message's headers; it allows continued headers, but it was checking with
6669 isspace() after an embedded newline instead of explicitly looking for a
6672 21. If all the "regular" hosts to which an address was routed had passed their
6673 expiry times, and had not reached their retry times, the address was
6674 bounced, even if fallback hosts were defined. Now Exim should go on to try
6677 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6678 equivalent code in the SMTP transport. Some hosts send humungous responses
6679 to HELO/EHLO, more than 1024 it seems.
6681 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6682 gives compiler warnings in some environments. I've now done it "properly",
6685 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6686 (because of problems with the built-in one) was declared to return uschar *
6687 instead of char *, causing compiler failure.
6689 25. Fixed a file descriptor leak when processing alias/forward files.
6691 26. Fixed a minor format string issue in dbfn.c.
6693 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6695 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6696 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6699 29. When a sender address is verified, it is cached, to save repeating the test
6700 when there is more than one recipient in a message. However, when the
6701 verification involves a callout, it is possible for different callout
6702 options to be set for different recipients. It is too complicated to keep
6703 track of this in the cache, so now Exim always runs a verification when a
6704 callout is required, relying on the callout cache for the optimization.
6705 The overhead is duplication of the address routing, but this should not be
6708 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6709 to be verified with callout=postmaster, and the main callout worked but the
6710 postmaster check failed, the verification correctly failed. However, if a
6711 subsequent RCPT command asked for sender verification *without* the
6712 postmaster check, incorrect caching caused this verification also to fail,
6715 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6716 it was not caching the DNS options (qualify_single, search_parents) that
6717 were used when the lookup failed. A subsequent lookup with different
6718 options therefore always gave the same answer, though there were cases
6719 where it should not have. (Example: a "domains = !$mx_any" option on a
6720 dnslookup router: the "domains" option is always processed without any
6721 widening, but the router might have qualify_single set.) Now Exim uses the
6722 cached value only when the same options are set.
6724 32. Added John Jetmore's "exipick" utility to the distribution.
6726 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6727 than a timeout (e.g. a certificate is required, and is not provided), an
6728 Exim server now closes the connection immediately. Previously it waited for
6729 the client to close - but if the client is SSL, it seems that they each
6730 wait for each other, leading to a delay before one of them times out.
6732 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6733 maintained 0.8.x compatibility because I don't think many are using it, and
6734 it is clearly obsolete.
6736 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6739 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6740 client certificate was expired. A simple patch fixes this, though I don't
6741 understand the full logic of why the verify callback is called multiple
6744 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6745 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6746 which causes problems with some clients (such as the Certicom SSL Plus
6747 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6748 disables the coutermeasure allowing Eudora to connect."
6750 38. Exim was not checking that a write() to a log file succeeded. This could
6751 lead to Bad Things if a log got too big, in particular if it hit a file
6752 size limit. Exim now panics and dies if it cannot write to a log file, just
6753 as it does if it cannot open a log file.
6755 39. Modified OS/Makefile-Linux so that it now contains
6757 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6759 The two -D definitions ensure that Exim is compiled with large file
6760 support, which makes it possible to handle log files that are bigger than
6763 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6764 instance) a domain was checked against a named list that involved a lookup,
6765 causing $domain_data to be set, then another domain was checked against the
6766 same list, then the first domain was re-checked, the value of $domain_data
6767 after the final check could be wrong. In particular, if the second check
6768 failed, it could be set empty. This bug probably also applied to
6771 41. The strip_trailing_dot option was not being applied to the address given
6772 with the -f command-line option.
6774 42. The code for reading a message's header from the spool was incrementing
6775 $received_count, but never initializing it. This meant that the value was
6776 incorrect (doubled) while delivering a message in the same process in which
6777 it was received. In the most common configuration of Exim, this never
6778 happens - a fresh exec is done - but it can happen when
6779 deliver_drop_privilege is set.
6781 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6782 it now includes up to 150 characters of the unexpected data in the log
6785 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6786 and building data strings. The size of both of these buffers was 10 000
6787 bytes - far larger than anybody would *ever* want, thought I. Needless to
6788 say, somebody hit the limit. I have increased the maximum line length to
6789 20 000 and the maximum data length of concatenated lines to 100 000. I have
6790 also fixed two bugs, because there was no checking on these buffers. Tsk,
6791 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6792 buffer is too small.
6794 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6795 lsearch lookups. Now it does.
6797 46. When parsing a route_list item in a manualroute router, a fixed-length
6798 buffer was used for the list of hosts. I made this 1024 bytes long,
6799 thinking that nobody would ever have a list of hosts that long. Wrong.
6800 Somebody had a whole pile of complicated expansion conditions, and the
6801 string was silently truncated, leading to an expansion error. It turns out
6802 that it is easier to change to an unlimited length (owing to other changes
6803 that have happened since this code was originally written) than to build
6804 structure for giving a limitation error. The length of the item that
6805 expands into the list of hosts is now unlimited.
6807 47. The lsearch lookup could not handle data where the length of text line was
6808 more than 4095 characters. Such lines were truncated, leading to shortened
6809 data being returned. It should now handle lines of any length.
6811 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6812 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6815 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6816 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6817 no longer happens in comment lines. A list of replacements is now placed
6818 at the head of all of the source files, except those whose only change is
6819 to replace PERL_COMMAND in the very first #! line.
6821 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6822 messages on the queue, with a bottom-up merge sort, using code contributed
6823 by Michael Haardt. This should make operations like -bp somewhat faster on
6824 large queues. It won't affect queue runners, except when queue_run_in_order
6827 51. Installed eximstats 1.31 in the distribution.
6829 52. Added support for SRV lookups to the dnslookup router.
6831 53. If an ACL referred to $message_body or $message_body_end, the value was not
6832 reset for any messages that followed in the same SMTP session.
6834 54. The store-handling optimization for building very long strings was not
6835 differentiating between the different store pools. I don't think this
6836 actually made any difference in practice, but I've tidied it.
6838 55. While running the routers to verify a sender address, $sender_address
6839 was still set to the sender address. This is wrong, because when routing to
6840 send a bounce to the sender, it would be empty. Therefore, I have changed
6841 it so that, while verifying a sender address, $sender_address is set to <>.
6842 (There is no change to what happens when verifying a recipient address.)
6844 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6845 target A or AAAA records (if not already returned) without resetting the
6846 qualify_single or search_parents options of the DNS resolver. These are
6847 inappropriate in this case because the targets of MX and SRV records must
6848 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6849 target that, when qualified, matched something in the local domain. These
6850 two options are now turned off when doing these lookups.
6852 57. It seems that at least some releases of Reiserfs (which does not have the
6853 concept of a fixed number of inodes) returns zero and not -1 for the
6854 number of available inodes. This interacted badly with check_spool_inodes,
6855 which assumed that -1 was the "no such thing" setting. What I have done is
6856 to check that the total number of inodes is greater than zero before doing
6857 the test of how many are available.
6859 58. When a "warn" ACL statement has a log_message modifier, the message is
6860 remembered, and not repeated. This is to avoid a lot of repetition when a
6861 message has many recipients that cause the same warning to be written.
6862 However, Exim was preserving the list of already written lines for an
6863 entire SMTP session, which doesn't seem right. The memory is now reset if a
6864 new message is started.
6866 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6867 debugging output unless log_rewrite was also set.
6869 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6870 of (int)(handle) when we know that handle contains (void *)(-1).
6872 61. The Exim daemon panic-logs an error return when it closes the incoming
6873 connection. However "connection reset by peer" seems to be common, and
6874 isn't really an error worthy of noting specially, so that particular error
6877 62. When Exim is trying to find all the local interfaces, it used to panic and
6878 die if the ioctl to get the interface flags failed. However, it seems that
6879 on at least one OS (Solaris 9) it is possible to have an interface that is
6880 included in the list of interfaces, but for which you get a failure error
6881 for this call. This happens when the interface is not "plumbed" into a
6882 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6883 failure of the "get flags" call assumes that the interface is down.
6885 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6886 makes life easier for people who are doing arithmetic on fields extracted
6887 from dates, where you often get leading zeros that should not be
6888 interpreted as octal.
6890 64. Added qualify_domain to the redirect router, to override the global
6893 65. If a pathologically long header line contained very many addresses (the
6894 report of this problem mentioned 10 000) and each of them was rewritten,
6895 Exim could use up a very large amount of memory. (It kept on making new
6896 copies of the header line as it rewrote, and never released the old ones.)
6897 At the expense of a bit more processing, the header rewriting function has
6898 been changed so that it no longer eats memory in this way.
6900 66. The generation of the Received: header has been moved from the time that a
6901 message starts to be received, to the time that it finishes. The timestamp
6902 in the Received: header should now be very close to that of the <= log
6903 line. There are two side-effects of this change:
6905 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6906 logged header lines no longer include the local Received: line, because
6907 it has not yet been created. The same applies to a copy of the message
6908 that is returned to a non-SMTP sender when a message is rejected.
6910 (b) When a filter file is tested using -bf, no additional Received: header
6911 is added to the test message. After some thought, I decided that this
6914 This change does not affect the value of $received_for. It is still set
6915 after address rewriting, but before local_scan() is called.
6917 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6919 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6920 gave an unhelpful panic error message, and a defer error. I have managed to
6921 change this behaviour so that it now rejects any supplied certificate,
6922 which seems right, as the list of acceptable certificates is empty.
6924 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6925 gave an unhelpful defer error. I have not managed to make this reject any
6926 supplied certificates, but the error message it gives is "no certificate
6927 supplied", which is not helpful.
6929 70. exigrep's output now also includes lines that are not associated with any
6930 message, but which match the given pattern. Implemented by a patch from
6931 Martin Sluka, which also tidied up the Perl a bit.
6933 71. Recipient callout verification, like sender verification, was using <> in
6934 the MAIL FROM command. This isn't really the right thing, since the actual
6935 sender may affect whether the remote host accepts the recipient or not. I
6936 have changed it to use the actual sender in the callout; this means that
6937 the cache record is now keyed on a recipient/sender pair, not just the
6938 recipient address. There doesn't seem to be a real danger of callout loops,
6939 since a callout by the remote host to check the sender would use <>.
6940 [SEE ABOVE: changed after hitting problems.]
6942 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6943 temporary errors. However, in the case of such a code being given after
6944 the end of a data transmission (i.e. after ".") Exim was failing to write
6945 a retry record for the message. (Yes, there was some broken host that was
6946 actually sending 8xx at this point.)
6948 73. An unknown lookup type in a host list could cause Exim to panic-die when
6949 the list was checked. (An example that provoked this was putting <; in the
6950 middle of a list instead of at the start.) If this happened during a DATA
6951 ACL check, a -D file could be left lying around. This kind of configuration
6952 error no longer causes Exim to die; instead it causes a defer error. The
6953 incident is still logged to the main and panic logs.
6955 74. Buglet left over from Exim 3 conversion. The message "too many messages
6956 in one connection" was written to the rejectlog but not the mainlog, except
6957 when address rewriting (yes!) was being logged.
6959 75. Added write_rejectlog option.
6961 76. When a system filter was run not as root (that is, when system_filter_user
6962 was set), the values of the $n variables were not being returned to the
6963 main process; thus, they were not subsequently available in the $sn
6966 77. Added +return_path_on_delivery log selector.
6968 78. A connection timeout was being treated differently from recipients deferred
6969 when testing hosts_max_try with a message that was older than the host's
6970 retry timeout. (The host should not be counted, thus allowing all hosts to
6971 be tried at least once before bouncing.) This may have been the cause of an
6972 occasionally reported bug whereby a message would remain on the queue
6973 longer than the retry timeout, but would be bounced if a delivery was
6974 forced. I say "may" because I never totally pinned down the problem;
6975 setting up timeout/retry tests is difficult. See also the next item.
6977 79. The ultimate address timeout was not being applied to errors that involved
6978 a combination of host plus message (for example, a timeout on a MAIL
6979 command). When an address resolved to a number of possible hosts, and they
6980 were not all tried for each delivery (e.g. because of hosts_max_try), a
6981 message could remain on the queue longer than the retry timeout.
6983 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6986 81. Fixed an obscure SMTP outgoing bug which required at least the following
6987 conditions: (a) there was another message waiting for the same server;
6988 (b) the server returned 5xx to all RCPT commands in the first message so
6989 that the message was not completed; (c) the server dropped the connection
6990 or gave a negative response to the RSET that Exim sends to abort the
6991 transaction. The observed case was a dropped connection after DATA that had
6992 been sent in pipelining mode. That is, the server had advertised PIPELINING
6993 but was not implementing it correctly. The effect of the bug was incorrect
6994 behaviour, such as trying another host, and this could lead to a crash.
7000 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7001 and daemon.c were passed as pointers to ints; they should have been
7002 pointers to socklen_t variables (which are typically unsigned ints).
7004 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7007 3. Fixed a really odd bug that affected only the testing scheme; patching a
7008 certain fixed string in the binary changed the value of another string that
7009 happened to be identical to the end of the original first string.
7011 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7012 name", it returns that address as the IP address. On some operating
7013 systems (e.g. Solaris), it also passes back the IP address string as the
7014 "host name". However, on others (e.g. Linux), it passes back an empty
7015 string. Exim wasn't checking for this, and was changing the host name to an
7016 empty string, assuming it had been canonicalized.
7018 5. Although rare, it is permitted to have more than one PTR record for a given
7019 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7020 all the names associated with an address, because they do in Solaris.
7021 However, it seems that they do not in Linux for data that comes from the
7022 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7023 I found this out when I moved to a new Linux workstation and tried to run
7024 the Exim test suite.
7026 To get round this problem I have changed the code so that it now does its
7027 own call to the DNS to look up PTR records when searching for a host name.
7028 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7029 addresses that are only in /etc/hosts are still found.
7031 This behaviour is, however, controlled by an option called host_lookup_
7032 order, which defaults to "bydns:byaddr". If people want to use the other
7033 order, or indeed, just use one or the other means of lookup, they can
7034 specify it in this variable.
7036 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7037 some operating systems, this comes back from gethostbyaddr() as an empty
7038 string, and this is what Exim used to test for. However, it seems that in
7039 other systems, "." is yielded. Exim now tests for this case too.
7041 7. The values of check_spool_space and check_log_space are now held internally
7042 as a number of kilobytes instead of an absolute number of bytes. If a
7043 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7044 kilobyte. This means that much larger values can be stored.
7046 8. Exim monitor: an attempt to get the action menu when not actually pointing
7047 at a message produces an empty menu entitled "No message selected". This
7048 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7049 no entries in it ("Shell widget menu has zero width and/or height"). So I
7050 have added a single, blank menu entry in this case.
7052 9. Added ${quote_local_part.
7054 10. MIME decoding is now applied to the contents of Subject: header lines when
7057 11. Now that a reference to $sender_host_address automatically causes a reverse
7058 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7059 host lookup before query-style lookups in lists that might use this
7060 variable. This has therefore been abolished, and the "net-" prefix is no
7061 longer necessary for query-style lookups.
7063 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7064 have been a typo for LFLAGS, so it has been changed.
7066 13. The install script calls Exim with "-C /dev/null" in order to find the
7067 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7068 to be output. However, since Exim outputs its version number before the
7069 error, it didn't break the script. It just looked ugly. I fixed this by
7070 always allowing "-C /dev/null" if the caller is root.
7072 14. Ignore overlarge ACL variable number when reading spool file - insurance
7073 against a later release with more variables having written the file.
7075 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7076 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7077 this, as well as the form without the "IPv6" on the front.
7079 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7080 OS/Makefile-Darwin file.
7082 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7083 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7085 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7086 variable. However, some OS use ints for inodes. I've added cast to long int
7087 to get rid of the compiler warning.
7089 19. I had forgotten to lock out "/../" in configuration file names when
7090 ALT_CONFIG_PREFIX was set.
7092 20. Routers used for verification do not need to specify transports. However,
7093 if such a router generated a host list, and callout was configured, Exim
7094 crashed, because it could not find a port number from the (non-existent)
7095 transport. It now assumes port 25 in this circumstance.
7097 21. Added the -t option to exigrep.
7099 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7100 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7101 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7102 them is set, LOOKUP_LSEARCH is forced.
7104 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7106 24. Added sender and host information to the "rejected by local_scan()" log
7107 line; previously there was no indication of these.
7109 25. Added .include_if_exists.
7111 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7112 Linux. It turns out that not all file systems support this. Apparently some
7113 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7114 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7115 not supported on the file descriptor, is now ignored when Exim is trying to
7116 sync a directory. This applies only to Linux.
7118 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7120 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7121 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7124 29. Added support for maildirsize files from supplied patch (modified a bit).
7126 30. The use of :fail: followed by an empty string could lead Exim to respond to
7127 sender verification failures with (e.g.):
7129 550 Verification failed for <xxx>
7130 550 Sender verify failed
7132 where the first response line was missing the '-' that indicates it is not
7133 the final line of the response.
7135 31. The loop for finding the name of the user that called Exim had a hardwired
7136 limit of 10; it now uses the value of finduser_retries, which is used for
7137 all other user lookups.
7139 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7142 33. Exim was neglecting to zero errno before one call of strtol() when
7143 expanding a string and expecting an integer value. On some systems this
7144 resulted in spurious "integer overflow" errors. Also, it was casting the
7145 result into an int without checking.
7147 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7148 did not work. The code looks as if it has *never* worked, though it appears
7149 to have been documented since at least release 1.62. I have made it work.
7151 35. The "timeout_DNS" error in retry rules, also documented since at least
7152 1.62, also never worked. As it isn't clear exactly what this means, and
7153 clearly it isn't a major issue, I have abolished the feature by treating it
7154 as "timeout", and writing a warning to the main and panic logs.
7156 36. The display of retry rules for -brt wasn't always showing the error code
7159 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7160 timeout_connect_A, timeout_connect_MX.
7162 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7163 to the empty sender.
7165 39. The daemon was not analysing the content of -oX till after it had closed
7166 stderr and disconnected from the controlling terminal. This meant that any
7167 syntax errors were only noted on the panic log, and the return code from
7168 the command was 0. By re-arranging the code a little, I've made the
7169 decoding happen first, so such errors now appear on stderr, and the return
7170 code is 1. However, the actual setting up of the sockets still happens in
7171 the disconnected process, so errors there are still only recorded on the
7174 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7175 connections (as happens on some IP stacks) was logged at start up time as
7176 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7177 it from "IPv6 and IPv4", which means that two separate sockets are being
7180 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7181 says whether AF_INET or AF_INET6 was passed as an argument.
7183 42. Exiwhat output was messed up when time zones were included in log
7186 43. Exiwhat now gives more information about the daemon's listening ports,
7187 and whether -tls-on-connect was used.
7189 44. The "port" option of the smtp transport is now expanded.
7191 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7192 silently ignored. Now an error message is written to the main and panic
7195 46. There's a new ACL modifier called "logwrite" which writes to a log file
7196 as soon as it is encountered.
7198 47. Added $local_user_uid and $local_user_gid at routing time.
7200 48. Exim crashed when trying to verify a sender address that was being
7203 49. Exim was recognizing only a space character after ".include". It now also
7204 recognizes a tab character.
7206 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7207 extracting the relevant information from the specification. The man page no
7208 longer contains scrambled data for the -d option, and I've added a section
7209 at the front about calling Exim under different names.
7211 51. Added "extra_headers" argument to the "mail" command in filter files.
7213 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7216 53. Installed eximstats 1.29.
7218 54. Added transport_filter_timeout as a generic transport option.
7220 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7221 Cc: header lines. This was required by RFC 822, but it not required by RFC
7224 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7225 incoming messages that did not have them. Now it does so only if the
7226 message originates locally, that is, if there is no associated remote host
7227 address. When Resent- header lines are present, this applies to the Resent-
7228 lines rather than the non-Resent- lines.
7230 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7231 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7233 58. Messages for configuration errors now include the name of the main
7234 configuration file - useful now that there may be more than one file in a
7235 list (.included file names were always shown).
7237 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7238 for those rare installations that do not start the daemon as root or run it
7239 setuid root. I've cut out the call to initgroups() if the daemon is not
7242 60. The Exim user and group can now be bound into the binary as text strings
7243 that are looked up at the start of Exim's processing.
7245 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7247 62. Added $mailstore_basename variable.
7249 63. Installed patch to sieve.c from Michael Haardt.
7251 64. When Exim failed to open the panic log after failing to open the main log,
7252 the original message it was trying to log was written to stderr and debug
7253 output, but if they were not available (the usual case in production), it
7254 was lost. Now it is written to syslog before the two lines that record the
7255 failures to open the logs.
7257 65. Users' Exim filters run in subprocesses under the user's uid. It is
7258 possible for a "deliver" command or an alias in a "personal" command to
7259 provoke an address rewrite. If logging of address rewriting is configured,
7260 this fails because the process is not running as root or exim. There may be
7261 a better way of dealing with this, but for the moment (because 4.30 needs
7262 to be released), I have disabled address rewrite logging when running a
7263 filter in a non-root, non-exim process.
7269 1. The buildconfig auxiliary program wasn't quoting the value set for
7270 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7271 not defined. This bug was masked in 4.22 by the effect that was fixed in
7274 2. Some messages that were rejected after a message id was allocated were
7275 shown as "incomplete" by exigrep. It no longer does this for messages that
7276 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7278 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7279 have allow_domain_literals set, the ID did not get logged in the <= line.
7280 Domain literals are now always recognized in Message-ID: header lines.
7282 4. The first argument for a ${extract expansion item is the key name or field
7283 number. Leading and trailing spaces in this item were not being ignored,
7284 causing some misleading effects.
7286 5. When deliver_drop_privilege was set, single queue runner processes started
7287 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7288 same command in the process it spins off) were not dropping privilege.
7290 6. When the daemon running as "exim" started a queue runner, it always
7291 re-executed Exim in the spun-off process. This is a waste of effort when
7292 deliver_drop_privilege is set. The new process now just calls the
7293 queue-runner function directly.
7299 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7302 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7303 ignored. Though the use of -oP was forcing the writing of a pid file, it
7304 was always written to the default place.
7306 3. If the message "no IP address found for host xxxx" is generated during
7307 incoming verification, it is now followed by identification of the incoming
7308 connection (so you can more easily find what provoked it).
7310 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7312 5. Added some features to "harden" Exim a bit more against certain attacks:
7314 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7315 be put in Local/Makefile. This is like the never_users runtime option,
7316 but it cannot be overridden. The default setting is "root".
7318 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7319 prefix string with which any file named in a -C command line option
7322 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7323 is retained for -C and -D only if the caller of Exim is root. Without
7324 it, the exim user may also use -C and -D and retain privilege.
7326 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7327 command line option is disabled.
7329 6. Macro names set by the -D option must start with an upper case letter, just
7330 like macro names defined in the configuration file.
7332 7. Added "dereference=" facility to LDAP.
7334 8. Two instances of the typo "uknown" in the source files are fixed.
7336 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7337 the Configure-Makefile script screwed up while processing it.
7339 10. Incorporated PCRE 4.4.
7341 11. The SMTP synchronization check was not operating right at the start of an
7342 SMTP session. For example, it could not catch a HELO sent before the client
7343 waited for the greeting. There is now a check for outstanding input at the
7344 point when the greeting is written. Because of the duplex, asynchronous
7345 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7346 way, but not yet received, when the check is performed.
7348 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7349 on TCP/IP sockets, because this apparently causes some broken clients to
7352 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7353 unchanged) from the Cygwin maintainer.
7355 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7356 was supported instead of testing for mbx. Effectively a typo.
7358 15. The spa authenticator server code was not checking that the input it
7359 received was valid base64.
7361 16. The debug output line for the "set" modifier in ACLs was not showing the
7362 name of the variable that was being set.
7364 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7366 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7367 lookup on its own. Something else was needed to trigger the lookup. For
7368 example, a match in host_lookup or the need for a host name in a host list.
7369 Now, if $sender_host_name is referenced and the host name has not yet been
7370 looked up, a lookup is performed. If the lookup fails, the variable remains
7371 empty, and $host_lookup_failed is set to "1".
7373 19. Added "eqi" as a case-independent comparison operator.
7375 20. The saslauthd authentication condition could segfault if neither service
7376 nor realm was specified.
7378 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7379 error message that was logged was misleading, and incoming SMTP
7380 connections were dropped. The message is now more accurate, and temporary
7381 errors are given to SMTP connections.
7383 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7384 (except RSET) with a 421 error, until QUIT is received. However, it was
7385 failing to send a response to QUIT.
7387 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7388 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7389 ACL is now run after the helo_try_verify_hosts code.
7391 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7392 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7393 case-independent, so other case variants are also recognized). Apparently
7394 some systems use these upper case variants.
7396 25. If more than two messages were waiting for the same host, and a transport
7397 filter was specified for the transport, Exim sent two messages over the
7398 same TCP/IP connection, and then failed with "socket operation on non-
7399 socket" when it tried to send the third.
7401 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7403 27. The extern definition of crypt16() in expand.c was not being excluded when
7404 the OS had its own crypt16() function.
7406 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7407 as a preferred synonym for return_size_limit, both as an option and as an
7410 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7412 30. Changed the default configuration ACL to relax the local part checking rule
7413 for addresses that are not in any local domains. For these addresses,
7414 slashes and pipe symbols are allowed within local parts, but the sequence
7415 /../ is explicitly forbidden.
7417 31. SPA server authentication was not clearing the challenge buffer before
7420 32. log_message in a "warn" ACL statement was writing to the reject log as
7421 well as to the main log, which contradicts the documentation and doesn't
7422 seem right (because no rejection is happening). So I have stopped it.
7424 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7425 However, I am unable to do any testing of this.
7427 34. Fixed an infelicity in the appendfile transport. When checking directories
7428 for a mailbox, to see if any needed to be created, it was accidentally
7429 using path names with one or more superfluous leading slashes; tracing
7430 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7432 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7433 contents are added to the log line that is written for every discarded
7434 recipient. (Previously a log_message setting was ignored.)
7436 36. The ${quote: operator now quotes the string if it is empty.
7438 37. The install script runs exim in order to find its version number. If for
7439 some reason other than non-existence or emptiness, which it checks, it
7440 could not run './exim', it was installing it with an empty version number,
7441 i.e. as "exim-". This error state is now caught, and the installation is
7444 38. An argument was missing from the function that creates an error message
7445 when Exim fails to connect to the socket for saslauthd authentication.
7446 This could cause Exim to crash, or give a corrupted message.
7448 39. Added isip, isip4, and isip6 to ${if conditions.
7450 40. The ACL variables $acl_xx are now saved with the message, and can be
7451 accessed later in routers, transports, and filters.
7453 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7454 strings in the file are not string-expanded.
7456 42. If a MAIL command specified a SIZE value that was too large to fit into an
7457 int variable, the check against message_size_limit failed. Such values are
7458 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7459 day this will have to be increased, but I don't think I want to be around
7460 when emails are that large.
7467 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7468 iconv() is not standard in FreeBSD.
7470 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7471 IPv6 enabled. The observed symptom was a segmentation fault on return from
7472 the function os_common_find_running_interfaces() in src/os.c.
7474 3. In the check_special_case() function in daemon.c I had used "errno" as an
7475 argument name, which causes warnings on some systems. This was basically a
7476 typo, since it was named "eno" in the comments!
7478 4. The code that waits for the clock to tick (at a resolution of some fraction
7479 of a second) so as to ensure message-id uniqueness was always waiting for
7480 at least one whole tick, when it could have waited for less. [This is
7481 almost certainly not relevant at current processor speeds, where it is
7482 unlikely to ever wait at all. But we try to future-proof.]
7484 5. The function that sleeps for a time interval that includes fractions of a
7485 second contained a race. It did not block SIGALRM between setting the
7486 timer, and suspending (a couple of lines later). If the interval was short
7487 and the sigsuspend() was delayed until after it had expired, the suspension
7488 never ended. On busy systems this could lead to processes getting stuck for
7491 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7492 process, before it forks any delivery processes. The open lookup caching
7493 mechanism meant that the open file or database connection was passed into
7494 the delivery process. The problem was that delivery processes always tidy
7495 up cached lookup data. This could cause a problem for the next delivery
7496 process started by the queue runner, because the external queue runner
7497 process does not know about the closure. So the next delivery process
7498 still has data in the lookup cache. In the case of a file lookup, there was
7499 no problem because closing a file descriptor in a subprocess doesn't affect
7500 the parent. However, if the lookup was caching a connection to a database,
7501 the connection was closed, and the second delivery process was likely to
7502 see errors such as "PGSQL: query failed: server closed the connection
7503 unexpectedly". The problem has been fixed by closing all cached lookups
7504 in a queue runner before running a delivery process.
7506 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7507 seem to have the "const" qualifier which it has on other OS. I've
7510 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7511 *of the same type* (client or server) with the same public name that an
7512 error should be diagnosed.
7514 9. When Exim looked up a host name for an IP address, but failed to find the
7515 original IP address when looking up the host name (a safety check), it
7516 output the message "<ip address> does not match any IP for NULL", which was
7517 confusing, to say the least. The bug was that the host name should have
7518 appeared instead of "NULL".
7520 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7521 user that is built into the binary, and the -C or -D options is used, root
7522 privilege is dropped before the configuration file is read. In addition,
7523 logging is switched to stderr instead of the normal log files. If the
7524 configuration then re-defines the Exim user, the unprivileged environment
7525 is probably not what is expected, so Exim logs a panic warning message (but
7528 However, if deliver_drop_privilege is set, the unprivileged state may well
7529 be exactly what is intended, so the warning has been cut out in that case,
7530 and Exim is allowed to try to write to its normal log files.
7536 1. smtp_return_error_details was not giving details for temporary sender
7537 or receiver verification errors.
7539 2. Diagnose a configuration error if two authenticators have the same public
7542 3. Exim used not to create the message log file for a message until the first
7543 delivery attempt. This could be confusing when incoming messages were held
7544 for policy or load reasons. The message log file is now created at the time
7545 the message is received, and an initial "Received" line is written to it.
7547 4. The automatically generated man page for command line options had a minor
7548 bug that caused no ill effects; however, a more serious problem was that
7549 the procedure for building the man page automatically didn't always
7550 operate. Consequently, release 4.20 contains an out-of-date version. This
7551 shouldn't happen again.
7553 5. When building Exim with embedded Perl support, the script that builds the
7554 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7555 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7557 6. The freeze_tell option was not being used for messages that were frozen on
7558 arrival, either by an ACL or by local_scan().
7560 7. Added the smtp_incomplete_transaction log selector.
7562 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7563 was accepting AUTH without a new EHLO.
7565 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7566 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7569 10. Logging of TCP/IP connections (when configured) now happens in the main
7570 daemon process instead of the child process, so that the TCP/IP connection
7571 count is more accurate (but it can never be perfect).
7573 11. The use of "drop" in a nested ACL was not being handled correctly in the
7574 outer ACL. Now, if condition failure induced by the nested "drop" causes
7575 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7576 or "require"), the connection is dropped.
7578 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7579 that yield "discard" can now be used with an "accept" or a "discard" verb,
7580 but an error is generated for any others (because I can't see a useful way
7581 to define what should happen).
7583 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7584 are now processed in the same way as lines in the Exim configuration file.
7585 In particular, continuation lines are supported.
7587 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7589 15. Added -ti meaning -t -i.
7591 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7592 domains, and warn by logging if others are found.
7594 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7595 returned from GIFCONF when Exim is trying to find the list of interfaces on
7596 a host. The code in os.c has been modified to copy each ifreq to an aligned
7597 structure in all cases.
7599 Also, in some cases, the returned ifreq's were being copied to a 'struct
7600 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7601 means the last couple of bytes of an IPv6 address could be chopped if the
7602 ifreq contained only a normal sockaddr (14 bytes storage).
7604 18. Named domain lists were not supported in the hosts_treat_as_local option.
7605 An entry such as +xxxx was not recognized, and was treated as a literal
7608 19. Ensure that header lines added by a DATA ACL are included in the reject log
7609 if the ACL subsequently rejects the message.
7611 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7612 MD5 (which is deprecated).
7614 21. When testing a filter file using -bf, Exim was writing a message when it
7615 took the sender from a "From " line in the message, but it was not doing so
7616 when it took $return_path from a Return-Path: header line. It now does.
7618 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7619 with a valid header line field name (a series of printing characters
7620 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7622 23. Changed "disc" in the source to "disk" to conform to the documentation and
7623 the book and for uniformity.
7625 24. Ignore Sendmail's -Ooption=value command line item.
7627 25. When execve() failed while trying to run a command in a pipe transport,
7628 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7629 could be confused with a return value of 69 from the command itself. This
7630 has been changed to 127, the value the shell returns if it is asked to run
7631 a non-existent command. The wording for the related log line suggests a
7632 non-existent command as the problem.
7634 26. If received_header_text expands to an empty string, do not add a Received:
7635 header line to the message. (Well, it adds a token one on the spool, but
7636 marks it "old" so that it doesn't get used or transmitted.)
7638 27. Installed eximstats 1.28 (addition of -nt option).
7640 28. There was no check for failure on the call to getsockname() in the daemon
7641 code. This can fail if there is a shortage of resources on the system, with
7642 ENOMEM, for example. A temporary error is now given on failure.
7644 29. Contrary to the C standard, it seems that in some environments, the
7645 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7646 program. Exim now does this explicitly; it affects the formatting of
7647 timestamps using strftime().
7649 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7650 complaints. I've now initialized all the variables, to avoid this.
7652 32. Header lines added by a system filter were not being "seen" during
7653 transport-time rewrites.
7655 33. The info_callback() function passed to OpenSSL is set up with type void
7656 (*)(SSL *, int, int), as described somewhere. However, when calling the
7657 function (actually a macro) that sets it up, the type void(*)() is
7658 expected. I've put in a cast to prevent warnings from picky compilers.
7660 34. If a DNS black list lookup found a CNAME record, but there were no A
7661 records associated with the domain it pointed at, Exim crashed.
7663 35. If a DNS black list lookup returned more than one A record, Exim ignored
7664 all but the first. It now scans all returned addresses if a particular IP
7665 value is being sought. In this situation, the contents of the
7666 $dnslist_value variable are a list of all the addresses, separated by a
7669 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7670 to accept a host name and all its aliases if the forward lookup for any of
7671 them yielded the IP address of the incoming connection. Now it accepts only
7672 those names whose forward lookup yields the correct IP address. Any other
7673 names are discarded. This closes a loophole whereby a rogue DNS
7674 administrator could create reverse DNS records to break through a
7675 wildcarded host restriction in an ACL.
7677 37. If a user filter or a system filter that ran in a subprocess used any of
7678 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7679 the wrong values were passed to the pipe command ($thisaddress had the
7680 value of $0, $0 had the value of $1, etc). This bug was introduced by
7681 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7683 38. Improved the line breaking for long SMTP error messages from ACLs.
7684 Previously, if there was no break point between 40 and 75 characters, Exim
7685 left the rest of the message alone. Two changes have been made: (a) I've
7686 reduced the minimum length to 35 characters; (b) if it can't find a break
7687 point between 35 and 75 characters, it looks ahead and uses the first one
7688 that it finds. This may give the occasional overlong line, but at least the
7689 remaining text gets split now.
7691 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7692 file descriptors might be low, and that setting 1000 would always raise it.
7693 It turns out that in some environments, the limit is already over 1000 and
7694 that lowering it causes trouble. So now Exim takes care not to decrease it.
7696 40. When delivering a message, the value of $return_path is set to $sender_
7697 address at the start of routing (routers may change the value). By an
7698 oversight, this default was not being set up when an address was tested by
7699 -bt or -bv, which affected the outcome if any router or filter referred to
7702 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7703 log only when "message" or "log_message" are set. However, if one of the
7704 conditions was an address verification, or a call to a nested ACL, the
7705 messages generated by the underlying test were being passed through. This
7706 no longer happens. The underlying message is available in $acl_verify_
7707 message for both "message" and "log_message" expansions, so it can be
7708 passed through if needed.
7710 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7711 new expansion $bh_ to give the encoded byte string without charset
7712 translation. Translation happens only if iconv() is available; HAVE_ICONV
7713 indicates this at build time. HEADERS_CHARSET gives the charset to
7714 translate to; headers_charset can change it in the configuration, and
7715 "headers charset" can change it in an individual filter file.
7717 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7718 that creates RFC 2047 encoded "words" labels them as that charset instead
7719 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7720 expansion operator; (ii) when Exim creates a From: line for a local
7721 message; (iii) when a header line is rewritten to include a "phrase" part.
7723 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7724 buggy, causing it to skip the first lines of messages whose message ID
7725 ended in 'D'. This would not have bitten before Exim release 4.14, because
7726 message IDs were unlikely to end in 'D' before then. The effect was to have
7727 incorrect size information for certain domains.
7729 45. #include "config.h" was missing at the start of the crypt16.c module. This
7730 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7733 46. If there was a timeout during a "random" callout check, Exim treated it as
7734 a failure of the random address, and carried on sending RSET and the real
7735 address. If the delay was just some slowness somewhere, the response to the
7736 original RCPT would be taken as a response to RSET and so on, causing
7737 mayhem of various kinds.
7739 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7740 when I implemented it. It didn't allow for the fact that some option values
7741 may legitimately be negative (e.g. size_addition), and it didn't even do
7742 the right test for positive values.
7744 48. Domain names in DNS records are case-independent. Exim always looks them up
7745 in lower case. Some resolvers return domain names in exactly the case they
7746 appear in the zone file, that is, they may contain uppercase letters. Not
7747 all resolvers do this - some return always lower case. Exim was treating a
7748 change of case by a resolver as a change of domain, similar to a widening
7749 of a domain abbreviation. This triggered its re-routing code and so it was
7750 trying to route what was effectively the same domain again. This normally
7751 caused routing to fail (because the router wouldn't handle the domain
7752 twice). Now Exim checks for this case specially, and just changes the
7753 casing of the domain that it ultimately uses when it transmits the message
7756 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7759 50. If a filter generated a file delivery with a non-absolute name (possible if
7760 no home directory exists for the router), the forbid_file option was not
7763 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7764 the existing equality matching.
7766 52. Exim was using ints instead of ino_t variables in some places where it was
7767 dealing with inode numbers.
7769 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7770 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7771 called TMPDIR, and if it finds it is different, it changes its value.
7773 54. The smtp_printf() function is now made available to local_scan() so
7774 additional output lines can be written before returning. There is also an
7775 smtp_fflush() function to enable the detection of a dropped connection.
7776 The variables smtp_input and smtp_batched_input are exported to
7779 55. Changed the default runtime configuration: the message "Unknown user"
7780 has been removed from the ACL, and instead placed on the localuser router,
7781 using the cannot_route_message feature. This means that any verification
7782 failures that generate their own messages won't get overridden. Similarly,
7783 the "Unrouteable address" message that was in the ACL for unverifiable
7784 relay addresses has also been removed.
7786 56. Added hosts_avoid_esmtp to the smtp transport.
7788 57. The exicyclog script was not checking for the esoteric option
7789 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7790 will work only if exicyclog is run under the appropriate euid.
7792 58. Following a discussion on the list, the rules by which Exim recognises line
7793 endings on incoming messages have been changed. The -dropcr and drop_cr
7794 options are now no-ops, retained only for backwards compatibility. The
7795 following line terminators are recognized: LF CRLF CR. However, special
7796 processing applies to CR:
7798 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7799 nor a local message in the state where . is a terminator.
7801 (ii) If a bare CR is encountered in a header line, an extra space is added
7802 after the line terminator so as not to end the header. The reasoning
7803 behind this is that bare CRs in header lines are most likely either
7804 to be mistakes, or people trying to play silly games.
7806 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7807 was being incorrectly given as 18 bytes larger than it should have been.
7808 This is a VOB (very old bug).
7810 60. This may never have affected anything current, but just in case it has:
7811 When the local host is found other than at the start of a list of hosts,
7812 the local host, those with the same MX, and any that follow, are discarded.
7813 When the list in question was part of a longer list of hosts, the following
7814 hosts (not currently being processed) were also being discarded. This no
7815 longer happens. I'm not sure if this situation could ever has previously
7818 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7821 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7822 header. This is something that is recommended in a new Internet Draft, and
7823 is something that is documented as being done by Sendmail. There are two
7824 possible values. For messages generated by the autoreply transport, Exim
7827 Auto-Submitted: auto-replied
7829 whereas for all other generated messages (e.g. bounces) it adds
7831 Auto-Submitted: auto-generated
7833 63. The "personal" condition in filters now includes a test for the
7834 Auto-Submitted: header. If it contains the string "auto-" the message it
7835 not considered personal.
7837 64. Added rcpt_include_affixes as a generic transport option.
7839 65. Added queue_only_override (default true).
7841 66. Added the syslog_duplication option.
7843 67. If what should have been the first header line of a message consisted of
7844 a space followed by a colon, Exim was mis-interpreting it as a header line.
7845 It isn't of course - it is syntactically invalid and should therefore be
7846 treated as the start of the message body. The misbehaviour could have
7847 caused a number of strange effects, including loss of data in subsequent
7848 header lines, and spool format errors.
7850 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7851 client host had authenticated. This control can now be exercised by an ACL
7852 for more flexibility.
7854 69. By default, callouts do not happen when testing with -bh. There is now a
7855 variant, -bhc, which does actually run the callout code, including
7856 consulting and updating the callout cache.
7858 70. Added support for saslauthd authentication, courtesy of Alexander
7861 71. If statvfs() failed on the spool or log directories while checking their
7862 size for availability, Exim confusingly gave the error "space shortage".
7863 Furthermore, in debugging mode it crashed with a floating point exception.
7864 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7865 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7866 serious problem, Exim now writes to the main and panic logs when this
7867 happens, with details of the failure. It then refuses to accept the
7868 incoming message, giving the message "spool directory problem" or "log
7869 directory problem" with a 421 code for SMTP messages.
7871 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7872 0, 1, and 2 exist, because some OS complain for execs without them (see
7873 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7874 descriptors. However, the code omitted to check that the open succeeded,
7875 causing mysterious errors if for some reason the permissions on /dev/null
7876 got screwed. Now Exim writes a message to the main and panic logs, and
7877 bombs out if it can't open /dev/null.
7879 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7880 interact so that it is all more flexible. It is supposed to remain
7881 backwards compatible. Also added extra_local_interfaces.
7883 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7884 to bomb out with an assertion failure - to the client this appears as a
7885 connection drop. This problem occurs in the part of the code that was taken
7886 from the Samba project. Fortunately, the assertion is in a very simple
7887 function, so I have fixed this by reproducing the function inline in the
7888 one place where it is called, and arranging for authentication to fail
7889 instead of killing the process with assert().
7891 75. The SPA client code was not working when the server requested OEM rather
7892 than Unicode encoding.
7894 76. Added code to make require_files with a specific uid setting more usable in
7895 the case where statting the file as root fails - usually a non-root-mounted
7896 NFS file system. When this happens and the failure is EACCES, Exim now
7897 forks a subprocess and does the per-uid checking as the relevant uid.
7899 77. Added process_log_path.
7901 78. If log_file_path was not explicitly set, a setting of check_log_space or
7902 check_log_inodes was ignored.
7904 79. If a space check for the spool or log partitions fails, the incident is now
7905 logged. Of course, in the latter case the data may get lost...
7907 80. Added the %p formatting code to string_format() so that it can be used to
7908 print addresses in debug_print(). Adjusted all the address printing in the
7909 debugging in store.c to use %p rather than %d.
7911 81. There was a concern that a line of code in smtp_in.c could overflow a
7912 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7913 initially expressed, the concern was not well-founded, because trailing
7914 spaces are removed early. However, if the trailing spaces were followed by
7915 a NULL, they did not get removed, so the overflow was possible. Two fixes
7918 (a) I re-wrote the offending code in a cleaner fashion.
7919 (b) If an incoming SMTP command contains a NULL character, it is rejected
7922 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7923 runs initgroups(), so that if the Exim user is in any additional groups,
7924 they will be used during message reception.
7930 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob