1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
9 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
10 related applications. Previously an "H" was used where available info
11 says that "M" should be, so change to match.
13 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
14 as arguments, so an implementation trying to copy these into a local
15 buffer was taking a taint-enforcement trap. Fix by using dynamically
16 created buffers. Similar fix for radius expansion condition.
18 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
19 reasonable, eg. to count headers. Fix by using dynamically created
20 buffers rather than a local. Do similar fixes for ACL actions "dcc",
21 "log_reject_target", "malware" and "spam"; the arguments are expanded
22 so could be handling tainted values.
24 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
25 broken the (no-op) support for this sendmail command. Restore it
26 to doing nothing, silently, and returning good status.
28 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
29 record path was given (or the default used) without a leading directory
30 path, an error occurred on trying to open it. Use the transport's working
33 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
34 transport. Previously it was the name on the DNS A-record; use instead
35 the head of the CNAME chain leading there (if there is one). This seems
36 to align better with RFC 6125.
38 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
39 smtp_accept_max_per_host allocated resources which were not released
40 when the limit was exceeded. This eventually crashed the daemon. Fix
41 by adding a relase action in that path.
43 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
44 expanded; previously using tainted values was rejected. Fix by using
45 dynamically-created buffers.
47 JH/09 Relax restrictions on ACL verify condition needing access to message
48 headers. Previously they were only permitted in data and non-smtp ACLs;
49 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
50 not_blind, header_sender and header_names_ascii verification.
52 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
53 Previously a macro used one argument twice; when called with the
54 argument as an expression having side-effects, incorrect operation
55 resulted. Use an inlineable function.
57 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
58 held open for a verify callout. Previously this wan not accounted for
59 and a corrupt onward SMTP conversation resulted.
61 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
62 Previously they were embedded naively in the constructed address; when
63 needed, strip the quoting and quote the entire local_part.
64 Also make the inbound_srs expansion condition handle quoting.
66 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
67 excluded, not matching the documentation.
69 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
70 was given for the sqlite_dbfile a trap resulted.
72 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
73 "name" argument resulted in a trap. There is no reason to disallow such;
74 this was a coding error.
76 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
77 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
78 spent suspended, ignoring the Posix definition. Previously we assumed
79 it did and a constant offset from real time could be used as a correction.
80 Change to using the same clock source for the start-of-message and the
81 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
82 exists, just to get a clock slightly more aligned to reality.
84 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
85 RFC says it is optional some validators care. The missing char was not
86 intended but triggered by a line-wrap alignement. Discovery and fix by
87 Guillaume Outters, hacked on by JH.
89 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
90 name being quoted was tainted a trap would be taken. Fix by using
91 dynamicaly created buffers. The routine could have been called by a
92 rewrite with the "h" flag, by using the "-F" command-line option, or
93 by using a "name=" option on a control=submission ACL modifier.
95 JH/19 SPF: change the Authentication-Results expansion component to give
96 smtp.helo when the sender domain is empty. Previously it gave
99 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
100 not in 127.0.0.0/8 to help in spotting list domains taken over by a
101 domain-parking registrar.
103 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
104 Previously when a whitespace character was specified it was not inserted
105 after removing the newline.
107 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
108 the domain part of the recipient address. This overrides any tls_sni
109 option set, which was previously used.
111 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
114 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
115 is_tainted() had an off-by-one error in the overenthusiastic direction.
116 Find and fix by Gavan. Although NetBSD is not a supported platform for
117 4.94 this bug could affect other platforms.
119 PP/01 Fix default prime selection to be consistent.
120 One path used ike23 still, instead of exim.dev.20160529.3; now both
121 execution flows will use the same DH primes (currently
122 exim.dev.20160529.3).
124 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
125 option in smtp transport, to match the documentation. Previously
126 verification was not being done in some cases where it should have been.
128 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
129 than one server was defined and depending on the platform memory layout
130 details, an internal consistency trap could be hit while walking the list
133 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
134 files. The value is used by the authresults expansion item. Previously
135 if this was used in a router or transport, a crash could result.
137 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
138 have a sufficient incidence of EINTR returns from select that an
139 interaction with other operations done by the main daemon loop exposed
140 a bug in the error-handling. This was benign apart from the log
143 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
144 for consistency with delivered "=>" and failed "**" lines. While we're
145 there, handle PRX and TFO.
147 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
148 applied. This resulted, if any header-line rewrite rules were configured,
149 in a panic-log trigerrable by sending a message with a long address in
150 a header. Fix by increaing the arbitrary limit to larger than a single
151 (dewrapped) 5322 header line maximum size.
153 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
154 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
155 RFC 6648 which deprecates X- options in protocols as a general practice.
156 Changeover between the implementations is handled by the mechanisms
159 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
160 a remote which uses callout/hold. Previously the local was queued.
162 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
165 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
166 log line. The asterisk is now consistently appended to the remote IP
167 (and port, if given), and will also be provided on defer and fail log
168 lines. Previously it could be placed on the local IP if that was being
169 logged, and was only provided on delivery lines.
175 JH/01 Avoid costly startup code when not strictly needed. This reduces time
176 for some exim process initialisations. It does mean that the logging
177 of TLS configuration problems is only done for the daemon startup.
179 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
181 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
182 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
185 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
186 DKIM signing is being done. Previously a transport_filter would always
187 disable CHUNKING, falling back to traditional DATA.
189 JH/05 Regard command-line recipients as tainted.
191 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
193 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
194 PAM library frees one of the arguments given to it, despite the
195 documentation. Therefore a plain malloc must be used.
197 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
198 on-stack buffers were used, resulting in a taint trap when DSN information
199 copied from a received message was written into the buffer.
201 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
202 the ordering of its ARC headers. This caused a crash.
204 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
205 a new record was being constructed with information from the peer, a trap
208 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
209 installation would get error messages from DMARC verify, when it hit the
210 nonexistent file indicated by the default. Distros wanting DMARC enabled
211 should both provide the file and set the option.
212 Also enforce no DMARC verification for command-line sourced messages.
214 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
215 could, depending on the platform, hang at the STARTTLS response.
217 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
218 message on a connection. Previously if one message had ARC headers and
219 the following one did not, a crash could result when adding an
220 Authentication-Results: header.
222 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
223 Exim main code and Exim-related utities. The introduction of taint
224 tracking also did many adjustments to string handling. Since then, eximon
225 frequently terminated with an assert failure.
227 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
228 check for 452 responses. This slightly helps the inefficieny of doing
229 a large alias-expansion into a recipient-limited target. The max_rcpt
230 transport option still applies (and at the current default, will override
231 the new feature). The check is done for either cause of synch, and forces
232 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
233 connection. The new facility is not tunable at this time.
235 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
236 library live data was being used, so the results became garbage. Make
237 copies while it is still usable.
239 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
240 on delivery deferred (==) and failed (**) lines (if a delivery was
241 attemtped). Previously it was only on completion (=>) lines.
243 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
244 for the expansion of the server_scram_iter and server_scram_salt options.
246 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
247 are now specifically given a NO_DATA response without hitting the system
248 resolver. The library goes on to do the now-standard TXT lookup.
249 Use of dnsdb lookups is not affected.
251 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
252 only retrieve the errormessage once. Previously two calls to dlerror()
253 were used, and the second one (for mainlog/paniclog) retrieved null
256 JH/20 Taint checking: disallow use of tainted data for
257 - the appendfile transport file and directory options
258 - the pipe transport command
259 - the autoreply transport file, log and once options
260 - file names used by the redirect router (including filter files)
262 - paths used by single-key lookups
263 Previously this was permitted.
265 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
266 adjusted the size of a major service buffer; this failed because the
267 buffer was in use at the time. Change to a compile-time increase in the
268 buffer size, when this authenticator is compiled into exim.
270 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
271 previous fast-mode was untenable in the face of glibs using mmap to
272 support larger malloc requests.
274 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
275 New values supported, if defined on system where compiled:
276 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
277 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
279 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
280 running a limited number of proceses in parallel, a benefit is gained. The
281 amount varies with the platform hardware and load. The use of the option
282 queue_run_in_order means we cannot do this, as ordering becomes
285 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
286 had introduced a string-copy (for ensuring NUL-termination) which was not
287 appropriate for that case, which can include embedded NUL bytes in the
288 block of data. Investigation showed the copy to actually be needless, the
289 data being length-specified.
291 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
292 done during a receiving connection, and both used TLS, global info was
293 used rather than per-connection info for tracking the state of data
294 queued for transmission. This could result in a connection hang.
296 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
297 Previously, when delivering serveral messages down a single connection
298 only the first would provide a SIZE. This was due to the size information
299 not being properly tracked.
301 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
302 TAI (at 37 seconds currently), pretend to be in UTC for time-related
303 expansion and logging. Previously, spurious values such as a future
304 minute could be seen.
306 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
307 it could crash from a null-deref. This could also affect the
308 ${addresses: } operator and ${readsock } item.
310 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
311 message following a mime one, the variable was not reset.
313 JH/30 When an pipelined-connect fails at the first response, assume incorrect
314 cached capability (perhaps the peer reneged?) and immediately retry in
317 JH/31 Fix spurious detection of timeout while writing to transport filter.
319 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
320 an attempt to copy the string was made before checking it.
322 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
323 taint of the lookup key was maintained; we now regard the presence in the
324 filesystem as sufficient validation.
326 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
327 argument is supplied.
329 JH/35 The dsearch lookup now requires that the directory is an absolute path.
330 Previously this was not checked, and nonempty relative paths made an
331 access under Exim's current working directory.
333 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
334 Previously no event was raised.
336 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
337 parameter supplied by the sender MAIL FROM command. Previously it was
338 ignored, and only the check_spool_space option value for the required
341 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
342 the size of the signing public-key. Previously it was instead giving
343 the size of the signature hash.
345 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
346 the default. See the (new) dkim_verify_min_keysizes option.
348 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
349 and an ACL use a lookup for checking either the local_part or domain,
350 stale data could be accessed. Ensure that variable references are
351 dropped between messages.
353 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
354 by the client was not checked as pointing within response data before
355 being used. A malicious client could thus cause an out-of-bounds read and
356 possibly gain authentication. Fix by adding the check.
358 JH/42 Internationalisation: change the default for downconversion in the smtp
359 transport to be "if needed". Previously it was "as previously set" for
360 the message, which usually meant "if needed" for message-submission but
361 "no" for everything else. However, MTAs have been seen using SMTPUTF8
362 even when the envelope addresses did not need it, resulting in forwarding
363 failures to non-supporting MTAs. A downconvert in such cases will be
364 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
365 transport. The change does mean that addresses needing conversion will
366 be converted when previously a delivery failure would occur.
368 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
369 response was received it would be used unchecked in a fail-DSN, violating
370 standards on line-length limits. Truncate if needed.
372 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
373 doesn't work. (Additionally add a new main config option to configure the
380 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
381 side, to decode a TLS 1.3 packet capture.
383 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
384 Previously the default library behaviour applied, sending two, each in
387 JH/03 Debug output for ACL now gives the config file name and line number for
390 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
392 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
394 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
395 buffer overrun for (non-chunking) other transports.
397 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
398 TLS1.3, means that a server rejecting a client certificate is not visible
399 to the client until the first read of encrypted data (typically the
400 response to EHLO). Add detection for that case and treat it as a failed
401 TLS connection attempt, so that the normal retry-in-clear can work (if
402 suitably configured).
404 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
405 and/or domain. Found and fixed by Jason Betts.
407 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
408 configuration). If a CNAME target was not a wellformed name pattern, a
411 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
412 the OS reports them interleaved with other addresses.
414 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
415 used both for input and for a verify callout, both encrypted, SMTP
416 responses being sent by the server could be lost. This resulted in
417 dropped connections and sometimes bounces generated by a peer sending
420 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
421 string. Previously it was possible to cause undefined behaviour in a
422 library routine (usually a crash). Found by "zerons".
424 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
427 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
428 API was removed, so update to use the newer ones.
430 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
431 any timeout set, is taking a long time. Previously we would hang on to a
432 rotated logfile "forever" if the input was arriving with long gaps
433 (a previous attempt to fix addressed lack, for a long time, of initial
436 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
437 shared (NFS) environment. The length of the tempfile name is now
438 4 + 16 ("hdr.$message_exim_id") which might break on file
439 systems which restrict the file name length to lower values.
442 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
443 shared (NFS) environment.
445 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
446 did for all versions <4.90). Notably -M, -m, --invert, -I may be
449 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
450 on some platforms for bit 31.
452 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
453 to changes apparently associated with TLS1.3 handling some of the APIs
454 previously used were either nonfunctional or inappropriate. Strings
455 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
456 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
457 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
458 This affects log line X= elements, the $tls_{in,out}_cipher variables,
459 and the use of specific cipher names in the encrypted= ACL condition.
461 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
463 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
464 verification result was not updated unless hosts_require_ocsp applied.
466 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
467 queue_list_requires_admin set to false, non-admin users were denied the
470 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
471 directory-of-certs mode. Previously they were advertised despite the
474 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
475 A single TCP connection by a client will now hold a TLS connection open
476 for multiple message deliveries, by default. Previously the default was to
479 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
480 default. If built with the facility, DANE will be used. The facility
481 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
483 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
484 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
485 must be defined and you must still, unless you define DISABLE_TLS, manage
486 the the include-dir and library-file requirements that go with that
487 choice. Non-TLS builds are still supported.
489 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
492 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
493 default. If the platform supports and has the facility enabled, it will
494 be requested on all coneections.
496 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
497 controlled by the build-time option SUPPORT_PIPE_CONNECT.
499 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
501 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
502 success-DSN messages. Previously the From: header was always the default
503 one for these; the option was ignored.
505 JH/28 Fix the timeout on smtp response to apply to the whole response.
506 Previously it was reset for every read, so a teergrubing peer sending
507 single bytes within the time limit could extend the connection for a
508 long time. Credit to Qualsys Security Advisory Team for the discovery.
510 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
511 delivery address, which leaked information of the results of local
512 forwarding. Change to the original envelope recipient address, per
515 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
516 requested. Previously not bounce was generated and a log entry of
517 error ignored was made.
519 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
521 JH/32 Introduce a general tainting mechanism for values read from the input
522 channel, and values derived from them. Refuse to expand any tainted
523 values, to catch one form of exploit.
525 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
526 was unused and the unexpanded text used for the test. Found and
527 fixed by Ruben Jenster.
529 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
530 an attempt to use a TLS library read routine dereffed a nul pointer,
533 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
534 them in our smtp response.
536 JH/36 Have the general router option retry_use_local_part default to true when
537 any of the restrictive preconditions are set (to anything). Previously it
538 was only for check_local user. The change removes one item of manual
539 configuration which is required for proper retries when a remote router
540 handles a subset of addresses for a domain.
542 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
543 link count into consideration.
545 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
546 caused the extension of big_buffer, the following lines were ignored.
548 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
549 accordance with RFC 2308. Previously there was no expiry, so a longlived
550 receive process (eg. due to ACL delays) versus a short SOA value could
553 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
555 JH/39 Promote DMARC support to mainline.
557 JH/40 Bug 2452: Add a References: header to DSNs.
559 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
560 parameters. The relevant library call is documented as "Deprecated: This
561 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
562 3.6.0, DH parameters are negotiated following RFC7919."
564 HS/06 Change the default of dnssec_request_domains to "*"
566 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
567 carried on and emitted a BDAT command, even when PIPELINING was not
570 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
571 buffer was used for the filename, resulting in a trap when tainted
572 arguments (eg. $domain) were used.
574 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
575 recommended to avoid a possible server-load attack. The feature can be
576 re-enabled via the openssl_options main cofiguration option.
578 JH/45 local_scan API: documented the current smtp_printf() call. This changed
579 for version 4.90 - adding a "more data" boolean to the arguments.
580 Bumped the ABI version number also, this having been missed previously;
581 release versions 4.90 to 4.92.3 inclusive were effectively broken in
582 respect of usage of smtp_printf() by either local_scan code or libraries
583 accessed via the ${dlfunc } expansion item. Both will need coding
584 adjustment for any calls to smtp_printf() to match the new function
585 signature; a FALSE value for the new argument is always safe.
587 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
588 the file-offset (which the Linux syscall does, and exim expects); this
589 resulted in an indefinite loop.
591 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
592 to do ARC verification. The Authentication-Results: header line added
593 by the configuration then had no ARC item.
599 JH/01 Remove code calling the customisable local_scan function, unless a new
600 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
602 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
603 non-signal-safe functions being used.
605 JH/03 Bug 2269: When presented with a received message having a stupidly large
606 number of DKIM-Signature headers, disable DKIM verification to avoid
607 a resource-consumption attack. The limit is set at twenty.
609 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
610 report of oldest_pass in ${authres } in consequence, and separate out
611 some descriptions of reasons for verification fail.
613 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
614 files in the spool were present and unlocked. A queue-runner could spot
615 them, resulting in a duplicate delivery. Fix that by doing the unlock
616 after the unlink. Investigation by Tim Stewart. Take the opportunity to
617 add more error-checking on spoolfile handling while that code is being
620 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
621 No known attacks, no CVE, this is defensive hardening.
623 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
624 a queue-runner could start a delivery while other operations were ongoing.
625 Cutthrough delivery was a common victim, resulting in duplicate delivery.
626 Found and investigated by Tim Stewart. Fix by using the open message data
627 file handle rather than opening another, and not locally closing it (which
628 releases a lock) for that case, while creating the temporary .eml format
629 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
631 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
632 $sender_verify_failure/$recipient_verify_failure to "random".
634 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
637 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
638 Previously this would segfault.
640 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
643 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
644 like zero, since the resolver should be doing this for us, But we need one
645 as a CNAME but no MX presence gets the CNAME returned; we need to check
646 that doesn't point to an MX to declare it "no MX returned" rather than
647 "error, loop". A new main option is added so the older capability of
648 following some limited number of chain links is maintained.
650 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
652 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
653 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
654 now-deprecated earlier definitions used only specified the range up to TLS
655 1.2 (in the older-version library docs).
657 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
659 JH/15 Rework TLS client-side context management. Stop using a global, and
660 explicitly pass a context around. This enables future use of TLS for
661 connections to service-daemons (eg. malware scanning) while a client smtp
662 connection is using TLS; with cutthrough connections this is quite likely.
664 JH/16 Fix ARC verification to do AS checks in reverse order.
666 JH/17 Support a "tls" option on the ${readsocket } expansion item.
668 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
669 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
670 Previously the "utf8" would be re-prepended for every additional message.
672 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
673 Previously thery were accepted, resulting in issues when attempting to
674 forward messages to a non-supporting MTA.
676 PP/02 Let -n work with printing macros too, not just options.
678 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
679 one parent address was copied, and bogus data was used at delivery-logging
680 time. Either a crash (after delivery) or bogus log data could result.
681 Discovery and analysis by Tim Stewart.
683 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
684 Previously if the string ended mid-character, we did not insert the
685 promised '?' replacement.
687 PP/04 Documentation: current string operators work on bytes, not codepoints.
689 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
690 should pack well giving a smaller memory footprint so better caching and
691 therefore performance. Group the declarations where this can't be done so
692 that the byte-sized flag variables are not interspersed among pointer
693 variables, giving a better chance of good packing by the compiler.
695 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
696 non-null, to avoid issues with sites running BATV. Previously reports were
697 sent with an empty envelope sender so looked like bounces.
699 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
700 The ignore_error flag wasn't being returned from the filter subprocess so
701 was not set for later routers. Investigation and fix by Matthias Kurz.
703 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
704 and a msg:complete for the whole, when a message is manually removed using
705 -Mrm. Developement by Matthias Kurz, hacked on by JH.
707 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
708 a "Gnu special" function, asprintf() in the DB utility binary builds; I
709 hope that is portable enough.
711 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
712 requiring a known-CA anchor certificate; make it now rely entirely on the
713 TLSA as an anchor. Checking the name on the leaf cert against the name
714 on the A-record for the host is still done for TA (but not for EE mode).
716 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
717 included in delivery lines for non-proxied connections, when compiled with
718 SUPPORT_SOCKS and running with proxy logging enabled.
720 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
721 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
722 move the existing event to fire before the normal logging of message
723 failure so that custom logging is bracketed by normal logging.
725 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
726 msg:fail:internal event. Developement by Matthias Kurz.
728 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
729 far too small for todays use of crypto signatures stored there. Go all
730 the way to the max DNS message size of 64kB, even though this might be
731 overmuch for IOT constrained device use.
733 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
734 copy a string over itself. The library routine is documented as not
735 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
737 JH/32 For main options check_spool_space and check_inode_space, where the
738 platform supports 64b integers, support more than the previous 2^31 kB
739 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
740 the previous G, M, k.
742 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
743 $authenticated_fail_id variable on authentication failure. Previously
746 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
747 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
748 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
749 GNUTLS_SEC_PARAM_MEDIUM.
751 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
752 side. Previously we would continue as if no SNI had been received.
754 JH/36 Harden the handling of string-lists. When a list consisted of a sole
755 "<" character, which should be a list-separator specification, we walked
756 off past the nul-terimation.
758 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
759 causes) even when the retry time is not yet met. Previously they were
760 not, meaning that when (say) an account was over-quota and temp-rejecting,
761 and multiple senders' messages were queued, only one sender would get
762 notified on each configured delay_warning cycle.
764 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
766 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
767 cipher-suites, an error can be left on the stack even for a succeeding
768 accept; this results in impossible error messages when a later operation
771 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
772 return error codes indicating retry. Under TLS1.3 this becomes required.
774 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
775 it only wrote the new authenticators, resulting in a lack of tracking of
776 peer changes of ESMTP extensions until the next cache flush.
778 JH/41 Fix the loop reading a message header line to check for integer overflow,
779 and more-often against header_maxsize. Previously a crafted message could
780 induce a crash of the recive process; now the message is cleanly rejected.
782 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
783 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
789 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
790 When redis_servers is set to a list of > 1 element, and the Redis servers
791 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
792 case of MOVED into a DEFER case instead, thus moving the query onto the
793 next server in the list. For a cluster of N elements, all N servers must
794 be defined in redis_servers.
796 GF/02 Catch and remove uninitialized value warning in exiqsumm
797 Check for existence of @ARGV before looking at $ARGV[0]
799 JH/01 Replace the store_release() internal interface with store_newblock(),
800 which internalises the check required to safely use the old one, plus
801 the allocate and data copy operations duplicated in both (!) of the
802 extant use locations.
804 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
805 modifier. This matches the restriction on the commandline.
807 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
808 Previously only the last row was returned.
810 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
811 we assumed that tags in the header were well-formed, and parsed the
812 element content after inspecting only the first char of the tag.
813 Assumptions at that stage could crash the receive process on malformed
816 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
817 While running the DKIM ACL we operate on the Permanent memory pool so that
818 variables created with "set" persist to the DATA ACL. Also (at any time)
819 DNS lookups that fail create cache records using the Permanent pool. But
820 expansions release any allocations made on the current pool - so a dnsdb
821 lookup expansion done in the DKIM ACL releases the memory used for the
822 DNS negative-cache, and bad things result. Solution is to switch to the
823 Main pool for expansions.
824 While we're in that code, add checks on the DNS cache during store_reset,
825 active in the testsuite.
826 Problem spotted, and debugging aided, by Wolfgang Breyha.
828 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
829 When none of the hosts presented to a transport match an already-open
830 connection, close it and proceed with the list. Previously we would
831 queue the message. Spotted by Lena with Yahoo, probably involving
834 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
835 Previously a spurious "250 OK id=" response was appended to the proper
838 JH/08 The "support for" informational output now, which built with Content
839 Scanning support, has a line for the malware scanner interfaces compiled
840 in. Interface can be individually included or not at build time.
842 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
843 by the template makefile "src/EDITME". The "STREAM" support for an older
844 ClamAV interface method is removed.
846 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
847 rows affected is given instead).
849 JH/11 The runtime Berkeley DB library version is now additionally output by
850 "exim -d -bV". Previously only the compile-time version was shown.
852 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
853 SMTP connection. Previously, when one had more recipients than the
854 first, an abortive onward connection was made. Move to full support for
855 multiple onward connections in sequence, handling cutthrough connection
856 for all multi-message initiating connections.
858 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
859 routers. Previously, a multi-recipient message would fail to match the
860 onward-connection opened for the first recipient, and cause its closure.
862 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
863 a timeout on read on a GnuTLS initiating connection, resulting in the
864 initiating connection being dropped. This mattered most when the callout
865 was marked defer_ok. Fix to keep the two timeout-detection methods
868 JH/15 Relax results from ACL control request to enable cutthrough, in
869 unsupported situations, from error to silently (except under debug)
870 ignoring. This covers use with PRDR, frozen messages, queue-only and
873 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
875 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
876 metadata, resulting in a crash in free().
878 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
879 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
880 Broken also in d185889f4, with init system revamp.
882 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
883 Previously we abruptly closed the connection after reading a malware-
884 found indication; now we go on to read the "scan ok" response line,
887 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
888 ACL. Previously, a crash would result.
890 JH/19 Speed up macro lookups during configuration file read, by skipping non-
891 macro text after a replacement (previously it was only once per line) and
892 by skipping builtin macros when searching for an uppercase lead character.
894 JH/20 DANE support moved from Experimental to mainline. The Makefile control
895 for the build is renamed.
897 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
898 was allocated for every new TLS startup, meaning one per message. Fix
899 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
901 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
902 reported the original. Fix to report (as far as possible) the ACL
903 result replacing the original.
905 JH/23 Fix memory leak during multi-message connections using STARTTLS under
906 OpenSSL. Certificate information is loaded for every new TLS startup,
907 and the resources needed to be freed.
909 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
911 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
914 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
915 DATA response info to the (existing) per-recipient response info for
916 the "C=" log element. It can have useful tracking info from the
917 destination system. Patch from Simon Arlott.
919 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
920 length value. Previously this would segfault.
922 HS/02 Support Avast multiline protoocol, this allows passing flags to
923 newer versions of the scanner.
925 JH/28 Ensure that variables possibly set during message acceptance are marked
926 dead before release of memory in the daemon loop. This stops complaints
927 about them when the debug_store option is enabled. Discovered specifically
928 for sender_rate_period, but applies to a whole set of variables.
929 Do the same for the queue-runner and queue-list loops, for variables set
930 from spool message files. Do the same for the SMTP per-message loop, for
931 certain variables indirectly set in ACL operations.
933 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
934 as a multi-recipient message from a mailinglist manager). The coding had
935 an arbitrary cutoff number of characters while checking for more input;
936 enforced by writing a NUL into the buffer. This corrupted long / fast
937 input. The problem was exposed more widely when more pipelineing of SMTP
938 responses was introduced, and one Exim system was feeding another.
939 The symptom is log complaints of SMTP syntax error (NUL chars) on the
940 receiving system, and refused recipients seen by the sending system
941 (propating to people being dropped from mailing lists).
942 Discovered and pinpointed by David Carter.
944 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
945 replaced by the ${authresults } expansion.
947 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
949 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
950 allows proper process termination in container environments.
952 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
953 Previously the "final dot" had a newline after it; ensure it is CR,LF.
955 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
956 and "err_perm", deprecated since 4.83 when the RFC-defined words
957 "temperror" and "permerror" were introduced.
959 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
960 transport-filters or DKIM-signing. The restriction was lost in the
961 consolidation of verify-callout and delivery SMTP handling.
962 Extend the restriction to also cover ARC-signing.
964 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
965 in defer=pass mode supply a 450 to the initiator. Previously the message
968 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
969 tls_require_ciphers is used as before.
971 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
972 "pass_unscanned". Only tmpfails from the scanner are written to
973 the paniclog, as they may require admin intervention (permission
974 denied, license issues). Other scanner errors (like decompression
975 bombs) do not cause a paniclog entry.
977 JH/36 Fix reinitialisation of DKIM logging variable between messages.
978 Previously it was possible to log spurious information in receive log
981 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
982 triggered odd behaviour from Outlook Express clients.
984 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
987 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
988 since the IETF WG has not yet settled on that versus the original
989 "bare" representation.
991 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
992 Previously the millisecond value corrupted the output.
993 Fix also for syslog_pid=no and log_selector +pid, for which the pid
994 corrupted the output.
1000 JH/01 Rework error string handling in TLS interface so that the caller in
1001 more cases is responsible for logging. This permits library-sourced
1002 string to be attached to addresses during delivery, and collapses
1003 pairs of long lines into single ones.
1005 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1006 during configuration. Wildcards are allowed and expanded.
1008 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1009 This permits better logging.
1011 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1012 do not close it down and have the child transport start it up again on
1013 the passed-on TCP connection. Instead, proxy the child (and any
1014 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1015 affected: the continued delivery log lines do not have any DNSSEC, TLS
1016 Certificate or OCSP information. TLS cipher information is still logged.
1018 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1019 identical IP addresses on different listening ports. Will also affect
1022 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1023 add noisy ifdef guards to special-case this sillyness.
1024 Patch from Bernd Kuhls.
1026 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1027 than 255 are no longer allowed.
1029 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1030 Disable the session-cache too, which might reduce our load. Since we
1031 currrectly use a new context for every connection, both as server and
1032 client, there is no benefit for these.
1033 GnuTLS appears to not support tickets server-side by default (we don't
1034 call gnutls_session_ticket_enable_server()) but client side is enabled
1035 by default on recent versions (3.1.3 +) unless the PFS priority string
1038 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1039 <https://reproducible-builds.org/specs/source-date-epoch/>.
1041 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1042 the check for any unsuccessful recipients did not notice the limit, and
1043 erroneously found still-pending ones.
1045 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1046 MSG_MORE. Only in-clear (not on TLS connections).
1048 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1049 transport-filter is involved we can buffer the headers in memory for
1050 creating the signature, and read the spool data file once for the
1051 signature and again for transmission.
1053 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1054 4.77 as the kernel support then wasn't solid, having issues in 64bit
1055 mode. Now, it's been long enough. Add support for FreeBSD also.
1057 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1058 case where the routing stage had gathered several addresses to send to
1059 a host before calling the transport for the first, we previously failed
1060 to close down TLS in the old transport process before passing the TCP
1061 connection to the new process. The new one sent a STARTTLS command
1062 which naturally failed, giving a failed delivery and bloating the retry
1063 database. Investigation and fix prototype from Wolfgang Breyha.
1065 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1066 false-negatives in the check that the sender had not preempted a response
1067 or prompt from Exim (running as a server), due to that code's lack of
1068 awareness of the SMTP input buffering.
1070 PP/04 Add commandline_checks_require_admin option.
1071 Exim drops privileges sanely, various checks such as -be aren't a
1072 security problem, as long as you trust local users with access to their
1073 own account. When invoked by services which pass untrusted data to
1074 Exim, this might be an issue. Set this option in main configuration
1075 AND make fixes to the calling application, such as using `--` to stop
1078 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1079 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1080 responses to those, into a single TLS record each way (this usually means
1081 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1084 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1085 affects you only if you're dancing at the edge of the param size limits.
1086 If you are, and this message makes sense to you, then: raise the
1087 configured limit or use OpenSSL 1.1. Nothing we can do for older
1090 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1091 cmdline element to get the documented default one. Previously it was
1094 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1095 are unsuable for non-message-specific reasons - eg. connection timeout,
1096 banner-time rejection.
1098 JH/16 Fix logging of delivery remote port, when specified by router, under
1101 PP/06 Repair manualroute's ability to take options in any order, even if one
1102 is the name of a transport.
1105 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1107 JH/17 Change the list-building routines interface to use the expanding-string
1108 triplet model, for better allocation and copying behaviour.
1110 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1111 Previously it was constructed the first time a possibly-matching string
1112 was met in the configuration file input during startup; now it is done
1115 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1116 compatible one, to avoid the (poorly documented) possibility of a config
1117 file in the working directory redirecting the DB files, possibly correpting
1118 some existing file. CVE-2017-10140 assigned for BDB.
1120 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1121 cache-hot. Previously, although the result was properly cached, the
1122 initial verify call returned a defer.
1124 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1125 the main verify for receipient in uncached-mode.
1127 JH/22 Retire historical build files to an "unsupported" subdir. These are
1128 defined as "ones for which we have no current evidence of testing".
1130 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1131 if present. Previously it was ignored.
1133 JH/24 Start using specified-initialisers in C structure init coding. This is
1134 a C99 feature (it's 2017, so now considered safe).
1136 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1137 if was a fixed-sized field and bitmask ops via macros; it is now more
1140 PP/07 GitHub PR 56: Apply MariaDB build fix.
1141 Patch provided by Jaroslav Škarvada.
1143 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1144 during Coverity cleanups [4.87 JH/47]
1145 Diagnosis and fix provided by Michael Fischer v. Mollard.
1147 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1148 the right size to place the terminating semicolon on its own folded
1149 line, the header hash was calculated to an incorrect value thanks to
1150 the (relaxed) space the fold became.
1152 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1153 and confused the parent.
1155 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1156 which could crash as a result. This could lead to undeliverable messages.
1158 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1161 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1162 matters on fast-turnover and PID-randomising systems, which were getting
1163 out-of-order delivery.
1165 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1166 a possibly-overlapping copy. The symptom was that "Remote host closed
1167 connection in response to HELO" was logged instead of the actual 4xx
1170 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1171 Previously only that bufferd was discarded, resulting in SYMTP command
1174 JH/32 DKIM: when a message has multiple signatures matching an identity given
1175 in dkim_verify_signers, run the dkim acl once for each. Previously only
1176 one run was done. Bug 2189.
1178 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1179 "panic the current process" to "deliberately defer". The panic log is
1180 still written with the problem list name; the mail and reject logs now
1181 get a temp-reject line for the message that was being handled, saying
1182 something like "domains check lookup or other defer". The SMTP 451
1183 message is still "Temporary local problem".
1185 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1186 A crafted sequence of BDAT commands could result in in-use memory beeing
1187 freed. CVE-2017-16943.
1189 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1190 from SMTP input. Previously it was always done; now only done for DATA
1191 and not BDAT commands. CVE-2017-16944.
1193 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1194 to the message (such as an overlong header line). Previously this was
1195 not done and we did not exit BDAT mode. Followon from the previous item
1196 though a different problem.
1202 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1203 than -2003 did; needs libidn2 in addition to libidn.
1205 JH/02 The path option on a pipe transport is now expanded before use.
1207 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1208 Patch provided by "Björn", documentation fix added too.
1210 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1211 missing a wire-to-host endian conversion.
1213 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1214 close after a BDAT command line could be taken as a following command,
1215 giving a synch failure. Fix by only checking for synch immediately
1216 before acknowledging the chunk.
1218 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1219 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1220 macro. Patches provided by Josh Soref.
1222 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1223 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1224 by IANA as a service extension. However, John Klensin suggests that we
1227 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1228 the dkim code may be unix-mode line endings rather than smtp wire-format
1229 CRLF, so prepend a CR to any bare LF.
1231 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1232 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1234 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1235 the first were themselves being wrongly included in the feed into dkim
1236 processing; with most chunk sizes in use this resulted in an incorrect
1237 body hash calculated value.
1239 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1240 DKIM signature block, for verification. Although advised against by
1241 standards it is specifically not ruled illegal.
1243 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1245 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1246 missing a body hash (the bh= tag).
1248 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1249 It seems that HAProxy sends the Proxy Protocol information in clear and
1250 only then does a TLS startup, so do the same.
1252 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1253 TCP connections (such as for Spamd) unless the daemon successfully set
1254 Fast Open mode on its listening sockets. This fixes breakage seen on
1255 too-old kernels or those not configured for Fast Open, at the cost of
1256 requiring both directions being enabled for TFO, and TFO never being used
1257 by non-daemon-related Exim processes.
1259 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1260 endings, at least on the first header line. Try to canonify any that get
1261 past that check, despite the cost.
1263 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1264 now limited to an arbitrary five deep, while parsing addresses with the
1265 strip_excess_angle_brackets option enabled.
1267 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1268 instead leave the unprompted TLS handshake in socket buffer for the
1269 TLS library to consume.
1271 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1273 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1275 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1276 region in one operation, for speed, and this leaves assigned pointers
1277 dangling. Add checks run only under the testsuite which checks all
1278 variables at a store-reset and panics on a dangling pointer; add code
1279 explicitly nulling out all the variables discovered. Fixes one known
1280 bug: a transport crash, where a dangling pointer for $sending_ip_address
1281 originally assigned in a verify callout, is re-used.
1283 PP/06 Drop '.' from @INC in various Perl scripts.
1285 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1287 PP/08 Reduce a number of compilation warnings under clang; building with
1288 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1289 should be warning-free.
1291 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1293 HS/01 Fix portability problems introduced by PP/08 for platforms where
1294 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1296 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1297 chunk. This allows us to accept broken chunked messages. We need a more
1298 general solution here.
1300 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1301 already-broken messages in the queue.
1303 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1305 JH/19 Fix reference counting bug in routing-generated-address tracking.
1311 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1312 supports it and a size is available (ie. the sending peer gave us one).
1314 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1315 years of being deprecated). The replacements are the ACLs
1316 acl_smtp_mime and acl_not_smtp_mime.
1318 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1319 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1320 or even an in-clear connection were permitted. Now, if the host lookup
1321 was dnssec and dane was requested then the host is only used if the
1322 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1323 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1324 if one fails this test.
1325 This means that a poorly-configured remote DNS will make it incommunicado;
1326 but it protects against a DNS-interception attack on it.
1328 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1329 non-noisy when a race steals the message being considered.
1331 JH/05 If main configuration option tls_certificate is unset, generate a
1332 self-signed certificate for inbound TLS connections.
1334 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1335 in rewrites and routers.
1337 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1338 and logged a warning sing 4.83; now they are a configuration file error.
1340 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1341 (lacking @domain). Apply the same qualification processing as RCPT.
1343 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1345 JH/10 Support ${sha256:} applied to a string (as well as the previous
1348 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1349 a cutthrough deliver is pending, as we always want to make a connection.
1350 This also avoids re-routing the message when later placing the cutthrough
1351 connection after a verify cache hit.
1352 Do not update it with the verify result either.
1354 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1355 when routing results in more than one destination address.
1357 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1358 signing (which inhibits the cutthrough capability). Previously only
1359 the presence of an option was tested; now an expansion evaluating as
1360 empty is permissible (obviously it should depend only on data available
1361 when the cutthrough connection is made).
1363 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1364 the relevant preceding SMTP command did not note the pipelining mode.
1366 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1367 Previously they were not counted.
1369 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1370 as one having no matching records. Previously we deferred the message
1371 that needed the lookup.
1373 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1374 distinguished as "(=".
1376 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1377 for missing MX records. Previously it only worked for missing A records.
1379 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1381 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1382 after the data-go-ahead and data-ack. Patch from Jason Betts.
1384 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1385 even for a "none" policy. Patch from Tony Meyer.
1387 JH/22 Fix continued use of a connection for further deliveries. If a port was
1388 specified by a router, it must also match for the delivery to be
1391 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1392 When one of the recipients of a message was unsuitable for the connection
1393 (has no matching addresses), we lost track of needing to mark it
1394 deferred. As a result mail would be lost.
1396 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1398 JH/25 Decoding ACL controls is now done using a binary search; the source code
1399 takes up less space and should be simpler to maintain. Merge the ACL
1400 condition decode tables also, with similar effect.
1402 JH/26 Fix problem with one_time used on a redirect router which returned the
1403 parent address unchanged. A retry would see the parent address marked as
1404 delivered, so not attempt the (identical) child. As a result mail would
1407 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1408 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1409 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1412 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1413 Main config options check_{log,spool}_{inodes,space} are now
1414 100 inodes, 10MB unless set otherwise in the configuration.
1416 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1417 Previously it only applied to the main-section connection policy
1420 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1422 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1423 by me. Added RFC7919 DH primes as an alternative.
1425 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1426 are not in the system include path.
1428 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1429 GnuTLS, when a session startup failed (eg because the client disconnected)
1430 Exim did stdio operations after fclose. This was exposed by a recent
1431 change which nulled out the file handle after the fclose.
1433 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1434 signed directly by the cert-signing cert, rather than an intermediate
1435 OCSP-signing cert. This is the model used by LetsEncrypt.
1437 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1439 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1440 an incoming connection.
1442 HS/02 Bug 1802: Do not half-close the connection after sending a request
1445 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1446 fallback to "prime256v1".
1448 JH/34 SECURITY: Use proper copy of DATA command in error message.
1449 Could leak key material. Remotely exploitable. CVE-2016-9963.
1455 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1456 and 3.4.4 - once the server is enabled to respond to an OCSP request
1457 it does even when not requested, resulting in a stapling non-aware
1458 client dropping the TLS connection.
1460 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1461 support variable-length bit vectors. No functional change.
1463 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1464 The I= interface field on outgoing lines is now after the H= remote
1465 host field, same as incoming lines. There is a separate
1466 outgoing_interface log selector which allows you to disable the
1469 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1470 If not running log_selector +smtp_connection the mainlog would be held
1471 open indefinitely after a "too many connections" event, including to a
1472 deleted file after a log rotate. Leave the per net connection logging
1473 leaving it open for efficiency as that will be quickly detected by the
1474 check on the next write.
1476 HS/01 Bug 1671: Fix post transport crash.
1477 Processing the wait-<transport> messages could crash the delivery
1478 process if the message IDs didn't exist for some reason. When
1479 using 'split_spool_directory=yes' the construction of the spool
1480 file name failed already, exposing the same netto behaviour.
1482 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1483 mime_regex ACL conditions.
1485 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1486 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1487 to HELO, local diagnostic string.
1489 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1490 log line to debug. Even when configured with a tls authenticator many
1491 client connections are expected to not authenticate in this way, so
1492 an authenticate fail is not an error.
1494 HS/02 Add the Exim version string to the process info. This way exiwhat
1495 gives some more detail about the running daemon.
1497 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1498 matter for fast-change records such as DNSBLs.
1500 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1501 retry record, even if constant. There may be multiple transports with
1502 different interface settings and the retry behaviour needs to be kept
1505 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1507 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1509 JH/10 Bug 840: fix log_defer_output option of pipe transport
1511 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1512 hosts_randomize. This matters a lot when combined with mua_wrapper.
1514 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1515 ${quote_pgsql:<string>} operator.
1517 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1519 JH/14 Change header file name for freeradius-client. Relevant if compiling
1520 with Radius support; from the Gentoo tree and checked under Fedora.
1522 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1524 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1525 option transport_filter, meaning no filtering.
1527 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1529 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1530 defaults to "*" (all hosts). The variable is now available when not built
1531 with TLS, default unset, mainly to enable keeping the testsuite sane.
1532 If a server certificate is not supplied (via tls_certificate) an error is
1533 logged, and clients will find TLS connections fail on startup. Presumably
1534 they will retry in-clear.
1535 Packagers of Exim are strongly encouraged to create a server certificate
1536 at installation time.
1538 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1539 with the $config_file variable.
1541 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1542 in transport context, after the attempt, and per-recipient. The latter type
1543 is per host attempted. The event data is the error message, and the errno
1544 information encodes the lookup type (A vs. MX) used for the (first) host,
1545 and the trailing two digits of the smtp 4xx response.
1547 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1548 to write to mainlog (or rejectlog, paniclog) in the window between file
1549 creation and permissions/ownership being changed. Particularly affects
1550 installations where exicyclog is run as root, rather than exim user;
1551 result is that the running daemon panics and dies.
1553 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1555 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1556 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1557 "pri" and "weight". Note that the previous implicit priority given by the
1558 list order is no longer honoured.
1560 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1561 for DKIM processing.
1563 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1564 by defining SUPPORT_SOCKS.
1566 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1567 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1568 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1569 variables are renamed to proxy_{local,external}_{address,port}.
1571 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1572 for a build by defining SUPPORT_I18N
1574 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1575 of the query string, and make ${quote_redis:} do that quoting.
1577 JH/29 Move Events support from Experimental to mainline, enabled by default
1578 and removable for a build by defining DISABLE_EVENT.
1580 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1582 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1583 cached by the daemon.
1585 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1586 by defining LOOKUP_REDIS. The libhiredis library is required.
1588 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1589 keys are given for lookup.
1591 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1592 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1593 only supported when built with TLS support. The PolarSSL SHA routines
1594 are still used when the TLS library is too old for convenient support.
1596 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1597 openssl_options), for security. OpenSSL forces this from version 1.1.0
1598 server-side so match that on older versions.
1600 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1601 allocation for $value could be released as the expansion processing
1602 concluded, but leaving the global pointer active for it.
1604 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1605 and to use the domains and local_parts ACL conditions.
1607 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1608 incorrectly not doubled on cutthrough transmission, hence seen as a
1609 body-termination at the receiving system - resulting in truncated mails.
1610 Commonly the sender saw a TCP-level error, and retransmitted the message
1611 via the normal store-and-forward channel. This could result in duplicates
1612 received - but deduplicating mailstores were liable to retain only the
1613 initial truncated version.
1615 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1617 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1619 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1620 we're in there, support oversigning also; bug 1309.
1622 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1624 HS/04 Add support for keep_environment and add_environment options.
1626 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1627 either intentional arithmetic overflow during PRNG, or testing config-
1630 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1631 delivery resulted in actual delivery. Cancel cutthrough before DATA
1634 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1635 rejects a recipient: pass the reject to the originator.
1637 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1638 Many were false-positives and ignorable, but it's worth fixing the
1641 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1642 for the new environment-manipulation done at startup. Move the routines
1643 from being local to tls.c to being global via the os.c file.
1645 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1646 an extract embedded as result-arg for a map, the first arg for extract
1647 is unavailable so we cannot tell if this is a numbered or keyed
1648 extraction. Accept either.
1654 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1657 JH/02 The smtp transport option "multi_domain" is now expanded.
1659 JH/03 The smtp transport now requests PRDR by default, if the server offers
1662 JH/04 Certificate name checking on server certificates, when exim is a client,
1663 is now done by default. The transport option tls_verify_cert_hostnames
1664 can be used to disable this per-host. The build option
1665 EXPERIMENTAL_CERTNAMES is withdrawn.
1667 JH/05 The value of the tls_verify_certificates smtp transport and main options
1668 default to the word "system" to access the system default CA bundle.
1669 For GnuTLS, only version 3.0.20 or later.
1671 JH/06 Verification of the server certificate for a TLS connection is now tried
1672 (but not required) by default. The verification status is now logged by
1673 default, for both outbound TLS and client-certificate supplying inbound
1676 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1679 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1680 Status Notification (bounce) messages are now MIME format per RFC 3464.
1681 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1682 under the control of the dsn_advertise_hosts option, and routers may
1683 have a dsn_lasthop option.
1685 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1686 default, modifiable by a malware= option. The list separator for
1687 the options can now be changed in the usual way. Bug 68.
1689 JH/10 The smtp_receive_timeout main option is now expanded before use.
1691 JH/11 The incoming_interface log option now also enables logging of the
1692 local interface on delivery outgoing connections.
1694 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1695 if the interface and destination host and port all match.
1697 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1700 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1701 Patch from Andrew Lewis.
1703 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1704 now supports optional time-restrictions, weighting, and priority
1705 modifiers per server. Patch originally by <rommer@active.by>.
1707 JH/16 The spamd_address main option now supports a mixed list of local
1708 and remote servers. Remote servers can be IPv6 addresses, and
1709 specify a port-range.
1711 JH/17 Bug 68: The spamd_address main option now supports an optional
1712 timeout value per server.
1714 JH/18 Bug 1581: Router and transport options headers_add/remove can
1715 now have the list separator specified.
1717 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1720 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1723 JH/21 Support for the A6 type of dns record is withdrawn.
1725 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1726 rather than the verbs used.
1728 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1729 from 255 to 1024 chars.
1731 JH/24 Verification callouts now attempt to use TLS by default.
1733 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1734 are generic router options now. The defaults didn't change.
1736 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1737 Original patch from Alexander Shikoff, worked over by JH.
1739 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1740 files if ZCAT_COMMAND is not executable.
1742 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1744 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1746 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1747 Normally benign, it bites when the pair was led to by a CNAME;
1748 modern usage is to not canonicalize the domain to a CNAME target
1749 (and we were inconsistent anyway for A-only vs AAAA+A).
1751 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1753 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1754 when evaluating $sender_host_dnssec.
1756 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1757 $sender_helo_dnssec variable.
1759 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1761 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1763 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1765 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1766 documented as working, but never had. Support all but $spam_report.
1768 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1769 added for tls authenticator.
1771 HS/03 Add perl_taintmode main config option
1777 TL/01 When running the test suite, the README says that variables such as
1778 no_msglog_check are global and can be placed anywhere in a specific
1779 test's script, however it was observed that placement needed to be near
1780 the beginning for it to behave that way. Changed the runtest perl
1781 script to read through the entire script once to detect and set these
1782 variables, reset to the beginning of the script, and then run through
1783 the script parsing/test process like normal.
1785 TL/02 The BSD's have an arc4random API. One of the functions to induce
1786 adding randomness was arc4random_stir(), but it has been removed in
1787 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1788 function when detected.
1790 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1791 cause callback expansion.
1793 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1794 syntax errors in an expansion can be treated as a string instead of
1795 logging or causing an error, due to the internal use of bool_lax
1796 instead of bool when processing it.
1798 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1799 server certificates when making smtp deliveries.
1801 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1803 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1805 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1807 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1808 Merged patch from Sebastian Wiedenroth.
1810 JH/05 Fix results-pipe from transport process. Several recipients, combined
1811 with certificate use, exposed issues where response data items split
1812 over buffer boundaries were not parsed properly. This eventually
1813 resulted in duplicates being sent. This issue only became common enough
1814 to notice due to the introduction of connection certificate information,
1815 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1817 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1818 size buffer was used, resulting in syntax errors when an expansion
1821 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1822 version 3.3.6 or later.
1824 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1825 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1826 both become "event_action", the variables become $event_name, $event_data
1827 and $event_defer_errno. There is a new variable $verify_mode, usable in
1828 routers, transports and related events. The tls:cert event is now also
1829 raised for inbound connections, if the main configuration event_action
1832 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1833 early OCSP support, but no stapling (appears to be less than 1.0.0).
1835 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1836 server certificate names available under the smtp transport option
1837 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1840 JH/10 Time-related extraction expansions from certificates now use the main
1841 option "timezone" setting for output formatting, and are consistent
1842 between OpenSSL and GnuTLS compilations. Bug 1541.
1844 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1845 encoded parameter in the incoming message. Bug 1558.
1847 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1848 include certificate info, eximon was claiming there were spoolfile
1851 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1853 JH/14 Log delivery-related information more consistently, using the sequence
1854 "H=<name> [<ip>]" wherever possible.
1856 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1857 are problematic for Debian distribution, omit them from the release
1860 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1862 JH/16 Fix string representation of time values on 64bit time_t architectures.
1865 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1866 output list separator was used.
1871 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1872 checkers that were complaining about end of non-void function with no
1875 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1876 This was a regression introduced in 4.83 by another bugfix.
1878 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1880 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1881 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1887 TF/01 Correctly close the server side of TLS when forking for delivery.
1889 When a message was received over SMTP with TLS, Exim failed to clear up
1890 the incoming connection properly after forking off the child process to
1891 deliver the message. In some situations the subsequent outgoing
1892 delivery connection happened to have the same fd number as the incoming
1893 connection previously had. Exim would try to use TLS and fail, logging
1894 a "Bad file descriptor" error.
1896 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1897 utilities have not been installed.
1899 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1900 temporary space as the ACL may create new global variables.
1902 TL/01 LDAP support uses per connection or global context settings, depending
1903 upon the detected version of the libraries at build time.
1905 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1906 to extract and use the src ip:port in logging and expansions as if it
1907 were a direct connection from the outside internet. PPv2 support was
1908 updated based on HAProxy spec change in May 2014.
1910 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1912 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1913 Properly escape header and check for NULL return.
1915 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1918 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1920 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1921 characters in header names, implemented as a verify condition.
1922 Contributed by Michael Fischer v. Mollard.
1924 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1925 results permerror and temperror. Previous values are deprecated but
1926 still accepted. In a future release, err_perm and err_temp will be
1927 completely removed, which will be a backward incompatibility if the
1928 ACL tests for either of these two old results. Patch contributed by
1929 user bes-internal on the mailing list.
1931 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1933 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1934 selectors, in both main and reject logs.
1936 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1939 JH/07 Add malware type "sock" for talking to simple daemon.
1941 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1943 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1944 routers/transports under cutthrough routing.
1946 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1947 numbers. Touch up "bool" conditional to keep the same definition.
1949 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1951 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1953 JH/12 Expand items in router/transport headers_add or headers_remove lists
1954 individually rather than the list as a whole. Bug 1452.
1956 Required for reasonable handling of multiple headers_ options when
1957 they may be empty; requires that headers_remove items with embedded
1958 colons must have them doubled (or the list-separator changed).
1960 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1961 view the policy declared in the DMARC record. Currently, $dmarc_status
1962 is a combined value of both the record presence and the result of the
1965 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1967 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1968 dnslookup router and the smtp transport (applying to the forward
1971 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1972 of ldap servers used for a specific lookup. Patch provided by Heiko
1975 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1976 New variable $lookup_dnssec_authenticated for observability.
1978 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1979 Patch submitted by Lars Timman.
1981 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1983 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1984 Requires trusted mode and valid format message id, aborts otherwise.
1985 Patch contributed by Heiko Schlichting.
1987 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1988 certextract with support for various fields. Bug 1358.
1990 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1991 is requested by default, modifiable by smtp transport option
1994 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1995 operate on certificate variables to give certificate fingerprints
1996 Also new ${sha256:cert_variable}.
1998 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2000 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2001 Christian Aistleitner.
2003 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2005 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2006 file. Patch from Wolfgang Breyha.
2008 JH/25 Expand the coverage of the delivery $host and $host_address to
2009 client authenticators run in verify callout. Bug 1476.
2011 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2012 align with daemon_smtp_ports. Bug 72.
2014 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2015 support and error reporting did not work properly.
2017 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2018 and is readable. Patch from Andrew Colin Kissa.
2020 TL/14 Enhance documentation of ${run expansion and how it parses the
2021 commandline after expansion, particularly in the case when an
2022 unquoted variable expansion results in an empty value.
2024 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2026 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2027 by renaming away from it. Addresses GH issue 15, reported by
2030 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2032 TL/15 SECURITY: prevent double expansion in math comparison functions
2033 (can expand unsanitized data). Not remotely exploitable.
2040 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2042 PP/02 Make -n do something, by making it not do something.
2043 When combined with -bP, the name of an option is not output.
2045 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2048 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2049 $sender_host_name and config options to manage this, and basic check
2052 PP/05 DSCP support for outbound connections and control modifier for inbound.
2054 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2055 (Only plugin which currently uses this is kerberos4, which nobody should
2056 be using, but we should make it available and other future plugins might
2057 conceivably use it, even though it would break NAT; stuff *should* be
2058 using channel bindings instead).
2060 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2061 name; added for Sendmail compatibility; requires admin caller.
2062 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2063 just ignore it); requires trusted caller.
2064 Also parse but ignore: -Ac -Am -X<logfile>
2067 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2069 TL/02 Add +smtp_confirmation as a default logging option.
2071 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2072 Patch by Magnus Holmgren from 2007-02-20.
2074 TL/04 Bugzilla 1281 - Spec typo.
2075 Bugzilla 1283 - Spec typo.
2076 Bugzilla 1290 - Spec grammar fixes.
2078 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2080 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2082 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2083 mailing list by by Dmitry Isaikin.
2085 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2087 JH/02 Support "G" suffix to numbers in ${if comparisons.
2089 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2091 NM/01 Bugzilla 1197 - Spec typo
2092 Bugzilla 1196 - Spec examples corrections
2094 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2096 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2097 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2100 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2101 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2103 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2104 "acl {{name}{arg}...}", and optional args on acl condition
2107 JH/05 Permit multiple router/transport headers_add/remove lines.
2109 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2111 JH/07 Avoid using a waiting database for a single-message-only transport.
2112 Performance patch from Paul Fisher. Bugzilla 1262.
2114 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2117 JH/09 Add $headers_added variable, with content from use of ACL modifier
2118 add_header (but not yet added to the message). Bugzilla 199.
2120 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2121 Pulled from Bugzilla 817 by Wolfgang Breyha.
2123 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2125 (nb: this is the same fix as in Exim 4.80.1)
2127 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2130 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2133 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2135 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2136 advertise SMTP AUTH mechanism to us, instead of a generic
2137 protocol violation error. Also, make Exim more robust to bad
2138 data from the Dovecot auth socket.
2140 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2142 When a queue runner is handling a message, Exim first routes the
2143 recipient addresses, during which it prunes them based on the retry
2144 hints database. After that it attempts to deliver the message to
2145 any remaining recipients. It then updates the hints database using
2148 So if a recipient address works intermittently, it can get repeatedly
2149 deferred at routing time. The retry hints record remains fresh so the
2150 address never reaches the final cutoff time.
2152 This is a fairly common occurrence when a user is bumping up against
2153 their storage quota. Exim had some logic in its local delivery code
2154 to deal with this. However it did not apply to per-recipient defers
2155 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2157 This change adds a proper retry rule check during routing so that the
2158 final cutoff time is checked against the message's age. We only do
2159 this check if there is an address retry record and there is not a
2160 domain retry record; this implies that previous attempts to handle
2161 the address had the retry_use_local_parts option turned on. We use
2162 this as an approximation for the destination being like a local
2163 delivery, as in LMTP.
2165 I suspect this new check makes the old local delivery cutoff check
2166 redundant, but I have not verified this so I left the code in place.
2168 TF/02 Correct gecos expansion when From: is a prefix of the username.
2170 Test 0254 submits a message to Exim with the header
2174 When I ran the test suite under the user fanf2, Exim expanded
2175 the header to contain my full name, whereas it should have added
2176 a Resent-Sender: header. It erroneously treats any prefix of the
2177 username as equal to the username.
2179 This change corrects that bug.
2181 GF/01 DCC debug and logging tidyup
2182 Error conditions log to paniclog rather than rejectlog.
2183 Debug lines prefixed by "DCC: " to remove any ambiguity.
2185 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2187 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2188 Bug spotted by Jeremy Harris; was flawed since initial commit.
2189 Would have resulted in OCSP responses post-SNI triggering an Exim
2190 NULL dereference and crash.
2192 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2194 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2195 Bug detection, analysis and fix by Samuel Thibault.
2196 Bugzilla 1331, Debian bug #698092.
2198 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2200 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2201 Server implementation by Todd Lyons, client by JH.
2202 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2203 config variable "prdr_enable" controls whether the server
2204 advertises the facility. If the client requests PRDR a new
2205 acl_data_smtp_prdr ACL is called once for each recipient, after
2206 the body content is received and before the acl_smtp_data ACL.
2207 The client is controlled by both of: a hosts_try_prdr option
2208 on the smtp transport, and the server advertisement.
2209 Default client logging of deliveries and rejections involving
2210 PRDR are flagged with the string "PRDR".
2212 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2213 fclose(). Diagnosis by Todd Lyons.
2215 PP/17 Update configure.default to handle IPv6 localhost better.
2216 Patch by Alain Williams (plus minor tweaks).
2219 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2220 This is now consistent with GnuTLS, and is now documented: the
2221 previous undocumented portable approach to treating the option as
2222 unset was to force an expansion failure. That still works, and
2223 an empty string is now equivalent.
2225 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2226 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2227 not performing validation itself.
2229 PP/20 Added force_command boolean option to pipe transport.
2230 Patch from Nick Koston, of cPanel Inc.
2232 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2235 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2237 PP/21 Fix eximon continuous updating with timestamped log-files.
2238 Broken in a format-string cleanup in 4.80, missed when I repaired the
2239 other false fix of the same issue.
2240 Report and fix from Heiko Schlichting.
2243 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2244 Report from Prashanth Katuri.
2246 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2247 It's SecureTransport, so affects any MacOS clients which use the
2248 system-integrated TLS libraries, including email clients.
2250 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2251 using a MIME ACL for non-SMTP local injection.
2252 Report and assistance in diagnosis by Warren Baker.
2254 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2256 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2258 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2259 last id that failed so it may be referenced in subsequent ACL's.
2261 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2264 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2265 ldap library initialization, allowing self-signed CA's to be
2266 used. Also properly sets require_cert option later in code by
2267 using NULL (global ldap config) instead of ldap handle (per
2268 session). Bug diagnosis and testing by alxgomz.
2270 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2271 the src/util/ subdirectory.
2273 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2274 renamed to Transport Post Delivery Action by Jeremy Harris, as
2277 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2278 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2279 redis_servers = needs to be configured which will be used by the redis
2280 lookup. Patch from Warren Baker, of The Packet Hub.
2282 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2284 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2285 hostname or reverse DNS when processing a host list. Used suggestions
2286 from multiple comments on this bug.
2288 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2290 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2291 Missed a few lines, added it to make the runtest require no keyboard
2294 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2295 contains upper case chars. Make router use caseful_local_part.
2297 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2298 support when GnuTLS has been built with p11-kit.
2304 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2306 This, or similar/improved, will also be change PP/11 of 4.82.
2312 PP/01 Handle short writes when writing local log-files.
2313 In practice, only affects FreeBSD (8 onwards).
2314 Bugzilla 1053, with thanks to Dmitry Isaikin.
2316 NM/01 Bugzilla 949 - Documentation tweak
2318 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2321 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2323 PP/02 Implemented gsasl authenticator.
2325 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2327 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2328 `pkg-config foo` for cflags/libs.
2330 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2331 with rest of GSASL and with heimdal_gssapi.
2333 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2334 `pkg-config foo` for cflags/libs for the TLS implementation.
2336 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2337 properties get this fed in as external SSF. A number of robustness
2338 and debugging improvements to the cyrus_sasl authenticator.
2340 PP/08 cyrus_sasl server now expands the server_realm option.
2342 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2343 Patch by Jeremy Harris.
2345 PP/10 Added dbmjz lookup type.
2347 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2349 PP/12 MAIL args handles TAB as well as SP, for better interop with
2350 non-compliant senders.
2351 Analysis and variant patch by Todd Lyons.
2353 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2354 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2355 Patch from Dirk Mueller <dmueller@suse.com>
2357 PP/13 tls_peerdn now print-escaped for spool files.
2358 Observed some $tls_peerdn in wild which contained \n, which resulted
2359 in spool file corruption.
2361 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2362 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2363 or write after TLS renegotiation, which otherwise led to messages
2366 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2367 as a tracking header (ie: a signed header comes before the signature).
2368 Patch from Wolfgang Breyha.
2370 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2371 comma-sep list; embedded commas doubled.
2373 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2375 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2377 Report and patch from Dmitry Banschikov.
2379 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2380 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2381 failures. We appear to now support TLS1.1+ with Exim.
2383 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2384 lets Exim select keys and certificates based upon TLS SNI from client.
2385 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2386 before an outbound SMTP session. New log_selector, +tls_sni.
2388 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2389 NULL dereference. Report and patch from Alun Jones.
2391 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2392 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2393 Not seeing resolver debug output on NetBSD, but suspect this is a
2394 resolver implementation change.
2396 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2397 Left warnings. Added "eximon gdb" invocation mode.
2399 PP/21 Defaulting "accept_8bitmime" to true, not false.
2401 PP/22 Added -bw for inetd wait mode support.
2403 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2404 locate the relevant includes and libraries. Made this the default.
2406 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2407 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2409 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2410 This may cause build issues on older platforms.
2412 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2413 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2414 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2415 Added SNI support via GnuTLS too.
2416 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2418 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2420 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2421 Applied second patch from Janne, implementing suggestion to default
2422 multiple-strings-in-record handling to match SPF spec.
2424 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2426 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2427 read-only, out of scope).
2428 Patch from Wolfgang Breyha, report from Stuart Northfield.
2430 PP/29 Fix three issues highlighted by clang analyser static analysis.
2431 Only crash-plausible issue would require the Cambridge-specific
2432 iplookup router and a misconfiguration.
2433 Report from Marcin Mirosław.
2435 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2437 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2438 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2439 As part of this, removing so much warning spew let me fix some minor
2440 real issues in debug logging.
2442 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2443 assignment on my part. Fixed.
2445 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2446 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2447 Janne Snabb (who went above and beyond: thank you).
2449 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2450 string otherwise requires a connection and a bunch more work and it's
2451 relatively easy to get wrong. Should also expose TLS library linkage
2454 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2455 64-bit ${eval} (JH/03).
2457 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2458 GNU libc to support some of the 64-bit stuff, should not lead to
2459 conflicts. Defined before os.h is pulled in, so if a given platform
2460 needs to override this, it can.
2462 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2463 protection layer was required, which is not implemented.
2464 Bugzilla 1254, patch from Wolfgang Breyha.
2466 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2467 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2468 tls_dhparam take prime identifiers. Also unbreak combination of
2469 OpenSSL+DH_params+TLSSNI.
2471 PP/39 Disable SSLv2 by default in OpenSSL support.
2477 PP/01 Solaris build fix for Oracle's LDAP libraries.
2478 Bugzilla 1109, patch from Stephen Usher.
2480 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2482 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2485 TF/02 Fix a couple more cases where we did not log the error message
2486 when unlink() failed. See also change 4.74-TF/03.
2488 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2489 lock up or crash if it happened to be inside a call to libc when it
2490 got a SIGUSR1 from exiwhat.
2492 The SIGUSR1 handler appends the current process status to the process
2493 log which is later printed by exiwhat. It used to use the general
2494 purpose logging code to do this, but several functions it calls are
2495 not safe for signals.
2497 The new output code in the SIGUSR1 handler is specific to the process
2498 log, and simple enough that it's easy to inspect for signal safety.
2499 Removing some special cases also simplifies the general logging code.
2500 Removing the spurious timestamps from the process log simplifies
2503 TF/04 Improved ratelimit ACL condition.
2505 The /noupdate option has been deprecated in favour of /readonly which
2506 has clearer semantics. The /leaky, /strict, and /readonly update modes
2507 are mutually exclusive. The update mode is no longer included in the
2508 database key; it just determines when the database is updated. (This
2509 means that when you upgrade Exim will forget old rate measurements.)
2511 Exim now checks that the per_* options are used with an update mode that
2512 makes sense for the current ACL. For example, when Exim is processing a
2513 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2514 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2515 must specify per_mail/readonly. If you omit the update mode it defaults to
2516 /leaky where that makes sense (as before) or /readonly where required.
2518 The /noupdate option is now undocumented but still supported for
2519 backwards compatibility. It is equivalent to /readonly except that in
2520 ACLs where /readonly is required you may specify /leaky/noupdate or
2521 /strict/noupdate which are treated the same as /readonly.
2523 A useful new feature is the /count= option. This is a generalization
2524 of the per_byte option, so that you can measure the throughput of other
2525 aggregate values. For example, the per_byte option is now equivalent
2526 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2528 The per_rcpt option has been generalized using the /count= mechanism
2529 (though it's more complicated than the per_byte equivalence). When it is
2530 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2531 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2532 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2533 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2534 Note that using per_rcpt with a non-readonly update mode in more than
2535 one ACL will cause the recipients to be double-counted. (The per_mail
2536 and per_byte options don't have this problem.)
2538 The handling of very low rates has changed slightly. If the computed rate
2539 is less than the event's count (usually one) then this event is the first
2540 after a long gap. In this case the rate is set to the same as this event's
2541 count, so that the first message of a spam run is counted properly.
2543 The major new feature is a mechanism for counting the rate of unique
2544 events. The new per_addr option counts the number of different
2545 recipients that someone has sent messages to in the last time period. It
2546 behaves like per_rcpt if all the recipient addresses are different, but
2547 duplicate recipient addresses do not increase the measured rate. Like
2548 the /count= option this is a general mechanism, so the per_addr option
2549 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2550 example, measure the rate that a client uses different sender addresses
2551 with the options per_mail/unique=$sender_address. There are further
2552 details in the main documentation.
2554 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2556 TF/06 Removed a few PCRE remnants.
2558 TF/07 Automatically extract Exim's version number from tags in the git
2559 repository when doing development or release builds.
2561 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2562 Bugzilla 879. Patch from Paul Fisher.
2564 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2565 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2568 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2570 PP/05 Variable $av_failed, true if the AV scanner deferred.
2571 Bugzilla 1078. Patch from John Horne.
2573 PP/06 Stop make process more reliably on build failure.
2574 Bugzilla 1087. Patch from Heiko Schlittermann.
2576 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2577 Bugzilla 1089. Patch from Heiko Schlittermann.
2579 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2580 Bugzilla 1131. Patch from Holger Weiß.
2582 PP/09 Handle IPv6 addresses with SPF.
2583 Bugzilla 860. Patch from Wolfgang Breyha.
2585 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2587 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2590 PP/11 match_* no longer expand right-hand-side by default.
2591 New compile-time build option, EXPAND_LISTMATCH_RHS.
2592 New expansion conditions, "inlist", "inlisti".
2594 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2596 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2598 PP/14 fix log_write() format string regression from TF/03.
2599 Bugzilla 1152. Patch from Dmitry Isaikin.
2605 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2607 PP/02 Harmonised TLS library version reporting; only show if debugging.
2608 Layout now matches that introduced for other libraries in 4.74 PP/03.
2610 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2612 PP/04 New "dns_use_edns0" global option.
2614 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2617 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2618 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2620 TK/01 Updated PolarSSL code to 0.14.2.
2621 Bugzilla 1097. Patch from Andreas Metzler.
2623 PP/07 Catch divide-by-zero in ${eval:...}.
2624 Fixes bugzilla 1102.
2626 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2629 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2630 format-string attack -- SECURITY: remote arbitrary code execution.
2632 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2633 time unintentionally subject to list matching rules, letting the header
2634 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2635 arbitrary string expansion). This allowed for information disclosure.
2637 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2638 INT_MIN/-1 -- value coerced to INT_MAX.
2644 NM/01 Workaround for PCRE version dependency in version reporting
2647 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2648 This fixes portability to compilers other than gcc, notably
2649 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2651 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2652 makefiles for portability to HP-UX and POSIX correctness.
2654 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2655 Also via indented variable definition in the Makefile.
2656 (Debugging by Oliver Heesakkers).
2658 PP/02 Restore caching of spamd results with expanded spamd_address.
2659 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2661 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2662 Improves build reliability. Fix from: Frank Elsner
2664 NM/02 Fix wide character breakage in the rfc2047 coding
2665 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2667 NM/03 Allow underscore in dnslist lookups
2668 Fixes bug 1026. Patch from Graeme Fowler
2670 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2671 Code patches from Adam Ciarcinski of NetBSD.
2673 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2676 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2677 is logged, to avoid truncation. Patch from John Horne.
2679 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2680 Patch from Jakob Hirsch.
2682 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2683 SQL string expansion failure details.
2684 Patch from Andrey Oktyabrski.
2686 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2687 Patch from Simon Arlott.
2689 PP/09 New lookups functionality failed to compile on old gcc which rejects
2690 extern declarations in function scope.
2691 Patch from Oliver Fleischmann
2693 PP/10 Use sig_atomic_t for flags set from signal handlers.
2694 Check getgroups() return and improve debugging.
2695 Fixed developed for diagnosis in bug 927 (which turned out to be
2698 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2699 Patch from Mark Zealey.
2701 PP/12 Bugzilla 1056: Improved spamd server selection.
2702 Patch from Mark Zealey.
2704 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2705 Based on patch from Heiko Schlittermann.
2707 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2708 Patch from Uwe Doering, sign-off by Michael Haardt.
2710 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2711 variable declaration deep within a block. Bug and patch from
2714 PP/15 lookups-Makefile IRIX compatibility coercion.
2716 PP/16 Make DISABLE_DKIM build knob functional.
2718 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2719 Patch by Simon Arlott
2721 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2722 variable argument macros. Our copy now differs from upstream.
2728 TF/01 Failure to get a lock on a hints database can have serious
2729 consequences so log it to the panic log.
2731 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2732 controlled using the smtp_confirmation log selector.
2734 TF/03 Include the error message when we fail to unlink a spool file.
2736 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2737 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2738 for maintaining out-of-tree patches for some time.
2740 PP/01 Bugzilla 139: Documentation and portability issues.
2741 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2742 Handle per-OS dynamic-module compilation flags.
2744 PP/02 Let /dev/null have normal permissions.
2745 The 4.73 fixes were a little too stringent and complained about the
2746 permissions on /dev/null. Exempt it from some checks.
2747 Reported by Andreas M. Kirchwitz.
2749 PP/03 Report version information for many libraries, including
2750 Exim version information for dynamically loaded libraries. Created
2751 version.h, now support a version extension string for distributors
2752 who patch heavily. Dynamic module ABI change.
2754 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2755 privilege escalation vulnerability whereby the Exim run-time user
2756 can cause root to append content of the attacker's choosing to
2759 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2762 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2763 If dropping privileges for untrusted macros, we disabled normal logging
2764 on the basis that it would fail; for the Exim run-time user, this is not
2765 the case, and it resulted in successful deliveries going unlogged.
2766 Fixed. Reported by Andreas Metzler.
2772 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2773 only prepend for the Resent-* case. Fixes regression introduced in
2774 Exim 4.70 by NM/22 for Bugzilla 607.
2776 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2777 increasing numbers of administrators be bitten by this.
2779 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2781 PP/03 Bugzilla 994: added openssl_options main configuration option.
2783 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2785 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2787 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2789 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2790 without a peer certificate, leading to a segfault because of an
2791 assumption that peers always have certificates. Be a little more
2792 paranoid. Problem reported by Martin Tscholak.
2794 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2795 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2796 NB: ClamAV planning to remove STREAM in "middle of 2010".
2797 CL also introduces -bmalware, various -d+acl logging additions and
2798 more caution in buffer sizes.
2800 PP/09 Implemented reverse_ip expansion operator.
2802 PP/10 Bugzilla 937: provide a "debug" ACL control.
2804 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2806 PP/12 Bugzilla 973: Implement --version.
2808 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2810 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2812 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2814 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2815 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2816 ignore trailing whitespace.
2818 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2820 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2823 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2824 Notification from Dr Andrew Aitchison.
2826 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2827 ExtendedDetectionInfo response format.
2828 Notification from John Horne.
2830 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2833 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2834 XSL and documented dependency on system catalogs, with examples of how
2837 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2840 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2841 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2842 configuration file which is writeable by the Exim user or group.
2844 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2845 of configuration files to cover files specified with the -C option if
2846 they are going to be used with root privileges, not just the default
2849 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2850 option (effectively making it always true).
2852 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2853 files to be used while preserving root privileges.
2855 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2856 that rogue child processes cannot use them.
2858 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2859 run-time user, instead of root.
2861 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2862 Exim run-time user without dropping privileges.
2864 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2865 result string, instead of calling string_vformat() twice with the same
2868 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2869 for other users. Others should always drop root privileges if they use
2870 -C on the command line, even for a whitelisted configure file.
2872 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2874 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2880 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2881 $data_path, and $header_path variables; fixed documentation bugs and
2884 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2885 exipick to access non-standard spools, including the "frozen" queue
2888 NM/01 Bugzilla 965: Support mysql stored procedures.
2889 Patch from Alain Williams
2891 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2893 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2894 Patch from Andreas Metzler
2896 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2897 Patch from Kirill Miazine
2899 NM/05 Bugzilla 671: Added umask to procmail example.
2901 JJ/03 installed exipick 20100323.0, fixing doc bug
2903 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2904 directory. Notification and patch from Dan Rosenberg.
2906 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2908 TK/02 Improve log output when DKIM signing operation fails.
2910 MH/01 Treat the transport option dkim_domain as a colon separated
2911 list, not as a single string, and sign the message with each element,
2912 omitting multiple occurences of the same signer.
2914 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2915 Bugzilla 985, 986. Patch by Simon Arlott
2917 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2918 Patch by Simon Arlott
2920 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2921 MBX locking. Notification from Dan Rosenberg.
2927 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2929 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2931 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2933 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2935 NM/04 Bugzilla 915: Flag broken perl installation during build.
2941 TK/01 Added patch by Johannes Berg that expands the main option
2942 "spamd_address" if it starts with a dollar sign.
2944 TK/02 Write list of recipients to X-Envelope-Sender header when building
2945 the mbox-format spool file for content scanning (suggested by Jakob
2948 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2949 (http://www.dcc-servers.net/) support via dccifd. Activated by
2950 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2952 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2953 by Mark Daniel Reidel <mr@df.eu>.
2955 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2956 When building exim an external PCRE library is now needed -
2957 PCRE is a system library on the majority of modern systems.
2958 See entry on PCRE_LIBS in EDITME file.
2960 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2961 conversation. Added nologin parameter to request.
2962 Patch contributed by Kirill Miazine.
2964 TF/01 Do not log submission mode rewrites if they do not change the address.
2966 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2968 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2969 log files in place. Contributed by Roberto Lima.
2971 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2973 TF/03 Bugzilla 615: When checking the local_parts router precondition
2974 after a local_part_suffix or local_part_prefix option, Exim now
2975 does not use the address's named list lookup cache, since this
2976 contains cached lookups for the whole local part.
2978 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2979 Robert Millan. Documentation is in experimental-spec.txt.
2981 TF/04 Bugzilla 668: Fix parallel build (make -j).
2983 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2985 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2986 Patch provided by Jan Srzednicki.
2988 TF/05 Leading white space used to be stripped from $spam_report which
2989 wrecked the formatting. Now it is preserved.
2991 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2992 that they are available at delivery time.
2994 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2996 TF/08 TLS error reporting now respects the incoming_interface and
2997 incoming_port log selectors.
2999 TF/09 Produce a more useful error message if an SMTP transport's hosts
3000 setting expands to an empty string.
3002 NM/06 Bugzilla 744: EXPN did not work under TLS.
3003 Patch provided by Phil Pennock.
3005 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3006 Patch provided by Richard Godbee.
3008 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3009 acl_smtp_notquit, added index entry.
3011 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3012 Patch provided by Eugene Bujak.
3014 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3015 accept(). Patch provided by Maxim Dounin.
3017 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3018 Patch provided by Phil Pennock.
3020 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3022 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3023 Patch provided by Brad "anomie" Jorsch.
3025 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3026 Patch provided by Dean Brooks.
3028 TK/05 Add native DKIM support (does not depend on external libraries).
3030 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3031 Patch provided by Graeme Fowler.
3033 NM/16 Bugzilla 851: Documentation example syntax fix.
3035 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3037 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3040 NM/19 Bugzilla 745: TLS version reporting.
3041 Patch provided by Phil Pennock.
3043 NM/20 Bugzilla 167: bool: condition support.
3044 Patch provided by Phil Pennock.
3046 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3047 clients. Patch provided by Phil Pennock.
3049 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3050 Patch provided by Brad "anomie" Jorsch.
3052 NM/23 Bugzilla 687: Fix misparses in eximstats.
3053 Patch provided by Heiko Schlittermann.
3055 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3056 Patch provided by Heiko Schlittermann.
3058 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3059 plus update to original patch.
3061 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3063 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3064 Patch provided by David Brownlee.
3066 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3068 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3070 NM/30 Bugzilla 888: TLS documentation bugfixes.
3072 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3074 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3075 Unlike the original bugzilla I have changed all shell scripts in src tree.
3077 NM/33 Bugzilla 898: Transport filter timeout fix.
3078 Patch by Todd Rinaldo.
3080 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3081 Patch by Serge Demonchaux.
3083 NM/35 Bugzilla 39: Base64 decode bug fixes.
3084 Patch by Jakob Hirsch.
3086 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3088 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3090 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3092 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3098 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3099 ALT-N's libdkim that I have put here:
3100 http://duncanthrax.net/exim-experimental/
3102 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3103 were called 'true' and it seems that C99 defines that as a reserved
3104 keyword to be used with 'bool' variable types. That means you could
3105 not include C99-style headers which use bools without triggering
3106 build errors in sieve.c.
3108 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3109 as mailq or other aliases. Changed the --help handling significantly
3110 to do whats expected. exim_usage() emits usage/help information.
3112 SC/01 Added the -bylocaldomain option to eximstats.
3114 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3116 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3118 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3124 PH/01 Another patch from the Sieve maintainer.
3126 PH/02 When an IPv6 address is converted to a string for single-key lookup
3127 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3128 dots are used instead of colons so that keys in lsearch files need not
3129 contain colons. This was done some time before quoting was made available
3130 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3131 (notated using the quote facility) so as to distinguish them from IPv4
3132 keys. This meant that lookups for IP addresses in host lists did not work
3133 for iplsearch lookups.
3135 This has been fixed by arranging for IPv6 addresses to be expressed with
3136 colons if the lookup type is iplsearch. This is not incompatible, because
3137 previously such lookups could never work.
3139 The situation is now rather anomalous, since one *can* have colons in
3140 ordinary lsearch keys. However, making the change in all cases is
3141 incompatible and would probably break a number of configurations.
3143 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3146 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3147 conversion specification without a maximum field width, thereby enabling
3148 a rogue spamd server to cause a buffer overflow. While nobody in their
3149 right mind would setup Exim to query an untrusted spamd server, an
3150 attacker that gains access to a server running spamd could potentially
3151 exploit this vulnerability to run arbitrary code as the Exim user.
3153 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3154 $primary_hostname instead of what libspf2 thinks the hosts name is.
3156 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3157 a directory entry by the name of the lookup key. Previously, if a
3158 symlink pointed to a non-existing file or a file in a directory that
3159 Exim lacked permissions to read, a lookup for a key matching that
3160 symlink would fail. Now it is enough that a matching directory entry
3161 exists, symlink or not. (Bugzilla 503.)
3163 PH/03 The body_linecount and body_zerocount variables are now exported in the
3166 PH/04 Added the $dnslist_matched variable.
3168 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3169 This means they are set thereafter only if the connection becomes
3172 PH/06 Added the client_condition to authenticators so that some can be skipped
3173 by clients under certain conditions.
3175 PH/07 The error message for a badly-placed control=no_multiline_responses left
3176 "_responses" off the end of the name.
3178 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3180 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3181 (without spaces) instead of just copying the configuration text.
3183 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3185 PH/11 Added $max_received_linelength.
3187 PH/12 Added +ignore_defer and +include_defer to host lists.
3189 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3190 way in which PCRE > 7.0 is built.
3192 PH/14 Implemented queue_only_load_latch.
3194 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3195 MAIL command. The effect was to mangle the value on 64-bit systems.
3197 PH/16 Another patch from the Sieve maintainer.
3199 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3201 PH/18 If a system quota error occurred while trying to create the file for
3202 a maildir delivery, the message "Mailbox is full" was not appended to the
3203 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3204 only to a quota excession during the actual writing of the file.
3206 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3207 characters?) which causes problems in log lines. The DN values are now
3208 passed through string_printing() before being added to log lines.
3210 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3211 and InterBase are left for another time.)
3213 PH/21 Added message_body_newlines option.
3215 PH/22 Guard against possible overflow in moan_check_errorcopy().
3217 PH/23 POSIX allows open() to be a macro; guard against that.
3219 PH/24 If the recipient of an error message contained an @ in the local part
3220 (suitably quoted, of course), incorrect values were put in $domain and
3221 $local_part during the evaluation of errors_copy.
3227 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3228 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3231 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3232 issue a MAIL command.
3234 PH/02 In an ACL statement such as
3236 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3238 if a client was not listed at all, or was listed with a value other than
3239 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3240 the condition was not true (as it should be), so access was not denied.
3241 The bug was that the ! inversion was incorrectly passed on to the second
3242 item. This has been fixed.
3244 PH/03 Added additional dnslists conditions == and =& which are different from
3245 = and & when the dns lookup returns more than one IP address.
3247 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3248 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3250 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3251 FSYNC, which compiles an option called disable_fsync that allows for
3252 bypassing fsync(). The documentation is heavily laced with warnings.
3254 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3256 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3257 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3258 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3259 including adding "make clean"; (3) Added -fPIC when compiling the test
3260 dynamically loaded module, to get rid of a warning.
3262 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3263 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3264 = 0s. The bug is otherwise harmless.
3266 PH/07 There was a bug in the dovecot authenticator such that the value of
3267 $auth1 could be overwritten, and so not correctly preserved, after a
3268 successful authentication. This usually meant that the value preserved by
3269 the server_setid option was incorrect.
3271 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3273 PH/09 Installed PCRE release 7.0.
3275 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3276 run for batched SMTP input. It is now run at the start of every message
3277 in the batch. While fixing this I discovered that the process information
3278 (output by running exiwhat) was not always getting set for -bs and -bS
3279 input. This is fixed, and it now also says "batched" for BSMTP.
3281 PH/11 Added control=no_pipelining.
3283 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3284 patch, slightly modified), and move the expansion of helo_data till after
3285 the connection is made in the smtp transport (so it can use these
3288 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3290 PH/14 Added log_selector = +pid.
3292 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3294 PH/16 Add ${if forany and ${if forall.
3296 PH/17 Added dsn_from option to vary the From: line in DSNs.
3298 PH/18 Flush SMTP output before performing a callout, unless control =
3299 no_callout_flush is set.
3301 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3302 was true (the default) a successful delivery failed to delete the retry
3303 item, thus causing premature timeout of the address. The bug is now
3306 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3308 PH/21 Long custom messages for fakedefer and fakereject are now split up
3309 into multiline responses in the same way that messages for "deny" and
3310 other ACL rejections are.
3312 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3313 with slight modification.
3315 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3316 draft, changing the syntax and factoring some duplicate code".
3318 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3319 for deliveries of the second and subsequent messages over the same SMTP
3322 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3323 ${reduce, with only minor "tidies".
3325 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3327 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3328 expansion side effects.
3330 PH/27 When a message times out after an over-quota error from an Exim-imposed
3331 quota, the bounce message says "mailbox is full". This message was not
3332 being given when it was a system quota that was exceeded. It now should
3335 MH/03 Made $recipients available in local_scan(). local_scan() already has
3336 better access to the recipient list through recipients_list[], but
3337 $recipients can be useful in postmaster-provided expansion strings.
3339 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3340 in the case of a MAIL command with additional options following the
3341 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3342 were accidentally chopped off.
3344 PH/29 SMTP synchronization checks are implemented when a command is read -
3345 there is a check that no more input is waiting when there shouldn't be
3346 any. However, for some commands, a delay in an ACL can mean that it is
3347 some time before the response is written. In this time, more input might
3348 arrive, invalidly. So now there are extra checks after an ACL has run for
3349 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3350 pipelining has not been advertised.
3352 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3354 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3355 returned in the SMTP response when smtp_return_error_details was false.
3356 This has been fixed.
3358 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3359 instead of the C I/O that was originally supplied, because problems were
3360 reported on Solaris.
3362 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3363 Exim which did not show up earlier: it was assuming that a call to
3364 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3365 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3366 was a macro that became an assignment, so it seemed to work. This has
3367 changed to a proper function call with a void return, hence the compile
3368 error. Exim's code has been fixed.
3370 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3373 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3375 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3377 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3378 because it turns out to be too big or there is a timeout) there is no
3379 "Completed" line in the log. When some messages of this type were
3380 selected by exigrep, they were listed as "not completed". Others were
3381 picked up by some special patterns. I have improved the selection
3382 criteria to be more general.
3384 PH/38 The host_find_failed option in the manualroute router can now be set
3385 to "ignore", to completely ignore a host whose IP address cannot be
3386 found. If all hosts are ignored, the behaviour is controlled by the new
3387 host_all_ignored option.
3389 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3390 homing or because of multiple MX records with /mx) generated more than
3391 one IP address, and the following item turned out to be the local host,
3392 all the secondary addresses of the first item were incorrectly removed
3393 from the list, along with the local host and any following hosts (which
3394 is what is supposed to happen).
3396 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3397 whoever called Exim into the -H file. In the case of the daemon it was
3398 behaving confusingly. When first started, it used values for whoever
3399 started the daemon, but after a SIGHUP it used the Exim user (because it
3400 calls itself on a restart). I have changed the code so that it now always
3403 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3404 message are rejected with the same error (e.g. no authentication or bad
3405 sender address), and a DATA command is nevertheless sent (as can happen
3406 with PIPELINING or a stupid MUA), the error message that was given to the
3407 RCPT commands is included in the rejection of the DATA command. This is
3408 intended to be helpful for MUAs that show only the final error to their
3411 PH/42 Another patch from the Sieve maintainer.
3413 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3414 Eximstats - Fixed some broken HTML links and added missing column headers
3416 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3417 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3419 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3421 PH/43 Yet another patch from the Sieve maintainer.
3423 PH/44 I found a way to check for a TCP/IP connection going away before sending
3424 the response to the final '.' that terminates a message, but only in the
3425 case where the client has not sent further data following the '.'
3426 (unfortunately, this is allowed). However, in many cases there won't be
3427 any further data because there won't be any more messages to send. A call
3428 to select() can be used: if it shows that the input is "ready", there is
3429 either input waiting, or the socket has been closed. An attempt to read
3430 the next input character can distinguish the two cases. Previously, Exim
3431 would have sent an OK response which the client would never have see.
3432 This could lead to message repetition. This fix should cure that, at
3433 least in a lot of common cases.
3435 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3436 advertised in response to EHLO.
3442 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3443 fixed by 4.65/MH/01 (is this a record?) are fixed:
3445 (i) An empty string was always treated as zero by the numeric comparison
3446 operators. This behaviour has been restored.
3448 (ii) It is documented that the numeric comparison operators always treat
3449 their arguments as decimal numbers. This was broken in that numbers
3450 starting with 0 were being interpreted as octal.
3452 While fixing these problems I realized that there was another issue that
3453 hadn't been noticed. Values of message_size_limit (both the global option
3454 and the transport option) were treated as octal if they started with 0.
3455 The documentation was vague. These values are now always treated as
3456 decimal, and I will make that clear in the documentation.
3462 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3463 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3466 MH/01 Don't check that the operands of numeric comparison operators are
3467 integers when their expansion is in "skipping" mode (fixes bug
3468 introduced by 4.64-PH/07).
3470 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3471 child addresses, Exim now panics and dies. Previously, because the count
3472 is held in a short int, deliveries were likely to be lost. As such a
3473 large number of recipients for a single message is ridiculous
3474 (performance will be very, very poor), I have chosen to impose a limit
3475 rather than extend the field.
3481 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3482 leftover -K file (the existence of which was triggered by #402).
3483 While we were at it, introduced process PID as part of the -K
3484 filename. This should rule out race conditions when creating
3487 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3488 processing considerably. Previous code took too long for large mails,
3489 triggering a timeout which in turn triggers #401.
3491 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3492 in the DK code in transports.c. sendfile() is not really portable,
3493 hence the _LINUX specificness.
3495 TF/01 In the add_headers option to the mail command in an Exim filter,
3496 there was a bug that Exim would claim a syntax error in any
3497 header after the first one which had an odd number of characters
3500 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3501 callout verification, Exim cached a "reject" for the entire domain. This
3502 is correct for most verifications, but it is not correct for a recipient
3503 verification with use_sender or use_postmaster set, because in that case
3504 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3505 case of MAIL FROM:<> rejection from other early rejections (e.g.
3506 rejection of HELO). When verifying a recipient using a non-null MAIL
3507 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3508 Whatever the result of the callout, the value of the domain cache is
3509 left unchanged (for any other kind of callout, getting as far as trying
3510 RCPT means that the domain itself is ok).
3512 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3515 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3516 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3517 session, the daemon ignores ECONNECT errors and logs others; it now
3518 ignores EPIPE as well.
3520 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3521 (quoted-printable decoding).
3523 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3524 later the small subsequent patch to fix an introduced bug.
3526 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3528 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3530 PH/08 An error is now given if message_size_limit is specified negative.
3532 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3533 to be given (somewhat) arbitrary names.
3535 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3538 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3539 miscellaneous code fixes
3541 PH/10 Added the log_reject_target ACL modifier to specify where to log
3544 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3545 hostname. This is wrong, because it relates to the incoming message (and
3546 probably the interface on which it is arriving) and not to the outgoing
3547 callout (which could be using a different interface). This has been
3548 changed to use the value of the helo_data option from the smtp transport
3549 instead - this is what is used when a message is actually being sent. If
3550 there is no remote transport (possible with a router that sets up host
3551 addresses), $smtp_active_hostname is used.
3553 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3554 tweaks were necessary in order to get it to work (see also 21 below):
3555 (a) The code assumed that strncpy() returns a negative number on buffer
3556 overflow, which isn't the case. Replaced with Exim's string_format()
3558 (b) There were several signed/unsigned issues. I just did the minimum
3559 hacking in of casts. There is scope for a larger refactoring.
3560 (c) The code used strcasecmp() which is not a standard C function.
3561 Replaced with Exim's strcmpic() function.
3562 (d) The code set only $1; it now sets $auth1 as well.
3563 (e) A simple test gave the error "authentication client didn't specify
3564 service in request". It would seem that Dovecot has changed its
3565 interface. Fortunately there's a specification; I followed it and
3566 changed what the client sends and it appears to be working now.
3568 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3571 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3572 address A is aliased to B and C, where B exists and C does not. Without
3573 -v the output is "A verified" because verification stops after a
3574 successful redirection if more than one address is generated. However,
3575 with -v the child addresses are also verified. Exim was outputting "A
3576 failed to verify" and then showing the successful verification for C,
3577 with its parentage. It now outputs "B failed to verify", showing B's
3578 parentage before showing the successful verification of C.
3580 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3581 look up a TXT record in a specific list after matching in a combined
3584 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3585 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3586 they consult the DNS. I had assumed they would set it the way they
3587 wanted; and indeed my experiments on Linux seem to show that in some
3588 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3589 To be on the safe side, however, I have now made the interface to
3590 host_find_byname() similar to host_find_bydns(), with an argument
3591 containing the DNS resolver options. The host_find_byname() function now
3592 sets these options at its start, just as host_find_bydns() does. The smtp
3593 transport options dns_qualify_single and dns_search_parents are passed to
3594 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3595 of host_find_byname() use the default settings of RES_DEFNAMES
3596 (qualify_single) but not RES_DNSRCH (search_parents).
3598 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3599 spool_read_header() do less string testing, by means of a preliminary
3600 switch on the second character of optional "-foo" lines. (This is
3601 overdue, caused by the large number of possibilities that now exist.
3602 Originally there were few.) While I was there, I also converted the
3603 str(n)cmp tests so they don't re-test the leading "-" and the first
3604 character, in the hope this might squeeze out yet more improvement.
3606 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3607 flag allowing group syntax was set by the header_syntax check but not
3608 turned off, possible causing trouble later; (2) The flag was not being
3609 set at all for the header_verify test, causing "group"-style headers to
3610 be rejected. I have now set it in this case, and also caused header_
3611 verify to ignore an empty address taken from a group. While doing this, I
3612 came across some other cases where the code for allowing group syntax
3613 while scanning a header line wasn't quite right (mostly, not resetting
3614 the flag correctly in the right place). These bugs could have caused
3615 trouble for malformed header lines. I hope it is now all correct.
3617 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3618 with the "reply" argument non-NULL. The code, however (which originally
3619 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3620 but it didn't always do it. This confused somebody who was copying the
3621 code for some other use. I have removed all the tests.
3623 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3624 feature that was used to support insecure browsers during the U.S. crypto
3625 embargo. It requires special client support, and Exim is probably the
3626 only MTA that supported it -- and would never use it because real RSA is
3627 always available. This code has been removed, because it had the bad
3628 effect of slowing Exim down by computing (never used) parameters for the
3629 RSA_EXPORT functionality.
3631 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3632 authenticator to fail if there's a tab character in the incoming data
3633 (there should never be unless someone is messing about, as it's supposed
3634 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3635 if the connection is using TLS or if the remote IP is the same as the
3636 local IP, and the "valid-client-cert option" if a client certificate has
3639 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3640 authenticators. This can be used for authorization after authentication
3641 succeeds. (In the case of plaintext, it servers for both authentication
3644 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3645 if any retry times were supplied.
3647 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3648 connection, where there is no client IP address to check. In this
3649 situation, the verify now always succeeds.
3651 PH/25 Applied John Jetmore's -Mset patch.
3653 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3655 PH/27 In a string expansion for a processed (not raw) header when multiple
3656 headers of the same name were present, leading whitespace was being
3657 removed from all of them, but trailing whitespace was being removed only
3658 from the last one. Now trailing whitespace is removed from each header
3659 before concatenation. Completely empty headers in a concatenation (as
3660 before) are ignored.
3662 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3663 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3665 PH/29 [Removed. This was a change that I later backed out, and forgot to
3666 correct the ChangeLog entry (that I had efficiently created) before
3667 committing the later change.]
3669 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3670 address errors (4xx response to RCPT) over the same connection as other
3671 messages routed to the same hosts. Such deliveries are always "forced",
3672 so retry times are not inspected. This resulted in far too many retries
3673 for the affected addresses. The effect occurred only when there were more
3674 hosts than the hosts_max_try setting in the smtp transport when it had
3675 the 4xx errors. Those hosts that it had tried were not added to the list
3676 of hosts for which the message was waiting, so if all were tried, there
3677 was no problem. Two fixes have been applied:
3679 (i) If there are any address or message errors in an SMTP delivery, none
3680 of the hosts (tried or untried) are now added to the list of hosts
3681 for which the message is waiting, so the message should not be a
3682 candidate for sending over the same connection that was used for a
3683 successful delivery of some other message. This seems entirely
3684 reasonable: after all the message is NOT "waiting for some host".
3685 This is so "obvious" that I'm not sure why it wasn't done
3686 previously. Hope I haven't missed anything, but it can't do any
3687 harm, as the worst effect is to miss an optimization.
3689 (ii) If, despite (i), such a delivery is accidentally attempted, the
3690 routing retry time is respected, so at least it doesn't keep
3691 hammering the server.
3693 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3694 in ${readsocket because some servers need this prod.
3696 PH/32 Added some extra debug output when updating a wait-xxx database.
3698 PH/33 The hint "could be header name not terminated by colon", which has been
3699 given for certain expansion errors for a long time, was not being given
3700 for the ${if def:h_colon_omitted{... case.
3702 PH/34 The spec says: "With one important exception, whenever a domain list is
3703 being scanned, $domain contains the subject domain." There was at least
3704 one case where this was not true.
3706 PH/35 The error "getsockname() failed: connection reset by peer" was being
3707 written to the panic log as well as the main log, but it isn't really
3708 panic-worthy as it just means the connection died rather early on. I have
3709 removed the panic log writing for the ECONNRESET error when getsockname()
3712 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3713 runs only) independently of the message's sender address. This meant
3714 that, if the 4xx error was in fact related to the sender, a different
3715 message to the same recipient with a different sender could confuse
3716 things. In particular, this can happen when sending to a greylisting
3717 server, but other circumstances could also provoke similar problems.
3718 I have changed the default so that the retry time for these errors is now
3719 based a combination of the sender and recipient addresses. This change
3720 can be overridden by setting address_retry_include_sender=false in the
3723 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3724 remote server are returned as part of bounce messages. This was not
3725 happening for LMTP over a pipe (the lmtp transport), but now it is the
3726 same for both kinds of LMTP.
3728 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3729 in header lines that were in fact CNAMEs. This is no longer the case.
3731 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3732 and queue runs started by the daemon processed all messages. This has
3733 been fixed so that -R and -S can now usefully be given with -q<time>.
3735 PH/40 Import PCRE release 6.7 (fixes some bugs).
3737 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3739 PH/42 Give an error if -q is specified more than once.
3741 PH/43 Renamed the variables $interface_address and $interface_port as
3742 $received_ip_address and $received_port, to make it clear that these
3743 values apply to message reception, and not to the outgoing interface when
3744 a message is delivered. (The old names remain recognized, of course.)
3746 PH/44 There was no timeout on the connect() call when using a Unix domain
3747 socket in the ${readsocket expansion. There now is.
3749 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3750 be meaningful with "accept".
3752 SC/01 Eximstats V1.43
3753 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3755 SC/02 Eximstats V1.44
3756 Use a glob alias rather than an array ref in the generated
3757 parser. This improves both readability and performance.
3759 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3760 Collect SpamAssassin and rejection statistics.
3761 Don't display local sender or destination tables unless
3762 there is data to show.
3763 Added average volumes into the top table text output.
3765 SC/04 Eximstats V1.46
3766 Collect data on the number of addresses (recipients)
3767 as well as the number of messages.
3769 SC/05 Eximstats V1.47
3770 Added 'Message too big' to the list of mail rejection
3771 reasons (thanks to Marco Gaiarin).
3773 SC/06 Eximstats V1.48
3774 Mainlog lines which have GMT offsets and are too short to
3775 have a flag are now skipped.
3777 SC/07 Eximstats V1.49 (Alain Williams)
3778 Added the -emptyok flag.
3780 SC/08 Eximstats V1.50
3781 Fixes for obtaining the IP address from reject messages.
3783 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3784 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3785 whitespace changes from 4.64-PH/27
3787 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3790 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3791 are found, allow negative numbers in numeric criteria)
3793 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3795 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3796 to match changes made in 4.64-PH/43
3798 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3800 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3801 "domain" is actually an IP address, and force a failure. This locks out
3802 those revolvers/nameservers that support "A-for-A" lookups, in
3803 contravention of the specifications.
3805 PH/48 When a host name was looked up from an IP address, and the subsequent
3806 forward lookup of the name timed out, the host name was left in
3807 $sender_host_name, contrary to the specification.
3809 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3810 restricted to single-key lookups, Exim was not diagnosing an error if
3811 * or *@ was used with a query-style lookup.
3813 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3815 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3816 long ago, but noone interested enough thought of it. Let's just say that
3817 the "1.1" means that there are some new functions that weren't there at
3818 some point in the past.
3820 PH/51 Error processing for expansion failure of helo_data from an smtp
3821 transport during callout processing was broken.
3823 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3824 tested/used via the -bh/-bhc/-bs options.
3826 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3827 bug, fixed in subsequent PCRE releases).
3829 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3830 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3832 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3838 SC/01 Use a glob alias rather than an array ref in eximstats generated
3839 parser. This improves both readability and performance.
3841 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3842 Don't display local sender or destination tables in eximstats unless
3843 there is data to show.
3844 Added average volumes into the eximstats top table text output.
3846 SC/03 Collect data on the number of addresses (recipients) as well
3847 as the number of messages in eximstats.
3849 TF/01 Correct an error in the documentation for the redirect router. Exim
3850 does (usually) call initgroups() when daemonizing.
3852 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3853 with consistent privilege compared to when running as a daemon.
3855 TF/03 Note in the spec that $authenticated_id is not set for local
3856 submissions from trusted users.
3858 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3859 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3861 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3862 by adding some example configuration directives to the default
3863 configuration file. A little bit of work is required to uncomment the
3864 directives and define how usernames and passwords are checked, but
3865 there is now a framework to start from.
3867 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3868 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3869 without this. I don't know how relevant this is to other LDAP libraries.
3871 PH/02 Add the verb name to the "unknown ACL verb" error.
3873 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3875 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3877 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3878 directory not expanded when it should be if an expanded home directory
3879 was set for the address (which is overridden by the transport).
3881 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3884 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3885 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3886 because it is too late at that time, and has no effect.
3888 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3889 security issue with \' (bugzilla #107). I could not use the
3890 PQescapeStringConn() function, because it needs a PGconn value as one of
3893 PH/08 When testing addresses using -bt, indicate those final addresses that
3894 are duplicates that would not cause an additional delivery. At least one
3895 person was confused, thinking that -bt output corresponded to deliveries.
3896 (Suppressing duplicates isn't a good idea as you lose the information
3897 about possibly different redirections that led to the duplicates.)
3899 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3900 systems where poll() doesn't work, in particular OS X.
3902 PH/10 Added more information to debugging output for retry time not reached.
3904 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3905 operations in malware.c.
3907 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3910 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3911 syslog_duplication set false, log lines that would normally be written
3912 both the the main log and to the reject log were not written to syslog at
3915 PH/14 In the default configuration, change the use of "message" in ACL warn
3916 statements to "add_header".
3918 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3919 not followed by a command (e.g. "seen endif").
3921 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3922 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3925 PH/17 Added extra conditions to the default value of delay_warning_condition
3929 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3930 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3931 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3934 The Auto-Submitted: and various List- headers are standardised, whereas I
3935 don't think Precedence: ever was.
3937 PH/18 Refactored debugging code in route_finduser() to show more information,
3938 in particular, the error code if getpwnam() issues one.
3940 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3941 This is apparently needed in addition to the PH/07 change above to avoid
3942 any possible encoding problems.
3944 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3945 but not after initializing Perl.
3947 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3948 output them only if debugging. By default they are written stderr,
3949 apparently, which is not desirable.
3951 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3954 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3957 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3959 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3960 authenticated or an ident call has been made. Suppress the default
3961 values for $authenticated_id and $authenticated_sender (but permit -oMai
3962 and -oMas) when testing with -bh.
3964 PH/24 Re-jigged the order of the tests in the default configuration so that the
3965 tests for valid domains and recipients precede the DNS black list and CSA
3966 tests, on the grounds that those ones are more expensive.
3968 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3969 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3970 This bug exists in every version of Exim that I still have, right back to
3973 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3974 However, an attempt to turn on case-sensitivity in a regex key by
3975 including (?-i) didn't work because the subject string was already
3976 lowercased, and the effects were non-intuitive. It turns out that a
3977 one-line patch can be used to allow (?-i) to work as expected.
3983 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3984 other effects) broke the use of negated acl sub-conditions.
3986 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3989 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3990 "Deny" causes Exim to reject the incoming connection with a 554 error.
3991 Unfortunately, if there is a major crisis, such as a disk failure,
3992 tcp-wrappers gives "deny", whereas what one would like would be some
3993 kind of temporary error. A kludge has been added to help with this.
3994 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3995 554 error is used if errno is still zero or contains ENOENT (which occurs
3996 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3999 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4001 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4002 errors. Otherwise a message that provokes a temporary error (when other
4003 messages do not) can cause a whole host to time out.
4005 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4006 addresses were routed directly to files or pipes from a redirect router.
4007 File deliveries just didn't batch; pipe deliveries might have suffered
4010 PH/06 A failure to get a lock for a hints database would erroneously always say
4011 "Failed to get write lock", even when it was really a read lock.
4013 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4014 of 0600. This has been changed to use the value of the lockfile_mode
4015 option (which defaults to 0600).
4017 PH/08 Applied small patch from the Sieve maintainer.
4019 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4020 folder from quota calculations, a direct delivery into this folder messed
4021 up the contents of the maildirsize file. This was because the regex was
4022 used only to exclude .Trash (or whatever) when the size of the mailbox
4023 was calculated. There was no check that a delivery was happening into an
4024 excluded directory. This bug has been fixed by ignoring all quota
4025 processing for deliveries into excluded directories.
4027 PH/10 Added the maildirfolder_create_regex option to appendfile.
4033 PH/01 The code for finding all the local interface addresses on a FreeBSD
4034 system running IPv6 was broken. This may well have applied to all BSD
4035 systems, as well as to others that have similar system calls. The broken
4036 code found IPv4 interfaces correctly, but gave incorrect values for the
4037 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4038 that it would not match correctly against @[] and not recognize the IPv6
4041 PH/02 The ipliteral router was not recognizing addresses of the form user@
4042 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4044 PH/03 Added disable_ipv6.
4046 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4047 decoded form, because it is most often used to construct To: headers
4048 lines in autoreplies, and the decoded form may well be syntactically
4049 invalid. However, $reply_address has leading white space removed, and all
4050 newlines turned into spaces so that the autoreply transport does not
4053 PH/05 If group was specified without a user on a router, and no group or user
4054 was specified on a transport, the group from the router was ignored.
4056 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4057 for visible compile-time settings that can be used to change these
4058 numbers, for those that want even more. Backwards compatibility with old
4059 spool files has been maintained. However, going back to a previous Exim
4060 release will lost any variables that are in spool files.
4062 PH/07 Two small changes when running in the test harness: increase delay when
4063 passing a TCP/IP connection to a new process, in case the original
4064 process has to generate a bounce, and remove special handling of
4065 127.0.0.2 (sic), which is no longer necessary.
4067 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4068 be the same on different OS.
4070 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4073 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4074 whether --show-vars was specified or not
4076 JJ/02 exipick: Added support for new ACL variable spool format introduced
4079 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4080 syntactically invalid From: or Reply-to: line, and a filter used this to
4081 generate an autoreply, and therefore failed to obtain an address for the
4082 autoreply, Exim could try to deliver to a non-existent relative file
4083 name, causing unrelated and misleading errors. What now happens is that
4084 it logs this as a hard delivery error, but does not attempt to create a
4087 PH/11 The exinext utility has a -C option for testing purposes, but although
4088 the given file was scanned by exinext itself; it wasn't being passed on
4089 when Exim was called.
4091 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4092 an end-of-file indication when reading a command response.
4094 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4095 compiled. In many other places in Exim, IPv6 addresses are always
4096 recognized, so I have changed this. It also means that IPv4 domain
4097 literals of the form [IPV4:n.n.n.n] are now always recognized.
4099 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4100 used if the router is not running as root, for example, when verifying at
4101 ACL time, or when using -bh. The debugging output from this situation was
4102 non-existent - all you got was a failure to exec. I have made two
4105 (a) Failures to set uid/gid, the current directory, or a process leader
4106 in a subprocess such as that created by queryprogram now generate
4107 suitable debugging output when -d is set.
4109 (b) The queryprogram router detects when it is not running as root,
4110 outputs suitable debugging information if -d is set, and then runs
4111 the subprocess without attempting to change uid/gid.
4113 PH/15 Minor change to Makefile for building test_host (undocumented testing
4116 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4117 additional section of a DNS packet that returns MX or SRV records.
4118 Instead, it always explicitly searches for A/AAAA records. This avoids
4119 major problems that occur when a DNS server includes only records of one
4120 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4121 fixed another bug: if SRV records were looked up and the corresponding
4122 address records were *not* found in the additional section, the port
4123 values from the SRV records were lost.
4125 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4126 using the correct key (the original address) when searching the retry
4127 rules in order to find which one to use for generating the retry hint.
4129 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4130 adding the default one. Similarly, if it contains a Reply-To: header, the
4131 errors_reply_to option, if set, is not used.
4133 PH/19 When calculating a retry time, Exim used to measure the "time since
4134 failure" by looking at the "first failed" field in the retry record. Now
4135 it does not use this if it is later than than the arrival time of the
4136 message. Instead it uses the arrival time. This makes for better
4137 behaviour in cases where some deliveries succeed, thus re-setting the
4138 "first failed" field. An example is a quota failure for a huge message
4139 when small messages continue to be delivered. Without this change, the
4140 "time since failure" will always be short, possible causing more frequent
4141 delivery attempts for the huge message than are intended.
4142 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4144 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4145 $1, $2, $3) because the numerical variables can be reset during some
4146 expansion items (e.g. "match"), thereby losing the authentication data.
4148 PH/21 Make -bV show the size of off_t variables so that the test suite can
4149 decide whether to run tests for quotas > 2G.
4151 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4152 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4153 filecount value is greater than 2G or if a quota value is greater than 2G
4154 on a system where the size of off_t is not greater than 4, a panic error
4157 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4158 never match. The debug and -bh output now contains an explicit error
4159 message indicating a malformed IPv4 address or mask.
4161 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4162 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4163 PH/23 above applies.
4165 PH/25 Do not write to syslog when running in the test harness. The only
4166 occasion when this arises is a failure to open the main or panic logs
4167 (for which there is an explicit test).
4169 PH/26 Added the /no_tell option to "control=freeze".
4171 PH/27 If a host name lookup failed very early in a connection, for example, if
4172 the IP address matched host_lookup and the reverse lookup yielded a name
4173 that did not have a forward lookup, an error message of the form "no IP
4174 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4175 could be logged. Now it outputs the IP address instead of "NULL".
4177 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4178 allows the sender and the authenticated sender to be set when
4179 submitting a message from within Exim. Since child_open_exim() is
4180 documented for local_scan(), the new function should be too.
4182 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4183 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4184 results in an empty string is now treated as unset.
4186 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4188 PH/31 Added sender_verify_fail logging option.
4190 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4191 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4192 tidied the source and removed it altogether.
4194 PH/33 When a queue run was abandoned because the load average was too high, a
4195 log line was always written; now it is written only if the queue_run log
4196 selector is set. In addition, the log line for abandonment now contains
4197 information about the queue run such as the pid. This is always present
4198 in "start" and "stop" lines but was omitted from the "abandon" line.
4200 PH/34 Omit spaces between a header name and the colon in the error message that
4201 is given when verify = headers_syntax fails (if there are lots of them,
4202 the message gets confusing).
4204 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4205 names, as there are now some PTR records that contain slashes. This check
4206 is only to protect against broken name servers that fall over on strange
4207 characters, so the fact that it applies to all lookups doesn't matter.
4209 PH/36 Now that the new test suite is complete, we can remove some of the
4210 special code in Exim that was needed for the old test suite. For example,
4211 sorting DNS records because real resolvers return them in an arbitrary
4212 order. The new test suite's fake resolver always returns records in the
4215 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4216 bounces) except when queue_only is set, to avoid logging races between
4217 the different processes.
4219 PH/38 Panic-die if .include specifies a non-absolute path.
4221 PH/39 A tweak to the "H" retry rule from its user.
4223 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4224 a label. They prevented compilation on older perls.
4226 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4227 a warning to be raised on newish perls.
4229 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4230 on queue. Changes to match documented behaviour of showing count of
4231 messages matching specified criteria.
4233 PH/40 Changed the default ident timeout from 30s to 5s.
4235 PH/41 Added support for the use of login_cap features, on those BSD systems
4236 that have them, for controlling the resources used by pipe deliveries.
4238 PH/42 The content-scanning code uses fopen() to create files in which to put
4239 message data. Previously it was not paying any attention to the mode of
4240 the files. Exim runs with umask(0) because the rest of the code creates
4241 files with open(), and sets the required mode explicitly. Thus, these
4242 files were ending up world-writeable. This was not a big issue, because,
4243 being within the spool directory, they were not world-accessible. I have
4244 created a function called modefopen, which takes an additional mode
4245 argument. It sets umask(777), creates the file, chmods it to the required
4246 mode, then resets the umask. All the relevant calls to fopen() in the
4247 content scanning code have been changed to use this function.
4249 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4250 to 24 hours. This avoids potential overflow problems when processing G
4251 and H retry rules. I suspect nobody ever tinkers with this value.
4253 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4255 PH/45 When the plaintext authenticator is running as a client, the server's
4256 challenges are checked to ensure they are valid base64 strings. By
4257 default, the authentication attempt is cancelled if an invalid string is
4258 received. Setting client_ignore_invalid_base64 true ignores these errors.
4259 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4260 they are received. Thus, the responses can be made to depend on the
4261 challenges. If an invalid string is ignored, an empty string is placed in
4264 PH/46 Messages that are created by the autoreply transport now contains a
4265 References: header, in accordance with RFCs 2822 and 3834.
4267 PH/47 Added authenticated_sender_force to the smtp transport.
4269 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4271 PH/49 Installed latest patch from the Sieve maintainer.
4273 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4274 was also set, the appendfile transport was unnecessarily scanning a
4275 directory of message files (e.g. for maildir delivery) to find the count
4276 of files (along with the size), even though it did not need this
4277 information. It now does the scan only if it needs to find either the
4278 size of the count of files.
4280 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4282 PH/52 Two bugs concerned with error handling when the smtp transport is
4285 (i) Exim was not creating retry information for temporary errors given
4286 for individual recipients after the DATA command when the smtp transport
4287 was used in LMTP mode. This meant that they could be retried too
4288 frequently, and not timed out correctly.
4290 (ii) Exim was setting the flag that allows error details to be returned
4291 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4292 recipients that were returned after the DATA command.
4294 PH/53 This is related to PH/52, but is more general: for any failing address,
4295 when detailed error information was permitted to be returned to the
4296 sender, but the error was temporary, then after the final timeout, only
4297 "retry timeout exceeded" was returned. Now it returns the full error as
4298 well as "retry timeout exceeded".
4300 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4301 do this, and (what is worse) MTAs that accept it.
4303 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4304 will now be deprecated.
4306 PH/56 New os.c-cygwin from the Cygwin maintainer.
4308 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4309 formats (previously only available in exim formats via -bpr, -bpru,
4310 and -bpra. Now also available in native and exiqgrep formats)
4312 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4313 with very large, slow to parse queues
4315 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4317 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4319 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4320 responses to authentication challenges, though it was showing the
4321 challenges; (ii) I've removed the CR characters from the debug output for
4324 PH/58 Allow for the insertion of a newline as well as a space when a string
4325 is turned into more than one encoded-word during RFC 2047 encoding. The
4326 Sieve code now uses this.
4328 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4329 data_4xx, lost_connection, tls_required.
4331 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4332 message was being sent as an SMTP response.
4334 PH/61 Add -l and -k options to exicyclog.
4336 PH/62 When verifying, if an address was redirected to one new address, so that
4337 verification continued, and the new address failed or deferred after
4338 having set something in $address_data, the value of $address_data was not
4339 passed back to the ACL. This was different to the case when no
4340 redirection occurred. The value is now passed back in both cases.
4342 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4343 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4344 use login_cap.h, so on its own it isn't the distinguishing feature. The
4345 new name refers directly to the setclassresources() function.
4347 PH/65 Added configuration files for NetBSD3.
4349 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4351 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4352 is preferred over IPv4.
4354 PH/68 The bounce_return_message and bounce_return_body options were not being
4355 honoured for bounces generated during the reception of non-SMTP messages.
4356 In particular, this applied to messages rejected by the ACL. This bug has
4357 been fixed. However, if bounce_return_message is true and bounce_return_
4358 body is false, the headers that are returned for a non-SMTP message
4359 include only those that have been read before the error was detected.
4360 (In the case of an ACL rejection, they have all been read.)
4362 PH/69 The HTML version of the specification is now built in a directory called
4363 spec_html instead of spec.html, because the latter looks like a path with
4364 a MIME-type, and this confuses some software.
4366 PH/70 Catch two compiler warnings in sieve.c.
4368 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4369 function verify_get_ident() calls ip_connect() to connect a socket, but
4370 if the "connect()" function timed out, ip_connect() used to close the
4371 socket. However, verify_get_ident() also closes the socket later, and in
4372 between Exim writes to the log, which may get opened at this point. When
4373 the socket was closed in ip_connect(), the log could get the same file
4374 descriptor number as the socket. This naturally causes chaos. The fix is
4375 not to close the socket in ip_connect(); the socket should be closed by
4376 the function that creates it. There was only one place in the code where
4377 this was missing, in the iplookup router, which I don't think anybody now
4378 uses, but I've fixed it anyway.
4380 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4381 well as to direct DNS lookups. Otherwise the handling of names in host
4382 lists is inconsistent and therefore confusing.
4388 PH/01 Two changes to the default runtime configuration:
4390 (1) Move the checks for relay_from_hosts and authenticated clients from
4391 after to before the (commented out) DNS black list checks.
4393 (2) Add control=submission to the relay_from_hosts and authenticated
4394 clients checks, on the grounds that messages accepted by these
4395 statements are most likely to be submissions.
4397 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4399 (1) Generate an error if the third argument for the ${prvs expansion is
4402 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4405 (3) Reset the variables that are obtained from the first argument of
4406 ${prvscheck and used in the second argument before leaving the code,
4407 because their memory is reclaimed, so using them afterwards may do
4410 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4411 one (it's much easier than Tom thought :-).
4413 (5) Because of (4), we can now allow for the use of $prvscheck_result
4414 inside the third argument.
4416 PH/03 For some reason, the default setting of PATH when running a command from
4417 a pipe transport was just "/usr/bin". I have changed it to
4420 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4421 anything to be listed in the output from -bV.
4423 PH/05 When a filter generated an autoreply, the entire To: header line was
4424 quoted in the delivery log line, like this:
4426 => >A.N.Other <ano@some.domain> <original@ddress> ...
4428 This has been changed so that it extracts the operative address. There
4429 may be more than one such address. If so, they are comma-separated, like
4432 => >ano@some.domain,ona@other.domain <original@ddress> ...
4434 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4435 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4436 not being looked up in the rDNS to get a host name, Exim was showing the
4437 IP address twice in Received: lines, even though the IP addresses were
4438 identical. For example:
4440 Received: from [1.2.3.4] (helo=[1.2.3.4])
4442 However, if the real host name was known, it was omitting the HELO data
4443 if it matched the actual IP address. This has been tidied up so that it
4444 doesn't show the same IP address twice.
4446 PH/07 When both +timestamp and +memory debugging was on, the value given by
4447 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4448 called by the debug printing, thereby overwriting the timestamp buffer.
4449 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4451 PH/08 When the original message was included in an autoreply transport, it
4452 always said "this is a copy of the message, including all the headers",
4453 even if body_only or headers_only was set. It now gives an appropriate
4456 PH/09 Applied a patch from the Sieve maintainer which:
4458 o fixes some comments
4459 o adds the (disabled) notify extension core
4460 o adds some debug output for the result of if/elsif tests
4461 o points to the current vacation draft in the documentation
4462 and documents the missing references header update
4466 o fixes a bug in processing the envelope test (when testing
4467 multiple envelope elements, the last element determined the
4470 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4471 Electronic Mail") by including:
4473 Auto-submitted: auto-generated
4475 in the messages that it generates (bounce messages and others, such as
4476 warnings). In the case of bounce messages for non-SMTP messages, there was
4477 also a typo: it was using "Auto_submitted" (underscore instead of
4478 hyphen). Since every message generated by Exim is necessarily in response
4479 to another message, thes have all been changed to:
4481 Auto-Submitted: auto-replied
4483 in accordance with these statements in the RFC:
4485 The auto-replied keyword:
4487 - SHOULD be used on messages sent in direct response to another
4488 message by an automatic process,
4490 - MUST NOT be used on manually-generated messages,
4492 - MAY be used on Delivery Status Notifications (DSNs) and Message
4493 Disposition Notifications (MDNs),
4495 - MUST NOT be used on messages generated by automatic or periodic
4496 processes, except for messages which are automatic responses to
4499 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4500 to the default Received: header definition.
4502 PH/12 Added log selector acl_warn_skipped (default on).
4504 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4505 variables because (a) they are in the wrong storage pool and (b) even if
4506 they were copied, it wouldn't work properly because of the caching.
4508 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4509 checking when decoding. Apparently there are clients that generate
4510 overlong encoded strings. Why am I not surprised?
4512 PH/15 If the first argument of "${if match_address" was not empty, but did not
4513 contain an "@" character, Exim crashed. Now it writes a panic log message
4514 and treats the condition as false.
4516 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4518 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4519 extension "envelope-auth". The code is finished and in agreement with
4520 other implementations, but there is no documentation so far and in fact,
4521 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4522 not changing the active code.
4524 Print executed "if" and "elsif" statements when debugging is used. This
4525 helps a great deal to understand what a filter does.
4527 Document more things not specified clearly in RFC3028. I had all this
4528 sorted out, when out of a sudden new issues came to my mind. Oops."
4530 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4533 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4534 canonical form (as documented). However, after a host name lookup from
4535 the IP address, check_host() was doing a simple string comparison with
4536 addresses acquired from the DNS when checking that the found name did
4537 have the original IP as one of its addresses. Since any found IPv6
4538 addresses are likely to be in abbreviated form, the comparison could
4539 fail. Luckily, there already exists a function for doing the comparison
4540 by converting both addresses to binary, so now that is used instead of
4541 the text comparison.
4543 PH/20 There was another similar case to PH/19, when a complete host name was
4544 given in a host list; looking up its IP address could give an abbreviated
4545 form, whereas the current host's name might or might not be abbreviated.
4546 The same fix has been applied.
4552 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4553 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4556 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4557 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4559 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4561 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4562 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4563 Also, for "Auto-Submitted", treat anything other than "no" as
4564 non-personal, in accordance with RFC 3834. (Previously it treated
4565 anything starting "auto-" as non-personal.)
4567 TF/01 The control=submission/name=... option had a problem with syntax
4568 errors if the name included a slash character. The /name= option
4569 now slurps the rest of the string, so it can include any characters
4570 but it must come last in the list of options (after /sender_retain
4573 PH/05 Some modifications to the interface to the fake nameserver for the new
4581 TK/01 Added the "success_on_redirect" address verification option. See
4582 NewStuff for rationale and an example.
4584 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4586 PH/02 Patch to exigrep to allow it to work on syslog lines.
4588 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4589 fread() to skip over the body file's header line, because in Cygwin the
4590 header line is locked and is inaccessible.
4592 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4593 co-exist for some time) to make it clear that it is the Exim ID that is
4594 referenced, not the Message-ID: header line.
4596 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4597 string_format() function, because snprintf() does not exist on all
4600 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4601 ${stat: expansion item.
4603 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4604 protocol synchronization error", to keep the pedants happy.
4606 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4607 well as for IRIX systems, when gcc is being used. See the host.c source
4610 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4612 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4615 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4616 transport and to the smtp transport in LMTP mode.
4618 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4620 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4621 run a filter in a subprocess. This could lead to confusion in subsequent
4622 lookups in the parent process. There should also be a search_tidyup() at
4623 the end of the subprocess.
4625 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4626 only if the host matched helo_try_verify_hosts, which caused the
4627 verification to occur when the EHLO/HELO command was issued. The ACL just
4628 tested the remembered result. Now, if a previous verification attempt has
4629 not happened, "verify = helo" does it there and then.
4631 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4633 TK/03 Fix log output including CR from clamd.
4635 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4636 exist provoked a memory error which could cause a segfault.
4638 PH/15 Installed PCRE 6.2
4640 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4642 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4643 of the problem. Specifically, suggested +O2 rather than +O1 for the
4646 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4648 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4649 record, the port was not correctly logged when the outgoing_port log
4650 selector was set (it logged the transort's default port).
4652 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4653 fallback_hosts, and "hosts" in the smtp transport.
4655 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4656 host errors such as "Connection refused".
4658 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4659 authentication with radiusclient 0.4.9:
4661 - Error returned from rc_read_config was caught wrongly
4662 - Username/password not passed on to radius server due to wrong length.
4664 The presumption is that some radiusclient API changes for 4.51/PH/17
4665 were not taken care of correctly. The code is still untested by me (my
4666 Linux distribution still has 0.3.2 of radiusclient), but it was
4667 contributed by a Radius user.
4669 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4670 expanding the "port" option of the smtp transport.
4672 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4673 while reading a MIME header. Thanks to Tom Hughes for a patch.
4675 PH/24 Include config.h inside local_scan.h so that configuration settings are
4678 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4679 that in an ACL for RCPT (for example), you can examine exactly what was
4682 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4683 commands, but it was not correctly comparing the address with the actual
4684 client host address. Thus, it would show the EHLO address in Received:
4685 header lines when this was not necessary.
4687 PH/27 Added the % operator to ${eval:}.
4689 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4690 it should be ignoring failures (because with -C, for example, it has lost
4691 privilege). It wasn't ignoring creation failures other than "already
4694 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4697 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4698 because an input line was too long, either on its own, or by virtue of
4699 too many continuations, the temporary file was not being removed, and the
4700 return code was incorrect.
4702 PH/31 Missing "BOOL" in function definition in filtertest.c.
4704 PH/32 Applied Sieve patches from the maintainer.
4706 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4708 PH/33 Added "verify = not_blind".
4710 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4711 Local/Makefile (with some defaults set). These are used in built scripts
4712 such as exicyclog, but they have never been used in the exim_install
4713 script (though there are many overriding facilities there). I have
4714 arranged that the exim_install script now takes note of these two
4717 PH/35 Installed configuration files for Dragonfly.
4719 PH/36 When a locally submitted message by a trusted user did not contain a
4720 From: header, and the sender address was obtained from -f or from an SMTP
4721 MAIL command, and the trusted user did not use -F to supply a sender
4722 name, $originator_name was incorrectly used when constructing a From:
4723 header. Furthermore, $originator_name was used for submission mode
4724 messages from external hosts without From: headers in a similar way,
4725 which is clearly wrong.
4727 PH/37 Added control=suppress_local_fixups.
4729 PH/38 When log_selector = +received_sender was set, and the addition of the
4730 sender made the log line's construction buffer exactly full, or one byte
4731 less than full, an overflow happened when the terminating "\n" was
4734 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4735 when the result of a list match is failure because a DNS lookup failed.
4737 PH/40 RM_COMMAND is now used in the building process.
4739 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4740 the "build-* directories that it finds.
4742 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4743 address in a domain literal was a prefix of an interface address.
4745 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4746 when verifying a sender address, unless rewrite_headers is false.
4748 PH/44 Wrote a long comment about why errors_to addresses are verified as
4749 recipients, not senders.
4751 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4752 the ratelimit ACL was added.
4754 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4756 PH/46 Added extra information about PostgreSQL errors to the error string.
4758 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4759 avoiding the need to install special zones in a real server. This is
4760 backwards compatible; if it can't find the fake resolver, it drops back.
4761 Thus, both old and new test suites can be run.
4763 TF/02 Added util/ratelimit.pl
4765 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4766 clock is set back in time.
4768 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4769 Candler <B.Candler@pobox.com>.
4771 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4772 OK for addresses that are the result of redirections.
4774 PH/48 A number of further additions for the benefit of the new test suite,
4775 including a fake gethostbyname() that interfaces to the fake DNS resolver
4778 TF/06 The fix for widen_domains has also been applied to qualify_single and
4779 search_parents which are the other dnslookup options that can cause
4782 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4785 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4787 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4788 service" if it encountered an MX record with an empty target hostname.
4789 The message is now "an MX or SRV record indicated no SMTP service".
4791 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4792 if the DNS of the sending site is misconfigured. This is quite a
4793 common situation. This change restores the behaviour of treating a
4794 helo verification defer as a failure.
4796 PH/51 If self=fail was set on a router, the bounce message did not include the
4797 actual error message.
4803 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4805 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4806 command itself ended in error, the underlying message about the transport
4807 filter timeout was being overwritten with the pipe command error. Now the
4808 underlying error message should be appended to the second error message.
4810 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4812 PH/02 Reduce the amount of output that "make" produces by default. Full output
4813 can still be requested.
4815 PH/03 The warning log line about a condition test deferring for a "warn" verb
4816 was being output only once per connection, rather than after each
4817 occurrence (because it was using the same function as for successful
4818 "warn" verbs). This seems wrong, so I have changed it.
4820 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4821 it should not have, which might have caused a crash in the right
4822 circumstances, but probably never did.
4824 PH/04 Installed a modified version of Tony Finch's patch to make submission
4825 mode fix the return path as well as the Sender: header line, and to
4826 add a /name= option so that you can make the user's friendly name appear
4829 TF/03 Added the control = fakedefer ACL modifier.
4831 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4832 Mark Lowes for thorough testing.
4834 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4836 TK/03 Merged latest SRS patch from Miles Wilton.
4838 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4839 with the definition in sysexits.h (which is #included earlier).
4840 Fortunately, Exim does not actually use EX_OK. The code used to try to
4841 preserve the sysexits.h value, by assuming that macro definitions were
4842 scanned for macro replacements. I have been disabused of this notion,
4843 so now the code just undefines EX_OK before #including unistd.h.
4845 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4846 in the smtp transport. When a block could not be written in a single
4847 write() function, the timeout was being re-applied to each part-write.
4848 This seems wrong - if the receiver was accepting one byte at a time it
4849 would take for ever. The timeout is now adjusted when this happens. It
4850 doesn't have to be particularly precise.
4852 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4853 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4855 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4856 without a domain if the check to <postmaster@domain> fails.
4858 SC/01 Eximstats: added -xls and the ability to specify output files
4859 (patch written by Frank Heydlauf).
4861 SC/02 Eximstats: use FileHandles for outputting results.
4863 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4865 SC/04 Eximstats: fixed display of large numbers with -nvr option
4867 SC/05 Eximstats: fixed merging of reports with empty tables.
4869 SC/06 Eximstats: added the -include_original_destination flag
4871 SC/07 Eximstats: removed tabs and trailing whitespace.
4873 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4875 TK/06 MBOX spool code: Add real "From " MBOX separator line
4876 so the .eml file is really in mbox format (even though
4877 most programs do not really care). Patch from Alex Miller.
4879 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4880 The latter is generated from $received_to and is only set if the
4881 message has one envelope recipient. SA can use these headers,
4882 obviously out-of-the-box. Patch from Alex Miller.
4884 PH/08 The ${def test on a variable was returning false if the variable's
4885 value was "0", contrary to what the specification has always said!
4886 The result should be true unless the variable is empty.
4888 PH/09 The syntax error of a character other than { following "${if
4889 def:variable_name" (after optional whitespace) was not being diagnosed.
4890 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4891 accidental colon was present, for example, could give incorrect results.
4893 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4894 result is used (not including appendfile, where other changes are about
4897 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4898 This involved changing a lot of size variables from int to off_t. It
4899 should work with maildirs and everything.
4901 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4902 spamd dying while we are connected to it.
4904 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4907 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4908 function for BDB 4.3.
4910 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4912 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4913 the sources for PCRE have been re-arranged and more files are now
4916 PH/15 The code I had for printing potentially long long variables in PH/11
4917 above was not the best (it lost precision). The length of off_t variables
4918 is now inspected at build time, and an appropriate printing format (%ld
4919 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4920 to be "long long int" or "long int". This is needed for the internal
4921 formatting function string_vformat().
4923 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4924 the configuration file to be ":syslog", then the script "guesses" where
4925 the logs files are, rather than using the compiled in default. In our
4926 case the guess is not the same as the compiled default, so the script
4927 suddenly stopped working when I started to use syslog. The patch checks
4928 to see if log_file_path is "". If so, it attempts to read it from exim
4929 with no configuration file to get the compiled in version, before it
4930 falls back to the previous guessing code."
4932 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4933 implementing BATV in an Exim configuration. See NewStuff for the gory
4936 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4937 Makefile that are specific to HP-UX.
4939 PH/18 If the "use_postmaster" option was set for a recipient callout together
4940 with the "random" option, the postmaster address was used as the MAIL
4941 FROM address for the random test, but not for the subsequent recipient
4942 test. It is now used for both.
4944 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4945 patch removes a few documentation additions to RFC 3028, because the
4946 latest draft now contains them. It adds the new en;ascii-case comparator
4947 and a new error check for 8bit text in MIME parts. Comparator and
4948 require names are now matched exactly. I enabled the subaddress
4949 extension, but it is not well tested yet (read: it works for me)."
4951 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4952 rework some of the code of TK/09 above to avoid the hardwired use of
4953 "%lld" and "long long". Replaced the call to snprintf() with a call to
4956 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4957 records point to non-existent hosts", "retry timeout exceeded", and
4958 "retry time not reached for any host after a long failure period".
4960 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4961 experimental DomainKeys support:
4963 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4964 (2) On an error such as an illegally used "control", the wrong name for
4965 the control was given.
4967 These problems did NOT occur unless DomainKeys support was compiled.
4969 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4971 PH/24 Added ${if match_ip condition.
4973 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4974 databases so that it will be absolutely obvious if a crash occurs in the
4975 DB library. This is a regular occurrence (often caused by mis-matched
4978 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4979 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4980 picked up on a user's system that detects such things. There doesn't seem
4981 to be a gcc warning option for this - only an attribute that has to be
4982 put on the function's prototype. It seems that in Fedora Core 4 they have
4983 set this on a number of new functions. No doubt there will be more in due
4986 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4987 specify a transport. However, if an address that was verified by such a
4988 router was the subject of a callout, Exim crashed because it tried to
4989 read the rcpt_include_affixes from the non-existent transport. Now it
4990 just assumes that the setting of that option is false. This bug was
4991 introduced by 4.51/PH/31.
4993 PH/28 Changed -d+all to exclude +memory, because that information is very
4994 rarely of interest, but it makes the output a lot bigger. People tend to
4995 do -d+all out of habit.
4997 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4998 code in os-type was giving problems when libc.so lives in lib64, like on
5001 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5002 aren't the modern standard, and it seems that some systems' include files
5003 don't always have them. Exim was already checking for some of the newer
5004 ones like T_AAAA, and defining it itself. I've added checks for all the
5005 record types that Exim uses.
5007 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5008 not automatically generating a new one, as it is supposed to. This
5009 prevented TLS from working. If the file did exist, but contained invalid
5010 data, a new version was generated, as expected. It was only the case of a
5011 non-existent file that was broken.
5013 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5014 with a change in libdomainkeys > 0.64.
5016 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5017 from DNS. If the selector record carries the flag, it now has
5018 precedence over the domain-wide flag.
5020 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5022 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5023 the use of a port name that isn't defined in /etc/services) occurred, the
5024 message was deferred as in a normal delivery, and thus remained on the
5025 spool, instead of being failed because of the mua_wrapper setting. This
5026 is now fixed, and I tidied up some of the mua_wrapper messages at the
5029 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5030 the messages in a hash of arrays rather than using individual hashes.
5031 This is a bit cleaner and results in dramatic memory savings, albeit
5032 at a slight CPU cost.
5034 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5035 as requested by Marc Sherman.
5037 SC/10 Eximstats: added histograms for user specified patterns as requested
5040 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5042 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5043 fopen() in the content-scanning modules that did not already have it.
5049 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5050 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5052 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5054 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5056 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5057 to test Sieve filters that use "vacation".
5059 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5060 that changes the way the GnuTLS parameters are stored in the cache file.
5061 The new format can be generated externally. For backward compatibility,
5062 if the data in the cache doesn't make sense, Exim assumes it has read an
5063 old-format file, and it generates new data and writes a new file. This
5064 means that you can't go back to an older release without removing the
5067 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5068 work if there are any delivery delays because "one_time" forces the
5069 parent to be marked "delivered", so its unseen clone is never tried
5070 again. For this reason, Exim now forbids the simultaneous setting of
5073 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5074 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5075 entry if you want to know the details. The fix, however, neglected to
5076 consider the case when local delivery batching is involved. The test for
5077 "previously delivered" was not happening when checking to see if an
5078 address could be batched with a previous (undelivered) one; under
5079 certain circumstances this could lead to multiple deliveries to the same
5082 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5083 in its include files, and this causes problems building Exim.
5085 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5086 header_syntax) but Exim was just ignoring anything given after a slash.
5087 In particular, this caused confusion with an attempt to use "verify =
5088 reverse_host_lookup/defer_ok". An error is now given when options are
5089 supplied for verify items that do not have them. (Maybe reverse_host_
5090 lookup should have a defer_ok option, but that's a different point.)
5092 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5093 defined by RFC 821) to 2048, because there were problems with some AUTH
5094 commands, and RFC 1869 says the size should be increased for extended
5095 SMTP commands that take arguments.
5097 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5100 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5101 "unknown" error; now it says that the functionality isn't in the binary.
5103 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5104 an address' error message when a string expansion fails (syntax or
5105 whatever). Otherwise the password may appear in the log. Following change
5106 PH/42 below, there is no longer a chance of it appearing in a bounce
5109 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5111 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5112 of its addresses was ever tried. (Bugzilla bug #2.)
5114 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5115 the result incorrectly in the debug output. (It correctly added a newline
5116 to what was transported.)
5118 TF/01 Added $received_time.
5120 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5121 commented out examples of how to interface to a virus scanner and to
5122 SpamAssassin. Also added commented examples of av_scanner and
5123 spamd_address settings.
5125 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5126 and controls are allowed in which ACLs. There were a couple of minor
5127 errors. Some of the entries in the conditions table (which is a table of
5128 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5129 negation of where the condition IS allowed.
5131 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5133 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5134 header file does not have a version number, so I've had to invent a new
5135 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5136 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5137 radiusclient), but it was contributed by a Radius user.
5139 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5140 files or directories, for OpenSSL.
5142 PH/19 When an Exim process that is running as root has to create an Exim log
5143 file, it does so in a subprocess that runs as exim:exim so as to get the
5144 ownership right at creation (otherwise, other Exim processes might see
5145 the file with the wrong ownership). There was no test for failure of this
5146 fork() call, which would lead to the process getting stuck as it waited
5147 for a non-existent subprocess. Forks do occasionally fail when resources
5148 run out. I reviewed all the other calls to fork(); they all seem to check
5151 PH/20 When checking for unexpected SMTP input at connect time (before writing
5152 the banner), Exim was not dealing correctly with a non-positive return
5153 from the read() function. If the client had disconnected by this time,
5154 the result was a log entry for a synchronization error with an empty
5155 string after "input=" when read() returned zero. If read() returned -1
5156 (an event I could not check), uninitialized data bytes were printed.
5157 There were reports of junk text (parts of files, etc) appearing after
5160 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5162 PH/22 Added support for macro redefinition, and (re)definition in between
5163 driver and ACL definitions.
5165 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5166 forgetting to use the resulting value; it was using the unexpanded value.
5168 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5169 hadn't been configured. The fix is from Juergen Kreileder, who
5170 understands it better than I do:
5172 "Here's what I see happening with three configured cyrus_sasl
5173 authenticators configured (plain, login, cram-md5):
5175 On startup auth_cyrus_sasl_init() gets called for each of these.
5176 This means three calls to sasl_listmech() without a specified mech_list.
5177 => SASL tests which mechs of all available mechs actually work
5178 => three warnings about OTP not working
5179 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5181 With the patch, sasl_listmech() also gets called three times. But now
5182 SASL's mech_list option is set to the server_mech specified in the the
5183 authenticator. Or in other words, the answer from sasl_listmech()
5184 gets limited to just the mech you're testing for (which is different
5186 => the return list contains just 'plain' or 'login', 'cram-md5' or
5187 nothing depending on the value of ob->server_mech.
5189 I've just tested the patch: Authentication still works fine,
5190 unavailable mechs specified in the exim configuration are still
5191 caught, and the auth.log warnings about OTP are gone."
5193 PH/25 When debugging is enabled, the contents of the command line are added
5194 to the debugging output, even when log_selector=+arguments is not
5197 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5198 answer is "GNU", and only if the return is "GNU/something" is the answer
5201 PH/27 $acl_verify_message is now set immediately after the failure of a
5202 verification in an ACL, and so is available in subsequent modifiers. In
5203 particular, the message can be preserved by coding like this:
5205 warn !verify = sender
5206 set acl_m0 = $acl_verify_message
5208 Previously, $acl_verify_message was set only while expanding "message"
5209 and "log_message" when a very denied access.
5211 PH/28 Modified OS/os.c-Linux with
5213 -#ifndef OS_LOAD_AVERAGE
5214 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5216 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5217 nomenclature these days.)
5219 PH/29 Installed patch from the Sieve maintainer that adds the options
5220 sieve_useraddress and sieve_subaddress to the redirect router.
5222 PH/30 In these circumstances:
5223 . Two addresses routed to the same list of hosts;
5224 . First host does not offer TLS;
5225 . First host accepts first address;
5226 . First host gives temporary error to second address;
5227 . Second host offers TLS and a TLS session is established;
5228 . Second host accepts second address.
5229 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5230 and peerdn, if requested) that were in fact used only for the second
5233 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5234 attention to any local part prefix or suffix that was matched by the
5235 router that accepted the address. It now behaves in the same way as it
5236 does for delivery: the affixes are removed from the local part unless
5237 rcpt_include_affixes is set on the transport.
5239 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5240 timeout during the DATA phase of an incoming message.
5242 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5243 applied a patch sanctioned by the Sieve maintainer.
5245 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5246 the uid or gid is negative. A case of a negative gid caused this to be
5247 noticed. The fix allows for either to be negative.
5249 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5250 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5253 PH/36 The change PH/12 above was broken. Fixed it.
5255 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5256 the grounds that routing the same address twice would always produce the
5257 same answer. This might have been true once, but it is certainly no
5258 longer true now. Routing a child address may depend on the previous
5259 routing that produced that child. Some complicated redirection strategies
5260 went wrong when messages had multiple recipients, and made Exim's
5261 behaviour dependent on the order in which the addresses were given.
5263 I have moved the duplicate checking until after the routing is complete.
5264 Exim scans the addresses that are assigned to local and remote
5265 transports, and removes any duplicates. This means that more work will be
5266 done, as duplicates will always all be routed, but duplicates are
5267 presumably rare, so I don't expect this is of any significance.
5269 For deliveries to pipes, files, and autoreplies, the duplicate checking
5270 still happens during the routing process, since they are not going to be
5273 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5274 It corrects a timeout issue with spamd. This is Ian's comment: "The
5275 background is that sometimes spamd either never reads data from a
5276 connection it has accepted, or it never writes response data. The exiscan
5277 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5278 blindly assumes that writes won't block so it may never time out."
5280 PH/39 Allow G after quota size as well as K and M.
5282 PH/40 The value set for $authenticated_id in an authenticator may not contain
5283 binary zeroes or newlines because the value is written to log lines and
5284 to spool files. There was no check on this. Now the value is run through
5285 the string_printing() function so that such characters are converted to
5286 printable escape sequences.
5288 PH/41 $message_linecount is a new variable that contains the total number of
5289 lines in the message. Compare $body_linecount, which is the count for the
5292 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5293 bounce and delay warning messages, except in certain special cases, which
5296 (a) An SMTP error message from a remote host;
5297 (b) A message specified in a :fail: redirection;
5298 (c) A message specified in a "fail" command in a system filter;
5299 (d) A message specified in a FAIL return from the queryprogram router;
5300 (e) A message specified by the cannot_route_message router option.
5302 In these cases only, Exim does include the error details in bounce and
5303 warning messages. There are also a few cases where bland messages such
5304 as "unrouteable address" or "local delivery error" are given.
5306 PH/43 $value is now also set for the "else" part of a ${run expansion.
5308 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5309 being worked on, but at least Exim now implements the latest version to
5312 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5313 process to complete was treated as a delivery failure, a timeout while
5314 writing the message to the pipe was logged, but erroneously treated as a
5315 successful delivery. Such timeouts include transport filter timeouts. For
5316 consistency with the overall process timeout, these timeouts are now
5317 treated as errors, giving rise to delivery failures by default. However,
5318 there is now a new Boolean option for the pipe transport called
5319 timeout_defer, which, if set TRUE, converts the failures into defers for
5320 both kinds of timeout. A transport filter timeout is now identified in
5323 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5324 systems where "make" and "gmake" are different, calling "gmake" at top
5325 level broke things. I've arranged for the value of $(MAKE) to be passed
5326 from the Makefile to this script so that it can call the same version of
5330 A note about Exim versions 4.44 and 4.50
5331 ----------------------------------------
5333 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5334 changes of various kinds. As a consequence, a big documentation update was
5335 needed. This delayed the release for rather longer than seemed good, especially
5336 in the light of a couple of (minor) security issues. Therefore, the changes
5337 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5338 release. So 4.44 and 4.50 are in effect two different branches that both start
5341 I have left the 4.50 change log unchanged; it contains all the changes since
5342 4.43. The change log for 4.44 is below; many of its items are identical to
5343 those for 4.50. This seems to be the most sensible way to preserve the
5344 historical information.
5350 1. Minor wording change to the doc/README.SIEVE file.
5352 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5353 computation of the current number of files was incorrect.
5355 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5356 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5359 4. Give more explanation in the error message when the command for a transport
5360 filter fails to execute.
5362 5. There are several places where Exim runs a non-Exim command in a
5363 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5364 was being done only for the command run by the queryprogram router. It is
5365 now done for all such subprocesses. The other cases are: ${run, transport
5366 filters, and the commands run by the lmtp and pipe transports.
5368 6. Added CONFIGURE_GROUP build-time option.
5370 7. Some older OS have a limit of 256 on the maximum number of file
5371 descriptors. Exim was using setrlimit() to set 1000 as a large value
5372 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5373 systems. I've change it so that if it can't get 1000, it tries for 256.
5375 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5376 was an oversight, and furthermore, ever since the addition of extra
5377 controls (e.g. 4.43/32), the checks on when to allow different forms of
5378 "control" were broken. There should now be diagnostics for all cases when a
5379 control that does not make sense is encountered.
5381 9. Added the /retain_sender option to "control=submission".
5383 10. $recipients is now available in the predata ACL (oversight).
5385 11. Tidy the search cache before the fork to do a delivery from a message
5386 received from the command line. Otherwise the child will trigger a lookup
5387 failure and thereby defer the delivery if it tries to use (for example) a
5388 cached ldap connection that the parent has called unbind on.
5390 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5391 of $address_data from the recipient verification was clobbered by the
5392 sender verification.
5394 13. The value of address_data from a sender verification is now available in
5395 $sender_address_data in subsequent conditions in the ACL statement.
5397 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5399 15. Added a new option "connect=<time>" to callout options, to set a different
5402 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5403 was its contents. (It was OK if the option was not defined at all.)
5405 17. A "Completed" log line is now written for messages that are removed from
5406 the spool by the -Mrm option.
5408 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5409 information about exactly what failed.
5411 19. Added -dd to debug only the daemon process.
5413 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5414 handles timeouts, both on the server side and network timeouts. Renamed the
5415 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5417 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5418 It is now set to "smtps".
5420 22. $host_address is now set to the target address during the checking of
5421 ignore_target_hosts.
5423 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5424 being passed; this would have caused $sender_host_name to have been used if
5425 matching the list had actually called for a host name (not very likely,
5426 since this list is usually IP addresses). A host name is now passed as
5429 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5430 code that specifies a non-restarting handler (typically sigaction() in
5431 modern systems) in an attempt to fix a rare and obscure crash bug.
5433 25. Narrowed the window for a race in the daemon that could cause it to ignore
5434 SIGCHLD signals. This is not a major problem, because they are used only to
5435 wake it up if nothing else does.
5437 26. A malformed maildirsize file could cause Exim to calculate negative values
5438 for the mailbox size or file count. Odd effects could occur as a result.
5439 The maildirsize information is now recalculated if the size or filecount
5442 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5443 support for a long time. Removed HAVE_SYS_VFS_H.
5445 28. Installed the latest version of exipick from John Jetmore.
5447 29. In an address list, if the pattern was not a regular expression, an empty
5448 subject address (from a bounce message) matched only if the pattern was an
5449 empty string. Non-empty patterns were not even tested. This was the wrong
5450 because it is perfectly reasonable to use an empty address as part of a
5451 database query. An empty address is now tested by patterns that are
5452 lookups. However, all the other forms of pattern expect the subject to
5453 contain a local part and a domain, and therefore, for them, an empty
5454 address still always fails if the pattern is not itself empty.
5456 30. Exim went into a mad DNS loop when attempting to do a callout where the
5457 host was specified on an smtp transport, and looking it up yielded more
5458 than one IP address.
5460 31. Re-factored the code for checking spool and log partition space into a
5461 function that finds that data and another that does the check. The former
5462 is then used to implement four new variables: $spool_space, $log_space,
5463 $spool_inodes, and $log_inodes.
5465 32. The RFC2047 encoding function was originally intended for short strings
5466 such as real names; it was not keeping to the 75-character limit for
5467 encoded words that the RFC imposes. It now respects the limit, and
5468 generates multiple encoded words if necessary. To be on the safe side, I
5469 have increased the buffer size for the ${rfc2047: expansion operator from
5472 33. It is now permitted to omit both strings after an "if" condition; if the
5473 condition is true, the result is "true". As before, when the second string
5474 is omitted, a false condition yields an empty string. This makes it less
5475 cumbersome to write custom ACL and router conditions.
5477 34. Failure to deliver a bounce message always caused it to be frozen, even if
5478 there was an errors_to setting on the router. The errors_to setting is now
5481 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5482 canonical form (fully expanded) before being placed in
5483 $sender_host_address.
5485 36. The table in the code that translates DNS record types into text (T_A to
5486 "A" for instance) was missing entries for NS and CNAME. It is just possible
5487 that this could have caused confusion if both these types were looked up
5488 for the same domain, because the text type is used as part of Exim's
5489 per-process caching. But the chance of anyone hitting this buglet seems
5492 37. The dnsdb lookup has been extended in a number of ways.
5494 (1) There is a new type, "zns", which walks up the domain tree until it
5495 finds some nameserver records. It should be used with care.
5497 (2) There is a new type, "mxh", which is like "mx" except that it returns
5498 just the host names, not the priorities.
5500 (3) It is now possible to give a list of domains (or IP addresses) to be
5501 looked up. The behaviour when one of the lookups defers can be
5502 controlled by a keyword.
5504 (4) It is now possible to specify the separator character for use when
5505 multiple records are returned.
5507 38. The dnslists ACL condition has been extended: it is now possible to supply
5508 a list of IP addresses and/or domains to be looked up in a particular DNS
5511 39. Added log_selector=+queue_time_overall.
5513 40. When running the queue in the test harness, wait just a tad after forking a
5514 delivery process, to get repeatability of debugging output.
5516 41. Include certificate and key file names in error message when GnuTLS fails
5517 to set them up, because the GnuTLS error message doesn't include the name
5518 of the failing file when there is a problem reading it.
5520 42. Allow both -bf and -bF in the same test run.
5522 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5524 44. The "Exiscan patch" is now merged into the mainline Exim source.
5526 45. Sometimes the final signoff response after QUIT could fail to get
5527 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5528 < 0 before doing a fflush(). This bug looks as though it goes back to the
5529 introduction of TLS in release 3.20, but "sometimes" must have been rare
5530 because the tests only now provoked it.
5532 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5533 (this can affect the format of dates).
5535 47. exim_tidydb, when checking for the continued existence of a message for
5536 which it has found a message-specific retry record, was not finding
5537 messages that were in split spool directories. Consequently, it was
5538 deleting retry records that should have stayed in existence.
5540 48. Steve fixed some bugs in eximstats.
5542 49. The SPA authentication driver was not abandoning authentication and moving
5543 on to the next authenticator when an expansion was forced to fail,
5544 contradicting the general specification for all authenticators. Instead it
5545 was generating a temporary error. It now behaves as specified.
5547 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5548 (the order specifies the preference for clients). The order is now AES256,
5549 AES128, 3DES, ARCFOUR128.
5551 51. Small patch to Sieve code - explicitly set From: when generating an
5554 52. Exim crashed if a remote delivery caused a very long error message to be
5555 recorded - for instance if somebody sent an entire SpamAssassin report back
5556 as a large number of 550 error lines. This bug was coincidentally fixed by
5557 increasing the size of one of Exim's internal buffers (big_buffer) that
5558 happened as part of the Exiscan merge. However, to be on the safe side, I
5559 have made the code more robust (and fixed the comments that describe what
5562 53. Now that there can be additional text after "Completed" in log lines (if
5563 the queue_time_overall log selector is set), a one-byte patch to exigrep
5564 was needed to allow it to recognize "Completed" as not the last thing in
5567 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5568 patch that reportedly fixes this has been added. I am not expert enough to
5569 create a test for it. This is what the patch creator wrote:
5571 "I found a little strange behaviour of ldap code when working with
5572 Windows 2003 AD Domain, where users was placed in more than one
5573 Organization Units. When I tried to give exim partial DN, the exit code
5574 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5575 But simultaneously result of request was absolutely normal ldap result,
5576 so I produce this patch..."
5578 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5579 REFERENCE, so I have modified the code to exclude the patch when that macro
5582 55. Some experimental protocols are using DNS PTR records for new purposes. The
5583 keys for these records are domain names, not reversed IP addresses. The
5584 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5585 leaves it alone. Component reversal etc. now happens only for IP addresses.
5588 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5590 57. Double the size of the debug message buffer (to 2048) so that more of very
5591 long debug lines gets shown.
5593 58. The exicyclog utility now does better if the number of log files to keep
5594 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5596 59. Two changes related to the smtp_active_hostname option:
5598 (1) $smtp_active_hostname is now available as a variable.
5599 (2) The default for smtp_banner uses $smtp_active_hostname instead
5600 of $primary_hostname.
5602 60. The host_aton() function is supposed to be passed a string that is known
5603 to be a valid IP address. However, in the case of IPv6 addresses, it was
5604 not checking this. This is a hostage to fortune. Exim now panics and dies
5605 if the condition is not met. A case was found where this could be provoked
5606 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5607 components; fortuitously, this particular loophole had already been fixed
5608 by change 4.50/55 above.
5610 If there are any other similar loopholes, the new check in host_aton()
5611 itself should stop them being exploited. The report I received stated that
5612 data on the command line could provoke the exploit when Exim was running as
5613 exim, but did not say which command line option was involved. All I could
5614 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5615 running as the user.
5618 61. There was a buffer overflow vulnerability in the SPA authentication code
5619 (which came originally from the Samba project). I have added a test to the
5620 spa_base64_to_bits() function which I hope fixes it.
5623 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5624 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5626 63. The daemon start-up calls getloadavg() while still root for those OS that
5627 need the first call to be done as root, but it missed one case: when
5628 deliver_queue_load_max is set with deliver_drop_privilege. This is
5629 necessary for the benefit of the queue runner, because there is no re-exec
5630 when deliver_drop_privilege is set.
5632 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5633 This has been fixed.
5635 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5636 was in use, was not putting the data itself into the right store pool;
5637 consequently, it could be overwritten for a subsequent message in the same
5638 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5641 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5643 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5644 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5645 the function were treating the return as a boolean value, which happened to
5646 work because 0=false and not-0=true, but is not correct code.
5648 68. The host_aton() function was not handling scoped IPv6 addresses (those
5649 with, for example, "%eth0" on the end) correctly.
5651 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5652 negated items (that is, ~something) in unsigned ints. Some compilers
5653 apparently mutter when there is no cast.
5655 70. If an address verification called from an ACL failed, and did not produce a
5656 user-specific message (i.e. there was only a "system" message), nothing was
5657 put in $acl_verify_message. In this situation, it now puts the system
5660 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5661 session; change 4.31/43 added the unwanted input to the log line - except
5662 that it did not do this in the start of session case. It now does.
5664 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5665 This is wrong and can cause the other end to generate a synchronization
5666 error if it is another Exim or anything else that does the synchronization
5667 check. A QUIT command is no longer sent after a timeout.
5669 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5670 during host lookups.
5672 74. The defer_ok option of callout verification was not working if it was used
5673 when verifying addresses in header lines, that is, for this case:
5675 verify = header_sender/callout=defer_ok
5677 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5678 those file descriptors could be used for SMTP connections. If anything
5679 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5680 could be sent to the SMTP client, causing chaos. The daemon now opens
5681 stdin, stdout, and stderr to /dev/null when it puts itself into the
5684 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5685 log by default. The user can override this with suitable Perl magic.
5687 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5688 the log message when discard triggers, was not working for the DATA ACL or
5689 for the non-SMTP ACL.
5691 78. Error message wording change in sieve.c.
5693 79. If smtp_accept_max_per_host was set, the number of connections could be
5694 restricted to fewer than expected, because the daemon was trying to set up
5695 a new connection before checking whether the processes handling previous
5696 connections had finished. The check for completed processes is now done
5697 earlier. On busy systems, this bug wouldn't be noticed because something
5698 else would have woken the daemon, and it would have reaped the completed
5701 80. If a message was submitted locally by a user whose login name contained one
5702 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5703 It caused a spool format error. I have fixed the spool reading code. A
5704 related problem was that the "from" clause in the Received: line became
5705 illegal because of the space(s). It is now covered by ${quote_local_part.
5707 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5710 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5712 83. Patch from Sieve maintainer to fix unterminated string problem in
5713 "vacation" handling.
5715 84. Some minor changes to the Linux configuration files to help with other
5716 OS variants using glibc.
5718 85. One more patch for Sieve to update vacation handling to latest spec.
5721 ----------------------------------------------------
5722 See the note above about the 4.44 and 4.50 releases.
5723 ----------------------------------------------------
5729 1. Change 4.43/35 introduced a bug that caused file counts to be
5730 incorrectly computed when quota_filecount was set in an appendfile
5733 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5734 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5737 3. Give more explanation in the error message when the command for a transport
5738 filter fails to execute.
5740 4. There are several places where Exim runs a non-Exim command in a
5741 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5742 was being done only for the command run by the queryprogram router. It is
5743 now done for all such subprocesses. The other cases are: ${run, transport
5744 filters, and the commands run by the lmtp and pipe transports.
5746 5. Some older OS have a limit of 256 on the maximum number of file
5747 descriptors. Exim was using setrlimit() to set 1000 as a large value
5748 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5749 systems. I've change it so that if it can't get 1000, it tries for 256.
5751 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5752 was an oversight, and furthermore, ever since the addition of extra
5753 controls (e.g. 4.43/32), the checks on when to allow different forms of
5754 "control" were broken. There should now be diagnostics for all cases when a
5755 control that does not make sense is encountered.
5757 7. $recipients is now available in the predata ACL (oversight).
5759 8. Tidy the search cache before the fork to do a delivery from a message
5760 received from the command line. Otherwise the child will trigger a lookup
5761 failure and thereby defer the delivery if it tries to use (for example) a
5762 cached ldap connection that the parent has called unbind on.
5764 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5765 of $address_data from the recipient verification was clobbered by the
5766 sender verification.
5768 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5769 was its contents. (It was OK if the option was not defined at all.)
5771 11. A "Completed" log line is now written for messages that are removed from
5772 the spool by the -Mrm option.
5774 12. $host_address is now set to the target address during the checking of
5775 ignore_target_hosts.
5777 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5778 being passed; this would have caused $sender_host_name to have been used if
5779 matching the list had actually called for a host name (not very likely,
5780 since this list is usually IP addresses). A host name is now passed as
5783 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5784 code that specifies a non-restarting handler (typically sigaction() in
5785 modern systems) in an attempt to fix a rare and obscure crash bug.
5787 15. Narrowed the window for a race in the daemon that could cause it to ignore
5788 SIGCHLD signals. This is not a major problem, because they are used only to
5789 wake it up if nothing else does.
5791 16. A malformed maildirsize file could cause Exim to calculate negative values
5792 for the mailbox size or file count. Odd effects could occur as a result.
5793 The maildirsize information is now recalculated if the size or filecount
5796 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5797 support for a long time. Removed HAVE_SYS_VFS_H.
5799 18. Updated exipick to current release from John Jetmore.
5801 19. Allow an empty sender to be matched against a lookup in an address list.
5802 Previously the only cases considered were a regular expression, or an
5805 20. Exim went into a mad DNS lookup loop when doing a callout where the
5806 host was specified on the transport, if the DNS lookup yielded more than
5809 21. The RFC2047 encoding function was originally intended for short strings
5810 such as real names; it was not keeping to the 75-character limit for
5811 encoded words that the RFC imposes. It now respects the limit, and
5812 generates multiple encoded words if necessary. To be on the safe side, I
5813 have increased the buffer size for the ${rfc2047: expansion operator from
5816 22. Failure to deliver a bounce message always caused it to be frozen, even if
5817 there was an errors_to setting on the router. The errors_to setting is now
5820 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5821 canonical form (fully expanded) before being placed in
5822 $sender_host_address.
5824 24. Updated eximstats to version 1.33
5826 25. Include certificate and key file names in error message when GnuTLS fails
5827 to set them up, because the GnuTLS error message doesn't include the name
5828 of the failing file when there is a problem reading it.
5830 26. Expand error message when OpenSSL has problems setting up cert/key files.
5833 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5834 (this can affect the format of dates).
5836 28. exim_tidydb, when checking for the continued existence of a message for
5837 which it has found a message-specific retry record, was not finding
5838 messages that were in split spool directories. Consequently, it was
5839 deleting retry records that should have stayed in existence.
5841 29. eximstats updated to version 1.35
5842 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5843 1.35 - bugfix such that pie charts by volume are generated correctly
5845 30. The SPA authentication driver was not abandoning authentication and moving
5846 on to the next authenticator when an expansion was forced to fail,
5847 contradicting the general specification for all authenticators. Instead it
5848 was generating a temporary error. It now behaves as specified.
5850 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5851 (the order specifies the preference for clients). The order is now AES256,
5852 AES128, 3DES, ARCFOUR128.
5854 31. Small patch to Sieve code - explicitly set From: when generating an
5857 32. Exim crashed if a remote delivery caused a very long error message to be
5858 recorded - for instance if somebody sent an entire SpamAssassin report back
5859 as a large number of 550 error lines. This bug was coincidentally fixed by
5860 increasing the size of one of Exim's internal buffers (big_buffer) that
5861 happened as part of the Exiscan merge. However, to be on the safe side, I
5862 have made the code more robust (and fixed the comments that describe what
5865 33. Some experimental protocols are using DNS PTR records for new purposes. The
5866 keys for these records are domain names, not reversed IP addresses. The
5867 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5868 leaves it alone. Component reversal etc. now happens only for IP addresses.
5871 34. The host_aton() function is supposed to be passed a string that is known
5872 to be a valid IP address. However, in the case of IPv6 addresses, it was
5873 not checking this. This is a hostage to fortune. Exim now panics and dies
5874 if the condition is not met. A case was found where this could be provoked
5875 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5876 components; fortuitously, this particular loophole had already been fixed
5877 by change 4.50/55 or 4.44/33 above.
5879 If there are any other similar loopholes, the new check in host_aton()
5880 itself should stop them being exploited. The report I received stated that
5881 data on the command line could provoke the exploit when Exim was running as
5882 exim, but did not say which command line option was involved. All I could
5883 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5884 running as the user.
5887 35. There was a buffer overflow vulnerability in the SPA authentication code
5888 (which came originally from the Samba project). I have added a test to the
5889 spa_base64_to_bits() function which I hope fixes it.
5892 36. The daemon start-up calls getloadavg() while still root for those OS that
5893 need the first call to be done as root, but it missed one case: when
5894 deliver_queue_load_max is set with deliver_drop_privilege. This is
5895 necessary for the benefit of the queue runner, because there is no re-exec
5896 when deliver_drop_privilege is set.
5898 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5899 was in use, was not putting the data itself into the right store pool;
5900 consequently, it could be overwritten for a subsequent message in the same
5901 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5904 38. Sometimes the final signoff response after QUIT could fail to get
5905 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5906 < 0 before doing a fflush(). This bug looks as though it goes back to the
5907 introduction of TLS in release 3.20, but "sometimes" must have been rare
5908 because the tests only now provoked it.
5914 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5915 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5916 Now it returns an integer. A number of calls were still expecting a T/F
5917 return. Fortuitously, in all cases, the tests worked in OK situations,
5918 which is the norm. However, things would have gone wrong on any write
5919 failures on the smtp file descriptor. This function is used when sending
5920 messages over SMTP and also when doing verify callouts.
5922 2. When Exim is called to do synchronous delivery of a locally submitted
5923 message (the -odf or -odi options), it no longer closes stderr before doing
5926 3. Implemented the mua_wrapper option.
5928 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5930 5. Implemented the functions header_remove(), header_testname(),
5931 header_add_at_position(), and receive_remove_recipient(), and exported them
5934 6. If an ACL "warn" statement specified the addition of headers, Exim already
5935 inserted X-ACL-Warn: at the start if there was no header name. However, it
5936 was not making this test for the second and subsequent header lines if
5937 there were newlines in the string. This meant that an invalid header could
5938 be inserted if Exim was badly configured.
5940 7. Allow an ACL "warn" statement to add header lines at the start or after all
5941 the Received: headers, as well as at the end.
5943 8. Added the rcpt_4xx retry error code.
5945 9. Added postmaster_mailfrom=xxx to callout verification option.
5947 10. Added mailfrom=xxxx to the callout verification option, for verify=
5950 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5951 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5953 12. Inserted some casts to stop certain compilers warning when using pointer
5954 differences as field lengths or precisions in printf-type calls (mostly
5955 affecting debugging statements).
5957 13. Added optional readline() support for -be (dynamically loaded).
5959 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5960 same clock tick as a message's arrival, so that its received time was the
5961 same as the "first fail" time on the retry record, and that message
5962 remained on the queue past the ultimate address timeout, every queue runner
5963 would try a delivery (because it was past the ultimate address timeout) but
5964 after another failure, the ultimate address timeout, which should have then
5965 bounced the address, did not kick in. This was a "< instead of <=" error;
5966 in most cases the first failure would have been in the next clock tick
5967 after the received time, and all would be well.
5969 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5970 being recognized when the domain list was tested by the match_domain
5971 condition in an expansion string.
5973 16. Added the ${str2b64: operator.
5975 17. Exim was always calling setrlimit() to set a large limit for the number of
5976 processes, without checking whether the existing limit was already
5977 adequate. (It did check for the limit on file descriptors.) Furthermore,
5978 errors from getrlimit() and setrlimit() were being ignored. Now they are
5979 logged to the main and panic logs, but Exim does carry on, to try to do its
5980 job under whatever limits there are.
5982 18. Imported PCRE 5.0.
5984 19. Trivial typo in log message " temporarily refused connection" (the leading
5987 20. If the log selector return_path_on_delivery was set and an address was
5988 redirected to /dev/null, the delivery process crashed because it assumed
5989 that a return path would always be set for a "successful" delivery. In this
5990 case, the whole delivery is bypassed as an optimization, and therefore no
5993 21. Internal re-arrangement: the function for sending a challenge and reading
5994 a response while authentication was assuming a zero-terminated challenge
5995 string. It's now changed to take a pointer and a length, to allow for
5996 binary data in such strings.
5998 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6000 23. Exim was not respecting finduser_retries when seeking the login of the
6001 uid under which it was called; it was always trying 10 times. (The default
6002 setting of finduser_retries is zero.) Also, it was sleeping after the final
6003 failure, which is pointless.
6005 24. Implemented tls_on_connect_ports.
6007 25. Implemented acl_smtp_predata.
6009 26. If the domain in control=submission is set empty, Exim assumes that the
6010 authenticated id is a complete email address when it generates From: or
6011 Sender: header lines.
6013 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6014 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6015 chown and chgrp in /bin and hostname in /usr/bin.
6017 28. Exim was keeping the "process log" file open after each use, just as it
6018 does for the main log. This opens the possibility of it remaining open for
6019 long periods when the USR1 signal hits a daemon. Occasional processlog
6020 errors were reported, that could have been caused by this. Anyway, it seems
6021 much more sensible not to leave this file open at all, so that is what now
6024 29. The long-running daemon process does not normally write to the log once it
6025 has entered its main loop, and it closes the log before doing so. This is
6026 so that log files can straightforwardly be renamed and moved. However,
6027 there are a couple of unusual error situations where the daemon does write
6028 log entries, and I had neglected to close the log afterwards.
6030 30. The text of an SMTP error response that was received during a remote
6031 delivery was being truncated at 512 bytes. This is too short for some of
6032 the long messages that one sometimes sees. I've increased the limit to
6035 31. It is now possible to make retry rules that apply only when a message has a
6036 specific sender, in particular, an empty sender.
6038 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6039 it possible to be selective about when SMTP synchronization is enforced.
6041 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6043 32. Implemented hosts_connection_nolog.
6045 33. Added an ACL for QUIT.
6047 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6050 35. Added mailbox_size and mailbox_filecount to appendfile.
6052 36. Added control = no_multiline_responses to ACLs.
6054 37. There was a bug in the logic of the code that waits for the clock to tick
6055 in the case where the clock went backwards by a substantial amount such
6056 that the microsecond fraction of "now" was more than the microsecond
6057 fraction of "then" (but the whole seconds number was less).
6059 38. Added support for the libradius Radius client library this is found on
6060 FreeBSD (previously only the radiusclient library was supported).
6066 1. When certain lookups returned multiple values in the form name=value, the
6067 quoting of the values was not always being done properly. Specifically:
6068 (a) If the value started with a double quote, but contained no whitespace,
6070 (b) If the value contained whitespace other than a space character (i.e.
6071 tabs or newlines or carriage returns) it was not quoted.
6072 This fix has been applied to the mysql and pgsql lookups by writing a
6073 separate quoting function and calling it from the lookup code. The fix
6074 should probably also be applied to nisplus, ibase and oracle lookups, but
6075 since I cannot test any of those, I have not disturbed their existing code.
6077 2. A hit in the callout cache for a specific address caused a log line with no
6078 reason for rejecting RCPT. Now it says "Previous (cached) callout
6079 verification failure".
6081 3. There was an off-by-one bug in the queryprogram router. An over-long
6082 return line was truncated at 256 instead of 255 characters, thereby
6083 overflowing its buffer with the terminating zero. As well as fixing this, I
6084 have increased the buffer size to 1024 (and made a note to document this).
6086 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6087 when Exim is waiting for an SMTP response from a remote server, Exim
6088 restarts its select() call on the socket, thereby resetting its timeout.
6089 This is not a problem when such interrupts are rare. Somebody set up a cron
6090 job to run exiwhat every 2 minutes, which is less than the normal select()
6091 timeout (5 or 10 minutes). This meant that the select() timeout never
6092 kicked in because it was always reset. I have fixed this by comparing the
6093 time when an interrupt arrives with the time at the start of the first call
6094 to select(). If more time than the timeout has elapsed, the interrupt is
6095 treated as a timeout.
6097 5. Some internal re-factoring in preparation for the addition of Sieve
6098 extensions (by MH). In particular, the "personal" test is moved to a
6099 separate function, and given an option for scanning Cc: and Bcc: (which is
6100 not set for Exim filters).
6102 6. When Exim created an email address using the login of the caller as the
6103 local part (e.g. when creating a From: or Sender: header line), it was not
6104 quoting the local part when it contained special characters such as @.
6106 7. Installed new OpenBSD configuration files.
6108 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6109 try to make them clearer.
6111 9. Callout options, other than the timeout value, were being ignored when
6112 verifying sender addresses in header lines. For example, when using
6114 verify = header_sender/callout=no_cache
6116 the cache was (incorrectly) being used.
6118 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6119 only the Cygwin environment.
6121 11. When return_path_on_delivery was set as a log selector, if different remote
6122 addresses in the same message used different return paths and parallel
6123 remote delivery occurred, the wrong values would sometimes be logged.
6124 (Whenever a remote delivery process finished, the return path value from
6125 the most recently started remote delivery process was logged.)
6127 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6128 lines when AUTH and/or TLS are in use. This is the "received protocol"
6129 field. Exim used to use "asmtp" for authenticated SMTP, without any
6130 indication (in the protocol name) for TLS use. Now it follows the RFC and
6131 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6132 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6133 names appear in log lines as well as in Received: header lines.
6135 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6136 extensions, and comparison tests, and to fix some bugs.
6138 14. Changes to the "personal" filter test:
6140 (1) The test was buggy in that it was just doing the equivalent of
6141 "contains" tests on header lines. For example, if a user's address was
6142 anne@some.where, the "personal" test would incorrectly be true for
6144 To: susanne@some.where
6146 This test is now done by extracting each address from the header in turn,
6147 and checking the entire address. Other tests that are part of "personal"
6148 are now done using regular expressions (for example, to check local parts
6149 of addresses in From: header lines).
6151 (2) The list of non-personal local parts in From: addresses has been
6152 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6153 taken from the Sieve specification recommendations.
6155 (3) If the message contains any header line starting with "List-" it is
6156 treated as non-personal.
6158 (4) The test for "circular" in the Subject: header line has been removed
6159 because it now seems ill-conceived.
6161 15. Minor typos in src/EDITME comments corrected.
6163 16. Installed latest exipick from John Jetmore.
6165 17. If headers_add on a router specified a text string that was too long for
6166 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6167 of string_sprintf() is now avoided.
6169 18. $message_body_size was not set (it was always zero) when running the DATA
6170 ACL and the local_scan() function.
6172 19. For the "mail" command in an Exim filter, no default was being set for
6173 the once_repeat time, causing a random time value to be used if "once" was
6174 specified. (If the value happened to be <= 0, no repeat happened.) The
6175 default is now 0s, meaning "never repeat". The "vacation" command was OK
6176 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6177 (I found it when inspecting the code).
6179 20. There is now an overall timeout for performing a callout verification. It
6180 defaults to 4 times the callout timeout, which applies to individual SMTP
6181 commands during the callout. The overall timeout applies when there is more
6182 than one host that can be tried. The timeout is checked before trying the
6183 next host. This prevents very long delays if there are a large number of
6184 hosts and all are timing out (e.g. when the network connections are timing
6185 out). The value of the overall timeout can be changed by specifying an
6186 additional sub-option for "callout", called "maxwait". For example:
6188 verify = sender/callout=5s,maxwait=20s
6190 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6191 to the end before writing, but this should make it even safer).
6193 22. Exim was forgetting that it had advertised PIPELINING for the second and
6194 subsequent messages on an SMTP connection. It was also not resetting its
6195 memory on STARTTLS and an internal HELO.
6197 23. When Exim logs an SMTP synchronization error within a session, it now
6198 records whether PIPELINING has been advertised or not.
6200 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6201 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6202 rather than long int.
6204 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6206 26. Added the never_mail option to autoreply.
6212 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6213 crash if the getsockname() call failed; this can happen if a connection is
6214 closed very soon after it is established. The problem was simply in the
6215 order in which certain operations were done, causing Exim to try to write
6216 to the SMTP stream before it had set up the file descriptor. The bug has
6217 been fixed by making things happen in the correct order.
6223 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6224 before the connection was closed, thus losing the rejection response.
6226 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6227 some early Solaris releases, but causes trouble in current releases where
6228 socklen_t is defined.
6230 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6233 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6236 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6237 that releasing the top of it at the end releases what was used for sub-
6238 expansions (unless the block got too big). However, discard this block if
6239 the first thing is a variable or header, so that we can use its block when
6240 it is dynamic (useful for very large $message_headers, for example).
6242 6. Lookups now cache *every* query, not just the most recent. A new, separate
6243 store pool is used for this. It can be recovered when all lookup caches are
6244 flushed. Lookups now release memory at the end of their result strings.
6245 This has involved some general refactoring of the lookup sources.
6247 7. Some code has been added to the store_xxx() functions to reduce the amount
6248 of flapping under certain conditions.
6250 8. log_incoming_interface used to affect only the <= reception log lines. Now
6251 it causes the local interface and port to be added to several more SMTP log
6252 lines, for example "SMTP connection from", and rejection lines.
6254 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6256 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6258 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6259 could be overwritten at the end of the current message (or the start of a
6260 new message if it was set in a HELO ACL). The value is now preserved for
6261 the duration of the SMTP connection.
6263 12. If a transport had a headers_rewrite setting, and a matching header line
6264 contained an unqualified address, that address was qualified, even if it
6265 did not match any rewriting rules. The underlying bug was that the values
6266 of the flags that permit the existence of unqualified sender and recipient
6267 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6268 non-local messages, and by -bnq for local messages) were not being
6269 preserved with the message after it was received.
6271 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6272 "next input=" as part of the text comprising the host identity instead of
6273 the correct text. The code was using the same buffer for two different
6274 strings. However, depending on which order the printing function evaluated
6275 its arguments, the bug did not always show up. Under Linux, for example, my
6276 test suite worked just fine.
6278 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6279 doesn't work with some older versions of Perl. It has been changed to "my",
6280 which in any case is probably the better facility to use.
6282 15. A really picky compiler found some instances of statements for creating
6283 error messages that either had too many or two few arguments for the format
6286 16. The size of the buffer for calls to the DNS resolver has been increased
6287 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6288 for addresses that have a lot of PTR records. This alleviates a problem; it
6289 does not fully solve it.
6291 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6292 buffer now truncates the list and logs the incident, which is the same
6293 action as happens when Exim is looking up a host name and its aliases.
6294 Previously in this situation something unpredictable would happen;
6295 sometimes it was "internal error: store_reset failed".
6297 18. If a server dropped the connection unexpectedly when an Exim client was
6298 using GnuTLS and trying to read a response, the client delivery process
6299 crashed while trying to generate an error log message.
6301 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6302 string, for example:
6304 warn message = H1: something\nH2: something
6306 the text was added as a single header line from Exim's point of view
6307 though it ended up OK in the delivered message. However, searching for the
6308 second and subsequent header lines using $h_h2: did not work. This has been
6309 fixed. Similarly, if a system filter added multiple headers in this way,
6310 the routers could not see them.
6312 20. Expanded the error message when iplsearch is called with an invalid key to
6313 suggest using net-iplsearch in a host list.
6315 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6316 ACLs are no longer actually imposed (and a message to that effect is
6319 22. If a "gecos" field in a passwd entry contained escaped characters, in
6320 particular, if it contained a \" sequence, Exim got it wrong when building
6321 a From: or a Sender: header from that name. A second bug also caused
6322 incorrect handling when an unquoted " was present following a character
6323 that needed quoting.
6325 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6326 was not being matched caselessly.
6328 24. Arranged for all hyphens in the exim.8 source to be escaped with
6331 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6332 properly. Recipient callout cache records were still being keyed to include
6333 the sender, even when use_sender was set false. This led to far more
6334 callouts that were necessary. The sender is no longer included in the key
6335 when use_sender is false.
6337 26. Added "control = submission" modifier to ACLs.
6339 27. Added the ${base62d: operator to decode base 62 numbers.
6341 28. dnsdb lookups can now access SRV records.
6343 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6344 the configuration file.
6346 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6347 (-v) mode. This makes the output for a verbose queue run more intelligible.
6349 31. Added a use_postmaster feature to recipient callouts.
6351 32. Added the $body_zerocount variable, containing the number of binary zero
6352 bytes in the message body.
6354 33. The time of last modification of the "new" subdirectory is now used as the
6355 "mailbox time last read" when there is a quota error for a maildir
6358 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6360 35. Added +ignore_unknown as a special item in host lists.
6362 36. Code for decoding IPv6 addresses in host lists is now included, even if
6363 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6364 address was recognized as an IP address, but was then not correctly decoded
6365 into binary, causing unexpected and incorrect effects when compared with
6372 1. Very minor rewording of debugging text in manualroute to say "list of
6373 hosts" instead of "hostlist".
6375 2. If verify=header_syntax was set, and a header line with an unqualified
6376 address (no domain) and a large number of spaces between the end of the
6377 name and the colon was received, the reception process suffered a buffer
6378 overflow, and (when I tested it) crashed. This was caused by some obsolete
6379 code that should have been removed. The fix is to remove it!
6381 3. When running in the test harness, delay a bit after writing a bounce
6382 message to get a bit more predictability in the log output.
6384 4. Added a call to search_tidyup() just before forking a reception process. In
6385 theory, someone could use a lookup in the expansion of smtp_accept_max_
6386 per_host which, without the tidyup, could leave open a database connection.
6388 5. Added the variables $recipient_data and $sender_data which get set from a
6389 lookup success in an ACL "recipients" or "senders" condition, or a router
6390 "senders" option, similar to $domain_data and $local_part_data.
6392 6. Moved the writing of debug_print from before to after the "senders" test
6395 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6396 problems for message scanning, either using a data ACL, or using
6397 local_scan() because the Received: header was not generated till after they
6398 were called (in order to set the time as the time of reception completion).
6399 I have revised the way this works. The header is now generated after the
6400 body is received, but before the ACL or local_scan() are called. After they
6401 are run, the timestamp in the header is updated.
6407 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6408 before starting a queue runner without re-exec. This happened only when
6409 deliver_drop_privilege was set or when the Exim user was set to root. The
6410 effect of the bug was that timeouts during subsequent deliveries caused
6411 crashes instead of being properly handled. The handler is now left at its
6412 default (and expected) setting.
6414 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6415 message, again when deliver_drop_privilege is set or Exim is run as root.
6416 The bug described in (1) was not present in this case, but the tidying up
6417 of the other signals was missing. I have made the two cases consistent.
6419 3. The ignore_target_hosts setting on a manualroute router was being ignored
6420 for hosts that were looked up using the /MX notation.
6422 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6425 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6426 operated on the sender address. After changing the $sender_address to <>
6427 for the sender address verify, Exim was re-instated it as the original
6428 (before rewriting) address, but remembering that it had rewritten it, so it
6429 wasn't rewriting it again. This bug also had the effect of breaking the
6430 sender address verification caching when the sender address was rewritten.
6432 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6433 This has been changed so that if the ip literal address matches
6434 ignore_target_hosts, the router declines.
6436 7. Added expansion conditions match_domain, match_address, and match_local_
6437 part (NOT match_host).
6439 8. The placeholder for the Received: header didn't have a length field set.
6441 9. Added code to Exim itself and to exim_lock to test for a specific race
6442 condition that could lead to file corruption when using MBX delivery. The
6443 issue is with the lockfile that is created in /tmp. If this file is removed
6444 after a process has opened it but before that process has acquired a lock,
6445 there is the potential for a second process to recreate the file and also
6446 acquire a lock. This could lead to two Exim processes writing to the file
6447 at the same time. The added code performs the same test as UW imapd; it
6448 checks after acquiring the lock that its file descriptor still refers to
6449 the same named file.
6451 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6452 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6455 11. Added the smtp_active_hostname option. If used, this will typically be made
6456 to depend on the incoming interface address. Because $interface_address is
6457 not set up until the daemon has forked a reception process, error responses
6458 that can happen earlier (such as "too many connections") no longer contain
6461 12. If an expansion in a condition on a "warn" statement fails because a lookup
6462 defers, the "warn" statement is abandoned, and the next ACL statement is
6463 processed. Previously this caused the whole ACL to be aborted.
6465 13. Added the iplsearch lookup type.
6467 14. Added ident_timeout as a log selector.
6469 15. Added tls_certificate_verified as a log selector.
6471 16. Added a global option tls_require_ciphers (compare the smtp transport
6472 option of the same name). This controls incoming TLS connections.
6474 17. I finally figured out how to make tls_require_ciphers do a similar thing
6475 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6476 before starting the TLS session.
6478 18. Tabs are now shown as \t in -bP output.
6480 19. If the log selector return_path_on_delivery was set, Exim crashed when
6481 bouncing a message because it had too many Received: header lines.
6483 20. If two routers both had headers_remove settings, and the first one included
6484 a superfluous trailing colon, the final name in the first list and the
6485 first name in the second list were incorrectly joined into one item (with a
6486 colon in the middle).
6492 1. Added -C and -D options to the exinext utility, mainly to make it easier
6493 to include in the automated testing, but these could be helpful when
6494 multiple configurations are in use.
6496 2. The exinext utility was not formatting the output nicely when there was
6497 an alternate port involved in the retry record key, nor when there was a
6498 message id as well (for retries that were specific to a specific message
6499 and a specific host). It was also confused by IPv6 addresses, because of
6500 the additional colons they contain. I have fixed the IPv4 problem, and
6501 patched it up to do a reasonable job for IPv6.
6503 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6504 delivery, the log line now contains "pipelined" if PIPELINING was used.
6506 4. An SMTP transport process used to panic and die if the bind() call to set
6507 an explicit outgoing interface failed. This has been changed; it is now
6508 treated in the same way as a connect() failure.
6510 5. A reference to $sender_host_name in the part of a conditional expansion
6511 that was being skipped was still causing a DNS lookup. This no longer
6514 6. The def: expansion condition was not recognizing references to header lines
6515 that used bh_ and bheader_.
6517 7. Added the _cache feature to named lists.
6519 8. The code for checking quota_filecount in the appendfile transport was
6520 allowing one more file than it should have been.
6522 9. For compatibility with Sendmail, the command line option
6530 and sets the incoming protocol and host name (for trusted callers). The
6531 host name and its colon can be omitted when only the protocol is to be set.
6532 Note the Exim already has two private options, -pd and -ps, that refer to
6533 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6534 "s", but I don't think that's a major issue.
6536 10. A number of refactoring changes to the code, none of which should affect
6539 (a) The number of logging options was getting close to filling up the
6540 32-bit word that was used as a bit map. I have split them into two classes:
6541 those that are passed in the argument to log_write(), and those that are
6542 only ever tested independently outside of that function. These are now in
6543 separate 32-bit words, so there is plenty of room for expansion again.
6544 There is no change in the user interface or the logging behaviour.
6546 (b) When building, for example, log lines, the code previously used a
6547 macro that called string_cat() twice, in order to add two strings. This is
6548 not really sufficiently general. Furthermore, there was one instance where
6549 it was actually wrong because one of the argument was used twice, and in
6550 one call a function was used. (As it happened, calling the function twice
6551 did not affect the overall behaviour.) The macro has been replaced by a
6552 function that can join an arbitrary number of extra strings onto a growing
6555 (c) The code for expansion conditions now uses a table and a binary chop
6556 instead of a serial search (which was left over from when there were very
6557 few conditions). Also, it now recognizes conditions like "pam" even when
6558 the relevant support is not compiled in: a suitably worded error message is
6559 given if an attempt is made to use such a condition.
6561 11. Added ${time_interval:xxxxx}.
6563 12. A bug was causing one of the ddress fields not to be passed back correctly
6564 from remote delivery subprocesses. The field in question was not being
6565 subsequently used, so this caused to problems in practice.
6567 13. Added new log selectors queue_time and deliver_time.
6569 14. Might have fixed a bug in maildirsizefile handling that threw up
6570 "unexpected character" debug warnings, and recalculated the data
6571 unnecessarily. In any case, I expanded the warning message to give more
6574 15. Added the message "Restricted characters in address" to the statements in
6575 the default ACL that block characters like @ and % in local parts.
6577 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6578 Three changes have been made:
6580 (a) There was a serious bug; a negative response to MAIL caused the whole
6581 recipient domain to be cached as invalid, thereby blocking all messages
6582 to all local parts at the same domain, from all senders. This bug has
6583 been fixed. The domain is no longer cached after a negative response to
6584 MAIL if the sender used is not empty.
6586 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6589 (c) A new callout option, "use_sender" has been added for people who want
6590 the modified behaviour.
6596 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6599 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6600 indeed breaks things for older releases.
6602 3. Added additional logging to the case where there is a problem reading data
6603 from a filter that is running in a subprocess using a pipe, in order to
6604 try to track down a specific problem.
6606 4. Testing facility fudge: when running in the test harness and attempting
6607 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6608 getting "No route to host". Convert this to a timeout.
6610 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6613 6. Some OS don't have socklen_t but use size_t instead. This affects the
6614 fifth argument of getsockopt() amongst other things. This is now
6615 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6616 can be set for individual OS. I have set it for SunOS5, OSF1, and
6617 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6618 some earlier ones do not.
6620 7. Change 4.30/15 was not doing the test caselessly.
6622 8. The standard form for an IPv6 address literal was being rejected by address
6623 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6624 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6625 this, as well as the form without the "IPv6" on the front (but only when
6626 address literals are enabled, of course).
6628 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6630 10. Exim crashed if a message with an empty sender address specified by -f
6631 encountered a router with an errors_to setting. This could be provoked only
6632 by a command such as
6636 where an empty string was supplied; "<>" did not hit this bug.
6638 11. Installed PCRE release 4.5.
6640 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6641 remained set. It is now erased.
6643 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6644 times from message ids (which are base 36 rather than the normal 62).
6646 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6647 were being counted as actual protocol errors, and logged if the log
6648 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6649 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6650 and DATA following a set of rejected RCPTs do not count as protocol errors.
6651 In other words, Exim assumes they were pipelined, though this may not
6652 actually be the case. Of course, in all cases the client gets an
6653 appropriate error code.
6655 15. If a lookup fails in an ACL condition, a message about the failure may
6656 be available; it is used if testing the ACL cannot continue, because most
6657 such messages specify what the cause of the deferral is. However, some
6658 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6659 that caused an old message to be retained and used if a later statement
6660 caused a defer, replacing the real cause of the deferral.
6662 16. If an IP address had so many PTR records that the DNS lookup buffer
6663 was not large enough to hold them, Exim could crash while trying to process
6664 the truncated data. It now detects and logs this case.
6666 17. Further to 4.21/58, another change has been made: if (and only if) the
6667 first line of a message (the first header line) ends with CRLF, a bare LF
6668 in a subsequent header line has a space inserted after it, so as not to
6669 terminate the header.
6671 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6672 unnecessarily, used atoi() instead of strtol(), and didn't check the
6673 termination when getting file sizes from file names by regex.
6675 19. Completely re-implemented the support for maildirsize files, in the light
6676 of a number of problems with the previous contributed implementation
6677 (4.30/29). In particular:
6679 . If the quota is zero, the maildirsize file is maintained, but no quota is
6682 . If the maildir directory does not exist, it is created before any attempt
6683 to write a maildirsize file.
6685 . The quota value in the file is just a cache; if the quota is changed in
6686 the transport, the new value overrides.
6688 . A regular expression is available for excluding directories from the
6691 20. The autoreply transport checks the characters in options that define the
6692 message's headers; it allows continued headers, but it was checking with
6693 isspace() after an embedded newline instead of explicitly looking for a
6696 21. If all the "regular" hosts to which an address was routed had passed their
6697 expiry times, and had not reached their retry times, the address was
6698 bounced, even if fallback hosts were defined. Now Exim should go on to try
6701 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6702 equivalent code in the SMTP transport. Some hosts send humungous responses
6703 to HELO/EHLO, more than 1024 it seems.
6705 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6706 gives compiler warnings in some environments. I've now done it "properly",
6709 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6710 (because of problems with the built-in one) was declared to return uschar *
6711 instead of char *, causing compiler failure.
6713 25. Fixed a file descriptor leak when processing alias/forward files.
6715 26. Fixed a minor format string issue in dbfn.c.
6717 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6719 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6720 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6723 29. When a sender address is verified, it is cached, to save repeating the test
6724 when there is more than one recipient in a message. However, when the
6725 verification involves a callout, it is possible for different callout
6726 options to be set for different recipients. It is too complicated to keep
6727 track of this in the cache, so now Exim always runs a verification when a
6728 callout is required, relying on the callout cache for the optimization.
6729 The overhead is duplication of the address routing, but this should not be
6732 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6733 to be verified with callout=postmaster, and the main callout worked but the
6734 postmaster check failed, the verification correctly failed. However, if a
6735 subsequent RCPT command asked for sender verification *without* the
6736 postmaster check, incorrect caching caused this verification also to fail,
6739 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6740 it was not caching the DNS options (qualify_single, search_parents) that
6741 were used when the lookup failed. A subsequent lookup with different
6742 options therefore always gave the same answer, though there were cases
6743 where it should not have. (Example: a "domains = !$mx_any" option on a
6744 dnslookup router: the "domains" option is always processed without any
6745 widening, but the router might have qualify_single set.) Now Exim uses the
6746 cached value only when the same options are set.
6748 32. Added John Jetmore's "exipick" utility to the distribution.
6750 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6751 than a timeout (e.g. a certificate is required, and is not provided), an
6752 Exim server now closes the connection immediately. Previously it waited for
6753 the client to close - but if the client is SSL, it seems that they each
6754 wait for each other, leading to a delay before one of them times out.
6756 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6757 maintained 0.8.x compatibility because I don't think many are using it, and
6758 it is clearly obsolete.
6760 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6763 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6764 client certificate was expired. A simple patch fixes this, though I don't
6765 understand the full logic of why the verify callback is called multiple
6768 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6769 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6770 which causes problems with some clients (such as the Certicom SSL Plus
6771 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6772 disables the coutermeasure allowing Eudora to connect."
6774 38. Exim was not checking that a write() to a log file succeeded. This could
6775 lead to Bad Things if a log got too big, in particular if it hit a file
6776 size limit. Exim now panics and dies if it cannot write to a log file, just
6777 as it does if it cannot open a log file.
6779 39. Modified OS/Makefile-Linux so that it now contains
6781 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6783 The two -D definitions ensure that Exim is compiled with large file
6784 support, which makes it possible to handle log files that are bigger than
6787 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6788 instance) a domain was checked against a named list that involved a lookup,
6789 causing $domain_data to be set, then another domain was checked against the
6790 same list, then the first domain was re-checked, the value of $domain_data
6791 after the final check could be wrong. In particular, if the second check
6792 failed, it could be set empty. This bug probably also applied to
6795 41. The strip_trailing_dot option was not being applied to the address given
6796 with the -f command-line option.
6798 42. The code for reading a message's header from the spool was incrementing
6799 $received_count, but never initializing it. This meant that the value was
6800 incorrect (doubled) while delivering a message in the same process in which
6801 it was received. In the most common configuration of Exim, this never
6802 happens - a fresh exec is done - but it can happen when
6803 deliver_drop_privilege is set.
6805 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6806 it now includes up to 150 characters of the unexpected data in the log
6809 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6810 and building data strings. The size of both of these buffers was 10 000
6811 bytes - far larger than anybody would *ever* want, thought I. Needless to
6812 say, somebody hit the limit. I have increased the maximum line length to
6813 20 000 and the maximum data length of concatenated lines to 100 000. I have
6814 also fixed two bugs, because there was no checking on these buffers. Tsk,
6815 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6816 buffer is too small.
6818 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6819 lsearch lookups. Now it does.
6821 46. When parsing a route_list item in a manualroute router, a fixed-length
6822 buffer was used for the list of hosts. I made this 1024 bytes long,
6823 thinking that nobody would ever have a list of hosts that long. Wrong.
6824 Somebody had a whole pile of complicated expansion conditions, and the
6825 string was silently truncated, leading to an expansion error. It turns out
6826 that it is easier to change to an unlimited length (owing to other changes
6827 that have happened since this code was originally written) than to build
6828 structure for giving a limitation error. The length of the item that
6829 expands into the list of hosts is now unlimited.
6831 47. The lsearch lookup could not handle data where the length of text line was
6832 more than 4095 characters. Such lines were truncated, leading to shortened
6833 data being returned. It should now handle lines of any length.
6835 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6836 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6839 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6840 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6841 no longer happens in comment lines. A list of replacements is now placed
6842 at the head of all of the source files, except those whose only change is
6843 to replace PERL_COMMAND in the very first #! line.
6845 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6846 messages on the queue, with a bottom-up merge sort, using code contributed
6847 by Michael Haardt. This should make operations like -bp somewhat faster on
6848 large queues. It won't affect queue runners, except when queue_run_in_order
6851 51. Installed eximstats 1.31 in the distribution.
6853 52. Added support for SRV lookups to the dnslookup router.
6855 53. If an ACL referred to $message_body or $message_body_end, the value was not
6856 reset for any messages that followed in the same SMTP session.
6858 54. The store-handling optimization for building very long strings was not
6859 differentiating between the different store pools. I don't think this
6860 actually made any difference in practice, but I've tidied it.
6862 55. While running the routers to verify a sender address, $sender_address
6863 was still set to the sender address. This is wrong, because when routing to
6864 send a bounce to the sender, it would be empty. Therefore, I have changed
6865 it so that, while verifying a sender address, $sender_address is set to <>.
6866 (There is no change to what happens when verifying a recipient address.)
6868 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6869 target A or AAAA records (if not already returned) without resetting the
6870 qualify_single or search_parents options of the DNS resolver. These are
6871 inappropriate in this case because the targets of MX and SRV records must
6872 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6873 target that, when qualified, matched something in the local domain. These
6874 two options are now turned off when doing these lookups.
6876 57. It seems that at least some releases of Reiserfs (which does not have the
6877 concept of a fixed number of inodes) returns zero and not -1 for the
6878 number of available inodes. This interacted badly with check_spool_inodes,
6879 which assumed that -1 was the "no such thing" setting. What I have done is
6880 to check that the total number of inodes is greater than zero before doing
6881 the test of how many are available.
6883 58. When a "warn" ACL statement has a log_message modifier, the message is
6884 remembered, and not repeated. This is to avoid a lot of repetition when a
6885 message has many recipients that cause the same warning to be written.
6886 However, Exim was preserving the list of already written lines for an
6887 entire SMTP session, which doesn't seem right. The memory is now reset if a
6888 new message is started.
6890 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6891 debugging output unless log_rewrite was also set.
6893 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6894 of (int)(handle) when we know that handle contains (void *)(-1).
6896 61. The Exim daemon panic-logs an error return when it closes the incoming
6897 connection. However "connection reset by peer" seems to be common, and
6898 isn't really an error worthy of noting specially, so that particular error
6901 62. When Exim is trying to find all the local interfaces, it used to panic and
6902 die if the ioctl to get the interface flags failed. However, it seems that
6903 on at least one OS (Solaris 9) it is possible to have an interface that is
6904 included in the list of interfaces, but for which you get a failure error
6905 for this call. This happens when the interface is not "plumbed" into a
6906 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6907 failure of the "get flags" call assumes that the interface is down.
6909 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6910 makes life easier for people who are doing arithmetic on fields extracted
6911 from dates, where you often get leading zeros that should not be
6912 interpreted as octal.
6914 64. Added qualify_domain to the redirect router, to override the global
6917 65. If a pathologically long header line contained very many addresses (the
6918 report of this problem mentioned 10 000) and each of them was rewritten,
6919 Exim could use up a very large amount of memory. (It kept on making new
6920 copies of the header line as it rewrote, and never released the old ones.)
6921 At the expense of a bit more processing, the header rewriting function has
6922 been changed so that it no longer eats memory in this way.
6924 66. The generation of the Received: header has been moved from the time that a
6925 message starts to be received, to the time that it finishes. The timestamp
6926 in the Received: header should now be very close to that of the <= log
6927 line. There are two side-effects of this change:
6929 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6930 logged header lines no longer include the local Received: line, because
6931 it has not yet been created. The same applies to a copy of the message
6932 that is returned to a non-SMTP sender when a message is rejected.
6934 (b) When a filter file is tested using -bf, no additional Received: header
6935 is added to the test message. After some thought, I decided that this
6938 This change does not affect the value of $received_for. It is still set
6939 after address rewriting, but before local_scan() is called.
6941 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6943 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6944 gave an unhelpful panic error message, and a defer error. I have managed to
6945 change this behaviour so that it now rejects any supplied certificate,
6946 which seems right, as the list of acceptable certificates is empty.
6948 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6949 gave an unhelpful defer error. I have not managed to make this reject any
6950 supplied certificates, but the error message it gives is "no certificate
6951 supplied", which is not helpful.
6953 70. exigrep's output now also includes lines that are not associated with any
6954 message, but which match the given pattern. Implemented by a patch from
6955 Martin Sluka, which also tidied up the Perl a bit.
6957 71. Recipient callout verification, like sender verification, was using <> in
6958 the MAIL FROM command. This isn't really the right thing, since the actual
6959 sender may affect whether the remote host accepts the recipient or not. I
6960 have changed it to use the actual sender in the callout; this means that
6961 the cache record is now keyed on a recipient/sender pair, not just the
6962 recipient address. There doesn't seem to be a real danger of callout loops,
6963 since a callout by the remote host to check the sender would use <>.
6964 [SEE ABOVE: changed after hitting problems.]
6966 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6967 temporary errors. However, in the case of such a code being given after
6968 the end of a data transmission (i.e. after ".") Exim was failing to write
6969 a retry record for the message. (Yes, there was some broken host that was
6970 actually sending 8xx at this point.)
6972 73. An unknown lookup type in a host list could cause Exim to panic-die when
6973 the list was checked. (An example that provoked this was putting <; in the
6974 middle of a list instead of at the start.) If this happened during a DATA
6975 ACL check, a -D file could be left lying around. This kind of configuration
6976 error no longer causes Exim to die; instead it causes a defer error. The
6977 incident is still logged to the main and panic logs.
6979 74. Buglet left over from Exim 3 conversion. The message "too many messages
6980 in one connection" was written to the rejectlog but not the mainlog, except
6981 when address rewriting (yes!) was being logged.
6983 75. Added write_rejectlog option.
6985 76. When a system filter was run not as root (that is, when system_filter_user
6986 was set), the values of the $n variables were not being returned to the
6987 main process; thus, they were not subsequently available in the $sn
6990 77. Added +return_path_on_delivery log selector.
6992 78. A connection timeout was being treated differently from recipients deferred
6993 when testing hosts_max_try with a message that was older than the host's
6994 retry timeout. (The host should not be counted, thus allowing all hosts to
6995 be tried at least once before bouncing.) This may have been the cause of an
6996 occasionally reported bug whereby a message would remain on the queue
6997 longer than the retry timeout, but would be bounced if a delivery was
6998 forced. I say "may" because I never totally pinned down the problem;
6999 setting up timeout/retry tests is difficult. See also the next item.
7001 79. The ultimate address timeout was not being applied to errors that involved
7002 a combination of host plus message (for example, a timeout on a MAIL
7003 command). When an address resolved to a number of possible hosts, and they
7004 were not all tried for each delivery (e.g. because of hosts_max_try), a
7005 message could remain on the queue longer than the retry timeout.
7007 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7010 81. Fixed an obscure SMTP outgoing bug which required at least the following
7011 conditions: (a) there was another message waiting for the same server;
7012 (b) the server returned 5xx to all RCPT commands in the first message so
7013 that the message was not completed; (c) the server dropped the connection
7014 or gave a negative response to the RSET that Exim sends to abort the
7015 transaction. The observed case was a dropped connection after DATA that had
7016 been sent in pipelining mode. That is, the server had advertised PIPELINING
7017 but was not implementing it correctly. The effect of the bug was incorrect
7018 behaviour, such as trying another host, and this could lead to a crash.
7024 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7025 and daemon.c were passed as pointers to ints; they should have been
7026 pointers to socklen_t variables (which are typically unsigned ints).
7028 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7031 3. Fixed a really odd bug that affected only the testing scheme; patching a
7032 certain fixed string in the binary changed the value of another string that
7033 happened to be identical to the end of the original first string.
7035 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7036 name", it returns that address as the IP address. On some operating
7037 systems (e.g. Solaris), it also passes back the IP address string as the
7038 "host name". However, on others (e.g. Linux), it passes back an empty
7039 string. Exim wasn't checking for this, and was changing the host name to an
7040 empty string, assuming it had been canonicalized.
7042 5. Although rare, it is permitted to have more than one PTR record for a given
7043 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7044 all the names associated with an address, because they do in Solaris.
7045 However, it seems that they do not in Linux for data that comes from the
7046 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7047 I found this out when I moved to a new Linux workstation and tried to run
7048 the Exim test suite.
7050 To get round this problem I have changed the code so that it now does its
7051 own call to the DNS to look up PTR records when searching for a host name.
7052 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7053 addresses that are only in /etc/hosts are still found.
7055 This behaviour is, however, controlled by an option called host_lookup_
7056 order, which defaults to "bydns:byaddr". If people want to use the other
7057 order, or indeed, just use one or the other means of lookup, they can
7058 specify it in this variable.
7060 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7061 some operating systems, this comes back from gethostbyaddr() as an empty
7062 string, and this is what Exim used to test for. However, it seems that in
7063 other systems, "." is yielded. Exim now tests for this case too.
7065 7. The values of check_spool_space and check_log_space are now held internally
7066 as a number of kilobytes instead of an absolute number of bytes. If a
7067 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7068 kilobyte. This means that much larger values can be stored.
7070 8. Exim monitor: an attempt to get the action menu when not actually pointing
7071 at a message produces an empty menu entitled "No message selected". This
7072 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7073 no entries in it ("Shell widget menu has zero width and/or height"). So I
7074 have added a single, blank menu entry in this case.
7076 9. Added ${quote_local_part.
7078 10. MIME decoding is now applied to the contents of Subject: header lines when
7081 11. Now that a reference to $sender_host_address automatically causes a reverse
7082 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7083 host lookup before query-style lookups in lists that might use this
7084 variable. This has therefore been abolished, and the "net-" prefix is no
7085 longer necessary for query-style lookups.
7087 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7088 have been a typo for LFLAGS, so it has been changed.
7090 13. The install script calls Exim with "-C /dev/null" in order to find the
7091 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7092 to be output. However, since Exim outputs its version number before the
7093 error, it didn't break the script. It just looked ugly. I fixed this by
7094 always allowing "-C /dev/null" if the caller is root.
7096 14. Ignore overlarge ACL variable number when reading spool file - insurance
7097 against a later release with more variables having written the file.
7099 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7100 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7101 this, as well as the form without the "IPv6" on the front.
7103 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7104 OS/Makefile-Darwin file.
7106 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7107 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7109 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7110 variable. However, some OS use ints for inodes. I've added cast to long int
7111 to get rid of the compiler warning.
7113 19. I had forgotten to lock out "/../" in configuration file names when
7114 ALT_CONFIG_PREFIX was set.
7116 20. Routers used for verification do not need to specify transports. However,
7117 if such a router generated a host list, and callout was configured, Exim
7118 crashed, because it could not find a port number from the (non-existent)
7119 transport. It now assumes port 25 in this circumstance.
7121 21. Added the -t option to exigrep.
7123 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7124 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7125 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7126 them is set, LOOKUP_LSEARCH is forced.
7128 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7130 24. Added sender and host information to the "rejected by local_scan()" log
7131 line; previously there was no indication of these.
7133 25. Added .include_if_exists.
7135 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7136 Linux. It turns out that not all file systems support this. Apparently some
7137 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7138 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7139 not supported on the file descriptor, is now ignored when Exim is trying to
7140 sync a directory. This applies only to Linux.
7142 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7144 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7145 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7148 29. Added support for maildirsize files from supplied patch (modified a bit).
7150 30. The use of :fail: followed by an empty string could lead Exim to respond to
7151 sender verification failures with (e.g.):
7153 550 Verification failed for <xxx>
7154 550 Sender verify failed
7156 where the first response line was missing the '-' that indicates it is not
7157 the final line of the response.
7159 31. The loop for finding the name of the user that called Exim had a hardwired
7160 limit of 10; it now uses the value of finduser_retries, which is used for
7161 all other user lookups.
7163 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7166 33. Exim was neglecting to zero errno before one call of strtol() when
7167 expanding a string and expecting an integer value. On some systems this
7168 resulted in spurious "integer overflow" errors. Also, it was casting the
7169 result into an int without checking.
7171 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7172 did not work. The code looks as if it has *never* worked, though it appears
7173 to have been documented since at least release 1.62. I have made it work.
7175 35. The "timeout_DNS" error in retry rules, also documented since at least
7176 1.62, also never worked. As it isn't clear exactly what this means, and
7177 clearly it isn't a major issue, I have abolished the feature by treating it
7178 as "timeout", and writing a warning to the main and panic logs.
7180 36. The display of retry rules for -brt wasn't always showing the error code
7183 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7184 timeout_connect_A, timeout_connect_MX.
7186 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7187 to the empty sender.
7189 39. The daemon was not analysing the content of -oX till after it had closed
7190 stderr and disconnected from the controlling terminal. This meant that any
7191 syntax errors were only noted on the panic log, and the return code from
7192 the command was 0. By re-arranging the code a little, I've made the
7193 decoding happen first, so such errors now appear on stderr, and the return
7194 code is 1. However, the actual setting up of the sockets still happens in
7195 the disconnected process, so errors there are still only recorded on the
7198 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7199 connections (as happens on some IP stacks) was logged at start up time as
7200 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7201 it from "IPv6 and IPv4", which means that two separate sockets are being
7204 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7205 says whether AF_INET or AF_INET6 was passed as an argument.
7207 42. Exiwhat output was messed up when time zones were included in log
7210 43. Exiwhat now gives more information about the daemon's listening ports,
7211 and whether -tls-on-connect was used.
7213 44. The "port" option of the smtp transport is now expanded.
7215 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7216 silently ignored. Now an error message is written to the main and panic
7219 46. There's a new ACL modifier called "logwrite" which writes to a log file
7220 as soon as it is encountered.
7222 47. Added $local_user_uid and $local_user_gid at routing time.
7224 48. Exim crashed when trying to verify a sender address that was being
7227 49. Exim was recognizing only a space character after ".include". It now also
7228 recognizes a tab character.
7230 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7231 extracting the relevant information from the specification. The man page no
7232 longer contains scrambled data for the -d option, and I've added a section
7233 at the front about calling Exim under different names.
7235 51. Added "extra_headers" argument to the "mail" command in filter files.
7237 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7240 53. Installed eximstats 1.29.
7242 54. Added transport_filter_timeout as a generic transport option.
7244 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7245 Cc: header lines. This was required by RFC 822, but it not required by RFC
7248 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7249 incoming messages that did not have them. Now it does so only if the
7250 message originates locally, that is, if there is no associated remote host
7251 address. When Resent- header lines are present, this applies to the Resent-
7252 lines rather than the non-Resent- lines.
7254 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7255 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7257 58. Messages for configuration errors now include the name of the main
7258 configuration file - useful now that there may be more than one file in a
7259 list (.included file names were always shown).
7261 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7262 for those rare installations that do not start the daemon as root or run it
7263 setuid root. I've cut out the call to initgroups() if the daemon is not
7266 60. The Exim user and group can now be bound into the binary as text strings
7267 that are looked up at the start of Exim's processing.
7269 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7271 62. Added $mailstore_basename variable.
7273 63. Installed patch to sieve.c from Michael Haardt.
7275 64. When Exim failed to open the panic log after failing to open the main log,
7276 the original message it was trying to log was written to stderr and debug
7277 output, but if they were not available (the usual case in production), it
7278 was lost. Now it is written to syslog before the two lines that record the
7279 failures to open the logs.
7281 65. Users' Exim filters run in subprocesses under the user's uid. It is
7282 possible for a "deliver" command or an alias in a "personal" command to
7283 provoke an address rewrite. If logging of address rewriting is configured,
7284 this fails because the process is not running as root or exim. There may be
7285 a better way of dealing with this, but for the moment (because 4.30 needs
7286 to be released), I have disabled address rewrite logging when running a
7287 filter in a non-root, non-exim process.
7293 1. The buildconfig auxiliary program wasn't quoting the value set for
7294 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7295 not defined. This bug was masked in 4.22 by the effect that was fixed in
7298 2. Some messages that were rejected after a message id was allocated were
7299 shown as "incomplete" by exigrep. It no longer does this for messages that
7300 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7302 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7303 have allow_domain_literals set, the ID did not get logged in the <= line.
7304 Domain literals are now always recognized in Message-ID: header lines.
7306 4. The first argument for a ${extract expansion item is the key name or field
7307 number. Leading and trailing spaces in this item were not being ignored,
7308 causing some misleading effects.
7310 5. When deliver_drop_privilege was set, single queue runner processes started
7311 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7312 same command in the process it spins off) were not dropping privilege.
7314 6. When the daemon running as "exim" started a queue runner, it always
7315 re-executed Exim in the spun-off process. This is a waste of effort when
7316 deliver_drop_privilege is set. The new process now just calls the
7317 queue-runner function directly.
7323 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7326 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7327 ignored. Though the use of -oP was forcing the writing of a pid file, it
7328 was always written to the default place.
7330 3. If the message "no IP address found for host xxxx" is generated during
7331 incoming verification, it is now followed by identification of the incoming
7332 connection (so you can more easily find what provoked it).
7334 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7336 5. Added some features to "harden" Exim a bit more against certain attacks:
7338 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7339 be put in Local/Makefile. This is like the never_users runtime option,
7340 but it cannot be overridden. The default setting is "root".
7342 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7343 prefix string with which any file named in a -C command line option
7346 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7347 is retained for -C and -D only if the caller of Exim is root. Without
7348 it, the exim user may also use -C and -D and retain privilege.
7350 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7351 command line option is disabled.
7353 6. Macro names set by the -D option must start with an upper case letter, just
7354 like macro names defined in the configuration file.
7356 7. Added "dereference=" facility to LDAP.
7358 8. Two instances of the typo "uknown" in the source files are fixed.
7360 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7361 the Configure-Makefile script screwed up while processing it.
7363 10. Incorporated PCRE 4.4.
7365 11. The SMTP synchronization check was not operating right at the start of an
7366 SMTP session. For example, it could not catch a HELO sent before the client
7367 waited for the greeting. There is now a check for outstanding input at the
7368 point when the greeting is written. Because of the duplex, asynchronous
7369 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7370 way, but not yet received, when the check is performed.
7372 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7373 on TCP/IP sockets, because this apparently causes some broken clients to
7376 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7377 unchanged) from the Cygwin maintainer.
7379 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7380 was supported instead of testing for mbx. Effectively a typo.
7382 15. The spa authenticator server code was not checking that the input it
7383 received was valid base64.
7385 16. The debug output line for the "set" modifier in ACLs was not showing the
7386 name of the variable that was being set.
7388 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7390 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7391 lookup on its own. Something else was needed to trigger the lookup. For
7392 example, a match in host_lookup or the need for a host name in a host list.
7393 Now, if $sender_host_name is referenced and the host name has not yet been
7394 looked up, a lookup is performed. If the lookup fails, the variable remains
7395 empty, and $host_lookup_failed is set to "1".
7397 19. Added "eqi" as a case-independent comparison operator.
7399 20. The saslauthd authentication condition could segfault if neither service
7400 nor realm was specified.
7402 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7403 error message that was logged was misleading, and incoming SMTP
7404 connections were dropped. The message is now more accurate, and temporary
7405 errors are given to SMTP connections.
7407 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7408 (except RSET) with a 421 error, until QUIT is received. However, it was
7409 failing to send a response to QUIT.
7411 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7412 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7413 ACL is now run after the helo_try_verify_hosts code.
7415 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7416 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7417 case-independent, so other case variants are also recognized). Apparently
7418 some systems use these upper case variants.
7420 25. If more than two messages were waiting for the same host, and a transport
7421 filter was specified for the transport, Exim sent two messages over the
7422 same TCP/IP connection, and then failed with "socket operation on non-
7423 socket" when it tried to send the third.
7425 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7427 27. The extern definition of crypt16() in expand.c was not being excluded when
7428 the OS had its own crypt16() function.
7430 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7431 as a preferred synonym for return_size_limit, both as an option and as an
7434 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7436 30. Changed the default configuration ACL to relax the local part checking rule
7437 for addresses that are not in any local domains. For these addresses,
7438 slashes and pipe symbols are allowed within local parts, but the sequence
7439 /../ is explicitly forbidden.
7441 31. SPA server authentication was not clearing the challenge buffer before
7444 32. log_message in a "warn" ACL statement was writing to the reject log as
7445 well as to the main log, which contradicts the documentation and doesn't
7446 seem right (because no rejection is happening). So I have stopped it.
7448 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7449 However, I am unable to do any testing of this.
7451 34. Fixed an infelicity in the appendfile transport. When checking directories
7452 for a mailbox, to see if any needed to be created, it was accidentally
7453 using path names with one or more superfluous leading slashes; tracing
7454 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7456 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7457 contents are added to the log line that is written for every discarded
7458 recipient. (Previously a log_message setting was ignored.)
7460 36. The ${quote: operator now quotes the string if it is empty.
7462 37. The install script runs exim in order to find its version number. If for
7463 some reason other than non-existence or emptiness, which it checks, it
7464 could not run './exim', it was installing it with an empty version number,
7465 i.e. as "exim-". This error state is now caught, and the installation is
7468 38. An argument was missing from the function that creates an error message
7469 when Exim fails to connect to the socket for saslauthd authentication.
7470 This could cause Exim to crash, or give a corrupted message.
7472 39. Added isip, isip4, and isip6 to ${if conditions.
7474 40. The ACL variables $acl_xx are now saved with the message, and can be
7475 accessed later in routers, transports, and filters.
7477 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7478 strings in the file are not string-expanded.
7480 42. If a MAIL command specified a SIZE value that was too large to fit into an
7481 int variable, the check against message_size_limit failed. Such values are
7482 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7483 day this will have to be increased, but I don't think I want to be around
7484 when emails are that large.
7491 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7492 iconv() is not standard in FreeBSD.
7494 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7495 IPv6 enabled. The observed symptom was a segmentation fault on return from
7496 the function os_common_find_running_interfaces() in src/os.c.
7498 3. In the check_special_case() function in daemon.c I had used "errno" as an
7499 argument name, which causes warnings on some systems. This was basically a
7500 typo, since it was named "eno" in the comments!
7502 4. The code that waits for the clock to tick (at a resolution of some fraction
7503 of a second) so as to ensure message-id uniqueness was always waiting for
7504 at least one whole tick, when it could have waited for less. [This is
7505 almost certainly not relevant at current processor speeds, where it is
7506 unlikely to ever wait at all. But we try to future-proof.]
7508 5. The function that sleeps for a time interval that includes fractions of a
7509 second contained a race. It did not block SIGALRM between setting the
7510 timer, and suspending (a couple of lines later). If the interval was short
7511 and the sigsuspend() was delayed until after it had expired, the suspension
7512 never ended. On busy systems this could lead to processes getting stuck for
7515 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7516 process, before it forks any delivery processes. The open lookup caching
7517 mechanism meant that the open file or database connection was passed into
7518 the delivery process. The problem was that delivery processes always tidy
7519 up cached lookup data. This could cause a problem for the next delivery
7520 process started by the queue runner, because the external queue runner
7521 process does not know about the closure. So the next delivery process
7522 still has data in the lookup cache. In the case of a file lookup, there was
7523 no problem because closing a file descriptor in a subprocess doesn't affect
7524 the parent. However, if the lookup was caching a connection to a database,
7525 the connection was closed, and the second delivery process was likely to
7526 see errors such as "PGSQL: query failed: server closed the connection
7527 unexpectedly". The problem has been fixed by closing all cached lookups
7528 in a queue runner before running a delivery process.
7530 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7531 seem to have the "const" qualifier which it has on other OS. I've
7534 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7535 *of the same type* (client or server) with the same public name that an
7536 error should be diagnosed.
7538 9. When Exim looked up a host name for an IP address, but failed to find the
7539 original IP address when looking up the host name (a safety check), it
7540 output the message "<ip address> does not match any IP for NULL", which was
7541 confusing, to say the least. The bug was that the host name should have
7542 appeared instead of "NULL".
7544 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7545 user that is built into the binary, and the -C or -D options is used, root
7546 privilege is dropped before the configuration file is read. In addition,
7547 logging is switched to stderr instead of the normal log files. If the
7548 configuration then re-defines the Exim user, the unprivileged environment
7549 is probably not what is expected, so Exim logs a panic warning message (but
7552 However, if deliver_drop_privilege is set, the unprivileged state may well
7553 be exactly what is intended, so the warning has been cut out in that case,
7554 and Exim is allowed to try to write to its normal log files.
7560 1. smtp_return_error_details was not giving details for temporary sender
7561 or receiver verification errors.
7563 2. Diagnose a configuration error if two authenticators have the same public
7566 3. Exim used not to create the message log file for a message until the first
7567 delivery attempt. This could be confusing when incoming messages were held
7568 for policy or load reasons. The message log file is now created at the time
7569 the message is received, and an initial "Received" line is written to it.
7571 4. The automatically generated man page for command line options had a minor
7572 bug that caused no ill effects; however, a more serious problem was that
7573 the procedure for building the man page automatically didn't always
7574 operate. Consequently, release 4.20 contains an out-of-date version. This
7575 shouldn't happen again.
7577 5. When building Exim with embedded Perl support, the script that builds the
7578 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7579 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7581 6. The freeze_tell option was not being used for messages that were frozen on
7582 arrival, either by an ACL or by local_scan().
7584 7. Added the smtp_incomplete_transaction log selector.
7586 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7587 was accepting AUTH without a new EHLO.
7589 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7590 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7593 10. Logging of TCP/IP connections (when configured) now happens in the main
7594 daemon process instead of the child process, so that the TCP/IP connection
7595 count is more accurate (but it can never be perfect).
7597 11. The use of "drop" in a nested ACL was not being handled correctly in the
7598 outer ACL. Now, if condition failure induced by the nested "drop" causes
7599 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7600 or "require"), the connection is dropped.
7602 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7603 that yield "discard" can now be used with an "accept" or a "discard" verb,
7604 but an error is generated for any others (because I can't see a useful way
7605 to define what should happen).
7607 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7608 are now processed in the same way as lines in the Exim configuration file.
7609 In particular, continuation lines are supported.
7611 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7613 15. Added -ti meaning -t -i.
7615 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7616 domains, and warn by logging if others are found.
7618 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7619 returned from GIFCONF when Exim is trying to find the list of interfaces on
7620 a host. The code in os.c has been modified to copy each ifreq to an aligned
7621 structure in all cases.
7623 Also, in some cases, the returned ifreq's were being copied to a 'struct
7624 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7625 means the last couple of bytes of an IPv6 address could be chopped if the
7626 ifreq contained only a normal sockaddr (14 bytes storage).
7628 18. Named domain lists were not supported in the hosts_treat_as_local option.
7629 An entry such as +xxxx was not recognized, and was treated as a literal
7632 19. Ensure that header lines added by a DATA ACL are included in the reject log
7633 if the ACL subsequently rejects the message.
7635 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7636 MD5 (which is deprecated).
7638 21. When testing a filter file using -bf, Exim was writing a message when it
7639 took the sender from a "From " line in the message, but it was not doing so
7640 when it took $return_path from a Return-Path: header line. It now does.
7642 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7643 with a valid header line field name (a series of printing characters
7644 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7646 23. Changed "disc" in the source to "disk" to conform to the documentation and
7647 the book and for uniformity.
7649 24. Ignore Sendmail's -Ooption=value command line item.
7651 25. When execve() failed while trying to run a command in a pipe transport,
7652 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7653 could be confused with a return value of 69 from the command itself. This
7654 has been changed to 127, the value the shell returns if it is asked to run
7655 a non-existent command. The wording for the related log line suggests a
7656 non-existent command as the problem.
7658 26. If received_header_text expands to an empty string, do not add a Received:
7659 header line to the message. (Well, it adds a token one on the spool, but
7660 marks it "old" so that it doesn't get used or transmitted.)
7662 27. Installed eximstats 1.28 (addition of -nt option).
7664 28. There was no check for failure on the call to getsockname() in the daemon
7665 code. This can fail if there is a shortage of resources on the system, with
7666 ENOMEM, for example. A temporary error is now given on failure.
7668 29. Contrary to the C standard, it seems that in some environments, the
7669 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7670 program. Exim now does this explicitly; it affects the formatting of
7671 timestamps using strftime().
7673 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7674 complaints. I've now initialized all the variables, to avoid this.
7676 32. Header lines added by a system filter were not being "seen" during
7677 transport-time rewrites.
7679 33. The info_callback() function passed to OpenSSL is set up with type void
7680 (*)(SSL *, int, int), as described somewhere. However, when calling the
7681 function (actually a macro) that sets it up, the type void(*)() is
7682 expected. I've put in a cast to prevent warnings from picky compilers.
7684 34. If a DNS black list lookup found a CNAME record, but there were no A
7685 records associated with the domain it pointed at, Exim crashed.
7687 35. If a DNS black list lookup returned more than one A record, Exim ignored
7688 all but the first. It now scans all returned addresses if a particular IP
7689 value is being sought. In this situation, the contents of the
7690 $dnslist_value variable are a list of all the addresses, separated by a
7693 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7694 to accept a host name and all its aliases if the forward lookup for any of
7695 them yielded the IP address of the incoming connection. Now it accepts only
7696 those names whose forward lookup yields the correct IP address. Any other
7697 names are discarded. This closes a loophole whereby a rogue DNS
7698 administrator could create reverse DNS records to break through a
7699 wildcarded host restriction in an ACL.
7701 37. If a user filter or a system filter that ran in a subprocess used any of
7702 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7703 the wrong values were passed to the pipe command ($thisaddress had the
7704 value of $0, $0 had the value of $1, etc). This bug was introduced by
7705 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7707 38. Improved the line breaking for long SMTP error messages from ACLs.
7708 Previously, if there was no break point between 40 and 75 characters, Exim
7709 left the rest of the message alone. Two changes have been made: (a) I've
7710 reduced the minimum length to 35 characters; (b) if it can't find a break
7711 point between 35 and 75 characters, it looks ahead and uses the first one
7712 that it finds. This may give the occasional overlong line, but at least the
7713 remaining text gets split now.
7715 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7716 file descriptors might be low, and that setting 1000 would always raise it.
7717 It turns out that in some environments, the limit is already over 1000 and
7718 that lowering it causes trouble. So now Exim takes care not to decrease it.
7720 40. When delivering a message, the value of $return_path is set to $sender_
7721 address at the start of routing (routers may change the value). By an
7722 oversight, this default was not being set up when an address was tested by
7723 -bt or -bv, which affected the outcome if any router or filter referred to
7726 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7727 log only when "message" or "log_message" are set. However, if one of the
7728 conditions was an address verification, or a call to a nested ACL, the
7729 messages generated by the underlying test were being passed through. This
7730 no longer happens. The underlying message is available in $acl_verify_
7731 message for both "message" and "log_message" expansions, so it can be
7732 passed through if needed.
7734 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7735 new expansion $bh_ to give the encoded byte string without charset
7736 translation. Translation happens only if iconv() is available; HAVE_ICONV
7737 indicates this at build time. HEADERS_CHARSET gives the charset to
7738 translate to; headers_charset can change it in the configuration, and
7739 "headers charset" can change it in an individual filter file.
7741 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7742 that creates RFC 2047 encoded "words" labels them as that charset instead
7743 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7744 expansion operator; (ii) when Exim creates a From: line for a local
7745 message; (iii) when a header line is rewritten to include a "phrase" part.
7747 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7748 buggy, causing it to skip the first lines of messages whose message ID
7749 ended in 'D'. This would not have bitten before Exim release 4.14, because
7750 message IDs were unlikely to end in 'D' before then. The effect was to have
7751 incorrect size information for certain domains.
7753 45. #include "config.h" was missing at the start of the crypt16.c module. This
7754 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7757 46. If there was a timeout during a "random" callout check, Exim treated it as
7758 a failure of the random address, and carried on sending RSET and the real
7759 address. If the delay was just some slowness somewhere, the response to the
7760 original RCPT would be taken as a response to RSET and so on, causing
7761 mayhem of various kinds.
7763 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7764 when I implemented it. It didn't allow for the fact that some option values
7765 may legitimately be negative (e.g. size_addition), and it didn't even do
7766 the right test for positive values.
7768 48. Domain names in DNS records are case-independent. Exim always looks them up
7769 in lower case. Some resolvers return domain names in exactly the case they
7770 appear in the zone file, that is, they may contain uppercase letters. Not
7771 all resolvers do this - some return always lower case. Exim was treating a
7772 change of case by a resolver as a change of domain, similar to a widening
7773 of a domain abbreviation. This triggered its re-routing code and so it was
7774 trying to route what was effectively the same domain again. This normally
7775 caused routing to fail (because the router wouldn't handle the domain
7776 twice). Now Exim checks for this case specially, and just changes the
7777 casing of the domain that it ultimately uses when it transmits the message
7780 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7783 50. If a filter generated a file delivery with a non-absolute name (possible if
7784 no home directory exists for the router), the forbid_file option was not
7787 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7788 the existing equality matching.
7790 52. Exim was using ints instead of ino_t variables in some places where it was
7791 dealing with inode numbers.
7793 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7794 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7795 called TMPDIR, and if it finds it is different, it changes its value.
7797 54. The smtp_printf() function is now made available to local_scan() so
7798 additional output lines can be written before returning. There is also an
7799 smtp_fflush() function to enable the detection of a dropped connection.
7800 The variables smtp_input and smtp_batched_input are exported to
7803 55. Changed the default runtime configuration: the message "Unknown user"
7804 has been removed from the ACL, and instead placed on the localuser router,
7805 using the cannot_route_message feature. This means that any verification
7806 failures that generate their own messages won't get overridden. Similarly,
7807 the "Unrouteable address" message that was in the ACL for unverifiable
7808 relay addresses has also been removed.
7810 56. Added hosts_avoid_esmtp to the smtp transport.
7812 57. The exicyclog script was not checking for the esoteric option
7813 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7814 will work only if exicyclog is run under the appropriate euid.
7816 58. Following a discussion on the list, the rules by which Exim recognises line
7817 endings on incoming messages have been changed. The -dropcr and drop_cr
7818 options are now no-ops, retained only for backwards compatibility. The
7819 following line terminators are recognized: LF CRLF CR. However, special
7820 processing applies to CR:
7822 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7823 nor a local message in the state where . is a terminator.
7825 (ii) If a bare CR is encountered in a header line, an extra space is added
7826 after the line terminator so as not to end the header. The reasoning
7827 behind this is that bare CRs in header lines are most likely either
7828 to be mistakes, or people trying to play silly games.
7830 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7831 was being incorrectly given as 18 bytes larger than it should have been.
7832 This is a VOB (very old bug).
7834 60. This may never have affected anything current, but just in case it has:
7835 When the local host is found other than at the start of a list of hosts,
7836 the local host, those with the same MX, and any that follow, are discarded.
7837 When the list in question was part of a longer list of hosts, the following
7838 hosts (not currently being processed) were also being discarded. This no
7839 longer happens. I'm not sure if this situation could ever has previously
7842 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7845 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7846 header. This is something that is recommended in a new Internet Draft, and
7847 is something that is documented as being done by Sendmail. There are two
7848 possible values. For messages generated by the autoreply transport, Exim
7851 Auto-Submitted: auto-replied
7853 whereas for all other generated messages (e.g. bounces) it adds
7855 Auto-Submitted: auto-generated
7857 63. The "personal" condition in filters now includes a test for the
7858 Auto-Submitted: header. If it contains the string "auto-" the message it
7859 not considered personal.
7861 64. Added rcpt_include_affixes as a generic transport option.
7863 65. Added queue_only_override (default true).
7865 66. Added the syslog_duplication option.
7867 67. If what should have been the first header line of a message consisted of
7868 a space followed by a colon, Exim was mis-interpreting it as a header line.
7869 It isn't of course - it is syntactically invalid and should therefore be
7870 treated as the start of the message body. The misbehaviour could have
7871 caused a number of strange effects, including loss of data in subsequent
7872 header lines, and spool format errors.
7874 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7875 client host had authenticated. This control can now be exercised by an ACL
7876 for more flexibility.
7878 69. By default, callouts do not happen when testing with -bh. There is now a
7879 variant, -bhc, which does actually run the callout code, including
7880 consulting and updating the callout cache.
7882 70. Added support for saslauthd authentication, courtesy of Alexander
7885 71. If statvfs() failed on the spool or log directories while checking their
7886 size for availability, Exim confusingly gave the error "space shortage".
7887 Furthermore, in debugging mode it crashed with a floating point exception.
7888 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7889 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7890 serious problem, Exim now writes to the main and panic logs when this
7891 happens, with details of the failure. It then refuses to accept the
7892 incoming message, giving the message "spool directory problem" or "log
7893 directory problem" with a 421 code for SMTP messages.
7895 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7896 0, 1, and 2 exist, because some OS complain for execs without them (see
7897 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7898 descriptors. However, the code omitted to check that the open succeeded,
7899 causing mysterious errors if for some reason the permissions on /dev/null
7900 got screwed. Now Exim writes a message to the main and panic logs, and
7901 bombs out if it can't open /dev/null.
7903 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7904 interact so that it is all more flexible. It is supposed to remain
7905 backwards compatible. Also added extra_local_interfaces.
7907 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7908 to bomb out with an assertion failure - to the client this appears as a
7909 connection drop. This problem occurs in the part of the code that was taken
7910 from the Samba project. Fortunately, the assertion is in a very simple
7911 function, so I have fixed this by reproducing the function inline in the
7912 one place where it is called, and arranging for authentication to fail
7913 instead of killing the process with assert().
7915 75. The SPA client code was not working when the server requested OEM rather
7916 than Unicode encoding.
7918 76. Added code to make require_files with a specific uid setting more usable in
7919 the case where statting the file as root fails - usually a non-root-mounted
7920 NFS file system. When this happens and the failure is EACCES, Exim now
7921 forks a subprocess and does the per-uid checking as the relevant uid.
7923 77. Added process_log_path.
7925 78. If log_file_path was not explicitly set, a setting of check_log_space or
7926 check_log_inodes was ignored.
7928 79. If a space check for the spool or log partitions fails, the incident is now
7929 logged. Of course, in the latter case the data may get lost...
7931 80. Added the %p formatting code to string_format() so that it can be used to
7932 print addresses in debug_print(). Adjusted all the address printing in the
7933 debugging in store.c to use %p rather than %d.
7935 81. There was a concern that a line of code in smtp_in.c could overflow a
7936 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7937 initially expressed, the concern was not well-founded, because trailing
7938 spaces are removed early. However, if the trailing spaces were followed by
7939 a NULL, they did not get removed, so the overflow was possible. Two fixes
7942 (a) I re-wrote the offending code in a cleaner fashion.
7943 (b) If an incoming SMTP command contains a NULL character, it is rejected
7946 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7947 runs initgroups(), so that if the Exim user is in any additional groups,
7948 they will be used during message reception.
7954 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob