1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
6 Since Exim version 4.94
7 -----------------------
9 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
10 as arguments, so an implementation trying to copy these into a local
11 buffer was taking a taint-enforcement trap. Fix by using dynamically
12 created buffers. Similar fix for radius expansion condition.
14 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
15 reasonable, eg. to count headers. Fix by using dynamically created
16 buffers rather than a local. Do similar fixes for ACL actions "dcc",
17 "log_reject_target", "malware" and "spam"; the arguments are expanded
18 so could be handling tainted values.
20 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
21 broken the (no-op) support for this sendmail command. Restore it
22 to doing nothing, silently, and returning good status.
24 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
25 record path was given (or the default used) without a leading directory
26 path, an error occurred on trying to open it. Use the transport's working
29 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
30 transport. Previously it was the name on the DNS A-record; use instead
31 the head of the CNAME chain leading there (if there is one). This seems
32 to align better with RFC 6125.
34 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
35 smtp_accept_max_per_host allocated resources which were not released
36 when the limit was exceeded. This eventually crashed the daemon. Fix
37 by adding a relase action in that path.
39 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
40 expanded; previously using tainted values was rejected. Fix by using
41 dynamically-created buffers.
43 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
44 Previously a macro used one argument twice; when called with the
45 argument as an expression having side-effects, incorrect operation
46 resulted. Use an inlineable function.
48 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
49 held open for a verify callout. Previously this wan not accounted for
50 and a corrupt onward SMTP conversation resulted.
52 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
53 excluded, not matching the documentation.
55 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
56 was given for the sqlite_dbfile a trap resulted.
58 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
59 "name" argument resulted in a trap. There is no reason to disallow such;
60 this was a coding error.
62 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
63 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
64 spent suspended, ignoring the Posix definition. Previously we assumed
65 it did and a constant offset from real time could be used as a correction.
66 Change to using the same clock source for the start-of-message and the
67 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
68 exists, just to get a clock slightly more aligned to reality.
70 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
71 RFC says it is optional some validators care. The missing char was not
72 intended but triggered by a line-wrap alignement. Discovery and fix by
73 Guillaume Outters, hacked on by JH.
75 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
76 name being quoted was tainted a trap would be taken. Fix by using
77 dynamicaly created buffers. The routine could have been called by a
78 rewrite with the "h" flag, by using the "-F" command-line option, or
79 by using a "name=" option on a control=submission ACL modifier.
81 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
82 Previously when a whitespace character was specified it was not inserted
83 after removing the newline.
85 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
86 is_tainted() had an off-by-one error in the overenthusiastic direction.
87 Find and fix by Gavan. Although NetBSD is not a supported platform for
88 4.94 this bug could affect other platforms.
90 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
91 is_tainted() had an off-by-one error in the overenthusiastic direction.
92 Find and fix by Gavan. Although NetBSD is not a supported platform for
93 4.94 this bug could affect other platforms.
94 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
95 Previously when a whitespace character was specified it was not inserted
96 after removing the newline.
98 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
99 the domain part of the recipient address. This overrides any tls_sni
100 option set, which was previously used.
102 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
105 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
106 than one server was defined and depending on the platform memory layout
107 details, an internal consistency trap could be hit while walking the list
110 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
111 files. The value is used by the authresults expansion item. Previously
112 if this was used in a router or transport, a crash could result.
114 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
115 applied. This resulted, if any header-line rewrite rules were configured,
116 in a panic-log trigerrable by sending a message with a long address in
117 a header. Fix by increaing the arbitrary limit to larger than a single
118 (dewrapped) 5322 header line maximum size.
120 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
121 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
122 RFC 6648 which deprecates X- options in protocols as a general practice.
123 Changeover between the implementations is handled by the mechanisms
126 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
127 a remote which uses callout/hold. Previously the local was queued.
129 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
132 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
134 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
135 authenticator client_send option. Previously the next char, after a pair
136 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
137 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
138 documentation. There is still no way to get a leading ^ immediately
139 after a NUL (ie. for the password of a PLAIN method authenticator.
141 JH/39 Bug 2691: fix $local_part_data. When the matching list element
142 referred to a file, bad data was returned. This likely also affected
149 JH/01 Avoid costly startup code when not strictly needed. This reduces time
150 for some exim process initialisations. It does mean that the logging
151 of TLS configuration problems is only done for the daemon startup.
153 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
155 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
156 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
159 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
160 DKIM signing is being done. Previously a transport_filter would always
161 disable CHUNKING, falling back to traditional DATA.
163 JH/05 Regard command-line receipients as tainted.
165 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
167 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
168 PAM library frees one of the arguments given to it, despite the
169 documentation. Therefore a plain malloc must be used.
171 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
172 on-stack buffers were used, resulting in a taint trap when DSN information
173 copied from a received message was written into the buffer.
175 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
176 the ordering of its ARC headers. This caused a crash.
178 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
179 a new record was being constructed with information from the peer, a trap
182 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
183 installation would get error messages from DMARC verify, when it hit the
184 nonexistent file indicated by the default. Distros wanting DMARC enabled
185 should both provide the file and set the option.
186 Also enforce no DMARC verification for command-line sourced messages.
188 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
189 could, depending on the platform, hang at the STARTTLS response.
191 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
192 message on a connection. Previously if one message had ARC headers and
193 the following one did not, a crash could result when adding an
194 Authentication-Results: header.
196 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
197 Exim main code and Exim-related utities. The introduction of taint
198 tracking also did many adjustments to string handling. Since then, eximon
199 frequently terminated with an assert failure.
201 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
202 check for 452 responses. This slightly helps the inefficieny of doing
203 a large alias-expansion into a recipient-limited target. The max_rcpt
204 transport option still applies (and at the current default, will override
205 the new feature). The check is done for either cause of synch, and forces
206 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
207 connection. The new facility is not tunable at this time.
209 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
210 library live data was being used, so the results became garbage. Make
211 copies while it is still usable.
213 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
214 on delivery deferred (==) and failed (**) lines (if a delivery was
215 attemtped). Previously it was only on completion (=>) lines.
217 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
218 for the expansion of the server_scram_iter and server_scram_salt options.
220 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
221 are now specifically given a NO_DATA response without hitting the system
222 resolver. The library goes on to do the now-standard TXT lookup.
223 Use of dnsdb lookups is not affected.
225 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
226 only retrieve the errormessage once. Previously two calls to dlerror()
227 were used, and the second one (for mainlog/paniclog) retrieved null
230 JH/20 Taint checking: disallow use of tainted data for
231 - the appendfile transport file and directory options
232 - the pipe transport command
233 - the autoreply transport file, log and once options
234 - file names used by the redirect router (including filter files)
236 - paths used by single-key lookups
237 Previously this was permitted.
239 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
240 adjusted the size of a major service buffer; this failed because the
241 buffer was in use at the time. Change to a compile-time increase in the
242 buffer size, when this authenticator is compiled into exim.
244 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
245 previous fast-mode was untenable in the face of glibs using mmap to
246 support larger malloc requests.
248 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
249 New values supported, if defined on system where compiled:
250 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
251 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
253 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
254 running a limited number of proceses in parallel, a benefit is gained. The
255 amount varies with the platform hardware and load. The use of the option
256 queue_run_in_order means we cannot do this, as ordering becomes
259 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
260 had introduced a string-copy (for ensuring NUL-termination) which was not
261 appropriate for that case, which can include embedded NUL bytes in the
262 block of data. Investigation showed the copy to actually be needless, the
263 data being length-specified.
265 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
266 done during a receiving connection, and both used TLS, global info was
267 used rather than per-connection info for tracking the state of data
268 queued for transmission. This could result in a connection hang.
270 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
271 Previously, when delivering serveral messages down a single connection
272 only the first would provide a SIZE. This was due to the size information
273 not being properly tracked.
275 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
276 TAI (at 37 seconds currently), pretend to be in UTC for time-related
277 expansion and logging. Previously, spurious values such as a future
278 minute could be seen.
280 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
281 it could crash from a null-deref. This could also affect the
282 ${addresses: } operator and ${readsock } item.
284 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
285 message following a mime one, the variable was not reset.
287 JH/30 When an pipelined-connect fails at the first response, assume incorrect
288 cached capability (perhaps the peer reneged?) and immediately retry in
291 JH/31 Fix spurious detection of timeout while writing to transport filter.
293 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
294 an attempt to copy the string was made before checking it.
296 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
297 taint of the lookup key was maintained; we now regard the presence in the
298 filesystem as sufficient validation.
300 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
301 argument is supplied.
303 JH/35 The dsearch lookup now requires that the directory is an absolute path.
304 Previously this was not checked, and nonempty relative paths made an
305 access under Exim's current working directory.
307 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
308 Previously no event was raised.
310 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
311 parameter supplied by the sender MAIL FROM command. Previously it was
312 ignored, and only the check_spool_space option value for the required
315 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
316 the size of the signing public-key. Previously it was instead giving
317 the size of the signature hash.
319 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
320 the default. See the (new) dkim_verify_min_keysizes option.
322 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
323 and an ACL use a lookup for checking either the local_part or domain,
324 stale data could be accessed. Ensure that variable references are
325 dropped between messages.
327 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
328 by the client was not checked as pointing within response data before
329 being used. A malicious client could thus cause an out-of-bounds read and
330 possibly gain authentication. Fix by adding the check.
332 JH/42 Internationalisation: change the default for downconversion in the smtp
333 transport to be "if needed". Previously it was "as previously set" for
334 the message, which usually meant "if needed" for message-submission but
335 "no" for everything else. However, MTAs have been seen using SMTPUTF8
336 even when the envelope addresses did not need it, resulting in forwarding
337 failures to non-supporting MTAs. A downconvert in such cases will be
338 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
339 transport. The change does mean that addresses needing conversion will
340 be converted when previously a delivery failure would occur.
342 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
343 response was received it would be used unchecked in a fail-DSN, violating
344 standards on line-length limits. Truncate if needed.
346 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
347 doesn't work. (Additionally add a new main config option to configure the
354 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
355 side, to decode a TLS 1.3 packet capture.
357 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
358 Previously the default library behaviour applied, sending two, each in
361 JH/03 Debug output for ACL now gives the config file name and line number for
364 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
366 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
368 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
369 buffer overrun for (non-chunking) other transports.
371 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
372 TLS1.3, means that a server rejecting a client certificate is not visible
373 to the client until the first read of encrypted data (typically the
374 response to EHLO). Add detection for that case and treat it as a failed
375 TLS connection attempt, so that the normal retry-in-clear can work (if
376 suitably configured).
378 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
379 and/or domain. Found and fixed by Jason Betts.
381 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
382 configuration). If a CNAME target was not a wellformed name pattern, a
385 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
386 the OS reports them interleaved with other addresses.
388 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
389 used both for input and for a verify callout, both encrypted, SMTP
390 responses being sent by the server could be lost. This resulted in
391 dropped connections and sometimes bounces generated by a peer sending
394 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
395 string. Previously it was possible to cause undefined behaviour in a
396 library routine (usually a crash). Found by "zerons".
398 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
401 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
402 API was removed, so update to use the newer ones.
404 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
405 any timeout set, is taking a long time. Previously we would hang on to a
406 rotated logfile "forever" if the input was arriving with long gaps
407 (a previous attempt to fix addressed lack, for a long time, of initial
410 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
411 shared (NFS) environment. The length of the tempfile name is now
412 4 + 16 ("hdr.$message_exim_id") which might break on file
413 systems which restrict the file name length to lower values.
416 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
417 shared (NFS) environment.
419 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
420 did for all versions <4.90). Notably -M, -m, --invert, -I may be
423 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
424 on some platforms for bit 31.
426 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
427 to changes apparently associated with TLS1.3 handling some of the APIs
428 previously used were either nonfunctional or inappropriate. Strings
429 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
430 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
431 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
432 This affects log line X= elements, the $tls_{in,out}_cipher variables,
433 and the use of specific cipher names in the encrypted= ACL condition.
435 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
437 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
438 verification result was not updated unless hosts_require_ocsp applied.
440 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
441 queue_list_requires_admin set to false, non-admin users were denied the
444 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
445 directory-of-certs mode. Previously they were advertised despite the
448 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
449 A single TCP connection by a client will now hold a TLS connection open
450 for multiple message deliveries, by default. Previously the default was to
453 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
454 default. If built with the facility, DANE will be used. The facility
455 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
457 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
458 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
459 must be defined and you must still, unless you define DISABLE_TLS, manage
460 the the include-dir and library-file requirements that go with that
461 choice. Non-TLS builds are still supported.
463 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
466 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
467 default. If the platform supports and has the facility enabled, it will
468 be requested on all coneections.
470 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
471 controlled by the build-time option SUPPORT_PIPE_CONNECT.
473 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
475 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
476 success-DSN messages. Previously the From: header was always the default
477 one for these; the option was ignored.
479 JH/28 Fix the timeout on smtp response to apply to the whole response.
480 Previously it was reset for every read, so a teergrubing peer sending
481 single bytes within the time limit could extend the connection for a
482 long time. Credit to Qualsys Security Advisory Team for the discovery.
484 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
485 delivery address, which leaked information of the results of local
486 forwarding. Change to the original envelope recipient address, per
489 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
490 requested. Previously not bounce was generated and a log entry of
491 error ignored was made.
493 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
495 JH/32 Introduce a general tainting mechanism for values read from the input
496 channel, and values derived from them. Refuse to expand any tainted
497 values, to catch one form of exploit.
499 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
500 was unused and the unexpanded text used for the test. Found and
501 fixed by Ruben Jenster.
503 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
504 an attempt to use a TLS library read routine dereffed a nul pointer,
507 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
508 them in our smtp response.
510 JH/36 Have the general router option retry_use_local_part default to true when
511 any of the restrictive preconditions are set (to anything). Previously it
512 was only for check_local user. The change removes one item of manual
513 configuration which is required for proper retries when a remote router
514 handles a subset of addresses for a domain.
516 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
517 link count into consideration.
519 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
520 caused the extension of big_buffer, the following lines were ignored.
522 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
523 accordance with RFC 2308. Previously there was no expiry, so a longlived
524 receive process (eg. due to ACL delays) versus a short SOA value could
527 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
529 JH/39 Promote DMARC support to mainline.
531 JH/40 Bug 2452: Add a References: header to DSNs.
533 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
534 parameters. The relevant library call is documented as "Deprecated: This
535 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
536 3.6.0, DH parameters are negotiated following RFC7919."
538 HS/06 Change the default of dnssec_request_domains to "*"
540 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
541 carried on and emitted a BDAT command, even when PIPELINING was not
544 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
545 buffer was used for the filename, resulting in a trap when tainted
546 arguments (eg. $domain) were used.
548 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
549 recommended to avoid a possible server-load attack. The feature can be
550 re-enabled via the openssl_options main cofiguration option.
552 JH/45 local_scan API: documented the current smtp_printf() call. This changed
553 for version 4.90 - adding a "more data" boolean to the arguments.
554 Bumped the ABI version number also, this having been missed previously;
555 release versions 4.90 to 4.92.3 inclusive were effectively broken in
556 respect of usage of smtp_printf() by either local_scan code or libraries
557 accessed via the ${dlfunc } expansion item. Both will need coding
558 adjustment for any calls to smtp_printf() to match the new function
559 signature; a FALSE value for the new argument is always safe.
561 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
562 the file-offset (which the Linux syscall does, and exim expects); this
563 resulted in an indefinite loop.
565 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
566 to do ARC verification. The Authentication-Results: header line added
567 by the configuration then had no ARC item.
573 JH/01 Remove code calling the customisable local_scan function, unless a new
574 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
576 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
577 non-signal-safe functions being used.
579 JH/03 Bug 2269: When presented with a received message having a stupidly large
580 number of DKIM-Signature headers, disable DKIM verification to avoid
581 a resource-consumption attack. The limit is set at twenty.
583 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
584 report of oldest_pass in ${authres } in consequence, and separate out
585 some descriptions of reasons for verification fail.
587 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
588 files in the spool were present and unlocked. A queue-runner could spot
589 them, resulting in a duplicate delivery. Fix that by doing the unlock
590 after the unlink. Investigation by Tim Stewart. Take the opportunity to
591 add more error-checking on spoolfile handling while that code is being
594 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
595 No known attacks, no CVE, this is defensive hardening.
597 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
598 a queue-runner could start a delivery while other operations were ongoing.
599 Cutthrough delivery was a common victim, resulting in duplicate delivery.
600 Found and investigated by Tim Stewart. Fix by using the open message data
601 file handle rather than opening another, and not locally closing it (which
602 releases a lock) for that case, while creating the temporary .eml format
603 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
605 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
606 $sender_verify_failure/$recipient_verify_failure to "random".
608 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
611 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
612 Previously this would segfault.
614 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
617 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
618 like zero, since the resolver should be doing this for us, But we need one
619 as a CNAME but no MX presence gets the CNAME returned; we need to check
620 that doesn't point to an MX to declare it "no MX returned" rather than
621 "error, loop". A new main option is added so the older capability of
622 following some limited number of chain links is maintained.
624 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
626 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
627 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
628 now-deprecated earlier definitions used only specified the range up to TLS
629 1.2 (in the older-version library docs).
631 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
633 JH/15 Rework TLS client-side context management. Stop using a global, and
634 explicitly pass a context around. This enables future use of TLS for
635 connections to service-daemons (eg. malware scanning) while a client smtp
636 connection is using TLS; with cutthrough connections this is quite likely.
638 JH/16 Fix ARC verification to do AS checks in reverse order.
640 JH/17 Support a "tls" option on the ${readsocket } expansion item.
642 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
643 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
644 Previously the "utf8" would be re-prepended for every additional message.
646 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
647 Previously thery were accepted, resulting in issues when attempting to
648 forward messages to a non-supporting MTA.
650 PP/02 Let -n work with printing macros too, not just options.
652 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
653 one parent address was copied, and bogus data was used at delivery-logging
654 time. Either a crash (after delivery) or bogus log data could result.
655 Discovery and analysis by Tim Stewart.
657 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
658 Previously if the string ended mid-character, we did not insert the
659 promised '?' replacement.
661 PP/04 Documentation: current string operators work on bytes, not codepoints.
663 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
664 should pack well giving a smaller memory footprint so better caching and
665 therefore performance. Group the declarations where this can't be done so
666 that the byte-sized flag variables are not interspersed among pointer
667 variables, giving a better chance of good packing by the compiler.
669 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
670 non-null, to avoid issues with sites running BATV. Previously reports were
671 sent with an empty envelope sender so looked like bounces.
673 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
674 The ignore_error flag wasn't being returned from the filter subprocess so
675 was not set for later routers. Investigation and fix by Matthias Kurz.
677 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
678 and a msg:complete for the whole, when a message is manually removed using
679 -Mrm. Developement by Matthias Kurz, hacked on by JH.
681 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
682 a "Gnu special" function, asprintf() in the DB utility binary builds; I
683 hope that is portable enough.
685 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
686 requiring a known-CA anchor certificate; make it now rely entirely on the
687 TLSA as an anchor. Checking the name on the leaf cert against the name
688 on the A-record for the host is still done for TA (but not for EE mode).
690 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
691 included in delivery lines for non-proxied connections, when compiled with
692 SUPPORT_SOCKS and running with proxy logging enabled.
694 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
695 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
696 move the existing event to fire before the normal logging of message
697 failure so that custom logging is bracketed by normal logging.
699 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
700 msg:fail:internal event. Developement by Matthias Kurz.
702 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
703 far too small for todays use of crypto signatures stored there. Go all
704 the way to the max DNS message size of 64kB, even though this might be
705 overmuch for IOT constrained device use.
707 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
708 copy a string over itself. The library routine is documented as not
709 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
711 JH/32 For main options check_spool_space and check_inode_space, where the
712 platform supports 64b integers, support more than the previous 2^31 kB
713 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
714 the previous G, M, k.
716 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
717 $authenticated_fail_id variable on authentication failure. Previously
720 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
721 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
722 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
723 GNUTLS_SEC_PARAM_MEDIUM.
725 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
726 side. Previously we would continue as if no SNI had been received.
728 JH/36 Harden the handling of string-lists. When a list consisted of a sole
729 "<" character, which should be a list-separator specification, we walked
730 off past the nul-terimation.
732 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
733 causes) even when the retry time is not yet met. Previously they were
734 not, meaning that when (say) an account was over-quota and temp-rejecting,
735 and multiple senders' messages were queued, only one sender would get
736 notified on each configured delay_warning cycle.
738 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
740 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
741 cipher-suites, an error can be left on the stack even for a succeeding
742 accept; this results in impossible error messages when a later operation
745 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
746 return error codes indicating retry. Under TLS1.3 this becomes required.
748 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
749 it only wrote the new authenticators, resulting in a lack of tracking of
750 peer changes of ESMTP extensions until the next cache flush.
752 JH/41 Fix the loop reading a message header line to check for integer overflow,
753 and more-often against header_maxsize. Previously a crafted message could
754 induce a crash of the recive process; now the message is cleanly rejected.
756 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
757 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
763 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
764 When redis_servers is set to a list of > 1 element, and the Redis servers
765 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
766 case of MOVED into a DEFER case instead, thus moving the query onto the
767 next server in the list. For a cluster of N elements, all N servers must
768 be defined in redis_servers.
770 GF/02 Catch and remove uninitialized value warning in exiqsumm
771 Check for existence of @ARGV before looking at $ARGV[0]
773 JH/01 Replace the store_release() internal interface with store_newblock(),
774 which internalises the check required to safely use the old one, plus
775 the allocate and data copy operations duplicated in both (!) of the
776 extant use locations.
778 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
779 modifier. This matches the restriction on the commandline.
781 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
782 Previously only the last row was returned.
784 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
785 we assumed that tags in the header were well-formed, and parsed the
786 element content after inspecting only the first char of the tag.
787 Assumptions at that stage could crash the receive process on malformed
790 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
791 While running the DKIM ACL we operate on the Permanent memory pool so that
792 variables created with "set" persist to the DATA ACL. Also (at any time)
793 DNS lookups that fail create cache records using the Permanent pool. But
794 expansions release any allocations made on the current pool - so a dnsdb
795 lookup expansion done in the DKIM ACL releases the memory used for the
796 DNS negative-cache, and bad things result. Solution is to switch to the
797 Main pool for expansions.
798 While we're in that code, add checks on the DNS cache during store_reset,
799 active in the testsuite.
800 Problem spotted, and debugging aided, by Wolfgang Breyha.
802 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
803 When none of the hosts presented to a transport match an already-open
804 connection, close it and proceed with the list. Previously we would
805 queue the message. Spotted by Lena with Yahoo, probably involving
808 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
809 Previously a spurious "250 OK id=" response was appended to the proper
812 JH/08 The "support for" informational output now, which built with Content
813 Scanning support, has a line for the malware scanner interfaces compiled
814 in. Interface can be individually included or not at build time.
816 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
817 by the template makefile "src/EDITME". The "STREAM" support for an older
818 ClamAV interface method is removed.
820 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
821 rows affected is given instead).
823 JH/11 The runtime Berkeley DB library version is now additionally output by
824 "exim -d -bV". Previously only the compile-time version was shown.
826 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
827 SMTP connection. Previously, when one had more receipients than the
828 first, an abortive onward connection was made. Move to full support for
829 multiple onward connections in sequence, handling cutthrough connection
830 for all multi-message initiating connections.
832 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
833 routers. Previously, a multi-recipient message would fail to match the
834 onward-connection opened for the first recipient, and cause its closure.
836 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
837 a timeout on read on a GnuTLS initiating connection, resulting in the
838 initiating connection being dropped. This mattered most when the callout
839 was marked defer_ok. Fix to keep the two timeout-detection methods
842 JH/15 Relax results from ACL control request to enable cutthrough, in
843 unsupported situations, from error to silently (except under debug)
844 ignoring. This covers use with PRDR, frozen messages, queue-only and
847 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
849 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
850 metadata, resulting in a crash in free().
852 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
853 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
854 Broken also in d185889f4, with init system revamp.
856 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
857 Previously we abruptly closed the connection after reading a malware-
858 found indication; now we go on to read the "scan ok" response line,
861 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
862 ACL. Previously, a crash would result.
864 JH/19 Speed up macro lookups during configuration file read, by skipping non-
865 macro text after a replacement (previously it was only once per line) and
866 by skipping builtin macros when searching for an uppercase lead character.
868 JH/20 DANE support moved from Experimental to mainline. The Makefile control
869 for the build is renamed.
871 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
872 was allocated for every new TLS startup, meaning one per message. Fix
873 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
875 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
876 reported the original. Fix to report (as far as possible) the ACL
877 result replacing the original.
879 JH/23 Fix memory leak during multi-message connections using STARTTLS under
880 OpenSSL. Certificate information is loaded for every new TLS startup,
881 and the resources needed to be freed.
883 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
885 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
888 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
889 DATA response info to the (existing) per-recipient response info for
890 the "C=" log element. It can have useful tracking info from the
891 destination system. Patch from Simon Arlott.
893 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
894 length value. Previously this would segfault.
896 HS/02 Support Avast multiline protoocol, this allows passing flags to
897 newer versions of the scanner.
899 JH/28 Ensure that variables possibly set during message acceptance are marked
900 dead before release of memory in the daemon loop. This stops complaints
901 about them when the debug_store option is enabled. Discovered specifically
902 for sender_rate_period, but applies to a whole set of variables.
903 Do the same for the queue-runner and queue-list loops, for variables set
904 from spool message files. Do the same for the SMTP per-message loop, for
905 certain variables indirectly set in ACL operations.
907 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
908 as a multi-recipient message from a mailinglist manager). The coding had
909 an arbitrary cutoff number of characters while checking for more input;
910 enforced by writing a NUL into the buffer. This corrupted long / fast
911 input. The problem was exposed more widely when more pipelineing of SMTP
912 responses was introduced, and one Exim system was feeding another.
913 The symptom is log complaints of SMTP syntax error (NUL chars) on the
914 receiving system, and refused recipients seen by the sending system
915 (propating to people being dropped from mailing lists).
916 Discovered and pinpointed by David Carter.
918 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
919 replaced by the ${authresults } expansion.
921 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
923 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
924 allows proper process termination in container environments.
926 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
927 Previously the "final dot" had a newline after it; ensure it is CR,LF.
929 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
930 and "err_perm", deprecated since 4.83 when the RFC-defined words
931 "temperror" and "permerror" were introduced.
933 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
934 transport-filters or DKIM-signing. The restriction was lost in the
935 consolidation of verify-callout and delivery SMTP handling.
936 Extend the restriction to also cover ARC-signing.
938 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
939 in defer=pass mode supply a 450 to the initiator. Previously the message
942 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
943 tls_require_ciphers is used as before.
945 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
946 "pass_unscanned". Only tmpfails from the scanner are written to
947 the paniclog, as they may require admin intervention (permission
948 denied, license issues). Other scanner errors (like decompression
949 bombs) do not cause a paniclog entry.
951 JH/36 Fix reinitialisation of DKIM logging variable between messages.
952 Previously it was possible to log spurious information in receive log
955 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
956 triggered odd behaviour from Outlook Express clients.
958 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
961 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
962 since the IETF WG has not yet settled on that versus the original
963 "bare" representation.
965 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
966 Previously the millisecond value corrupted the output.
967 Fix also for syslog_pid=no and log_selector +pid, for which the pid
968 corrupted the output.
974 JH/01 Rework error string handling in TLS interface so that the caller in
975 more cases is responsible for logging. This permits library-sourced
976 string to be attached to addresses during delivery, and collapses
977 pairs of long lines into single ones.
979 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
980 during configuration. Wildcards are allowed and expanded.
982 JH/02 Rework error string handling in DKIM to pass more info back to callers.
983 This permits better logging.
985 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
986 do not close it down and have the child transport start it up again on
987 the passed-on TCP connection. Instead, proxy the child (and any
988 subsequent ones) for TLS via a unix-domain socket channel. Logging is
989 affected: the continued delivery log lines do not have any DNSSEC, TLS
990 Certificate or OCSP information. TLS cipher information is still logged.
992 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
993 identical IP addresses on different listening ports. Will also affect
996 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
997 add noisy ifdef guards to special-case this sillyness.
998 Patch from Bernd Kuhls.
1000 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1001 than 255 are no longer allowed.
1003 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1004 Disable the session-cache too, which might reduce our load. Since we
1005 currrectly use a new context for every connection, both as server and
1006 client, there is no benefit for these.
1007 GnuTLS appears to not support tickets server-side by default (we don't
1008 call gnutls_session_ticket_enable_server()) but client side is enabled
1009 by default on recent versions (3.1.3 +) unless the PFS priority string
1012 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1013 <https://reproducible-builds.org/specs/source-date-epoch/>.
1015 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1016 the check for any unsuccessful recipients did not notice the limit, and
1017 erroneously found still-pending ones.
1019 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1020 MSG_MORE. Only in-clear (not on TLS connections).
1022 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1023 transport-filter is involved we can buffer the headers in memory for
1024 creating the signature, and read the spool data file once for the
1025 signature and again for transmission.
1027 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1028 4.77 as the kernel support then wasn't solid, having issues in 64bit
1029 mode. Now, it's been long enough. Add support for FreeBSD also.
1031 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1032 case where the routing stage had gathered several addresses to send to
1033 a host before calling the transport for the first, we previously failed
1034 to close down TLS in the old transport process before passing the TCP
1035 connection to the new process. The new one sent a STARTTLS command
1036 which naturally failed, giving a failed delivery and bloating the retry
1037 database. Investigation and fix prototype from Wolfgang Breyha.
1039 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1040 false-negatives in the check that the sender had not preempted a response
1041 or prompt from Exim (running as a server), due to that code's lack of
1042 awareness of the SMTP input buffering.
1044 PP/04 Add commandline_checks_require_admin option.
1045 Exim drops privileges sanely, various checks such as -be aren't a
1046 security problem, as long as you trust local users with access to their
1047 own account. When invoked by services which pass untrusted data to
1048 Exim, this might be an issue. Set this option in main configuration
1049 AND make fixes to the calling application, such as using `--` to stop
1052 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1053 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1054 responses to those, into a single TLS record each way (this usually means
1055 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1058 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1059 affects you only if you're dancing at the edge of the param size limits.
1060 If you are, and this message makes sense to you, then: raise the
1061 configured limit or use OpenSSL 1.1. Nothing we can do for older
1064 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1065 cmdline element to get the documented default one. Previously it was
1068 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1069 are unsuable for non-message-specific reasons - eg. connection timeout,
1070 banner-time rejection.
1072 JH/16 Fix logging of delivery remote port, when specified by router, under
1075 PP/06 Repair manualroute's ability to take options in any order, even if one
1076 is the name of a transport.
1079 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1081 JH/17 Change the list-building routines interface to use the expanding-string
1082 triplet model, for better allocation and copying behaviour.
1084 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1085 Previously it was constructed the first time a possibly-matching string
1086 was met in the configuration file input during startup; now it is done
1089 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1090 compatible one, to avoid the (poorly documented) possibility of a config
1091 file in the working directory redirecting the DB files, possibly correpting
1092 some existing file. CVE-2017-10140 assigned for BDB.
1094 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1095 cache-hot. Previously, although the result was properly cached, the
1096 initial verify call returned a defer.
1098 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1099 the main verify for receipient in uncached-mode.
1101 JH/22 Retire historical build files to an "unsupported" subdir. These are
1102 defined as "ones for which we have no current evidence of testing".
1104 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1105 if present. Previously it was ignored.
1107 JH/24 Start using specified-initialisers in C structure init coding. This is
1108 a C99 feature (it's 2017, so now considered safe).
1110 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1111 if was a fixed-sized field and bitmask ops via macros; it is now more
1114 PP/07 GitHub PR 56: Apply MariaDB build fix.
1115 Patch provided by Jaroslav Škarvada.
1117 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1118 during Coverity cleanups [4.87 JH/47]
1119 Diagnosis and fix provided by Michael Fischer v. Mollard.
1121 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1122 the right size to place the terminating semicolon on its own folded
1123 line, the header hash was calculated to an incorrect value thanks to
1124 the (relaxed) space the fold became.
1126 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1127 and confused the parent.
1129 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1130 which could crash as a result. This could lead to undeliverable messages.
1132 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1135 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1136 matters on fast-turnover and PID-randomising systems, which were getting
1137 out-of-order delivery.
1139 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1140 a possibly-overlapping copy. The symptom was that "Remote host closed
1141 connection in response to HELO" was logged instead of the actual 4xx
1144 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1145 Previously only that bufferd was discarded, resulting in SYMTP command
1148 JH/32 DKIM: when a message has multiple signatures matching an identity given
1149 in dkim_verify_signers, run the dkim acl once for each. Previously only
1150 one run was done. Bug 2189.
1152 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1153 "panic the current process" to "deliberately defer". The panic log is
1154 still written with the problem list name; the mail and reject logs now
1155 get a temp-reject line for the message that was being handled, saying
1156 something like "domains check lookup or other defer". The SMTP 451
1157 message is still "Temporary local problem".
1159 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1160 A crafted sequence of BDAT commands could result in in-use memory beeing
1161 freed. CVE-2017-16943.
1163 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1164 from SMTP input. Previously it was always done; now only done for DATA
1165 and not BDAT commands. CVE-2017-16944.
1167 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1168 to the message (such as an overlong header line). Previously this was
1169 not done and we did not exit BDAT mode. Followon from the previous item
1170 though a different problem.
1176 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1177 than -2003 did; needs libidn2 in addition to libidn.
1179 JH/02 The path option on a pipe transport is now expanded before use.
1181 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1182 Patch provided by "Björn", documentation fix added too.
1184 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1185 missing a wire-to-host endian conversion.
1187 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1188 close after a BDAT command line could be taken as a following command,
1189 giving a synch failure. Fix by only checking for synch immediately
1190 before acknowledging the chunk.
1192 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1193 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1194 macro. Patches provided by Josh Soref.
1196 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1197 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1198 by IANA as a service extension. However, John Klensin suggests that we
1201 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1202 the dkim code may be unix-mode line endings rather than smtp wire-format
1203 CRLF, so prepend a CR to any bare LF.
1205 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1206 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1208 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1209 the first were themselves being wrongly included in the feed into dkim
1210 processing; with most chunk sizes in use this resulted in an incorrect
1211 body hash calculated value.
1213 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1214 DKIM signature block, for verification. Although advised against by
1215 standards it is specifically not ruled illegal.
1217 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1219 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1220 missing a body hash (the bh= tag).
1222 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1223 It seems that HAProxy sends the Proxy Protocol information in clear and
1224 only then does a TLS startup, so do the same.
1226 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1227 TCP connections (such as for Spamd) unless the daemon successfully set
1228 Fast Open mode on its listening sockets. This fixes breakage seen on
1229 too-old kernels or those not configured for Fast Open, at the cost of
1230 requiring both directions being enabled for TFO, and TFO never being used
1231 by non-daemon-related Exim processes.
1233 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1234 endings, at least on the first header line. Try to canonify any that get
1235 past that check, despite the cost.
1237 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1238 now limited to an arbitrary five deep, while parsing addresses with the
1239 strip_excess_angle_brackets option enabled.
1241 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1242 instead leave the unprompted TLS handshake in socket buffer for the
1243 TLS library to consume.
1245 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1247 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1249 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1250 region in one operation, for speed, and this leaves assigned pointers
1251 dangling. Add checks run only under the testsuite which checks all
1252 variables at a store-reset and panics on a dangling pointer; add code
1253 explicitly nulling out all the variables discovered. Fixes one known
1254 bug: a transport crash, where a dangling pointer for $sending_ip_address
1255 originally assigned in a verify callout, is re-used.
1257 PP/06 Drop '.' from @INC in various Perl scripts.
1259 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1261 PP/08 Reduce a number of compilation warnings under clang; building with
1262 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1263 should be warning-free.
1265 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1267 HS/01 Fix portability problems introduced by PP/08 for platforms where
1268 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1270 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1271 chunk. This allows us to accept broken chunked messages. We need a more
1272 general solution here.
1274 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1275 already-broken messages in the queue.
1277 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1279 JH/19 Fix reference counting bug in routing-generated-address tracking.
1285 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1286 supports it and a size is available (ie. the sending peer gave us one).
1288 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1289 years of being deprecated). The replacements are the ACLs
1290 acl_smtp_mime and acl_not_smtp_mime.
1292 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1293 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1294 or even an in-clear connection were permitted. Now, if the host lookup
1295 was dnssec and dane was requested then the host is only used if the
1296 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1297 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1298 if one fails this test.
1299 This means that a poorly-configured remote DNS will make it incommunicado;
1300 but it protects against a DNS-interception attack on it.
1302 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1303 non-noisy when a race steals the message being considered.
1305 JH/05 If main configuration option tls_certificate is unset, generate a
1306 self-signed certificate for inbound TLS connections.
1308 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1309 in rewrites and routers.
1311 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1312 and logged a warning sing 4.83; now they are a configuration file error.
1314 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1315 (lacking @domain). Apply the same qualification processing as RCPT.
1317 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1319 JH/10 Support ${sha256:} applied to a string (as well as the previous
1322 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1323 a cutthrough deliver is pending, as we always want to make a connection.
1324 This also avoids re-routing the message when later placing the cutthrough
1325 connection after a verify cache hit.
1326 Do not update it with the verify result either.
1328 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1329 when routing results in more than one destination address.
1331 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1332 signing (which inhibits the cutthrough capability). Previously only
1333 the presence of an option was tested; now an expansion evaluating as
1334 empty is permissible (obviously it should depend only on data available
1335 when the cutthrough connection is made).
1337 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1338 the relevant preceding SMTP command did not note the pipelining mode.
1340 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1341 Previously they were not counted.
1343 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1344 as one having no matching records. Previously we deferred the message
1345 that needed the lookup.
1347 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1348 distinguished as "(=".
1350 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1351 for missing MX records. Previously it only worked for missing A records.
1353 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1355 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1356 after the data-go-ahead and data-ack. Patch from Jason Betts.
1358 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1359 even for a "none" policy. Patch from Tony Meyer.
1361 JH/22 Fix continued use of a connection for further deliveries. If a port was
1362 specified by a router, it must also match for the delivery to be
1365 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1366 When one of the recipients of a message was unsuitable for the connection
1367 (has no matching addresses), we lost track of needing to mark it
1368 deferred. As a result mail would be lost.
1370 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1372 JH/25 Decoding ACL controls is now done using a binary search; the source code
1373 takes up less space and should be simpler to maintain. Merge the ACL
1374 condition decode tables also, with similar effect.
1376 JH/26 Fix problem with one_time used on a redirect router which returned the
1377 parent address unchanged. A retry would see the parent address marked as
1378 delivered, so not attempt the (identical) child. As a result mail would
1381 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1382 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1383 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1386 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1387 Main config options check_{log,spool}_{inodes,space} are now
1388 100 inodes, 10MB unless set otherwise in the configuration.
1390 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1391 Previously it only applied to the main-section connection policy
1394 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1396 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1397 by me. Added RFC7919 DH primes as an alternative.
1399 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1400 are not in the system include path.
1402 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1403 GnuTLS, when a session startup failed (eg because the client disconnected)
1404 Exim did stdio operations after fclose. This was exposed by a recent
1405 change which nulled out the file handle after the fclose.
1407 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1408 signed directly by the cert-signing cert, rather than an intermediate
1409 OCSP-signing cert. This is the model used by LetsEncrypt.
1411 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1413 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1414 an incoming connection.
1416 HS/02 Bug 1802: Do not half-close the connection after sending a request
1419 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1420 fallback to "prime256v1".
1422 JH/34 SECURITY: Use proper copy of DATA command in error message.
1423 Could leak key material. Remotely exploitable. CVE-2016-9963.
1429 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1430 and 3.4.4 - once the server is enabled to respond to an OCSP request
1431 it does even when not requested, resulting in a stapling non-aware
1432 client dropping the TLS connection.
1434 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1435 support variable-length bit vectors. No functional change.
1437 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1438 The I= interface field on outgoing lines is now after the H= remote
1439 host field, same as incoming lines. There is a separate
1440 outgoing_interface log selector which allows you to disable the
1443 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1444 If not running log_selector +smtp_connection the mainlog would be held
1445 open indefinitely after a "too many connections" event, including to a
1446 deleted file after a log rotate. Leave the per net connection logging
1447 leaving it open for efficiency as that will be quickly detected by the
1448 check on the next write.
1450 HS/01 Bug 1671: Fix post transport crash.
1451 Processing the wait-<transport> messages could crash the delivery
1452 process if the message IDs didn't exist for some reason. When
1453 using 'split_spool_directory=yes' the construction of the spool
1454 file name failed already, exposing the same netto behaviour.
1456 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1457 mime_regex ACL conditions.
1459 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1460 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1461 to HELO, local diagnostic string.
1463 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1464 log line to debug. Even when configured with a tls authenticator many
1465 client connections are expected to not authenticate in this way, so
1466 an authenticate fail is not an error.
1468 HS/02 Add the Exim version string to the process info. This way exiwhat
1469 gives some more detail about the running daemon.
1471 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1472 matter for fast-change records such as DNSBLs.
1474 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1475 retry record, even if constant. There may be multiple transports with
1476 different interface settings and the retry behaviour needs to be kept
1479 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1481 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1483 JH/10 Bug 840: fix log_defer_output option of pipe transport
1485 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1486 hosts_randomize. This matters a lot when combined with mua_wrapper.
1488 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1489 ${quote_pgsql:<string>} operator.
1491 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1493 JH/14 Change header file name for freeradius-client. Relevant if compiling
1494 with Radius support; from the Gentoo tree and checked under Fedora.
1496 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1498 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1499 option transport_filter, meaning no filtering.
1501 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1503 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1504 defaults to "*" (all hosts). The variable is now available when not built
1505 with TLS, default unset, mainly to enable keeping the testsuite sane.
1506 If a server certificate is not supplied (via tls_certificate) an error is
1507 logged, and clients will find TLS connections fail on startup. Presumably
1508 they will retry in-clear.
1509 Packagers of Exim are strongly encouraged to create a server certificate
1510 at installation time.
1512 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1513 with the $config_file variable.
1515 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1516 in transport context, after the attempt, and per-recipient. The latter type
1517 is per host attempted. The event data is the error message, and the errno
1518 information encodes the lookup type (A vs. MX) used for the (first) host,
1519 and the trailing two digits of the smtp 4xx response.
1521 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1522 to write to mainlog (or rejectlog, paniclog) in the window between file
1523 creation and permissions/ownership being changed. Particularly affects
1524 installations where exicyclog is run as root, rather than exim user;
1525 result is that the running daemon panics and dies.
1527 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1529 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1530 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1531 "pri" and "weight". Note that the previous implicit priority given by the
1532 list order is no longer honoured.
1534 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1535 for DKIM processing.
1537 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1538 by defining SUPPORT_SOCKS.
1540 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1541 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1542 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1543 variables are renamed to proxy_{local,external}_{address,port}.
1545 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1546 for a build by defining SUPPORT_I18N
1548 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1549 of the query string, and make ${quote_redis:} do that quoting.
1551 JH/29 Move Events support from Experimental to mainline, enabled by default
1552 and removable for a build by defining DISABLE_EVENT.
1554 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1556 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1557 cached by the daemon.
1559 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1560 by defining LOOKUP_REDIS. The libhiredis library is required.
1562 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1563 keys are given for lookup.
1565 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1566 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1567 only supported when built with TLS support. The PolarSSL SHA routines
1568 are still used when the TLS library is too old for convenient support.
1570 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1571 openssl_options), for security. OpenSSL forces this from version 1.1.0
1572 server-side so match that on older versions.
1574 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1575 allocation for $value could be released as the expansion processing
1576 concluded, but leaving the global pointer active for it.
1578 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1579 and to use the domains and local_parts ACL conditions.
1581 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1582 incorrectly not doubled on cutthrough transmission, hence seen as a
1583 body-termination at the receiving system - resulting in truncated mails.
1584 Commonly the sender saw a TCP-level error, and retransmitted the message
1585 via the normal store-and-forward channel. This could result in duplicates
1586 received - but deduplicating mailstores were liable to retain only the
1587 initial truncated version.
1589 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1591 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1593 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1594 we're in there, support oversigning also; bug 1309.
1596 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1598 HS/04 Add support for keep_environment and add_environment options.
1600 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1601 either intentional arithmetic overflow during PRNG, or testing config-
1604 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1605 delivery resulted in actual delivery. Cancel cutthrough before DATA
1608 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1609 rejects a recipient: pass the reject to the originator.
1611 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1612 Many were false-positives and ignorable, but it's worth fixing the
1615 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1616 for the new environment-manipulation done at startup. Move the routines
1617 from being local to tls.c to being global via the os.c file.
1619 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1620 an extract embedded as result-arg for a map, the first arg for extract
1621 is unavailable so we cannot tell if this is a numbered or keyed
1622 extraction. Accept either.
1628 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1631 JH/02 The smtp transport option "multi_domain" is now expanded.
1633 JH/03 The smtp transport now requests PRDR by default, if the server offers
1636 JH/04 Certificate name checking on server certificates, when exim is a client,
1637 is now done by default. The transport option tls_verify_cert_hostnames
1638 can be used to disable this per-host. The build option
1639 EXPERIMENTAL_CERTNAMES is withdrawn.
1641 JH/05 The value of the tls_verify_certificates smtp transport and main options
1642 default to the word "system" to access the system default CA bundle.
1643 For GnuTLS, only version 3.0.20 or later.
1645 JH/06 Verification of the server certificate for a TLS connection is now tried
1646 (but not required) by default. The verification status is now logged by
1647 default, for both outbound TLS and client-certificate supplying inbound
1650 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1653 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1654 Status Notification (bounce) messages are now MIME format per RFC 3464.
1655 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1656 under the control of the dsn_advertise_hosts option, and routers may
1657 have a dsn_lasthop option.
1659 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1660 default, modifiable by a malware= option. The list separator for
1661 the options can now be changed in the usual way. Bug 68.
1663 JH/10 The smtp_receive_timeout main option is now expanded before use.
1665 JH/11 The incoming_interface log option now also enables logging of the
1666 local interface on delivery outgoing connections.
1668 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1669 if the interface and destination host and port all match.
1671 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1674 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1675 Patch from Andrew Lewis.
1677 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1678 now supports optional time-restrictions, weighting, and priority
1679 modifiers per server. Patch originally by <rommer@active.by>.
1681 JH/16 The spamd_address main option now supports a mixed list of local
1682 and remote servers. Remote servers can be IPv6 addresses, and
1683 specify a port-range.
1685 JH/17 Bug 68: The spamd_address main option now supports an optional
1686 timeout value per server.
1688 JH/18 Bug 1581: Router and transport options headers_add/remove can
1689 now have the list separator specified.
1691 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1694 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1697 JH/21 Support for the A6 type of dns record is withdrawn.
1699 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1700 rather than the verbs used.
1702 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1703 from 255 to 1024 chars.
1705 JH/24 Verification callouts now attempt to use TLS by default.
1707 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1708 are generic router options now. The defaults didn't change.
1710 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1711 Original patch from Alexander Shikoff, worked over by JH.
1713 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1714 files if ZCAT_COMMAND is not executable.
1716 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1718 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1720 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1721 Normally benign, it bites when the pair was led to by a CNAME;
1722 modern usage is to not canonicalize the domain to a CNAME target
1723 (and we were inconsistent anyway for A-only vs AAAA+A).
1725 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1727 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1728 when evaluating $sender_host_dnssec.
1730 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1731 $sender_helo_dnssec variable.
1733 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1735 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1737 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1739 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1740 documented as working, but never had. Support all but $spam_report.
1742 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1743 added for tls authenticator.
1745 HS/03 Add perl_taintmode main config option
1751 TL/01 When running the test suite, the README says that variables such as
1752 no_msglog_check are global and can be placed anywhere in a specific
1753 test's script, however it was observed that placement needed to be near
1754 the beginning for it to behave that way. Changed the runtest perl
1755 script to read through the entire script once to detect and set these
1756 variables, reset to the beginning of the script, and then run through
1757 the script parsing/test process like normal.
1759 TL/02 The BSD's have an arc4random API. One of the functions to induce
1760 adding randomness was arc4random_stir(), but it has been removed in
1761 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1762 function when detected.
1764 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1765 cause callback expansion.
1767 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1768 syntax errors in an expansion can be treated as a string instead of
1769 logging or causing an error, due to the internal use of bool_lax
1770 instead of bool when processing it.
1772 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1773 server certificates when making smtp deliveries.
1775 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1777 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1779 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1781 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1782 Merged patch from Sebastian Wiedenroth.
1784 JH/05 Fix results-pipe from transport process. Several recipients, combined
1785 with certificate use, exposed issues where response data items split
1786 over buffer boundaries were not parsed properly. This eventually
1787 resulted in duplicates being sent. This issue only became common enough
1788 to notice due to the introduction of connection certificate information,
1789 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1791 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1792 size buffer was used, resulting in syntax errors when an expansion
1795 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1796 version 3.3.6 or later.
1798 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1799 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1800 both become "event_action", the variables become $event_name, $event_data
1801 and $event_defer_errno. There is a new variable $verify_mode, usable in
1802 routers, transports and related events. The tls:cert event is now also
1803 raised for inbound connections, if the main configuration event_action
1806 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1807 early OCSP support, but no stapling (appears to be less than 1.0.0).
1809 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1810 server certificate names available under the smtp transport option
1811 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1814 JH/10 Time-related extraction expansions from certificates now use the main
1815 option "timezone" setting for output formatting, and are consistent
1816 between OpenSSL and GnuTLS compilations. Bug 1541.
1818 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1819 encoded parameter in the incoming message. Bug 1558.
1821 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1822 include certificate info, eximon was claiming there were spoolfile
1825 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1827 JH/14 Log delivery-related information more consistently, using the sequence
1828 "H=<name> [<ip>]" wherever possible.
1830 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1831 are problematic for Debian distribution, omit them from the release
1834 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1836 JH/16 Fix string representation of time values on 64bit time_t architectures.
1839 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1840 output list separator was used.
1845 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1846 checkers that were complaining about end of non-void function with no
1849 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1850 This was a regression introduced in 4.83 by another bugfix.
1852 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1854 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1855 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1861 TF/01 Correctly close the server side of TLS when forking for delivery.
1863 When a message was received over SMTP with TLS, Exim failed to clear up
1864 the incoming connection properly after forking off the child process to
1865 deliver the message. In some situations the subsequent outgoing
1866 delivery connection happened to have the same fd number as the incoming
1867 connection previously had. Exim would try to use TLS and fail, logging
1868 a "Bad file descriptor" error.
1870 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1871 utilities have not been installed.
1873 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1874 temporary space as the ACL may create new global variables.
1876 TL/01 LDAP support uses per connection or global context settings, depending
1877 upon the detected version of the libraries at build time.
1879 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1880 to extract and use the src ip:port in logging and expansions as if it
1881 were a direct connection from the outside internet. PPv2 support was
1882 updated based on HAProxy spec change in May 2014.
1884 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1886 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1887 Properly escape header and check for NULL return.
1889 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1892 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1894 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1895 characters in header names, implemented as a verify condition.
1896 Contributed by Michael Fischer v. Mollard.
1898 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1899 results permerror and temperror. Previous values are deprecated but
1900 still accepted. In a future release, err_perm and err_temp will be
1901 completely removed, which will be a backward incompatibility if the
1902 ACL tests for either of these two old results. Patch contributed by
1903 user bes-internal on the mailing list.
1905 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1907 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1908 selectors, in both main and reject logs.
1910 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1913 JH/07 Add malware type "sock" for talking to simple daemon.
1915 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1917 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1918 routers/transports under cutthrough routing.
1920 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1921 numbers. Touch up "bool" conditional to keep the same definition.
1923 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1925 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1927 JH/12 Expand items in router/transport headers_add or headers_remove lists
1928 individually rather than the list as a whole. Bug 1452.
1930 Required for reasonable handling of multiple headers_ options when
1931 they may be empty; requires that headers_remove items with embedded
1932 colons must have them doubled (or the list-separator changed).
1934 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1935 view the policy declared in the DMARC record. Currently, $dmarc_status
1936 is a combined value of both the record presence and the result of the
1939 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1941 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1942 dnslookup router and the smtp transport (applying to the forward
1945 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1946 of ldap servers used for a specific lookup. Patch provided by Heiko
1949 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1950 New variable $lookup_dnssec_authenticated for observability.
1952 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1953 Patch submitted by Lars Timman.
1955 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1957 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1958 Requires trusted mode and valid format message id, aborts otherwise.
1959 Patch contributed by Heiko Schlichting.
1961 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1962 certextract with support for various fields. Bug 1358.
1964 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1965 is requested by default, modifiable by smtp transport option
1968 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1969 operate on certificate variables to give certificate fingerprints
1970 Also new ${sha256:cert_variable}.
1972 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1974 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1975 Christian Aistleitner.
1977 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1979 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1980 file. Patch from Wolfgang Breyha.
1982 JH/25 Expand the coverage of the delivery $host and $host_address to
1983 client authenticators run in verify callout. Bug 1476.
1985 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1986 align with daemon_smtp_ports. Bug 72.
1988 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1989 support and error reporting did not work properly.
1991 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1992 and is readable. Patch from Andrew Colin Kissa.
1994 TL/14 Enhance documentation of ${run expansion and how it parses the
1995 commandline after expansion, particularly in the case when an
1996 unquoted variable expansion results in an empty value.
1998 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2000 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2001 by renaming away from it. Addresses GH issue 15, reported by
2004 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2006 TL/15 SECURITY: prevent double expansion in math comparison functions
2007 (can expand unsanitized data). Not remotely exploitable.
2014 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2016 PP/02 Make -n do something, by making it not do something.
2017 When combined with -bP, the name of an option is not output.
2019 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2022 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2023 $sender_host_name and config options to manage this, and basic check
2026 PP/05 DSCP support for outbound connections and control modifier for inbound.
2028 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2029 (Only plugin which currently uses this is kerberos4, which nobody should
2030 be using, but we should make it available and other future plugins might
2031 conceivably use it, even though it would break NAT; stuff *should* be
2032 using channel bindings instead).
2034 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2035 name; added for Sendmail compatibility; requires admin caller.
2036 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2037 just ignore it); requires trusted caller.
2038 Also parse but ignore: -Ac -Am -X<logfile>
2041 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2043 TL/02 Add +smtp_confirmation as a default logging option.
2045 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2046 Patch by Magnus Holmgren from 2007-02-20.
2048 TL/04 Bugzilla 1281 - Spec typo.
2049 Bugzilla 1283 - Spec typo.
2050 Bugzilla 1290 - Spec grammar fixes.
2052 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2054 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2056 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2057 mailing list by by Dmitry Isaikin.
2059 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2061 JH/02 Support "G" suffix to numbers in ${if comparisons.
2063 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2065 NM/01 Bugzilla 1197 - Spec typo
2066 Bugzilla 1196 - Spec examples corrections
2068 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2070 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2071 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2074 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2075 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2077 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2078 "acl {{name}{arg}...}", and optional args on acl condition
2081 JH/05 Permit multiple router/transport headers_add/remove lines.
2083 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2085 JH/07 Avoid using a waiting database for a single-message-only transport.
2086 Performance patch from Paul Fisher. Bugzilla 1262.
2088 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2091 JH/09 Add $headers_added variable, with content from use of ACL modifier
2092 add_header (but not yet added to the message). Bugzilla 199.
2094 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2095 Pulled from Bugzilla 817 by Wolfgang Breyha.
2097 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2099 (nb: this is the same fix as in Exim 4.80.1)
2101 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2104 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2107 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2109 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2110 advertise SMTP AUTH mechanism to us, instead of a generic
2111 protocol violation error. Also, make Exim more robust to bad
2112 data from the Dovecot auth socket.
2114 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2116 When a queue runner is handling a message, Exim first routes the
2117 recipient addresses, during which it prunes them based on the retry
2118 hints database. After that it attempts to deliver the message to
2119 any remaining recipients. It then updates the hints database using
2122 So if a recipient address works intermittently, it can get repeatedly
2123 deferred at routing time. The retry hints record remains fresh so the
2124 address never reaches the final cutoff time.
2126 This is a fairly common occurrence when a user is bumping up against
2127 their storage quota. Exim had some logic in its local delivery code
2128 to deal with this. However it did not apply to per-recipient defers
2129 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2131 This change adds a proper retry rule check during routing so that the
2132 final cutoff time is checked against the message's age. We only do
2133 this check if there is an address retry record and there is not a
2134 domain retry record; this implies that previous attempts to handle
2135 the address had the retry_use_local_parts option turned on. We use
2136 this as an approximation for the destination being like a local
2137 delivery, as in LMTP.
2139 I suspect this new check makes the old local delivery cutoff check
2140 redundant, but I have not verified this so I left the code in place.
2142 TF/02 Correct gecos expansion when From: is a prefix of the username.
2144 Test 0254 submits a message to Exim with the header
2148 When I ran the test suite under the user fanf2, Exim expanded
2149 the header to contain my full name, whereas it should have added
2150 a Resent-Sender: header. It erroneously treats any prefix of the
2151 username as equal to the username.
2153 This change corrects that bug.
2155 GF/01 DCC debug and logging tidyup
2156 Error conditions log to paniclog rather than rejectlog.
2157 Debug lines prefixed by "DCC: " to remove any ambiguity.
2159 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2161 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2162 Bug spotted by Jeremy Harris; was flawed since initial commit.
2163 Would have resulted in OCSP responses post-SNI triggering an Exim
2164 NULL dereference and crash.
2166 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2168 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2169 Bug detection, analysis and fix by Samuel Thibault.
2170 Bugzilla 1331, Debian bug #698092.
2172 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2174 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2175 Server implementation by Todd Lyons, client by JH.
2176 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2177 config variable "prdr_enable" controls whether the server
2178 advertises the facility. If the client requests PRDR a new
2179 acl_data_smtp_prdr ACL is called once for each recipient, after
2180 the body content is received and before the acl_smtp_data ACL.
2181 The client is controlled by both of: a hosts_try_prdr option
2182 on the smtp transport, and the server advertisement.
2183 Default client logging of deliveries and rejections involving
2184 PRDR are flagged with the string "PRDR".
2186 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2187 fclose(). Diagnosis by Todd Lyons.
2189 PP/17 Update configure.default to handle IPv6 localhost better.
2190 Patch by Alain Williams (plus minor tweaks).
2193 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2194 This is now consistent with GnuTLS, and is now documented: the
2195 previous undocumented portable approach to treating the option as
2196 unset was to force an expansion failure. That still works, and
2197 an empty string is now equivalent.
2199 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2200 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2201 not performing validation itself.
2203 PP/20 Added force_command boolean option to pipe transport.
2204 Patch from Nick Koston, of cPanel Inc.
2206 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2209 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2211 PP/21 Fix eximon continuous updating with timestamped log-files.
2212 Broken in a format-string cleanup in 4.80, missed when I repaired the
2213 other false fix of the same issue.
2214 Report and fix from Heiko Schlichting.
2217 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2218 Report from Prashanth Katuri.
2220 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2221 It's SecureTransport, so affects any MacOS clients which use the
2222 system-integrated TLS libraries, including email clients.
2224 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2225 using a MIME ACL for non-SMTP local injection.
2226 Report and assistance in diagnosis by Warren Baker.
2228 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2230 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2232 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2233 last id that failed so it may be referenced in subsequent ACL's.
2235 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2238 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2239 ldap library initialization, allowing self-signed CA's to be
2240 used. Also properly sets require_cert option later in code by
2241 using NULL (global ldap config) instead of ldap handle (per
2242 session). Bug diagnosis and testing by alxgomz.
2244 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2245 the src/util/ subdirectory.
2247 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2248 renamed to Transport Post Delivery Action by Jeremy Harris, as
2251 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2252 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2253 redis_servers = needs to be configured which will be used by the redis
2254 lookup. Patch from Warren Baker, of The Packet Hub.
2256 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2258 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2259 hostname or reverse DNS when processing a host list. Used suggestions
2260 from multiple comments on this bug.
2262 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2264 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2265 Missed a few lines, added it to make the runtest require no keyboard
2268 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2269 contains upper case chars. Make router use caseful_local_part.
2271 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2272 support when GnuTLS has been built with p11-kit.
2278 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2280 This, or similar/improved, will also be change PP/11 of 4.82.
2286 PP/01 Handle short writes when writing local log-files.
2287 In practice, only affects FreeBSD (8 onwards).
2288 Bugzilla 1053, with thanks to Dmitry Isaikin.
2290 NM/01 Bugzilla 949 - Documentation tweak
2292 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2295 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2297 PP/02 Implemented gsasl authenticator.
2299 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2301 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2302 `pkg-config foo` for cflags/libs.
2304 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2305 with rest of GSASL and with heimdal_gssapi.
2307 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2308 `pkg-config foo` for cflags/libs for the TLS implementation.
2310 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2311 properties get this fed in as external SSF. A number of robustness
2312 and debugging improvements to the cyrus_sasl authenticator.
2314 PP/08 cyrus_sasl server now expands the server_realm option.
2316 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2317 Patch by Jeremy Harris.
2319 PP/10 Added dbmjz lookup type.
2321 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2323 PP/12 MAIL args handles TAB as well as SP, for better interop with
2324 non-compliant senders.
2325 Analysis and variant patch by Todd Lyons.
2327 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2328 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2329 Patch from Dirk Mueller <dmueller@suse.com>
2331 PP/13 tls_peerdn now print-escaped for spool files.
2332 Observed some $tls_peerdn in wild which contained \n, which resulted
2333 in spool file corruption.
2335 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2336 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2337 or write after TLS renegotiation, which otherwise led to messages
2340 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2341 as a tracking header (ie: a signed header comes before the signature).
2342 Patch from Wolfgang Breyha.
2344 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2345 comma-sep list; embedded commas doubled.
2347 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2349 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2351 Report and patch from Dmitry Banschikov.
2353 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2354 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2355 failures. We appear to now support TLS1.1+ with Exim.
2357 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2358 lets Exim select keys and certificates based upon TLS SNI from client.
2359 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2360 before an outbound SMTP session. New log_selector, +tls_sni.
2362 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2363 NULL dereference. Report and patch from Alun Jones.
2365 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2366 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2367 Not seeing resolver debug output on NetBSD, but suspect this is a
2368 resolver implementation change.
2370 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2371 Left warnings. Added "eximon gdb" invocation mode.
2373 PP/21 Defaulting "accept_8bitmime" to true, not false.
2375 PP/22 Added -bw for inetd wait mode support.
2377 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2378 locate the relevant includes and libraries. Made this the default.
2380 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2381 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2383 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2384 This may cause build issues on older platforms.
2386 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2387 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2388 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2389 Added SNI support via GnuTLS too.
2390 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2392 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2394 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2395 Applied second patch from Janne, implementing suggestion to default
2396 multiple-strings-in-record handling to match SPF spec.
2398 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2400 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2401 read-only, out of scope).
2402 Patch from Wolfgang Breyha, report from Stuart Northfield.
2404 PP/29 Fix three issues highlighted by clang analyser static analysis.
2405 Only crash-plausible issue would require the Cambridge-specific
2406 iplookup router and a misconfiguration.
2407 Report from Marcin Mirosław.
2409 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2411 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2412 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2413 As part of this, removing so much warning spew let me fix some minor
2414 real issues in debug logging.
2416 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2417 assignment on my part. Fixed.
2419 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2420 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2421 Janne Snabb (who went above and beyond: thank you).
2423 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2424 string otherwise requires a connection and a bunch more work and it's
2425 relatively easy to get wrong. Should also expose TLS library linkage
2428 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2429 64-bit ${eval} (JH/03).
2431 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2432 GNU libc to support some of the 64-bit stuff, should not lead to
2433 conflicts. Defined before os.h is pulled in, so if a given platform
2434 needs to override this, it can.
2436 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2437 protection layer was required, which is not implemented.
2438 Bugzilla 1254, patch from Wolfgang Breyha.
2440 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2441 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2442 tls_dhparam take prime identifiers. Also unbreak combination of
2443 OpenSSL+DH_params+TLSSNI.
2445 PP/39 Disable SSLv2 by default in OpenSSL support.
2451 PP/01 Solaris build fix for Oracle's LDAP libraries.
2452 Bugzilla 1109, patch from Stephen Usher.
2454 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2456 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2459 TF/02 Fix a couple more cases where we did not log the error message
2460 when unlink() failed. See also change 4.74-TF/03.
2462 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2463 lock up or crash if it happened to be inside a call to libc when it
2464 got a SIGUSR1 from exiwhat.
2466 The SIGUSR1 handler appends the current process status to the process
2467 log which is later printed by exiwhat. It used to use the general
2468 purpose logging code to do this, but several functions it calls are
2469 not safe for signals.
2471 The new output code in the SIGUSR1 handler is specific to the process
2472 log, and simple enough that it's easy to inspect for signal safety.
2473 Removing some special cases also simplifies the general logging code.
2474 Removing the spurious timestamps from the process log simplifies
2477 TF/04 Improved ratelimit ACL condition.
2479 The /noupdate option has been deprecated in favour of /readonly which
2480 has clearer semantics. The /leaky, /strict, and /readonly update modes
2481 are mutually exclusive. The update mode is no longer included in the
2482 database key; it just determines when the database is updated. (This
2483 means that when you upgrade Exim will forget old rate measurements.)
2485 Exim now checks that the per_* options are used with an update mode that
2486 makes sense for the current ACL. For example, when Exim is processing a
2487 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2488 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2489 must specify per_mail/readonly. If you omit the update mode it defaults to
2490 /leaky where that makes sense (as before) or /readonly where required.
2492 The /noupdate option is now undocumented but still supported for
2493 backwards compatibility. It is equivalent to /readonly except that in
2494 ACLs where /readonly is required you may specify /leaky/noupdate or
2495 /strict/noupdate which are treated the same as /readonly.
2497 A useful new feature is the /count= option. This is a generalization
2498 of the per_byte option, so that you can measure the throughput of other
2499 aggregate values. For example, the per_byte option is now equivalent
2500 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2502 The per_rcpt option has been generalized using the /count= mechanism
2503 (though it's more complicated than the per_byte equivalence). When it is
2504 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2505 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2506 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2507 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2508 Note that using per_rcpt with a non-readonly update mode in more than
2509 one ACL will cause the recipients to be double-counted. (The per_mail
2510 and per_byte options don't have this problem.)
2512 The handling of very low rates has changed slightly. If the computed rate
2513 is less than the event's count (usually one) then this event is the first
2514 after a long gap. In this case the rate is set to the same as this event's
2515 count, so that the first message of a spam run is counted properly.
2517 The major new feature is a mechanism for counting the rate of unique
2518 events. The new per_addr option counts the number of different
2519 recipients that someone has sent messages to in the last time period. It
2520 behaves like per_rcpt if all the recipient addresses are different, but
2521 duplicate recipient addresses do not increase the measured rate. Like
2522 the /count= option this is a general mechanism, so the per_addr option
2523 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2524 example, measure the rate that a client uses different sender addresses
2525 with the options per_mail/unique=$sender_address. There are further
2526 details in the main documentation.
2528 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2530 TF/06 Removed a few PCRE remnants.
2532 TF/07 Automatically extract Exim's version number from tags in the git
2533 repository when doing development or release builds.
2535 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2536 Bugzilla 879. Patch from Paul Fisher.
2538 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2539 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2542 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2544 PP/05 Variable $av_failed, true if the AV scanner deferred.
2545 Bugzilla 1078. Patch from John Horne.
2547 PP/06 Stop make process more reliably on build failure.
2548 Bugzilla 1087. Patch from Heiko Schlittermann.
2550 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2551 Bugzilla 1089. Patch from Heiko Schlittermann.
2553 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2554 Bugzilla 1131. Patch from Holger Weiß.
2556 PP/09 Handle IPv6 addresses with SPF.
2557 Bugzilla 860. Patch from Wolfgang Breyha.
2559 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2561 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2564 PP/11 match_* no longer expand right-hand-side by default.
2565 New compile-time build option, EXPAND_LISTMATCH_RHS.
2566 New expansion conditions, "inlist", "inlisti".
2568 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2570 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2572 PP/14 fix log_write() format string regression from TF/03.
2573 Bugzilla 1152. Patch from Dmitry Isaikin.
2579 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2581 PP/02 Harmonised TLS library version reporting; only show if debugging.
2582 Layout now matches that introduced for other libraries in 4.74 PP/03.
2584 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2586 PP/04 New "dns_use_edns0" global option.
2588 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2591 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2592 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2594 TK/01 Updated PolarSSL code to 0.14.2.
2595 Bugzilla 1097. Patch from Andreas Metzler.
2597 PP/07 Catch divide-by-zero in ${eval:...}.
2598 Fixes bugzilla 1102.
2600 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2603 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2604 format-string attack -- SECURITY: remote arbitrary code execution.
2606 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2607 time unintentionally subject to list matching rules, letting the header
2608 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2609 arbitrary string expansion). This allowed for information disclosure.
2611 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2612 INT_MIN/-1 -- value coerced to INT_MAX.
2618 NM/01 Workaround for PCRE version dependency in version reporting
2621 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2622 This fixes portability to compilers other than gcc, notably
2623 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2625 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2626 makefiles for portability to HP-UX and POSIX correctness.
2628 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2629 Also via indented variable definition in the Makefile.
2630 (Debugging by Oliver Heesakkers).
2632 PP/02 Restore caching of spamd results with expanded spamd_address.
2633 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2635 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2636 Improves build reliability. Fix from: Frank Elsner
2638 NM/02 Fix wide character breakage in the rfc2047 coding
2639 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2641 NM/03 Allow underscore in dnslist lookups
2642 Fixes bug 1026. Patch from Graeme Fowler
2644 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2645 Code patches from Adam Ciarcinski of NetBSD.
2647 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2650 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2651 is logged, to avoid truncation. Patch from John Horne.
2653 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2654 Patch from Jakob Hirsch.
2656 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2657 SQL string expansion failure details.
2658 Patch from Andrey Oktyabrski.
2660 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2661 Patch from Simon Arlott.
2663 PP/09 New lookups functionality failed to compile on old gcc which rejects
2664 extern declarations in function scope.
2665 Patch from Oliver Fleischmann
2667 PP/10 Use sig_atomic_t for flags set from signal handlers.
2668 Check getgroups() return and improve debugging.
2669 Fixed developed for diagnosis in bug 927 (which turned out to be
2672 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2673 Patch from Mark Zealey.
2675 PP/12 Bugzilla 1056: Improved spamd server selection.
2676 Patch from Mark Zealey.
2678 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2679 Based on patch from Heiko Schlittermann.
2681 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2682 Patch from Uwe Doering, sign-off by Michael Haardt.
2684 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2685 variable declaration deep within a block. Bug and patch from
2688 PP/15 lookups-Makefile IRIX compatibility coercion.
2690 PP/16 Make DISABLE_DKIM build knob functional.
2692 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2693 Patch by Simon Arlott
2695 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2696 variable argument macros. Our copy now differs from upstream.
2702 TF/01 Failure to get a lock on a hints database can have serious
2703 consequences so log it to the panic log.
2705 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2706 controlled using the smtp_confirmation log selector.
2708 TF/03 Include the error message when we fail to unlink a spool file.
2710 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2711 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2712 for maintaining out-of-tree patches for some time.
2714 PP/01 Bugzilla 139: Documentation and portability issues.
2715 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2716 Handle per-OS dynamic-module compilation flags.
2718 PP/02 Let /dev/null have normal permissions.
2719 The 4.73 fixes were a little too stringent and complained about the
2720 permissions on /dev/null. Exempt it from some checks.
2721 Reported by Andreas M. Kirchwitz.
2723 PP/03 Report version information for many libraries, including
2724 Exim version information for dynamically loaded libraries. Created
2725 version.h, now support a version extension string for distributors
2726 who patch heavily. Dynamic module ABI change.
2728 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2729 privilege escalation vulnerability whereby the Exim run-time user
2730 can cause root to append content of the attacker's choosing to
2733 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2736 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2737 If dropping privileges for untrusted macros, we disabled normal logging
2738 on the basis that it would fail; for the Exim run-time user, this is not
2739 the case, and it resulted in successful deliveries going unlogged.
2740 Fixed. Reported by Andreas Metzler.
2746 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2747 only prepend for the Resent-* case. Fixes regression introduced in
2748 Exim 4.70 by NM/22 for Bugzilla 607.
2750 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2751 increasing numbers of administrators be bitten by this.
2753 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2755 PP/03 Bugzilla 994: added openssl_options main configuration option.
2757 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2759 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2761 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2763 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2764 without a peer certificate, leading to a segfault because of an
2765 assumption that peers always have certificates. Be a little more
2766 paranoid. Problem reported by Martin Tscholak.
2768 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2769 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2770 NB: ClamAV planning to remove STREAM in "middle of 2010".
2771 CL also introduces -bmalware, various -d+acl logging additions and
2772 more caution in buffer sizes.
2774 PP/09 Implemented reverse_ip expansion operator.
2776 PP/10 Bugzilla 937: provide a "debug" ACL control.
2778 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2780 PP/12 Bugzilla 973: Implement --version.
2782 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2784 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2786 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2788 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2789 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2790 ignore trailing whitespace.
2792 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2794 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2797 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2798 Notification from Dr Andrew Aitchison.
2800 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2801 ExtendedDetectionInfo response format.
2802 Notification from John Horne.
2804 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2807 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2808 XSL and documented dependency on system catalogs, with examples of how
2811 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2814 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2815 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2816 configuration file which is writeable by the Exim user or group.
2818 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2819 of configuration files to cover files specified with the -C option if
2820 they are going to be used with root privileges, not just the default
2823 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2824 option (effectively making it always true).
2826 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2827 files to be used while preserving root privileges.
2829 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2830 that rogue child processes cannot use them.
2832 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2833 run-time user, instead of root.
2835 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2836 Exim run-time user without dropping privileges.
2838 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2839 result string, instead of calling string_vformat() twice with the same
2842 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2843 for other users. Others should always drop root privileges if they use
2844 -C on the command line, even for a whitelisted configure file.
2846 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2848 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2854 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2855 $data_path, and $header_path variables; fixed documentation bugs and
2858 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2859 exipick to access non-standard spools, including the "frozen" queue
2862 NM/01 Bugzilla 965: Support mysql stored procedures.
2863 Patch from Alain Williams
2865 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2867 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2868 Patch from Andreas Metzler
2870 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2871 Patch from Kirill Miazine
2873 NM/05 Bugzilla 671: Added umask to procmail example.
2875 JJ/03 installed exipick 20100323.0, fixing doc bug
2877 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2878 directory. Notification and patch from Dan Rosenberg.
2880 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2882 TK/02 Improve log output when DKIM signing operation fails.
2884 MH/01 Treat the transport option dkim_domain as a colon separated
2885 list, not as a single string, and sign the message with each element,
2886 omitting multiple occurences of the same signer.
2888 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2889 Bugzilla 985, 986. Patch by Simon Arlott
2891 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2892 Patch by Simon Arlott
2894 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2895 MBX locking. Notification from Dan Rosenberg.
2901 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2903 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2905 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2907 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2909 NM/04 Bugzilla 915: Flag broken perl installation during build.
2915 TK/01 Added patch by Johannes Berg that expands the main option
2916 "spamd_address" if it starts with a dollar sign.
2918 TK/02 Write list of recipients to X-Envelope-Sender header when building
2919 the mbox-format spool file for content scanning (suggested by Jakob
2922 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2923 (http://www.dcc-servers.net/) support via dccifd. Activated by
2924 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2926 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2927 by Mark Daniel Reidel <mr@df.eu>.
2929 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2930 When building exim an external PCRE library is now needed -
2931 PCRE is a system library on the majority of modern systems.
2932 See entry on PCRE_LIBS in EDITME file.
2934 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2935 conversation. Added nologin parameter to request.
2936 Patch contributed by Kirill Miazine.
2938 TF/01 Do not log submission mode rewrites if they do not change the address.
2940 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2942 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2943 log files in place. Contributed by Roberto Lima.
2945 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2947 TF/03 Bugzilla 615: When checking the local_parts router precondition
2948 after a local_part_suffix or local_part_prefix option, Exim now
2949 does not use the address's named list lookup cache, since this
2950 contains cached lookups for the whole local part.
2952 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2953 Robert Millan. Documentation is in experimental-spec.txt.
2955 TF/04 Bugzilla 668: Fix parallel build (make -j).
2957 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2959 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2960 Patch provided by Jan Srzednicki.
2962 TF/05 Leading white space used to be stripped from $spam_report which
2963 wrecked the formatting. Now it is preserved.
2965 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2966 that they are available at delivery time.
2968 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2970 TF/08 TLS error reporting now respects the incoming_interface and
2971 incoming_port log selectors.
2973 TF/09 Produce a more useful error message if an SMTP transport's hosts
2974 setting expands to an empty string.
2976 NM/06 Bugzilla 744: EXPN did not work under TLS.
2977 Patch provided by Phil Pennock.
2979 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2980 Patch provided by Richard Godbee.
2982 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2983 acl_smtp_notquit, added index entry.
2985 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2986 Patch provided by Eugene Bujak.
2988 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2989 accept(). Patch provided by Maxim Dounin.
2991 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2992 Patch provided by Phil Pennock.
2994 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2996 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2997 Patch provided by Brad "anomie" Jorsch.
2999 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3000 Patch provided by Dean Brooks.
3002 TK/05 Add native DKIM support (does not depend on external libraries).
3004 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3005 Patch provided by Graeme Fowler.
3007 NM/16 Bugzilla 851: Documentation example syntax fix.
3009 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3011 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3014 NM/19 Bugzilla 745: TLS version reporting.
3015 Patch provided by Phil Pennock.
3017 NM/20 Bugzilla 167: bool: condition support.
3018 Patch provided by Phil Pennock.
3020 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3021 clients. Patch provided by Phil Pennock.
3023 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3024 Patch provided by Brad "anomie" Jorsch.
3026 NM/23 Bugzilla 687: Fix misparses in eximstats.
3027 Patch provided by Heiko Schlittermann.
3029 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3030 Patch provided by Heiko Schlittermann.
3032 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3033 plus update to original patch.
3035 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3037 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3038 Patch provided by David Brownlee.
3040 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3042 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3044 NM/30 Bugzilla 888: TLS documentation bugfixes.
3046 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3048 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3049 Unlike the original bugzilla I have changed all shell scripts in src tree.
3051 NM/33 Bugzilla 898: Transport filter timeout fix.
3052 Patch by Todd Rinaldo.
3054 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3055 Patch by Serge Demonchaux.
3057 NM/35 Bugzilla 39: Base64 decode bug fixes.
3058 Patch by Jakob Hirsch.
3060 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3062 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3064 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3066 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3072 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3073 ALT-N's libdkim that I have put here:
3074 http://duncanthrax.net/exim-experimental/
3076 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3077 were called 'true' and it seems that C99 defines that as a reserved
3078 keyword to be used with 'bool' variable types. That means you could
3079 not include C99-style headers which use bools without triggering
3080 build errors in sieve.c.
3082 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3083 as mailq or other aliases. Changed the --help handling significantly
3084 to do whats expected. exim_usage() emits usage/help information.
3086 SC/01 Added the -bylocaldomain option to eximstats.
3088 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3090 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3092 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3098 PH/01 Another patch from the Sieve maintainer.
3100 PH/02 When an IPv6 address is converted to a string for single-key lookup
3101 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3102 dots are used instead of colons so that keys in lsearch files need not
3103 contain colons. This was done some time before quoting was made available
3104 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3105 (notated using the quote facility) so as to distinguish them from IPv4
3106 keys. This meant that lookups for IP addresses in host lists did not work
3107 for iplsearch lookups.
3109 This has been fixed by arranging for IPv6 addresses to be expressed with
3110 colons if the lookup type is iplsearch. This is not incompatible, because
3111 previously such lookups could never work.
3113 The situation is now rather anomalous, since one *can* have colons in
3114 ordinary lsearch keys. However, making the change in all cases is
3115 incompatible and would probably break a number of configurations.
3117 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3120 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3121 conversion specification without a maximum field width, thereby enabling
3122 a rogue spamd server to cause a buffer overflow. While nobody in their
3123 right mind would setup Exim to query an untrusted spamd server, an
3124 attacker that gains access to a server running spamd could potentially
3125 exploit this vulnerability to run arbitrary code as the Exim user.
3127 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3128 $primary_hostname instead of what libspf2 thinks the hosts name is.
3130 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3131 a directory entry by the name of the lookup key. Previously, if a
3132 symlink pointed to a non-existing file or a file in a directory that
3133 Exim lacked permissions to read, a lookup for a key matching that
3134 symlink would fail. Now it is enough that a matching directory entry
3135 exists, symlink or not. (Bugzilla 503.)
3137 PH/03 The body_linecount and body_zerocount variables are now exported in the
3140 PH/04 Added the $dnslist_matched variable.
3142 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3143 This means they are set thereafter only if the connection becomes
3146 PH/06 Added the client_condition to authenticators so that some can be skipped
3147 by clients under certain conditions.
3149 PH/07 The error message for a badly-placed control=no_multiline_responses left
3150 "_responses" off the end of the name.
3152 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3154 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3155 (without spaces) instead of just copying the configuration text.
3157 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3159 PH/11 Added $max_received_linelength.
3161 PH/12 Added +ignore_defer and +include_defer to host lists.
3163 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3164 way in which PCRE > 7.0 is built.
3166 PH/14 Implemented queue_only_load_latch.
3168 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3169 MAIL command. The effect was to mangle the value on 64-bit systems.
3171 PH/16 Another patch from the Sieve maintainer.
3173 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3175 PH/18 If a system quota error occurred while trying to create the file for
3176 a maildir delivery, the message "Mailbox is full" was not appended to the
3177 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3178 only to a quota excession during the actual writing of the file.
3180 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3181 characters?) which causes problems in log lines. The DN values are now
3182 passed through string_printing() before being added to log lines.
3184 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3185 and InterBase are left for another time.)
3187 PH/21 Added message_body_newlines option.
3189 PH/22 Guard against possible overflow in moan_check_errorcopy().
3191 PH/23 POSIX allows open() to be a macro; guard against that.
3193 PH/24 If the recipient of an error message contained an @ in the local part
3194 (suitably quoted, of course), incorrect values were put in $domain and
3195 $local_part during the evaluation of errors_copy.
3201 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3202 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3205 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3206 issue a MAIL command.
3208 PH/02 In an ACL statement such as
3210 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3212 if a client was not listed at all, or was listed with a value other than
3213 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3214 the condition was not true (as it should be), so access was not denied.
3215 The bug was that the ! inversion was incorrectly passed on to the second
3216 item. This has been fixed.
3218 PH/03 Added additional dnslists conditions == and =& which are different from
3219 = and & when the dns lookup returns more than one IP address.
3221 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3222 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3224 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3225 FSYNC, which compiles an option called disable_fsync that allows for
3226 bypassing fsync(). The documentation is heavily laced with warnings.
3228 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3230 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3231 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3232 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3233 including adding "make clean"; (3) Added -fPIC when compiling the test
3234 dynamically loaded module, to get rid of a warning.
3236 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3237 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3238 = 0s. The bug is otherwise harmless.
3240 PH/07 There was a bug in the dovecot authenticator such that the value of
3241 $auth1 could be overwritten, and so not correctly preserved, after a
3242 successful authentication. This usually meant that the value preserved by
3243 the server_setid option was incorrect.
3245 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3247 PH/09 Installed PCRE release 7.0.
3249 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3250 run for batched SMTP input. It is now run at the start of every message
3251 in the batch. While fixing this I discovered that the process information
3252 (output by running exiwhat) was not always getting set for -bs and -bS
3253 input. This is fixed, and it now also says "batched" for BSMTP.
3255 PH/11 Added control=no_pipelining.
3257 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3258 patch, slightly modified), and move the expansion of helo_data till after
3259 the connection is made in the smtp transport (so it can use these
3262 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3264 PH/14 Added log_selector = +pid.
3266 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3268 PH/16 Add ${if forany and ${if forall.
3270 PH/17 Added dsn_from option to vary the From: line in DSNs.
3272 PH/18 Flush SMTP output before performing a callout, unless control =
3273 no_callout_flush is set.
3275 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3276 was true (the default) a successful delivery failed to delete the retry
3277 item, thus causing premature timeout of the address. The bug is now
3280 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3282 PH/21 Long custom messages for fakedefer and fakereject are now split up
3283 into multiline responses in the same way that messages for "deny" and
3284 other ACL rejections are.
3286 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3287 with slight modification.
3289 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3290 draft, changing the syntax and factoring some duplicate code".
3292 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3293 for deliveries of the second and subsequent messages over the same SMTP
3296 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3297 ${reduce, with only minor "tidies".
3299 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3301 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3302 expansion side effects.
3304 PH/27 When a message times out after an over-quota error from an Exim-imposed
3305 quota, the bounce message says "mailbox is full". This message was not
3306 being given when it was a system quota that was exceeded. It now should
3309 MH/03 Made $recipients available in local_scan(). local_scan() already has
3310 better access to the recipient list through recipients_list[], but
3311 $recipients can be useful in postmaster-provided expansion strings.
3313 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3314 in the case of a MAIL command with additional options following the
3315 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3316 were accidentally chopped off.
3318 PH/29 SMTP synchronization checks are implemented when a command is read -
3319 there is a check that no more input is waiting when there shouldn't be
3320 any. However, for some commands, a delay in an ACL can mean that it is
3321 some time before the response is written. In this time, more input might
3322 arrive, invalidly. So now there are extra checks after an ACL has run for
3323 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3324 pipelining has not been advertised.
3326 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3328 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3329 returned in the SMTP response when smtp_return_error_details was false.
3330 This has been fixed.
3332 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3333 instead of the C I/O that was originally supplied, because problems were
3334 reported on Solaris.
3336 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3337 Exim which did not show up earlier: it was assuming that a call to
3338 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3339 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3340 was a macro that became an assignment, so it seemed to work. This has
3341 changed to a proper function call with a void return, hence the compile
3342 error. Exim's code has been fixed.
3344 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3347 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3349 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3351 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3352 because it turns out to be too big or there is a timeout) there is no
3353 "Completed" line in the log. When some messages of this type were
3354 selected by exigrep, they were listed as "not completed". Others were
3355 picked up by some special patterns. I have improved the selection
3356 criteria to be more general.
3358 PH/38 The host_find_failed option in the manualroute router can now be set
3359 to "ignore", to completely ignore a host whose IP address cannot be
3360 found. If all hosts are ignored, the behaviour is controlled by the new
3361 host_all_ignored option.
3363 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3364 homing or because of multiple MX records with /mx) generated more than
3365 one IP address, and the following item turned out to be the local host,
3366 all the secondary addresses of the first item were incorrectly removed
3367 from the list, along with the local host and any following hosts (which
3368 is what is supposed to happen).
3370 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3371 whoever called Exim into the -H file. In the case of the daemon it was
3372 behaving confusingly. When first started, it used values for whoever
3373 started the daemon, but after a SIGHUP it used the Exim user (because it
3374 calls itself on a restart). I have changed the code so that it now always
3377 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3378 message are rejected with the same error (e.g. no authentication or bad
3379 sender address), and a DATA command is nevertheless sent (as can happen
3380 with PIPELINING or a stupid MUA), the error message that was given to the
3381 RCPT commands is included in the rejection of the DATA command. This is
3382 intended to be helpful for MUAs that show only the final error to their
3385 PH/42 Another patch from the Sieve maintainer.
3387 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3388 Eximstats - Fixed some broken HTML links and added missing column headers
3390 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3391 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3393 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3395 PH/43 Yet another patch from the Sieve maintainer.
3397 PH/44 I found a way to check for a TCP/IP connection going away before sending
3398 the response to the final '.' that terminates a message, but only in the
3399 case where the client has not sent further data following the '.'
3400 (unfortunately, this is allowed). However, in many cases there won't be
3401 any further data because there won't be any more messages to send. A call
3402 to select() can be used: if it shows that the input is "ready", there is
3403 either input waiting, or the socket has been closed. An attempt to read
3404 the next input character can distinguish the two cases. Previously, Exim
3405 would have sent an OK response which the client would never have see.
3406 This could lead to message repetition. This fix should cure that, at
3407 least in a lot of common cases.
3409 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3410 advertised in response to EHLO.
3416 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3417 fixed by 4.65/MH/01 (is this a record?) are fixed:
3419 (i) An empty string was always treated as zero by the numeric comparison
3420 operators. This behaviour has been restored.
3422 (ii) It is documented that the numeric comparison operators always treat
3423 their arguments as decimal numbers. This was broken in that numbers
3424 starting with 0 were being interpreted as octal.
3426 While fixing these problems I realized that there was another issue that
3427 hadn't been noticed. Values of message_size_limit (both the global option
3428 and the transport option) were treated as octal if they started with 0.
3429 The documentation was vague. These values are now always treated as
3430 decimal, and I will make that clear in the documentation.
3436 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3437 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3440 MH/01 Don't check that the operands of numeric comparison operators are
3441 integers when their expansion is in "skipping" mode (fixes bug
3442 introduced by 4.64-PH/07).
3444 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3445 child addresses, Exim now panics and dies. Previously, because the count
3446 is held in a short int, deliveries were likely to be lost. As such a
3447 large number of recipients for a single message is ridiculous
3448 (performance will be very, very poor), I have chosen to impose a limit
3449 rather than extend the field.
3455 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3456 leftover -K file (the existence of which was triggered by #402).
3457 While we were at it, introduced process PID as part of the -K
3458 filename. This should rule out race conditions when creating
3461 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3462 processing considerably. Previous code took too long for large mails,
3463 triggering a timeout which in turn triggers #401.
3465 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3466 in the DK code in transports.c. sendfile() is not really portable,
3467 hence the _LINUX specificness.
3469 TF/01 In the add_headers option to the mail command in an Exim filter,
3470 there was a bug that Exim would claim a syntax error in any
3471 header after the first one which had an odd number of characters
3474 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3475 callout verification, Exim cached a "reject" for the entire domain. This
3476 is correct for most verifications, but it is not correct for a recipient
3477 verification with use_sender or use_postmaster set, because in that case
3478 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3479 case of MAIL FROM:<> rejection from other early rejections (e.g.
3480 rejection of HELO). When verifying a recipient using a non-null MAIL
3481 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3482 Whatever the result of the callout, the value of the domain cache is
3483 left unchanged (for any other kind of callout, getting as far as trying
3484 RCPT means that the domain itself is ok).
3486 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3489 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3490 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3491 session, the daemon ignores ECONNECT errors and logs others; it now
3492 ignores EPIPE as well.
3494 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3495 (quoted-printable decoding).
3497 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3498 later the small subsequent patch to fix an introduced bug.
3500 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3502 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3504 PH/08 An error is now given if message_size_limit is specified negative.
3506 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3507 to be given (somewhat) arbitrary names.
3509 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3512 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3513 miscellaneous code fixes
3515 PH/10 Added the log_reject_target ACL modifier to specify where to log
3518 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3519 hostname. This is wrong, because it relates to the incoming message (and
3520 probably the interface on which it is arriving) and not to the outgoing
3521 callout (which could be using a different interface). This has been
3522 changed to use the value of the helo_data option from the smtp transport
3523 instead - this is what is used when a message is actually being sent. If
3524 there is no remote transport (possible with a router that sets up host
3525 addresses), $smtp_active_hostname is used.
3527 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3528 tweaks were necessary in order to get it to work (see also 21 below):
3529 (a) The code assumed that strncpy() returns a negative number on buffer
3530 overflow, which isn't the case. Replaced with Exim's string_format()
3532 (b) There were several signed/unsigned issues. I just did the minimum
3533 hacking in of casts. There is scope for a larger refactoring.
3534 (c) The code used strcasecmp() which is not a standard C function.
3535 Replaced with Exim's strcmpic() function.
3536 (d) The code set only $1; it now sets $auth1 as well.
3537 (e) A simple test gave the error "authentication client didn't specify
3538 service in request". It would seem that Dovecot has changed its
3539 interface. Fortunately there's a specification; I followed it and
3540 changed what the client sends and it appears to be working now.
3542 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3545 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3546 address A is aliased to B and C, where B exists and C does not. Without
3547 -v the output is "A verified" because verification stops after a
3548 successful redirection if more than one address is generated. However,
3549 with -v the child addresses are also verified. Exim was outputting "A
3550 failed to verify" and then showing the successful verification for C,
3551 with its parentage. It now outputs "B failed to verify", showing B's
3552 parentage before showing the successful verification of C.
3554 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3555 look up a TXT record in a specific list after matching in a combined
3558 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3559 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3560 they consult the DNS. I had assumed they would set it the way they
3561 wanted; and indeed my experiments on Linux seem to show that in some
3562 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3563 To be on the safe side, however, I have now made the interface to
3564 host_find_byname() similar to host_find_bydns(), with an argument
3565 containing the DNS resolver options. The host_find_byname() function now
3566 sets these options at its start, just as host_find_bydns() does. The smtp
3567 transport options dns_qualify_single and dns_search_parents are passed to
3568 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3569 of host_find_byname() use the default settings of RES_DEFNAMES
3570 (qualify_single) but not RES_DNSRCH (search_parents).
3572 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3573 spool_read_header() do less string testing, by means of a preliminary
3574 switch on the second character of optional "-foo" lines. (This is
3575 overdue, caused by the large number of possibilities that now exist.
3576 Originally there were few.) While I was there, I also converted the
3577 str(n)cmp tests so they don't re-test the leading "-" and the first
3578 character, in the hope this might squeeze out yet more improvement.
3580 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3581 flag allowing group syntax was set by the header_syntax check but not
3582 turned off, possible causing trouble later; (2) The flag was not being
3583 set at all for the header_verify test, causing "group"-style headers to
3584 be rejected. I have now set it in this case, and also caused header_
3585 verify to ignore an empty address taken from a group. While doing this, I
3586 came across some other cases where the code for allowing group syntax
3587 while scanning a header line wasn't quite right (mostly, not resetting
3588 the flag correctly in the right place). These bugs could have caused
3589 trouble for malformed header lines. I hope it is now all correct.
3591 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3592 with the "reply" argument non-NULL. The code, however (which originally
3593 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3594 but it didn't always do it. This confused somebody who was copying the
3595 code for some other use. I have removed all the tests.
3597 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3598 feature that was used to support insecure browsers during the U.S. crypto
3599 embargo. It requires special client support, and Exim is probably the
3600 only MTA that supported it -- and would never use it because real RSA is
3601 always available. This code has been removed, because it had the bad
3602 effect of slowing Exim down by computing (never used) parameters for the
3603 RSA_EXPORT functionality.
3605 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3606 authenticator to fail if there's a tab character in the incoming data
3607 (there should never be unless someone is messing about, as it's supposed
3608 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3609 if the connection is using TLS or if the remote IP is the same as the
3610 local IP, and the "valid-client-cert option" if a client certificate has
3613 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3614 authenticators. This can be used for authorization after authentication
3615 succeeds. (In the case of plaintext, it servers for both authentication
3618 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3619 if any retry times were supplied.
3621 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3622 connection, where there is no client IP address to check. In this
3623 situation, the verify now always succeeds.
3625 PH/25 Applied John Jetmore's -Mset patch.
3627 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3629 PH/27 In a string expansion for a processed (not raw) header when multiple
3630 headers of the same name were present, leading whitespace was being
3631 removed from all of them, but trailing whitespace was being removed only
3632 from the last one. Now trailing whitespace is removed from each header
3633 before concatenation. Completely empty headers in a concatenation (as
3634 before) are ignored.
3636 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3637 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3639 PH/29 [Removed. This was a change that I later backed out, and forgot to
3640 correct the ChangeLog entry (that I had efficiently created) before
3641 committing the later change.]
3643 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3644 address errors (4xx response to RCPT) over the same connection as other
3645 messages routed to the same hosts. Such deliveries are always "forced",
3646 so retry times are not inspected. This resulted in far too many retries
3647 for the affected addresses. The effect occurred only when there were more
3648 hosts than the hosts_max_try setting in the smtp transport when it had
3649 the 4xx errors. Those hosts that it had tried were not added to the list
3650 of hosts for which the message was waiting, so if all were tried, there
3651 was no problem. Two fixes have been applied:
3653 (i) If there are any address or message errors in an SMTP delivery, none
3654 of the hosts (tried or untried) are now added to the list of hosts
3655 for which the message is waiting, so the message should not be a
3656 candidate for sending over the same connection that was used for a
3657 successful delivery of some other message. This seems entirely
3658 reasonable: after all the message is NOT "waiting for some host".
3659 This is so "obvious" that I'm not sure why it wasn't done
3660 previously. Hope I haven't missed anything, but it can't do any
3661 harm, as the worst effect is to miss an optimization.
3663 (ii) If, despite (i), such a delivery is accidentally attempted, the
3664 routing retry time is respected, so at least it doesn't keep
3665 hammering the server.
3667 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3668 in ${readsocket because some servers need this prod.
3670 PH/32 Added some extra debug output when updating a wait-xxx database.
3672 PH/33 The hint "could be header name not terminated by colon", which has been
3673 given for certain expansion errors for a long time, was not being given
3674 for the ${if def:h_colon_omitted{... case.
3676 PH/34 The spec says: "With one important exception, whenever a domain list is
3677 being scanned, $domain contains the subject domain." There was at least
3678 one case where this was not true.
3680 PH/35 The error "getsockname() failed: connection reset by peer" was being
3681 written to the panic log as well as the main log, but it isn't really
3682 panic-worthy as it just means the connection died rather early on. I have
3683 removed the panic log writing for the ECONNRESET error when getsockname()
3686 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3687 runs only) independently of the message's sender address. This meant
3688 that, if the 4xx error was in fact related to the sender, a different
3689 message to the same recipient with a different sender could confuse
3690 things. In particular, this can happen when sending to a greylisting
3691 server, but other circumstances could also provoke similar problems.
3692 I have changed the default so that the retry time for these errors is now
3693 based a combination of the sender and recipient addresses. This change
3694 can be overridden by setting address_retry_include_sender=false in the
3697 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3698 remote server are returned as part of bounce messages. This was not
3699 happening for LMTP over a pipe (the lmtp transport), but now it is the
3700 same for both kinds of LMTP.
3702 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3703 in header lines that were in fact CNAMEs. This is no longer the case.
3705 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3706 and queue runs started by the daemon processed all messages. This has
3707 been fixed so that -R and -S can now usefully be given with -q<time>.
3709 PH/40 Import PCRE release 6.7 (fixes some bugs).
3711 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3713 PH/42 Give an error if -q is specified more than once.
3715 PH/43 Renamed the variables $interface_address and $interface_port as
3716 $received_ip_address and $received_port, to make it clear that these
3717 values apply to message reception, and not to the outgoing interface when
3718 a message is delivered. (The old names remain recognized, of course.)
3720 PH/44 There was no timeout on the connect() call when using a Unix domain
3721 socket in the ${readsocket expansion. There now is.
3723 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3724 be meaningful with "accept".
3726 SC/01 Eximstats V1.43
3727 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3729 SC/02 Eximstats V1.44
3730 Use a glob alias rather than an array ref in the generated
3731 parser. This improves both readability and performance.
3733 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3734 Collect SpamAssassin and rejection statistics.
3735 Don't display local sender or destination tables unless
3736 there is data to show.
3737 Added average volumes into the top table text output.
3739 SC/04 Eximstats V1.46
3740 Collect data on the number of addresses (recipients)
3741 as well as the number of messages.
3743 SC/05 Eximstats V1.47
3744 Added 'Message too big' to the list of mail rejection
3745 reasons (thanks to Marco Gaiarin).
3747 SC/06 Eximstats V1.48
3748 Mainlog lines which have GMT offsets and are too short to
3749 have a flag are now skipped.
3751 SC/07 Eximstats V1.49 (Alain Williams)
3752 Added the -emptyok flag.
3754 SC/08 Eximstats V1.50
3755 Fixes for obtaining the IP address from reject messages.
3757 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3758 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3759 whitespace changes from 4.64-PH/27
3761 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3764 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3765 are found, allow negative numbers in numeric criteria)
3767 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3769 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3770 to match changes made in 4.64-PH/43
3772 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3774 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3775 "domain" is actually an IP address, and force a failure. This locks out
3776 those revolvers/nameservers that support "A-for-A" lookups, in
3777 contravention of the specifications.
3779 PH/48 When a host name was looked up from an IP address, and the subsequent
3780 forward lookup of the name timed out, the host name was left in
3781 $sender_host_name, contrary to the specification.
3783 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3784 restricted to single-key lookups, Exim was not diagnosing an error if
3785 * or *@ was used with a query-style lookup.
3787 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3789 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3790 long ago, but noone interested enough thought of it. Let's just say that
3791 the "1.1" means that there are some new functions that weren't there at
3792 some point in the past.
3794 PH/51 Error processing for expansion failure of helo_data from an smtp
3795 transport during callout processing was broken.
3797 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3798 tested/used via the -bh/-bhc/-bs options.
3800 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3801 bug, fixed in subsequent PCRE releases).
3803 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3804 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3806 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3812 SC/01 Use a glob alias rather than an array ref in eximstats generated
3813 parser. This improves both readability and performance.
3815 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3816 Don't display local sender or destination tables in eximstats unless
3817 there is data to show.
3818 Added average volumes into the eximstats top table text output.
3820 SC/03 Collect data on the number of addresses (recipients) as well
3821 as the number of messages in eximstats.
3823 TF/01 Correct an error in the documentation for the redirect router. Exim
3824 does (usually) call initgroups() when daemonizing.
3826 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3827 with consistent privilege compared to when running as a daemon.
3829 TF/03 Note in the spec that $authenticated_id is not set for local
3830 submissions from trusted users.
3832 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3833 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3835 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3836 by adding some example configuration directives to the default
3837 configuration file. A little bit of work is required to uncomment the
3838 directives and define how usernames and passwords are checked, but
3839 there is now a framework to start from.
3841 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3842 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3843 without this. I don't know how relevant this is to other LDAP libraries.
3845 PH/02 Add the verb name to the "unknown ACL verb" error.
3847 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3849 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3851 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3852 directory not expanded when it should be if an expanded home directory
3853 was set for the address (which is overridden by the transport).
3855 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3858 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3859 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3860 because it is too late at that time, and has no effect.
3862 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3863 security issue with \' (bugzilla #107). I could not use the
3864 PQescapeStringConn() function, because it needs a PGconn value as one of
3867 PH/08 When testing addresses using -bt, indicate those final addresses that
3868 are duplicates that would not cause an additional delivery. At least one
3869 person was confused, thinking that -bt output corresponded to deliveries.
3870 (Suppressing duplicates isn't a good idea as you lose the information
3871 about possibly different redirections that led to the duplicates.)
3873 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3874 systems where poll() doesn't work, in particular OS X.
3876 PH/10 Added more information to debugging output for retry time not reached.
3878 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3879 operations in malware.c.
3881 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3884 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3885 syslog_duplication set false, log lines that would normally be written
3886 both the the main log and to the reject log were not written to syslog at
3889 PH/14 In the default configuration, change the use of "message" in ACL warn
3890 statements to "add_header".
3892 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3893 not followed by a command (e.g. "seen endif").
3895 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3896 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3899 PH/17 Added extra conditions to the default value of delay_warning_condition
3903 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3904 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3905 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3908 The Auto-Submitted: and various List- headers are standardised, whereas I
3909 don't think Precedence: ever was.
3911 PH/18 Refactored debugging code in route_finduser() to show more information,
3912 in particular, the error code if getpwnam() issues one.
3914 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3915 This is apparently needed in addition to the PH/07 change above to avoid
3916 any possible encoding problems.
3918 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3919 but not after initializing Perl.
3921 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3922 output them only if debugging. By default they are written stderr,
3923 apparently, which is not desirable.
3925 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3928 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3931 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3933 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3934 authenticated or an ident call has been made. Suppress the default
3935 values for $authenticated_id and $authenticated_sender (but permit -oMai
3936 and -oMas) when testing with -bh.
3938 PH/24 Re-jigged the order of the tests in the default configuration so that the
3939 tests for valid domains and recipients precede the DNS black list and CSA
3940 tests, on the grounds that those ones are more expensive.
3942 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3943 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3944 This bug exists in every version of Exim that I still have, right back to
3947 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3948 However, an attempt to turn on case-sensitivity in a regex key by
3949 including (?-i) didn't work because the subject string was already
3950 lowercased, and the effects were non-intuitive. It turns out that a
3951 one-line patch can be used to allow (?-i) to work as expected.
3957 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3958 other effects) broke the use of negated acl sub-conditions.
3960 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3963 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3964 "Deny" causes Exim to reject the incoming connection with a 554 error.
3965 Unfortunately, if there is a major crisis, such as a disk failure,
3966 tcp-wrappers gives "deny", whereas what one would like would be some
3967 kind of temporary error. A kludge has been added to help with this.
3968 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3969 554 error is used if errno is still zero or contains ENOENT (which occurs
3970 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3973 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3975 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3976 errors. Otherwise a message that provokes a temporary error (when other
3977 messages do not) can cause a whole host to time out.
3979 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3980 addresses were routed directly to files or pipes from a redirect router.
3981 File deliveries just didn't batch; pipe deliveries might have suffered
3984 PH/06 A failure to get a lock for a hints database would erroneously always say
3985 "Failed to get write lock", even when it was really a read lock.
3987 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3988 of 0600. This has been changed to use the value of the lockfile_mode
3989 option (which defaults to 0600).
3991 PH/08 Applied small patch from the Sieve maintainer.
3993 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3994 folder from quota calculations, a direct delivery into this folder messed
3995 up the contents of the maildirsize file. This was because the regex was
3996 used only to exclude .Trash (or whatever) when the size of the mailbox
3997 was calculated. There was no check that a delivery was happening into an
3998 excluded directory. This bug has been fixed by ignoring all quota
3999 processing for deliveries into excluded directories.
4001 PH/10 Added the maildirfolder_create_regex option to appendfile.
4007 PH/01 The code for finding all the local interface addresses on a FreeBSD
4008 system running IPv6 was broken. This may well have applied to all BSD
4009 systems, as well as to others that have similar system calls. The broken
4010 code found IPv4 interfaces correctly, but gave incorrect values for the
4011 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4012 that it would not match correctly against @[] and not recognize the IPv6
4015 PH/02 The ipliteral router was not recognizing addresses of the form user@
4016 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4018 PH/03 Added disable_ipv6.
4020 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4021 decoded form, because it is most often used to construct To: headers
4022 lines in autoreplies, and the decoded form may well be syntactically
4023 invalid. However, $reply_address has leading white space removed, and all
4024 newlines turned into spaces so that the autoreply transport does not
4027 PH/05 If group was specified without a user on a router, and no group or user
4028 was specified on a transport, the group from the router was ignored.
4030 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4031 for visible compile-time settings that can be used to change these
4032 numbers, for those that want even more. Backwards compatibility with old
4033 spool files has been maintained. However, going back to a previous Exim
4034 release will lost any variables that are in spool files.
4036 PH/07 Two small changes when running in the test harness: increase delay when
4037 passing a TCP/IP connection to a new process, in case the original
4038 process has to generate a bounce, and remove special handling of
4039 127.0.0.2 (sic), which is no longer necessary.
4041 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4042 be the same on different OS.
4044 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4047 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4048 whether --show-vars was specified or not
4050 JJ/02 exipick: Added support for new ACL variable spool format introduced
4053 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4054 syntactically invalid From: or Reply-to: line, and a filter used this to
4055 generate an autoreply, and therefore failed to obtain an address for the
4056 autoreply, Exim could try to deliver to a non-existent relative file
4057 name, causing unrelated and misleading errors. What now happens is that
4058 it logs this as a hard delivery error, but does not attempt to create a
4061 PH/11 The exinext utility has a -C option for testing purposes, but although
4062 the given file was scanned by exinext itself; it wasn't being passed on
4063 when Exim was called.
4065 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4066 an end-of-file indication when reading a command response.
4068 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4069 compiled. In many other places in Exim, IPv6 addresses are always
4070 recognized, so I have changed this. It also means that IPv4 domain
4071 literals of the form [IPV4:n.n.n.n] are now always recognized.
4073 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4074 used if the router is not running as root, for example, when verifying at
4075 ACL time, or when using -bh. The debugging output from this situation was
4076 non-existent - all you got was a failure to exec. I have made two
4079 (a) Failures to set uid/gid, the current directory, or a process leader
4080 in a subprocess such as that created by queryprogram now generate
4081 suitable debugging output when -d is set.
4083 (b) The queryprogram router detects when it is not running as root,
4084 outputs suitable debugging information if -d is set, and then runs
4085 the subprocess without attempting to change uid/gid.
4087 PH/15 Minor change to Makefile for building test_host (undocumented testing
4090 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4091 additional section of a DNS packet that returns MX or SRV records.
4092 Instead, it always explicitly searches for A/AAAA records. This avoids
4093 major problems that occur when a DNS server includes only records of one
4094 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4095 fixed another bug: if SRV records were looked up and the corresponding
4096 address records were *not* found in the additional section, the port
4097 values from the SRV records were lost.
4099 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4100 using the correct key (the original address) when searching the retry
4101 rules in order to find which one to use for generating the retry hint.
4103 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4104 adding the default one. Similarly, if it contains a Reply-To: header, the
4105 errors_reply_to option, if set, is not used.
4107 PH/19 When calculating a retry time, Exim used to measure the "time since
4108 failure" by looking at the "first failed" field in the retry record. Now
4109 it does not use this if it is later than than the arrival time of the
4110 message. Instead it uses the arrival time. This makes for better
4111 behaviour in cases where some deliveries succeed, thus re-setting the
4112 "first failed" field. An example is a quota failure for a huge message
4113 when small messages continue to be delivered. Without this change, the
4114 "time since failure" will always be short, possible causing more frequent
4115 delivery attempts for the huge message than are intended.
4116 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4118 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4119 $1, $2, $3) because the numerical variables can be reset during some
4120 expansion items (e.g. "match"), thereby losing the authentication data.
4122 PH/21 Make -bV show the size of off_t variables so that the test suite can
4123 decide whether to run tests for quotas > 2G.
4125 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4126 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4127 filecount value is greater than 2G or if a quota value is greater than 2G
4128 on a system where the size of off_t is not greater than 4, a panic error
4131 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4132 never match. The debug and -bh output now contains an explicit error
4133 message indicating a malformed IPv4 address or mask.
4135 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4136 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4137 PH/23 above applies.
4139 PH/25 Do not write to syslog when running in the test harness. The only
4140 occasion when this arises is a failure to open the main or panic logs
4141 (for which there is an explicit test).
4143 PH/26 Added the /no_tell option to "control=freeze".
4145 PH/27 If a host name lookup failed very early in a connection, for example, if
4146 the IP address matched host_lookup and the reverse lookup yielded a name
4147 that did not have a forward lookup, an error message of the form "no IP
4148 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4149 could be logged. Now it outputs the IP address instead of "NULL".
4151 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4152 allows the sender and the authenticated sender to be set when
4153 submitting a message from within Exim. Since child_open_exim() is
4154 documented for local_scan(), the new function should be too.
4156 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4157 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4158 results in an empty string is now treated as unset.
4160 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4162 PH/31 Added sender_verify_fail logging option.
4164 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4165 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4166 tidied the source and removed it altogether.
4168 PH/33 When a queue run was abandoned because the load average was too high, a
4169 log line was always written; now it is written only if the queue_run log
4170 selector is set. In addition, the log line for abandonment now contains
4171 information about the queue run such as the pid. This is always present
4172 in "start" and "stop" lines but was omitted from the "abandon" line.
4174 PH/34 Omit spaces between a header name and the colon in the error message that
4175 is given when verify = headers_syntax fails (if there are lots of them,
4176 the message gets confusing).
4178 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4179 names, as there are now some PTR records that contain slashes. This check
4180 is only to protect against broken name servers that fall over on strange
4181 characters, so the fact that it applies to all lookups doesn't matter.
4183 PH/36 Now that the new test suite is complete, we can remove some of the
4184 special code in Exim that was needed for the old test suite. For example,
4185 sorting DNS records because real resolvers return them in an arbitrary
4186 order. The new test suite's fake resolver always returns records in the
4189 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4190 bounces) except when queue_only is set, to avoid logging races between
4191 the different processes.
4193 PH/38 Panic-die if .include specifies a non-absolute path.
4195 PH/39 A tweak to the "H" retry rule from its user.
4197 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4198 a label. They prevented compilation on older perls.
4200 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4201 a warning to be raised on newish perls.
4203 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4204 on queue. Changes to match documented behaviour of showing count of
4205 messages matching specified criteria.
4207 PH/40 Changed the default ident timeout from 30s to 5s.
4209 PH/41 Added support for the use of login_cap features, on those BSD systems
4210 that have them, for controlling the resources used by pipe deliveries.
4212 PH/42 The content-scanning code uses fopen() to create files in which to put
4213 message data. Previously it was not paying any attention to the mode of
4214 the files. Exim runs with umask(0) because the rest of the code creates
4215 files with open(), and sets the required mode explicitly. Thus, these
4216 files were ending up world-writeable. This was not a big issue, because,
4217 being within the spool directory, they were not world-accessible. I have
4218 created a function called modefopen, which takes an additional mode
4219 argument. It sets umask(777), creates the file, chmods it to the required
4220 mode, then resets the umask. All the relevant calls to fopen() in the
4221 content scanning code have been changed to use this function.
4223 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4224 to 24 hours. This avoids potential overflow problems when processing G
4225 and H retry rules. I suspect nobody ever tinkers with this value.
4227 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4229 PH/45 When the plaintext authenticator is running as a client, the server's
4230 challenges are checked to ensure they are valid base64 strings. By
4231 default, the authentication attempt is cancelled if an invalid string is
4232 received. Setting client_ignore_invalid_base64 true ignores these errors.
4233 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4234 they are received. Thus, the responses can be made to depend on the
4235 challenges. If an invalid string is ignored, an empty string is placed in
4238 PH/46 Messages that are created by the autoreply transport now contains a
4239 References: header, in accordance with RFCs 2822 and 3834.
4241 PH/47 Added authenticated_sender_force to the smtp transport.
4243 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4245 PH/49 Installed latest patch from the Sieve maintainer.
4247 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4248 was also set, the appendfile transport was unnecessarily scanning a
4249 directory of message files (e.g. for maildir delivery) to find the count
4250 of files (along with the size), even though it did not need this
4251 information. It now does the scan only if it needs to find either the
4252 size of the count of files.
4254 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4256 PH/52 Two bugs concerned with error handling when the smtp transport is
4259 (i) Exim was not creating retry information for temporary errors given
4260 for individual recipients after the DATA command when the smtp transport
4261 was used in LMTP mode. This meant that they could be retried too
4262 frequently, and not timed out correctly.
4264 (ii) Exim was setting the flag that allows error details to be returned
4265 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4266 recipients that were returned after the DATA command.
4268 PH/53 This is related to PH/52, but is more general: for any failing address,
4269 when detailed error information was permitted to be returned to the
4270 sender, but the error was temporary, then after the final timeout, only
4271 "retry timeout exceeded" was returned. Now it returns the full error as
4272 well as "retry timeout exceeded".
4274 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4275 do this, and (what is worse) MTAs that accept it.
4277 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4278 will now be deprecated.
4280 PH/56 New os.c-cygwin from the Cygwin maintainer.
4282 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4283 formats (previously only available in exim formats via -bpr, -bpru,
4284 and -bpra. Now also available in native and exiqgrep formats)
4286 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4287 with very large, slow to parse queues
4289 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4291 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4293 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4294 responses to authentication challenges, though it was showing the
4295 challenges; (ii) I've removed the CR characters from the debug output for
4298 PH/58 Allow for the insertion of a newline as well as a space when a string
4299 is turned into more than one encoded-word during RFC 2047 encoding. The
4300 Sieve code now uses this.
4302 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4303 data_4xx, lost_connection, tls_required.
4305 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4306 message was being sent as an SMTP response.
4308 PH/61 Add -l and -k options to exicyclog.
4310 PH/62 When verifying, if an address was redirected to one new address, so that
4311 verification continued, and the new address failed or deferred after
4312 having set something in $address_data, the value of $address_data was not
4313 passed back to the ACL. This was different to the case when no
4314 redirection occurred. The value is now passed back in both cases.
4316 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4317 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4318 use login_cap.h, so on its own it isn't the distinguishing feature. The
4319 new name refers directly to the setclassresources() function.
4321 PH/65 Added configuration files for NetBSD3.
4323 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4325 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4326 is preferred over IPv4.
4328 PH/68 The bounce_return_message and bounce_return_body options were not being
4329 honoured for bounces generated during the reception of non-SMTP messages.
4330 In particular, this applied to messages rejected by the ACL. This bug has
4331 been fixed. However, if bounce_return_message is true and bounce_return_
4332 body is false, the headers that are returned for a non-SMTP message
4333 include only those that have been read before the error was detected.
4334 (In the case of an ACL rejection, they have all been read.)
4336 PH/69 The HTML version of the specification is now built in a directory called
4337 spec_html instead of spec.html, because the latter looks like a path with
4338 a MIME-type, and this confuses some software.
4340 PH/70 Catch two compiler warnings in sieve.c.
4342 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4343 function verify_get_ident() calls ip_connect() to connect a socket, but
4344 if the "connect()" function timed out, ip_connect() used to close the
4345 socket. However, verify_get_ident() also closes the socket later, and in
4346 between Exim writes to the log, which may get opened at this point. When
4347 the socket was closed in ip_connect(), the log could get the same file
4348 descriptor number as the socket. This naturally causes chaos. The fix is
4349 not to close the socket in ip_connect(); the socket should be closed by
4350 the function that creates it. There was only one place in the code where
4351 this was missing, in the iplookup router, which I don't think anybody now
4352 uses, but I've fixed it anyway.
4354 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4355 well as to direct DNS lookups. Otherwise the handling of names in host
4356 lists is inconsistent and therefore confusing.
4362 PH/01 Two changes to the default runtime configuration:
4364 (1) Move the checks for relay_from_hosts and authenticated clients from
4365 after to before the (commented out) DNS black list checks.
4367 (2) Add control=submission to the relay_from_hosts and authenticated
4368 clients checks, on the grounds that messages accepted by these
4369 statements are most likely to be submissions.
4371 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4373 (1) Generate an error if the third argument for the ${prvs expansion is
4376 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4379 (3) Reset the variables that are obtained from the first argument of
4380 ${prvscheck and used in the second argument before leaving the code,
4381 because their memory is reclaimed, so using them afterwards may do
4384 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4385 one (it's much easier than Tom thought :-).
4387 (5) Because of (4), we can now allow for the use of $prvscheck_result
4388 inside the third argument.
4390 PH/03 For some reason, the default setting of PATH when running a command from
4391 a pipe transport was just "/usr/bin". I have changed it to
4394 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4395 anything to be listed in the output from -bV.
4397 PH/05 When a filter generated an autoreply, the entire To: header line was
4398 quoted in the delivery log line, like this:
4400 => >A.N.Other <ano@some.domain> <original@ddress> ...
4402 This has been changed so that it extracts the operative address. There
4403 may be more than one such address. If so, they are comma-separated, like
4406 => >ano@some.domain,ona@other.domain <original@ddress> ...
4408 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4409 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4410 not being looked up in the rDNS to get a host name, Exim was showing the
4411 IP address twice in Received: lines, even though the IP addresses were
4412 identical. For example:
4414 Received: from [1.2.3.4] (helo=[1.2.3.4])
4416 However, if the real host name was known, it was omitting the HELO data
4417 if it matched the actual IP address. This has been tidied up so that it
4418 doesn't show the same IP address twice.
4420 PH/07 When both +timestamp and +memory debugging was on, the value given by
4421 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4422 called by the debug printing, thereby overwriting the timestamp buffer.
4423 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4425 PH/08 When the original message was included in an autoreply transport, it
4426 always said "this is a copy of the message, including all the headers",
4427 even if body_only or headers_only was set. It now gives an appropriate
4430 PH/09 Applied a patch from the Sieve maintainer which:
4432 o fixes some comments
4433 o adds the (disabled) notify extension core
4434 o adds some debug output for the result of if/elsif tests
4435 o points to the current vacation draft in the documentation
4436 and documents the missing references header update
4440 o fixes a bug in processing the envelope test (when testing
4441 multiple envelope elements, the last element determined the
4444 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4445 Electronic Mail") by including:
4447 Auto-submitted: auto-generated
4449 in the messages that it generates (bounce messages and others, such as
4450 warnings). In the case of bounce messages for non-SMTP messages, there was
4451 also a typo: it was using "Auto_submitted" (underscore instead of
4452 hyphen). Since every message generated by Exim is necessarily in response
4453 to another message, thes have all been changed to:
4455 Auto-Submitted: auto-replied
4457 in accordance with these statements in the RFC:
4459 The auto-replied keyword:
4461 - SHOULD be used on messages sent in direct response to another
4462 message by an automatic process,
4464 - MUST NOT be used on manually-generated messages,
4466 - MAY be used on Delivery Status Notifications (DSNs) and Message
4467 Disposition Notifications (MDNs),
4469 - MUST NOT be used on messages generated by automatic or periodic
4470 processes, except for messages which are automatic responses to
4473 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4474 to the default Received: header definition.
4476 PH/12 Added log selector acl_warn_skipped (default on).
4478 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4479 variables because (a) they are in the wrong storage pool and (b) even if
4480 they were copied, it wouldn't work properly because of the caching.
4482 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4483 checking when decoding. Apparently there are clients that generate
4484 overlong encoded strings. Why am I not surprised?
4486 PH/15 If the first argument of "${if match_address" was not empty, but did not
4487 contain an "@" character, Exim crashed. Now it writes a panic log message
4488 and treats the condition as false.
4490 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4492 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4493 extension "envelope-auth". The code is finished and in agreement with
4494 other implementations, but there is no documentation so far and in fact,
4495 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4496 not changing the active code.
4498 Print executed "if" and "elsif" statements when debugging is used. This
4499 helps a great deal to understand what a filter does.
4501 Document more things not specified clearly in RFC3028. I had all this
4502 sorted out, when out of a sudden new issues came to my mind. Oops."
4504 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4507 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4508 canonical form (as documented). However, after a host name lookup from
4509 the IP address, check_host() was doing a simple string comparison with
4510 addresses acquired from the DNS when checking that the found name did
4511 have the original IP as one of its addresses. Since any found IPv6
4512 addresses are likely to be in abbreviated form, the comparison could
4513 fail. Luckily, there already exists a function for doing the comparison
4514 by converting both addresses to binary, so now that is used instead of
4515 the text comparison.
4517 PH/20 There was another similar case to PH/19, when a complete host name was
4518 given in a host list; looking up its IP address could give an abbreviated
4519 form, whereas the current host's name might or might not be abbreviated.
4520 The same fix has been applied.
4526 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4527 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4530 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4531 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4533 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4535 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4536 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4537 Also, for "Auto-Submitted", treat anything other than "no" as
4538 non-personal, in accordance with RFC 3834. (Previously it treated
4539 anything starting "auto-" as non-personal.)
4541 TF/01 The control=submission/name=... option had a problem with syntax
4542 errors if the name included a slash character. The /name= option
4543 now slurps the rest of the string, so it can include any characters
4544 but it must come last in the list of options (after /sender_retain
4547 PH/05 Some modifications to the interface to the fake nameserver for the new
4555 TK/01 Added the "success_on_redirect" address verification option. See
4556 NewStuff for rationale and an example.
4558 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4560 PH/02 Patch to exigrep to allow it to work on syslog lines.
4562 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4563 fread() to skip over the body file's header line, because in Cygwin the
4564 header line is locked and is inaccessible.
4566 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4567 co-exist for some time) to make it clear that it is the Exim ID that is
4568 referenced, not the Message-ID: header line.
4570 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4571 string_format() function, because snprintf() does not exist on all
4574 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4575 ${stat: expansion item.
4577 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4578 protocol synchronization error", to keep the pedants happy.
4580 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4581 well as for IRIX systems, when gcc is being used. See the host.c source
4584 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4586 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4589 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4590 transport and to the smtp transport in LMTP mode.
4592 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4594 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4595 run a filter in a subprocess. This could lead to confusion in subsequent
4596 lookups in the parent process. There should also be a search_tidyup() at
4597 the end of the subprocess.
4599 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4600 only if the host matched helo_try_verify_hosts, which caused the
4601 verification to occur when the EHLO/HELO command was issued. The ACL just
4602 tested the remembered result. Now, if a previous verification attempt has
4603 not happened, "verify = helo" does it there and then.
4605 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4607 TK/03 Fix log output including CR from clamd.
4609 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4610 exist provoked a memory error which could cause a segfault.
4612 PH/15 Installed PCRE 6.2
4614 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4616 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4617 of the problem. Specifically, suggested +O2 rather than +O1 for the
4620 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4622 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4623 record, the port was not correctly logged when the outgoing_port log
4624 selector was set (it logged the transort's default port).
4626 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4627 fallback_hosts, and "hosts" in the smtp transport.
4629 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4630 host errors such as "Connection refused".
4632 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4633 authentication with radiusclient 0.4.9:
4635 - Error returned from rc_read_config was caught wrongly
4636 - Username/password not passed on to radius server due to wrong length.
4638 The presumption is that some radiusclient API changes for 4.51/PH/17
4639 were not taken care of correctly. The code is still untested by me (my
4640 Linux distribution still has 0.3.2 of radiusclient), but it was
4641 contributed by a Radius user.
4643 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4644 expanding the "port" option of the smtp transport.
4646 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4647 while reading a MIME header. Thanks to Tom Hughes for a patch.
4649 PH/24 Include config.h inside local_scan.h so that configuration settings are
4652 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4653 that in an ACL for RCPT (for example), you can examine exactly what was
4656 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4657 commands, but it was not correctly comparing the address with the actual
4658 client host address. Thus, it would show the EHLO address in Received:
4659 header lines when this was not necessary.
4661 PH/27 Added the % operator to ${eval:}.
4663 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4664 it should be ignoring failures (because with -C, for example, it has lost
4665 privilege). It wasn't ignoring creation failures other than "already
4668 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4671 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4672 because an input line was too long, either on its own, or by virtue of
4673 too many continuations, the temporary file was not being removed, and the
4674 return code was incorrect.
4676 PH/31 Missing "BOOL" in function definition in filtertest.c.
4678 PH/32 Applied Sieve patches from the maintainer.
4680 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4682 PH/33 Added "verify = not_blind".
4684 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4685 Local/Makefile (with some defaults set). These are used in built scripts
4686 such as exicyclog, but they have never been used in the exim_install
4687 script (though there are many overriding facilities there). I have
4688 arranged that the exim_install script now takes note of these two
4691 PH/35 Installed configuration files for Dragonfly.
4693 PH/36 When a locally submitted message by a trusted user did not contain a
4694 From: header, and the sender address was obtained from -f or from an SMTP
4695 MAIL command, and the trusted user did not use -F to supply a sender
4696 name, $originator_name was incorrectly used when constructing a From:
4697 header. Furthermore, $originator_name was used for submission mode
4698 messages from external hosts without From: headers in a similar way,
4699 which is clearly wrong.
4701 PH/37 Added control=suppress_local_fixups.
4703 PH/38 When log_selector = +received_sender was set, and the addition of the
4704 sender made the log line's construction buffer exactly full, or one byte
4705 less than full, an overflow happened when the terminating "\n" was
4708 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4709 when the result of a list match is failure because a DNS lookup failed.
4711 PH/40 RM_COMMAND is now used in the building process.
4713 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4714 the "build-* directories that it finds.
4716 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4717 address in a domain literal was a prefix of an interface address.
4719 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4720 when verifying a sender address, unless rewrite_headers is false.
4722 PH/44 Wrote a long comment about why errors_to addresses are verified as
4723 recipients, not senders.
4725 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4726 the ratelimit ACL was added.
4728 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4730 PH/46 Added extra information about PostgreSQL errors to the error string.
4732 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4733 avoiding the need to install special zones in a real server. This is
4734 backwards compatible; if it can't find the fake resolver, it drops back.
4735 Thus, both old and new test suites can be run.
4737 TF/02 Added util/ratelimit.pl
4739 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4740 clock is set back in time.
4742 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4743 Candler <B.Candler@pobox.com>.
4745 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4746 OK for addresses that are the result of redirections.
4748 PH/48 A number of further additions for the benefit of the new test suite,
4749 including a fake gethostbyname() that interfaces to the fake DNS resolver
4752 TF/06 The fix for widen_domains has also been applied to qualify_single and
4753 search_parents which are the other dnslookup options that can cause
4756 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4759 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4761 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4762 service" if it encountered an MX record with an empty target hostname.
4763 The message is now "an MX or SRV record indicated no SMTP service".
4765 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4766 if the DNS of the sending site is misconfigured. This is quite a
4767 common situation. This change restores the behaviour of treating a
4768 helo verification defer as a failure.
4770 PH/51 If self=fail was set on a router, the bounce message did not include the
4771 actual error message.
4777 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4779 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4780 command itself ended in error, the underlying message about the transport
4781 filter timeout was being overwritten with the pipe command error. Now the
4782 underlying error message should be appended to the second error message.
4784 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4786 PH/02 Reduce the amount of output that "make" produces by default. Full output
4787 can still be requested.
4789 PH/03 The warning log line about a condition test deferring for a "warn" verb
4790 was being output only once per connection, rather than after each
4791 occurrence (because it was using the same function as for successful
4792 "warn" verbs). This seems wrong, so I have changed it.
4794 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4795 it should not have, which might have caused a crash in the right
4796 circumstances, but probably never did.
4798 PH/04 Installed a modified version of Tony Finch's patch to make submission
4799 mode fix the return path as well as the Sender: header line, and to
4800 add a /name= option so that you can make the user's friendly name appear
4803 TF/03 Added the control = fakedefer ACL modifier.
4805 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4806 Mark Lowes for thorough testing.
4808 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4810 TK/03 Merged latest SRS patch from Miles Wilton.
4812 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4813 with the definition in sysexits.h (which is #included earlier).
4814 Fortunately, Exim does not actually use EX_OK. The code used to try to
4815 preserve the sysexits.h value, by assuming that macro definitions were
4816 scanned for macro replacements. I have been disabused of this notion,
4817 so now the code just undefines EX_OK before #including unistd.h.
4819 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4820 in the smtp transport. When a block could not be written in a single
4821 write() function, the timeout was being re-applied to each part-write.
4822 This seems wrong - if the receiver was accepting one byte at a time it
4823 would take for ever. The timeout is now adjusted when this happens. It
4824 doesn't have to be particularly precise.
4826 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4827 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4829 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4830 without a domain if the check to <postmaster@domain> fails.
4832 SC/01 Eximstats: added -xls and the ability to specify output files
4833 (patch written by Frank Heydlauf).
4835 SC/02 Eximstats: use FileHandles for outputting results.
4837 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4839 SC/04 Eximstats: fixed display of large numbers with -nvr option
4841 SC/05 Eximstats: fixed merging of reports with empty tables.
4843 SC/06 Eximstats: added the -include_original_destination flag
4845 SC/07 Eximstats: removed tabs and trailing whitespace.
4847 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4849 TK/06 MBOX spool code: Add real "From " MBOX separator line
4850 so the .eml file is really in mbox format (even though
4851 most programs do not really care). Patch from Alex Miller.
4853 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4854 The latter is generated from $received_to and is only set if the
4855 message has one envelope recipient. SA can use these headers,
4856 obviously out-of-the-box. Patch from Alex Miller.
4858 PH/08 The ${def test on a variable was returning false if the variable's
4859 value was "0", contrary to what the specification has always said!
4860 The result should be true unless the variable is empty.
4862 PH/09 The syntax error of a character other than { following "${if
4863 def:variable_name" (after optional whitespace) was not being diagnosed.
4864 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4865 accidental colon was present, for example, could give incorrect results.
4867 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4868 result is used (not including appendfile, where other changes are about
4871 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4872 This involved changing a lot of size variables from int to off_t. It
4873 should work with maildirs and everything.
4875 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4876 spamd dying while we are connected to it.
4878 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4881 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4882 function for BDB 4.3.
4884 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4886 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4887 the sources for PCRE have been re-arranged and more files are now
4890 PH/15 The code I had for printing potentially long long variables in PH/11
4891 above was not the best (it lost precision). The length of off_t variables
4892 is now inspected at build time, and an appropriate printing format (%ld
4893 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4894 to be "long long int" or "long int". This is needed for the internal
4895 formatting function string_vformat().
4897 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4898 the configuration file to be ":syslog", then the script "guesses" where
4899 the logs files are, rather than using the compiled in default. In our
4900 case the guess is not the same as the compiled default, so the script
4901 suddenly stopped working when I started to use syslog. The patch checks
4902 to see if log_file_path is "". If so, it attempts to read it from exim
4903 with no configuration file to get the compiled in version, before it
4904 falls back to the previous guessing code."
4906 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4907 implementing BATV in an Exim configuration. See NewStuff for the gory
4910 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4911 Makefile that are specific to HP-UX.
4913 PH/18 If the "use_postmaster" option was set for a recipient callout together
4914 with the "random" option, the postmaster address was used as the MAIL
4915 FROM address for the random test, but not for the subsequent recipient
4916 test. It is now used for both.
4918 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4919 patch removes a few documentation additions to RFC 3028, because the
4920 latest draft now contains them. It adds the new en;ascii-case comparator
4921 and a new error check for 8bit text in MIME parts. Comparator and
4922 require names are now matched exactly. I enabled the subaddress
4923 extension, but it is not well tested yet (read: it works for me)."
4925 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4926 rework some of the code of TK/09 above to avoid the hardwired use of
4927 "%lld" and "long long". Replaced the call to snprintf() with a call to
4930 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4931 records point to non-existent hosts", "retry timeout exceeded", and
4932 "retry time not reached for any host after a long failure period".
4934 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4935 experimental DomainKeys support:
4937 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4938 (2) On an error such as an illegally used "control", the wrong name for
4939 the control was given.
4941 These problems did NOT occur unless DomainKeys support was compiled.
4943 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4945 PH/24 Added ${if match_ip condition.
4947 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4948 databases so that it will be absolutely obvious if a crash occurs in the
4949 DB library. This is a regular occurrence (often caused by mis-matched
4952 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4953 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4954 picked up on a user's system that detects such things. There doesn't seem
4955 to be a gcc warning option for this - only an attribute that has to be
4956 put on the function's prototype. It seems that in Fedora Core 4 they have
4957 set this on a number of new functions. No doubt there will be more in due
4960 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4961 specify a transport. However, if an address that was verified by such a
4962 router was the subject of a callout, Exim crashed because it tried to
4963 read the rcpt_include_affixes from the non-existent transport. Now it
4964 just assumes that the setting of that option is false. This bug was
4965 introduced by 4.51/PH/31.
4967 PH/28 Changed -d+all to exclude +memory, because that information is very
4968 rarely of interest, but it makes the output a lot bigger. People tend to
4969 do -d+all out of habit.
4971 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4972 code in os-type was giving problems when libc.so lives in lib64, like on
4975 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4976 aren't the modern standard, and it seems that some systems' include files
4977 don't always have them. Exim was already checking for some of the newer
4978 ones like T_AAAA, and defining it itself. I've added checks for all the
4979 record types that Exim uses.
4981 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4982 not automatically generating a new one, as it is supposed to. This
4983 prevented TLS from working. If the file did exist, but contained invalid
4984 data, a new version was generated, as expected. It was only the case of a
4985 non-existent file that was broken.
4987 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4988 with a change in libdomainkeys > 0.64.
4990 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4991 from DNS. If the selector record carries the flag, it now has
4992 precedence over the domain-wide flag.
4994 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4996 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4997 the use of a port name that isn't defined in /etc/services) occurred, the
4998 message was deferred as in a normal delivery, and thus remained on the
4999 spool, instead of being failed because of the mua_wrapper setting. This
5000 is now fixed, and I tidied up some of the mua_wrapper messages at the
5003 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5004 the messages in a hash of arrays rather than using individual hashes.
5005 This is a bit cleaner and results in dramatic memory savings, albeit
5006 at a slight CPU cost.
5008 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5009 as requested by Marc Sherman.
5011 SC/10 Eximstats: added histograms for user specified patterns as requested
5014 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5016 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5017 fopen() in the content-scanning modules that did not already have it.
5023 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5024 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5026 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5028 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5030 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5031 to test Sieve filters that use "vacation".
5033 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5034 that changes the way the GnuTLS parameters are stored in the cache file.
5035 The new format can be generated externally. For backward compatibility,
5036 if the data in the cache doesn't make sense, Exim assumes it has read an
5037 old-format file, and it generates new data and writes a new file. This
5038 means that you can't go back to an older release without removing the
5041 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5042 work if there are any delivery delays because "one_time" forces the
5043 parent to be marked "delivered", so its unseen clone is never tried
5044 again. For this reason, Exim now forbids the simultaneous setting of
5047 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5048 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5049 entry if you want to know the details. The fix, however, neglected to
5050 consider the case when local delivery batching is involved. The test for
5051 "previously delivered" was not happening when checking to see if an
5052 address could be batched with a previous (undelivered) one; under
5053 certain circumstances this could lead to multiple deliveries to the same
5056 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5057 in its include files, and this causes problems building Exim.
5059 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5060 header_syntax) but Exim was just ignoring anything given after a slash.
5061 In particular, this caused confusion with an attempt to use "verify =
5062 reverse_host_lookup/defer_ok". An error is now given when options are
5063 supplied for verify items that do not have them. (Maybe reverse_host_
5064 lookup should have a defer_ok option, but that's a different point.)
5066 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5067 defined by RFC 821) to 2048, because there were problems with some AUTH
5068 commands, and RFC 1869 says the size should be increased for extended
5069 SMTP commands that take arguments.
5071 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5074 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5075 "unknown" error; now it says that the functionality isn't in the binary.
5077 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5078 an address' error message when a string expansion fails (syntax or
5079 whatever). Otherwise the password may appear in the log. Following change
5080 PH/42 below, there is no longer a chance of it appearing in a bounce
5083 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5085 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5086 of its addresses was ever tried. (Bugzilla bug #2.)
5088 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5089 the result incorrectly in the debug output. (It correctly added a newline
5090 to what was transported.)
5092 TF/01 Added $received_time.
5094 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5095 commented out examples of how to interface to a virus scanner and to
5096 SpamAssassin. Also added commented examples of av_scanner and
5097 spamd_address settings.
5099 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5100 and controls are allowed in which ACLs. There were a couple of minor
5101 errors. Some of the entries in the conditions table (which is a table of
5102 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5103 negation of where the condition IS allowed.
5105 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5107 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5108 header file does not have a version number, so I've had to invent a new
5109 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5110 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5111 radiusclient), but it was contributed by a Radius user.
5113 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5114 files or directories, for OpenSSL.
5116 PH/19 When an Exim process that is running as root has to create an Exim log
5117 file, it does so in a subprocess that runs as exim:exim so as to get the
5118 ownership right at creation (otherwise, other Exim processes might see
5119 the file with the wrong ownership). There was no test for failure of this
5120 fork() call, which would lead to the process getting stuck as it waited
5121 for a non-existent subprocess. Forks do occasionally fail when resources
5122 run out. I reviewed all the other calls to fork(); they all seem to check
5125 PH/20 When checking for unexpected SMTP input at connect time (before writing
5126 the banner), Exim was not dealing correctly with a non-positive return
5127 from the read() function. If the client had disconnected by this time,
5128 the result was a log entry for a synchronization error with an empty
5129 string after "input=" when read() returned zero. If read() returned -1
5130 (an event I could not check), uninitialized data bytes were printed.
5131 There were reports of junk text (parts of files, etc) appearing after
5134 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5136 PH/22 Added support for macro redefinition, and (re)definition in between
5137 driver and ACL definitions.
5139 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5140 forgetting to use the resulting value; it was using the unexpanded value.
5142 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5143 hadn't been configured. The fix is from Juergen Kreileder, who
5144 understands it better than I do:
5146 "Here's what I see happening with three configured cyrus_sasl
5147 authenticators configured (plain, login, cram-md5):
5149 On startup auth_cyrus_sasl_init() gets called for each of these.
5150 This means three calls to sasl_listmech() without a specified mech_list.
5151 => SASL tests which mechs of all available mechs actually work
5152 => three warnings about OTP not working
5153 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5155 With the patch, sasl_listmech() also gets called three times. But now
5156 SASL's mech_list option is set to the server_mech specified in the the
5157 authenticator. Or in other words, the answer from sasl_listmech()
5158 gets limited to just the mech you're testing for (which is different
5160 => the return list contains just 'plain' or 'login', 'cram-md5' or
5161 nothing depending on the value of ob->server_mech.
5163 I've just tested the patch: Authentication still works fine,
5164 unavailable mechs specified in the exim configuration are still
5165 caught, and the auth.log warnings about OTP are gone."
5167 PH/25 When debugging is enabled, the contents of the command line are added
5168 to the debugging output, even when log_selector=+arguments is not
5171 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5172 answer is "GNU", and only if the return is "GNU/something" is the answer
5175 PH/27 $acl_verify_message is now set immediately after the failure of a
5176 verification in an ACL, and so is available in subsequent modifiers. In
5177 particular, the message can be preserved by coding like this:
5179 warn !verify = sender
5180 set acl_m0 = $acl_verify_message
5182 Previously, $acl_verify_message was set only while expanding "message"
5183 and "log_message" when a very denied access.
5185 PH/28 Modified OS/os.c-Linux with
5187 -#ifndef OS_LOAD_AVERAGE
5188 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5190 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5191 nomenclature these days.)
5193 PH/29 Installed patch from the Sieve maintainer that adds the options
5194 sieve_useraddress and sieve_subaddress to the redirect router.
5196 PH/30 In these circumstances:
5197 . Two addresses routed to the same list of hosts;
5198 . First host does not offer TLS;
5199 . First host accepts first address;
5200 . First host gives temporary error to second address;
5201 . Second host offers TLS and a TLS session is established;
5202 . Second host accepts second address.
5203 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5204 and peerdn, if requested) that were in fact used only for the second
5207 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5208 attention to any local part prefix or suffix that was matched by the
5209 router that accepted the address. It now behaves in the same way as it
5210 does for delivery: the affixes are removed from the local part unless
5211 rcpt_include_affixes is set on the transport.
5213 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5214 timeout during the DATA phase of an incoming message.
5216 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5217 applied a patch sanctioned by the Sieve maintainer.
5219 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5220 the uid or gid is negative. A case of a negative gid caused this to be
5221 noticed. The fix allows for either to be negative.
5223 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5224 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5227 PH/36 The change PH/12 above was broken. Fixed it.
5229 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5230 the grounds that routing the same address twice would always produce the
5231 same answer. This might have been true once, but it is certainly no
5232 longer true now. Routing a child address may depend on the previous
5233 routing that produced that child. Some complicated redirection strategies
5234 went wrong when messages had multiple recipients, and made Exim's
5235 behaviour dependent on the order in which the addresses were given.
5237 I have moved the duplicate checking until after the routing is complete.
5238 Exim scans the addresses that are assigned to local and remote
5239 transports, and removes any duplicates. This means that more work will be
5240 done, as duplicates will always all be routed, but duplicates are
5241 presumably rare, so I don't expect this is of any significance.
5243 For deliveries to pipes, files, and autoreplies, the duplicate checking
5244 still happens during the routing process, since they are not going to be
5247 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5248 It corrects a timeout issue with spamd. This is Ian's comment: "The
5249 background is that sometimes spamd either never reads data from a
5250 connection it has accepted, or it never writes response data. The exiscan
5251 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5252 blindly assumes that writes won't block so it may never time out."
5254 PH/39 Allow G after quota size as well as K and M.
5256 PH/40 The value set for $authenticated_id in an authenticator may not contain
5257 binary zeroes or newlines because the value is written to log lines and
5258 to spool files. There was no check on this. Now the value is run through
5259 the string_printing() function so that such characters are converted to
5260 printable escape sequences.
5262 PH/41 $message_linecount is a new variable that contains the total number of
5263 lines in the message. Compare $body_linecount, which is the count for the
5266 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5267 bounce and delay warning messages, except in certain special cases, which
5270 (a) An SMTP error message from a remote host;
5271 (b) A message specified in a :fail: redirection;
5272 (c) A message specified in a "fail" command in a system filter;
5273 (d) A message specified in a FAIL return from the queryprogram router;
5274 (e) A message specified by the cannot_route_message router option.
5276 In these cases only, Exim does include the error details in bounce and
5277 warning messages. There are also a few cases where bland messages such
5278 as "unrouteable address" or "local delivery error" are given.
5280 PH/43 $value is now also set for the "else" part of a ${run expansion.
5282 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5283 being worked on, but at least Exim now implements the latest version to
5286 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5287 process to complete was treated as a delivery failure, a timeout while
5288 writing the message to the pipe was logged, but erroneously treated as a
5289 successful delivery. Such timeouts include transport filter timeouts. For
5290 consistency with the overall process timeout, these timeouts are now
5291 treated as errors, giving rise to delivery failures by default. However,
5292 there is now a new Boolean option for the pipe transport called
5293 timeout_defer, which, if set TRUE, converts the failures into defers for
5294 both kinds of timeout. A transport filter timeout is now identified in
5297 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5298 systems where "make" and "gmake" are different, calling "gmake" at top
5299 level broke things. I've arranged for the value of $(MAKE) to be passed
5300 from the Makefile to this script so that it can call the same version of
5304 A note about Exim versions 4.44 and 4.50
5305 ----------------------------------------
5307 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5308 changes of various kinds. As a consequence, a big documentation update was
5309 needed. This delayed the release for rather longer than seemed good, especially
5310 in the light of a couple of (minor) security issues. Therefore, the changes
5311 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5312 release. So 4.44 and 4.50 are in effect two different branches that both start
5315 I have left the 4.50 change log unchanged; it contains all the changes since
5316 4.43. The change log for 4.44 is below; many of its items are identical to
5317 those for 4.50. This seems to be the most sensible way to preserve the
5318 historical information.
5324 1. Minor wording change to the doc/README.SIEVE file.
5326 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5327 computation of the current number of files was incorrect.
5329 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5330 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5333 4. Give more explanation in the error message when the command for a transport
5334 filter fails to execute.
5336 5. There are several places where Exim runs a non-Exim command in a
5337 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5338 was being done only for the command run by the queryprogram router. It is
5339 now done for all such subprocesses. The other cases are: ${run, transport
5340 filters, and the commands run by the lmtp and pipe transports.
5342 6. Added CONFIGURE_GROUP build-time option.
5344 7. Some older OS have a limit of 256 on the maximum number of file
5345 descriptors. Exim was using setrlimit() to set 1000 as a large value
5346 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5347 systems. I've change it so that if it can't get 1000, it tries for 256.
5349 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5350 was an oversight, and furthermore, ever since the addition of extra
5351 controls (e.g. 4.43/32), the checks on when to allow different forms of
5352 "control" were broken. There should now be diagnostics for all cases when a
5353 control that does not make sense is encountered.
5355 9. Added the /retain_sender option to "control=submission".
5357 10. $recipients is now available in the predata ACL (oversight).
5359 11. Tidy the search cache before the fork to do a delivery from a message
5360 received from the command line. Otherwise the child will trigger a lookup
5361 failure and thereby defer the delivery if it tries to use (for example) a
5362 cached ldap connection that the parent has called unbind on.
5364 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5365 of $address_data from the recipient verification was clobbered by the
5366 sender verification.
5368 13. The value of address_data from a sender verification is now available in
5369 $sender_address_data in subsequent conditions in the ACL statement.
5371 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5373 15. Added a new option "connect=<time>" to callout options, to set a different
5376 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5377 was its contents. (It was OK if the option was not defined at all.)
5379 17. A "Completed" log line is now written for messages that are removed from
5380 the spool by the -Mrm option.
5382 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5383 information about exactly what failed.
5385 19. Added -dd to debug only the daemon process.
5387 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5388 handles timeouts, both on the server side and network timeouts. Renamed the
5389 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5391 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5392 It is now set to "smtps".
5394 22. $host_address is now set to the target address during the checking of
5395 ignore_target_hosts.
5397 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5398 being passed; this would have caused $sender_host_name to have been used if
5399 matching the list had actually called for a host name (not very likely,
5400 since this list is usually IP addresses). A host name is now passed as
5403 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5404 code that specifies a non-restarting handler (typically sigaction() in
5405 modern systems) in an attempt to fix a rare and obscure crash bug.
5407 25. Narrowed the window for a race in the daemon that could cause it to ignore
5408 SIGCHLD signals. This is not a major problem, because they are used only to
5409 wake it up if nothing else does.
5411 26. A malformed maildirsize file could cause Exim to calculate negative values
5412 for the mailbox size or file count. Odd effects could occur as a result.
5413 The maildirsize information is now recalculated if the size or filecount
5416 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5417 support for a long time. Removed HAVE_SYS_VFS_H.
5419 28. Installed the latest version of exipick from John Jetmore.
5421 29. In an address list, if the pattern was not a regular expression, an empty
5422 subject address (from a bounce message) matched only if the pattern was an
5423 empty string. Non-empty patterns were not even tested. This was the wrong
5424 because it is perfectly reasonable to use an empty address as part of a
5425 database query. An empty address is now tested by patterns that are
5426 lookups. However, all the other forms of pattern expect the subject to
5427 contain a local part and a domain, and therefore, for them, an empty
5428 address still always fails if the pattern is not itself empty.
5430 30. Exim went into a mad DNS loop when attempting to do a callout where the
5431 host was specified on an smtp transport, and looking it up yielded more
5432 than one IP address.
5434 31. Re-factored the code for checking spool and log partition space into a
5435 function that finds that data and another that does the check. The former
5436 is then used to implement four new variables: $spool_space, $log_space,
5437 $spool_inodes, and $log_inodes.
5439 32. The RFC2047 encoding function was originally intended for short strings
5440 such as real names; it was not keeping to the 75-character limit for
5441 encoded words that the RFC imposes. It now respects the limit, and
5442 generates multiple encoded words if necessary. To be on the safe side, I
5443 have increased the buffer size for the ${rfc2047: expansion operator from
5446 33. It is now permitted to omit both strings after an "if" condition; if the
5447 condition is true, the result is "true". As before, when the second string
5448 is omitted, a false condition yields an empty string. This makes it less
5449 cumbersome to write custom ACL and router conditions.
5451 34. Failure to deliver a bounce message always caused it to be frozen, even if
5452 there was an errors_to setting on the router. The errors_to setting is now
5455 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5456 canonical form (fully expanded) before being placed in
5457 $sender_host_address.
5459 36. The table in the code that translates DNS record types into text (T_A to
5460 "A" for instance) was missing entries for NS and CNAME. It is just possible
5461 that this could have caused confusion if both these types were looked up
5462 for the same domain, because the text type is used as part of Exim's
5463 per-process caching. But the chance of anyone hitting this buglet seems
5466 37. The dnsdb lookup has been extended in a number of ways.
5468 (1) There is a new type, "zns", which walks up the domain tree until it
5469 finds some nameserver records. It should be used with care.
5471 (2) There is a new type, "mxh", which is like "mx" except that it returns
5472 just the host names, not the priorities.
5474 (3) It is now possible to give a list of domains (or IP addresses) to be
5475 looked up. The behaviour when one of the lookups defers can be
5476 controlled by a keyword.
5478 (4) It is now possible to specify the separator character for use when
5479 multiple records are returned.
5481 38. The dnslists ACL condition has been extended: it is now possible to supply
5482 a list of IP addresses and/or domains to be looked up in a particular DNS
5485 39. Added log_selector=+queue_time_overall.
5487 40. When running the queue in the test harness, wait just a tad after forking a
5488 delivery process, to get repeatability of debugging output.
5490 41. Include certificate and key file names in error message when GnuTLS fails
5491 to set them up, because the GnuTLS error message doesn't include the name
5492 of the failing file when there is a problem reading it.
5494 42. Allow both -bf and -bF in the same test run.
5496 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5498 44. The "Exiscan patch" is now merged into the mainline Exim source.
5500 45. Sometimes the final signoff response after QUIT could fail to get
5501 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5502 < 0 before doing a fflush(). This bug looks as though it goes back to the
5503 introduction of TLS in release 3.20, but "sometimes" must have been rare
5504 because the tests only now provoked it.
5506 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5507 (this can affect the format of dates).
5509 47. exim_tidydb, when checking for the continued existence of a message for
5510 which it has found a message-specific retry record, was not finding
5511 messages that were in split spool directories. Consequently, it was
5512 deleting retry records that should have stayed in existence.
5514 48. Steve fixed some bugs in eximstats.
5516 49. The SPA authentication driver was not abandoning authentication and moving
5517 on to the next authenticator when an expansion was forced to fail,
5518 contradicting the general specification for all authenticators. Instead it
5519 was generating a temporary error. It now behaves as specified.
5521 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5522 (the order specifies the preference for clients). The order is now AES256,
5523 AES128, 3DES, ARCFOUR128.
5525 51. Small patch to Sieve code - explicitly set From: when generating an
5528 52. Exim crashed if a remote delivery caused a very long error message to be
5529 recorded - for instance if somebody sent an entire SpamAssassin report back
5530 as a large number of 550 error lines. This bug was coincidentally fixed by
5531 increasing the size of one of Exim's internal buffers (big_buffer) that
5532 happened as part of the Exiscan merge. However, to be on the safe side, I
5533 have made the code more robust (and fixed the comments that describe what
5536 53. Now that there can be additional text after "Completed" in log lines (if
5537 the queue_time_overall log selector is set), a one-byte patch to exigrep
5538 was needed to allow it to recognize "Completed" as not the last thing in
5541 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5542 patch that reportedly fixes this has been added. I am not expert enough to
5543 create a test for it. This is what the patch creator wrote:
5545 "I found a little strange behaviour of ldap code when working with
5546 Windows 2003 AD Domain, where users was placed in more than one
5547 Organization Units. When I tried to give exim partial DN, the exit code
5548 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5549 But simultaneously result of request was absolutely normal ldap result,
5550 so I produce this patch..."
5552 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5553 REFERENCE, so I have modified the code to exclude the patch when that macro
5556 55. Some experimental protocols are using DNS PTR records for new purposes. The
5557 keys for these records are domain names, not reversed IP addresses. The
5558 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5559 leaves it alone. Component reversal etc. now happens only for IP addresses.
5562 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5564 57. Double the size of the debug message buffer (to 2048) so that more of very
5565 long debug lines gets shown.
5567 58. The exicyclog utility now does better if the number of log files to keep
5568 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5570 59. Two changes related to the smtp_active_hostname option:
5572 (1) $smtp_active_hostname is now available as a variable.
5573 (2) The default for smtp_banner uses $smtp_active_hostname instead
5574 of $primary_hostname.
5576 60. The host_aton() function is supposed to be passed a string that is known
5577 to be a valid IP address. However, in the case of IPv6 addresses, it was
5578 not checking this. This is a hostage to fortune. Exim now panics and dies
5579 if the condition is not met. A case was found where this could be provoked
5580 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5581 components; fortuitously, this particular loophole had already been fixed
5582 by change 4.50/55 above.
5584 If there are any other similar loopholes, the new check in host_aton()
5585 itself should stop them being exploited. The report I received stated that
5586 data on the command line could provoke the exploit when Exim was running as
5587 exim, but did not say which command line option was involved. All I could
5588 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5589 running as the user.
5592 61. There was a buffer overflow vulnerability in the SPA authentication code
5593 (which came originally from the Samba project). I have added a test to the
5594 spa_base64_to_bits() function which I hope fixes it.
5597 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5598 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5600 63. The daemon start-up calls getloadavg() while still root for those OS that
5601 need the first call to be done as root, but it missed one case: when
5602 deliver_queue_load_max is set with deliver_drop_privilege. This is
5603 necessary for the benefit of the queue runner, because there is no re-exec
5604 when deliver_drop_privilege is set.
5606 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5607 This has been fixed.
5609 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5610 was in use, was not putting the data itself into the right store pool;
5611 consequently, it could be overwritten for a subsequent message in the same
5612 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5615 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5617 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5618 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5619 the function were treating the return as a boolean value, which happened to
5620 work because 0=false and not-0=true, but is not correct code.
5622 68. The host_aton() function was not handling scoped IPv6 addresses (those
5623 with, for example, "%eth0" on the end) correctly.
5625 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5626 negated items (that is, ~something) in unsigned ints. Some compilers
5627 apparently mutter when there is no cast.
5629 70. If an address verification called from an ACL failed, and did not produce a
5630 user-specific message (i.e. there was only a "system" message), nothing was
5631 put in $acl_verify_message. In this situation, it now puts the system
5634 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5635 session; change 4.31/43 added the unwanted input to the log line - except
5636 that it did not do this in the start of session case. It now does.
5638 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5639 This is wrong and can cause the other end to generate a synchronization
5640 error if it is another Exim or anything else that does the synchronization
5641 check. A QUIT command is no longer sent after a timeout.
5643 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5644 during host lookups.
5646 74. The defer_ok option of callout verification was not working if it was used
5647 when verifying addresses in header lines, that is, for this case:
5649 verify = header_sender/callout=defer_ok
5651 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5652 those file descriptors could be used for SMTP connections. If anything
5653 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5654 could be sent to the SMTP client, causing chaos. The daemon now opens
5655 stdin, stdout, and stderr to /dev/null when it puts itself into the
5658 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5659 log by default. The user can override this with suitable Perl magic.
5661 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5662 the log message when discard triggers, was not working for the DATA ACL or
5663 for the non-SMTP ACL.
5665 78. Error message wording change in sieve.c.
5667 79. If smtp_accept_max_per_host was set, the number of connections could be
5668 restricted to fewer than expected, because the daemon was trying to set up
5669 a new connection before checking whether the processes handling previous
5670 connections had finished. The check for completed processes is now done
5671 earlier. On busy systems, this bug wouldn't be noticed because something
5672 else would have woken the daemon, and it would have reaped the completed
5675 80. If a message was submitted locally by a user whose login name contained one
5676 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5677 It caused a spool format error. I have fixed the spool reading code. A
5678 related problem was that the "from" clause in the Received: line became
5679 illegal because of the space(s). It is now covered by ${quote_local_part.
5681 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5684 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5686 83. Patch from Sieve maintainer to fix unterminated string problem in
5687 "vacation" handling.
5689 84. Some minor changes to the Linux configuration files to help with other
5690 OS variants using glibc.
5692 85. One more patch for Sieve to update vacation handling to latest spec.
5695 ----------------------------------------------------
5696 See the note above about the 4.44 and 4.50 releases.
5697 ----------------------------------------------------
5703 1. Change 4.43/35 introduced a bug that caused file counts to be
5704 incorrectly computed when quota_filecount was set in an appendfile
5707 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5708 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5711 3. Give more explanation in the error message when the command for a transport
5712 filter fails to execute.
5714 4. There are several places where Exim runs a non-Exim command in a
5715 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5716 was being done only for the command run by the queryprogram router. It is
5717 now done for all such subprocesses. The other cases are: ${run, transport
5718 filters, and the commands run by the lmtp and pipe transports.
5720 5. Some older OS have a limit of 256 on the maximum number of file
5721 descriptors. Exim was using setrlimit() to set 1000 as a large value
5722 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5723 systems. I've change it so that if it can't get 1000, it tries for 256.
5725 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5726 was an oversight, and furthermore, ever since the addition of extra
5727 controls (e.g. 4.43/32), the checks on when to allow different forms of
5728 "control" were broken. There should now be diagnostics for all cases when a
5729 control that does not make sense is encountered.
5731 7. $recipients is now available in the predata ACL (oversight).
5733 8. Tidy the search cache before the fork to do a delivery from a message
5734 received from the command line. Otherwise the child will trigger a lookup
5735 failure and thereby defer the delivery if it tries to use (for example) a
5736 cached ldap connection that the parent has called unbind on.
5738 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5739 of $address_data from the recipient verification was clobbered by the
5740 sender verification.
5742 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5743 was its contents. (It was OK if the option was not defined at all.)
5745 11. A "Completed" log line is now written for messages that are removed from
5746 the spool by the -Mrm option.
5748 12. $host_address is now set to the target address during the checking of
5749 ignore_target_hosts.
5751 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5752 being passed; this would have caused $sender_host_name to have been used if
5753 matching the list had actually called for a host name (not very likely,
5754 since this list is usually IP addresses). A host name is now passed as
5757 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5758 code that specifies a non-restarting handler (typically sigaction() in
5759 modern systems) in an attempt to fix a rare and obscure crash bug.
5761 15. Narrowed the window for a race in the daemon that could cause it to ignore
5762 SIGCHLD signals. This is not a major problem, because they are used only to
5763 wake it up if nothing else does.
5765 16. A malformed maildirsize file could cause Exim to calculate negative values
5766 for the mailbox size or file count. Odd effects could occur as a result.
5767 The maildirsize information is now recalculated if the size or filecount
5770 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5771 support for a long time. Removed HAVE_SYS_VFS_H.
5773 18. Updated exipick to current release from John Jetmore.
5775 19. Allow an empty sender to be matched against a lookup in an address list.
5776 Previously the only cases considered were a regular expression, or an
5779 20. Exim went into a mad DNS lookup loop when doing a callout where the
5780 host was specified on the transport, if the DNS lookup yielded more than
5783 21. The RFC2047 encoding function was originally intended for short strings
5784 such as real names; it was not keeping to the 75-character limit for
5785 encoded words that the RFC imposes. It now respects the limit, and
5786 generates multiple encoded words if necessary. To be on the safe side, I
5787 have increased the buffer size for the ${rfc2047: expansion operator from
5790 22. Failure to deliver a bounce message always caused it to be frozen, even if
5791 there was an errors_to setting on the router. The errors_to setting is now
5794 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5795 canonical form (fully expanded) before being placed in
5796 $sender_host_address.
5798 24. Updated eximstats to version 1.33
5800 25. Include certificate and key file names in error message when GnuTLS fails
5801 to set them up, because the GnuTLS error message doesn't include the name
5802 of the failing file when there is a problem reading it.
5804 26. Expand error message when OpenSSL has problems setting up cert/key files.
5807 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5808 (this can affect the format of dates).
5810 28. exim_tidydb, when checking for the continued existence of a message for
5811 which it has found a message-specific retry record, was not finding
5812 messages that were in split spool directories. Consequently, it was
5813 deleting retry records that should have stayed in existence.
5815 29. eximstats updated to version 1.35
5816 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5817 1.35 - bugfix such that pie charts by volume are generated correctly
5819 30. The SPA authentication driver was not abandoning authentication and moving
5820 on to the next authenticator when an expansion was forced to fail,
5821 contradicting the general specification for all authenticators. Instead it
5822 was generating a temporary error. It now behaves as specified.
5824 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5825 (the order specifies the preference for clients). The order is now AES256,
5826 AES128, 3DES, ARCFOUR128.
5828 31. Small patch to Sieve code - explicitly set From: when generating an
5831 32. Exim crashed if a remote delivery caused a very long error message to be
5832 recorded - for instance if somebody sent an entire SpamAssassin report back
5833 as a large number of 550 error lines. This bug was coincidentally fixed by
5834 increasing the size of one of Exim's internal buffers (big_buffer) that
5835 happened as part of the Exiscan merge. However, to be on the safe side, I
5836 have made the code more robust (and fixed the comments that describe what
5839 33. Some experimental protocols are using DNS PTR records for new purposes. The
5840 keys for these records are domain names, not reversed IP addresses. The
5841 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5842 leaves it alone. Component reversal etc. now happens only for IP addresses.
5845 34. The host_aton() function is supposed to be passed a string that is known
5846 to be a valid IP address. However, in the case of IPv6 addresses, it was
5847 not checking this. This is a hostage to fortune. Exim now panics and dies
5848 if the condition is not met. A case was found where this could be provoked
5849 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5850 components; fortuitously, this particular loophole had already been fixed
5851 by change 4.50/55 or 4.44/33 above.
5853 If there are any other similar loopholes, the new check in host_aton()
5854 itself should stop them being exploited. The report I received stated that
5855 data on the command line could provoke the exploit when Exim was running as
5856 exim, but did not say which command line option was involved. All I could
5857 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5858 running as the user.
5861 35. There was a buffer overflow vulnerability in the SPA authentication code
5862 (which came originally from the Samba project). I have added a test to the
5863 spa_base64_to_bits() function which I hope fixes it.
5866 36. The daemon start-up calls getloadavg() while still root for those OS that
5867 need the first call to be done as root, but it missed one case: when
5868 deliver_queue_load_max is set with deliver_drop_privilege. This is
5869 necessary for the benefit of the queue runner, because there is no re-exec
5870 when deliver_drop_privilege is set.
5872 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5873 was in use, was not putting the data itself into the right store pool;
5874 consequently, it could be overwritten for a subsequent message in the same
5875 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5878 38. Sometimes the final signoff response after QUIT could fail to get
5879 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5880 < 0 before doing a fflush(). This bug looks as though it goes back to the
5881 introduction of TLS in release 3.20, but "sometimes" must have been rare
5882 because the tests only now provoked it.
5888 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5889 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5890 Now it returns an integer. A number of calls were still expecting a T/F
5891 return. Fortuitously, in all cases, the tests worked in OK situations,
5892 which is the norm. However, things would have gone wrong on any write
5893 failures on the smtp file descriptor. This function is used when sending
5894 messages over SMTP and also when doing verify callouts.
5896 2. When Exim is called to do synchronous delivery of a locally submitted
5897 message (the -odf or -odi options), it no longer closes stderr before doing
5900 3. Implemented the mua_wrapper option.
5902 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5904 5. Implemented the functions header_remove(), header_testname(),
5905 header_add_at_position(), and receive_remove_recipient(), and exported them
5908 6. If an ACL "warn" statement specified the addition of headers, Exim already
5909 inserted X-ACL-Warn: at the start if there was no header name. However, it
5910 was not making this test for the second and subsequent header lines if
5911 there were newlines in the string. This meant that an invalid header could
5912 be inserted if Exim was badly configured.
5914 7. Allow an ACL "warn" statement to add header lines at the start or after all
5915 the Received: headers, as well as at the end.
5917 8. Added the rcpt_4xx retry error code.
5919 9. Added postmaster_mailfrom=xxx to callout verification option.
5921 10. Added mailfrom=xxxx to the callout verification option, for verify=
5924 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5925 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5927 12. Inserted some casts to stop certain compilers warning when using pointer
5928 differences as field lengths or precisions in printf-type calls (mostly
5929 affecting debugging statements).
5931 13. Added optional readline() support for -be (dynamically loaded).
5933 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5934 same clock tick as a message's arrival, so that its received time was the
5935 same as the "first fail" time on the retry record, and that message
5936 remained on the queue past the ultimate address timeout, every queue runner
5937 would try a delivery (because it was past the ultimate address timeout) but
5938 after another failure, the ultimate address timeout, which should have then
5939 bounced the address, did not kick in. This was a "< instead of <=" error;
5940 in most cases the first failure would have been in the next clock tick
5941 after the received time, and all would be well.
5943 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5944 being recognized when the domain list was tested by the match_domain
5945 condition in an expansion string.
5947 16. Added the ${str2b64: operator.
5949 17. Exim was always calling setrlimit() to set a large limit for the number of
5950 processes, without checking whether the existing limit was already
5951 adequate. (It did check for the limit on file descriptors.) Furthermore,
5952 errors from getrlimit() and setrlimit() were being ignored. Now they are
5953 logged to the main and panic logs, but Exim does carry on, to try to do its
5954 job under whatever limits there are.
5956 18. Imported PCRE 5.0.
5958 19. Trivial typo in log message " temporarily refused connection" (the leading
5961 20. If the log selector return_path_on_delivery was set and an address was
5962 redirected to /dev/null, the delivery process crashed because it assumed
5963 that a return path would always be set for a "successful" delivery. In this
5964 case, the whole delivery is bypassed as an optimization, and therefore no
5967 21. Internal re-arrangement: the function for sending a challenge and reading
5968 a response while authentication was assuming a zero-terminated challenge
5969 string. It's now changed to take a pointer and a length, to allow for
5970 binary data in such strings.
5972 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5974 23. Exim was not respecting finduser_retries when seeking the login of the
5975 uid under which it was called; it was always trying 10 times. (The default
5976 setting of finduser_retries is zero.) Also, it was sleeping after the final
5977 failure, which is pointless.
5979 24. Implemented tls_on_connect_ports.
5981 25. Implemented acl_smtp_predata.
5983 26. If the domain in control=submission is set empty, Exim assumes that the
5984 authenticated id is a complete email address when it generates From: or
5985 Sender: header lines.
5987 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5988 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5989 chown and chgrp in /bin and hostname in /usr/bin.
5991 28. Exim was keeping the "process log" file open after each use, just as it
5992 does for the main log. This opens the possibility of it remaining open for
5993 long periods when the USR1 signal hits a daemon. Occasional processlog
5994 errors were reported, that could have been caused by this. Anyway, it seems
5995 much more sensible not to leave this file open at all, so that is what now
5998 29. The long-running daemon process does not normally write to the log once it
5999 has entered its main loop, and it closes the log before doing so. This is
6000 so that log files can straightforwardly be renamed and moved. However,
6001 there are a couple of unusual error situations where the daemon does write
6002 log entries, and I had neglected to close the log afterwards.
6004 30. The text of an SMTP error response that was received during a remote
6005 delivery was being truncated at 512 bytes. This is too short for some of
6006 the long messages that one sometimes sees. I've increased the limit to
6009 31. It is now possible to make retry rules that apply only when a message has a
6010 specific sender, in particular, an empty sender.
6012 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6013 it possible to be selective about when SMTP synchronization is enforced.
6015 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6017 32. Implemented hosts_connection_nolog.
6019 33. Added an ACL for QUIT.
6021 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6024 35. Added mailbox_size and mailbox_filecount to appendfile.
6026 36. Added control = no_multiline_responses to ACLs.
6028 37. There was a bug in the logic of the code that waits for the clock to tick
6029 in the case where the clock went backwards by a substantial amount such
6030 that the microsecond fraction of "now" was more than the microsecond
6031 fraction of "then" (but the whole seconds number was less).
6033 38. Added support for the libradius Radius client library this is found on
6034 FreeBSD (previously only the radiusclient library was supported).
6040 1. When certain lookups returned multiple values in the form name=value, the
6041 quoting of the values was not always being done properly. Specifically:
6042 (a) If the value started with a double quote, but contained no whitespace,
6044 (b) If the value contained whitespace other than a space character (i.e.
6045 tabs or newlines or carriage returns) it was not quoted.
6046 This fix has been applied to the mysql and pgsql lookups by writing a
6047 separate quoting function and calling it from the lookup code. The fix
6048 should probably also be applied to nisplus, ibase and oracle lookups, but
6049 since I cannot test any of those, I have not disturbed their existing code.
6051 2. A hit in the callout cache for a specific address caused a log line with no
6052 reason for rejecting RCPT. Now it says "Previous (cached) callout
6053 verification failure".
6055 3. There was an off-by-one bug in the queryprogram router. An over-long
6056 return line was truncated at 256 instead of 255 characters, thereby
6057 overflowing its buffer with the terminating zero. As well as fixing this, I
6058 have increased the buffer size to 1024 (and made a note to document this).
6060 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6061 when Exim is waiting for an SMTP response from a remote server, Exim
6062 restarts its select() call on the socket, thereby resetting its timeout.
6063 This is not a problem when such interrupts are rare. Somebody set up a cron
6064 job to run exiwhat every 2 minutes, which is less than the normal select()
6065 timeout (5 or 10 minutes). This meant that the select() timeout never
6066 kicked in because it was always reset. I have fixed this by comparing the
6067 time when an interrupt arrives with the time at the start of the first call
6068 to select(). If more time than the timeout has elapsed, the interrupt is
6069 treated as a timeout.
6071 5. Some internal re-factoring in preparation for the addition of Sieve
6072 extensions (by MH). In particular, the "personal" test is moved to a
6073 separate function, and given an option for scanning Cc: and Bcc: (which is
6074 not set for Exim filters).
6076 6. When Exim created an email address using the login of the caller as the
6077 local part (e.g. when creating a From: or Sender: header line), it was not
6078 quoting the local part when it contained special characters such as @.
6080 7. Installed new OpenBSD configuration files.
6082 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6083 try to make them clearer.
6085 9. Callout options, other than the timeout value, were being ignored when
6086 verifying sender addresses in header lines. For example, when using
6088 verify = header_sender/callout=no_cache
6090 the cache was (incorrectly) being used.
6092 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6093 only the Cygwin environment.
6095 11. When return_path_on_delivery was set as a log selector, if different remote
6096 addresses in the same message used different return paths and parallel
6097 remote delivery occurred, the wrong values would sometimes be logged.
6098 (Whenever a remote delivery process finished, the return path value from
6099 the most recently started remote delivery process was logged.)
6101 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6102 lines when AUTH and/or TLS are in use. This is the "received protocol"
6103 field. Exim used to use "asmtp" for authenticated SMTP, without any
6104 indication (in the protocol name) for TLS use. Now it follows the RFC and
6105 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6106 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6107 names appear in log lines as well as in Received: header lines.
6109 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6110 extensions, and comparison tests, and to fix some bugs.
6112 14. Changes to the "personal" filter test:
6114 (1) The test was buggy in that it was just doing the equivalent of
6115 "contains" tests on header lines. For example, if a user's address was
6116 anne@some.where, the "personal" test would incorrectly be true for
6118 To: susanne@some.where
6120 This test is now done by extracting each address from the header in turn,
6121 and checking the entire address. Other tests that are part of "personal"
6122 are now done using regular expressions (for example, to check local parts
6123 of addresses in From: header lines).
6125 (2) The list of non-personal local parts in From: addresses has been
6126 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6127 taken from the Sieve specification recommendations.
6129 (3) If the message contains any header line starting with "List-" it is
6130 treated as non-personal.
6132 (4) The test for "circular" in the Subject: header line has been removed
6133 because it now seems ill-conceived.
6135 15. Minor typos in src/EDITME comments corrected.
6137 16. Installed latest exipick from John Jetmore.
6139 17. If headers_add on a router specified a text string that was too long for
6140 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6141 of string_sprintf() is now avoided.
6143 18. $message_body_size was not set (it was always zero) when running the DATA
6144 ACL and the local_scan() function.
6146 19. For the "mail" command in an Exim filter, no default was being set for
6147 the once_repeat time, causing a random time value to be used if "once" was
6148 specified. (If the value happened to be <= 0, no repeat happened.) The
6149 default is now 0s, meaning "never repeat". The "vacation" command was OK
6150 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6151 (I found it when inspecting the code).
6153 20. There is now an overall timeout for performing a callout verification. It
6154 defaults to 4 times the callout timeout, which applies to individual SMTP
6155 commands during the callout. The overall timeout applies when there is more
6156 than one host that can be tried. The timeout is checked before trying the
6157 next host. This prevents very long delays if there are a large number of
6158 hosts and all are timing out (e.g. when the network connections are timing
6159 out). The value of the overall timeout can be changed by specifying an
6160 additional sub-option for "callout", called "maxwait". For example:
6162 verify = sender/callout=5s,maxwait=20s
6164 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6165 to the end before writing, but this should make it even safer).
6167 22. Exim was forgetting that it had advertised PIPELINING for the second and
6168 subsequent messages on an SMTP connection. It was also not resetting its
6169 memory on STARTTLS and an internal HELO.
6171 23. When Exim logs an SMTP synchronization error within a session, it now
6172 records whether PIPELINING has been advertised or not.
6174 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6175 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6176 rather than long int.
6178 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6180 26. Added the never_mail option to autoreply.
6186 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6187 crash if the getsockname() call failed; this can happen if a connection is
6188 closed very soon after it is established. The problem was simply in the
6189 order in which certain operations were done, causing Exim to try to write
6190 to the SMTP stream before it had set up the file descriptor. The bug has
6191 been fixed by making things happen in the correct order.
6197 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6198 before the connection was closed, thus losing the rejection response.
6200 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6201 some early Solaris releases, but causes trouble in current releases where
6202 socklen_t is defined.
6204 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6207 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6210 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6211 that releasing the top of it at the end releases what was used for sub-
6212 expansions (unless the block got too big). However, discard this block if
6213 the first thing is a variable or header, so that we can use its block when
6214 it is dynamic (useful for very large $message_headers, for example).
6216 6. Lookups now cache *every* query, not just the most recent. A new, separate
6217 store pool is used for this. It can be recovered when all lookup caches are
6218 flushed. Lookups now release memory at the end of their result strings.
6219 This has involved some general refactoring of the lookup sources.
6221 7. Some code has been added to the store_xxx() functions to reduce the amount
6222 of flapping under certain conditions.
6224 8. log_incoming_interface used to affect only the <= reception log lines. Now
6225 it causes the local interface and port to be added to several more SMTP log
6226 lines, for example "SMTP connection from", and rejection lines.
6228 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6230 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6232 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6233 could be overwritten at the end of the current message (or the start of a
6234 new message if it was set in a HELO ACL). The value is now preserved for
6235 the duration of the SMTP connection.
6237 12. If a transport had a headers_rewrite setting, and a matching header line
6238 contained an unqualified address, that address was qualified, even if it
6239 did not match any rewriting rules. The underlying bug was that the values
6240 of the flags that permit the existence of unqualified sender and recipient
6241 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6242 non-local messages, and by -bnq for local messages) were not being
6243 preserved with the message after it was received.
6245 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6246 "next input=" as part of the text comprising the host identity instead of
6247 the correct text. The code was using the same buffer for two different
6248 strings. However, depending on which order the printing function evaluated
6249 its arguments, the bug did not always show up. Under Linux, for example, my
6250 test suite worked just fine.
6252 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6253 doesn't work with some older versions of Perl. It has been changed to "my",
6254 which in any case is probably the better facility to use.
6256 15. A really picky compiler found some instances of statements for creating
6257 error messages that either had too many or two few arguments for the format
6260 16. The size of the buffer for calls to the DNS resolver has been increased
6261 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6262 for addresses that have a lot of PTR records. This alleviates a problem; it
6263 does not fully solve it.
6265 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6266 buffer now truncates the list and logs the incident, which is the same
6267 action as happens when Exim is looking up a host name and its aliases.
6268 Previously in this situation something unpredictable would happen;
6269 sometimes it was "internal error: store_reset failed".
6271 18. If a server dropped the connection unexpectedly when an Exim client was
6272 using GnuTLS and trying to read a response, the client delivery process
6273 crashed while trying to generate an error log message.
6275 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6276 string, for example:
6278 warn message = H1: something\nH2: something
6280 the text was added as a single header line from Exim's point of view
6281 though it ended up OK in the delivered message. However, searching for the
6282 second and subsequent header lines using $h_h2: did not work. This has been
6283 fixed. Similarly, if a system filter added multiple headers in this way,
6284 the routers could not see them.
6286 20. Expanded the error message when iplsearch is called with an invalid key to
6287 suggest using net-iplsearch in a host list.
6289 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6290 ACLs are no longer actually imposed (and a message to that effect is
6293 22. If a "gecos" field in a passwd entry contained escaped characters, in
6294 particular, if it contained a \" sequence, Exim got it wrong when building
6295 a From: or a Sender: header from that name. A second bug also caused
6296 incorrect handling when an unquoted " was present following a character
6297 that needed quoting.
6299 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6300 was not being matched caselessly.
6302 24. Arranged for all hyphens in the exim.8 source to be escaped with
6305 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6306 properly. Recipient callout cache records were still being keyed to include
6307 the sender, even when use_sender was set false. This led to far more
6308 callouts that were necessary. The sender is no longer included in the key
6309 when use_sender is false.
6311 26. Added "control = submission" modifier to ACLs.
6313 27. Added the ${base62d: operator to decode base 62 numbers.
6315 28. dnsdb lookups can now access SRV records.
6317 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6318 the configuration file.
6320 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6321 (-v) mode. This makes the output for a verbose queue run more intelligible.
6323 31. Added a use_postmaster feature to recipient callouts.
6325 32. Added the $body_zerocount variable, containing the number of binary zero
6326 bytes in the message body.
6328 33. The time of last modification of the "new" subdirectory is now used as the
6329 "mailbox time last read" when there is a quota error for a maildir
6332 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6334 35. Added +ignore_unknown as a special item in host lists.
6336 36. Code for decoding IPv6 addresses in host lists is now included, even if
6337 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6338 address was recognized as an IP address, but was then not correctly decoded
6339 into binary, causing unexpected and incorrect effects when compared with
6346 1. Very minor rewording of debugging text in manualroute to say "list of
6347 hosts" instead of "hostlist".
6349 2. If verify=header_syntax was set, and a header line with an unqualified
6350 address (no domain) and a large number of spaces between the end of the
6351 name and the colon was received, the reception process suffered a buffer
6352 overflow, and (when I tested it) crashed. This was caused by some obsolete
6353 code that should have been removed. The fix is to remove it!
6355 3. When running in the test harness, delay a bit after writing a bounce
6356 message to get a bit more predictability in the log output.
6358 4. Added a call to search_tidyup() just before forking a reception process. In
6359 theory, someone could use a lookup in the expansion of smtp_accept_max_
6360 per_host which, without the tidyup, could leave open a database connection.
6362 5. Added the variables $recipient_data and $sender_data which get set from a
6363 lookup success in an ACL "recipients" or "senders" condition, or a router
6364 "senders" option, similar to $domain_data and $local_part_data.
6366 6. Moved the writing of debug_print from before to after the "senders" test
6369 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6370 problems for message scanning, either using a data ACL, or using
6371 local_scan() because the Received: header was not generated till after they
6372 were called (in order to set the time as the time of reception completion).
6373 I have revised the way this works. The header is now generated after the
6374 body is received, but before the ACL or local_scan() are called. After they
6375 are run, the timestamp in the header is updated.
6381 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6382 before starting a queue runner without re-exec. This happened only when
6383 deliver_drop_privilege was set or when the Exim user was set to root. The
6384 effect of the bug was that timeouts during subsequent deliveries caused
6385 crashes instead of being properly handled. The handler is now left at its
6386 default (and expected) setting.
6388 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6389 message, again when deliver_drop_privilege is set or Exim is run as root.
6390 The bug described in (1) was not present in this case, but the tidying up
6391 of the other signals was missing. I have made the two cases consistent.
6393 3. The ignore_target_hosts setting on a manualroute router was being ignored
6394 for hosts that were looked up using the /MX notation.
6396 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6399 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6400 operated on the sender address. After changing the $sender_address to <>
6401 for the sender address verify, Exim was re-instated it as the original
6402 (before rewriting) address, but remembering that it had rewritten it, so it
6403 wasn't rewriting it again. This bug also had the effect of breaking the
6404 sender address verification caching when the sender address was rewritten.
6406 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6407 This has been changed so that if the ip literal address matches
6408 ignore_target_hosts, the router declines.
6410 7. Added expansion conditions match_domain, match_address, and match_local_
6411 part (NOT match_host).
6413 8. The placeholder for the Received: header didn't have a length field set.
6415 9. Added code to Exim itself and to exim_lock to test for a specific race
6416 condition that could lead to file corruption when using MBX delivery. The
6417 issue is with the lockfile that is created in /tmp. If this file is removed
6418 after a process has opened it but before that process has acquired a lock,
6419 there is the potential for a second process to recreate the file and also
6420 acquire a lock. This could lead to two Exim processes writing to the file
6421 at the same time. The added code performs the same test as UW imapd; it
6422 checks after acquiring the lock that its file descriptor still refers to
6423 the same named file.
6425 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6426 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6429 11. Added the smtp_active_hostname option. If used, this will typically be made
6430 to depend on the incoming interface address. Because $interface_address is
6431 not set up until the daemon has forked a reception process, error responses
6432 that can happen earlier (such as "too many connections") no longer contain
6435 12. If an expansion in a condition on a "warn" statement fails because a lookup
6436 defers, the "warn" statement is abandoned, and the next ACL statement is
6437 processed. Previously this caused the whole ACL to be aborted.
6439 13. Added the iplsearch lookup type.
6441 14. Added ident_timeout as a log selector.
6443 15. Added tls_certificate_verified as a log selector.
6445 16. Added a global option tls_require_ciphers (compare the smtp transport
6446 option of the same name). This controls incoming TLS connections.
6448 17. I finally figured out how to make tls_require_ciphers do a similar thing
6449 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6450 before starting the TLS session.
6452 18. Tabs are now shown as \t in -bP output.
6454 19. If the log selector return_path_on_delivery was set, Exim crashed when
6455 bouncing a message because it had too many Received: header lines.
6457 20. If two routers both had headers_remove settings, and the first one included
6458 a superfluous trailing colon, the final name in the first list and the
6459 first name in the second list were incorrectly joined into one item (with a
6460 colon in the middle).
6466 1. Added -C and -D options to the exinext utility, mainly to make it easier
6467 to include in the automated testing, but these could be helpful when
6468 multiple configurations are in use.
6470 2. The exinext utility was not formatting the output nicely when there was
6471 an alternate port involved in the retry record key, nor when there was a
6472 message id as well (for retries that were specific to a specific message
6473 and a specific host). It was also confused by IPv6 addresses, because of
6474 the additional colons they contain. I have fixed the IPv4 problem, and
6475 patched it up to do a reasonable job for IPv6.
6477 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6478 delivery, the log line now contains "pipelined" if PIPELINING was used.
6480 4. An SMTP transport process used to panic and die if the bind() call to set
6481 an explicit outgoing interface failed. This has been changed; it is now
6482 treated in the same way as a connect() failure.
6484 5. A reference to $sender_host_name in the part of a conditional expansion
6485 that was being skipped was still causing a DNS lookup. This no longer
6488 6. The def: expansion condition was not recognizing references to header lines
6489 that used bh_ and bheader_.
6491 7. Added the _cache feature to named lists.
6493 8. The code for checking quota_filecount in the appendfile transport was
6494 allowing one more file than it should have been.
6496 9. For compatibility with Sendmail, the command line option
6504 and sets the incoming protocol and host name (for trusted callers). The
6505 host name and its colon can be omitted when only the protocol is to be set.
6506 Note the Exim already has two private options, -pd and -ps, that refer to
6507 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6508 "s", but I don't think that's a major issue.
6510 10. A number of refactoring changes to the code, none of which should affect
6513 (a) The number of logging options was getting close to filling up the
6514 32-bit word that was used as a bit map. I have split them into two classes:
6515 those that are passed in the argument to log_write(), and those that are
6516 only ever tested independently outside of that function. These are now in
6517 separate 32-bit words, so there is plenty of room for expansion again.
6518 There is no change in the user interface or the logging behaviour.
6520 (b) When building, for example, log lines, the code previously used a
6521 macro that called string_cat() twice, in order to add two strings. This is
6522 not really sufficiently general. Furthermore, there was one instance where
6523 it was actually wrong because one of the argument was used twice, and in
6524 one call a function was used. (As it happened, calling the function twice
6525 did not affect the overall behaviour.) The macro has been replaced by a
6526 function that can join an arbitrary number of extra strings onto a growing
6529 (c) The code for expansion conditions now uses a table and a binary chop
6530 instead of a serial search (which was left over from when there were very
6531 few conditions). Also, it now recognizes conditions like "pam" even when
6532 the relevant support is not compiled in: a suitably worded error message is
6533 given if an attempt is made to use such a condition.
6535 11. Added ${time_interval:xxxxx}.
6537 12. A bug was causing one of the ddress fields not to be passed back correctly
6538 from remote delivery subprocesses. The field in question was not being
6539 subsequently used, so this caused to problems in practice.
6541 13. Added new log selectors queue_time and deliver_time.
6543 14. Might have fixed a bug in maildirsizefile handling that threw up
6544 "unexpected character" debug warnings, and recalculated the data
6545 unnecessarily. In any case, I expanded the warning message to give more
6548 15. Added the message "Restricted characters in address" to the statements in
6549 the default ACL that block characters like @ and % in local parts.
6551 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6552 Three changes have been made:
6554 (a) There was a serious bug; a negative response to MAIL caused the whole
6555 recipient domain to be cached as invalid, thereby blocking all messages
6556 to all local parts at the same domain, from all senders. This bug has
6557 been fixed. The domain is no longer cached after a negative response to
6558 MAIL if the sender used is not empty.
6560 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6563 (c) A new callout option, "use_sender" has been added for people who want
6564 the modified behaviour.
6570 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6573 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6574 indeed breaks things for older releases.
6576 3. Added additional logging to the case where there is a problem reading data
6577 from a filter that is running in a subprocess using a pipe, in order to
6578 try to track down a specific problem.
6580 4. Testing facility fudge: when running in the test harness and attempting
6581 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6582 getting "No route to host". Convert this to a timeout.
6584 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6587 6. Some OS don't have socklen_t but use size_t instead. This affects the
6588 fifth argument of getsockopt() amongst other things. This is now
6589 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6590 can be set for individual OS. I have set it for SunOS5, OSF1, and
6591 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6592 some earlier ones do not.
6594 7. Change 4.30/15 was not doing the test caselessly.
6596 8. The standard form for an IPv6 address literal was being rejected by address
6597 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6598 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6599 this, as well as the form without the "IPv6" on the front (but only when
6600 address literals are enabled, of course).
6602 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6604 10. Exim crashed if a message with an empty sender address specified by -f
6605 encountered a router with an errors_to setting. This could be provoked only
6606 by a command such as
6610 where an empty string was supplied; "<>" did not hit this bug.
6612 11. Installed PCRE release 4.5.
6614 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6615 remained set. It is now erased.
6617 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6618 times from message ids (which are base 36 rather than the normal 62).
6620 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6621 were being counted as actual protocol errors, and logged if the log
6622 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6623 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6624 and DATA following a set of rejected RCPTs do not count as protocol errors.
6625 In other words, Exim assumes they were pipelined, though this may not
6626 actually be the case. Of course, in all cases the client gets an
6627 appropriate error code.
6629 15. If a lookup fails in an ACL condition, a message about the failure may
6630 be available; it is used if testing the ACL cannot continue, because most
6631 such messages specify what the cause of the deferral is. However, some
6632 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6633 that caused an old message to be retained and used if a later statement
6634 caused a defer, replacing the real cause of the deferral.
6636 16. If an IP address had so many PTR records that the DNS lookup buffer
6637 was not large enough to hold them, Exim could crash while trying to process
6638 the truncated data. It now detects and logs this case.
6640 17. Further to 4.21/58, another change has been made: if (and only if) the
6641 first line of a message (the first header line) ends with CRLF, a bare LF
6642 in a subsequent header line has a space inserted after it, so as not to
6643 terminate the header.
6645 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6646 unnecessarily, used atoi() instead of strtol(), and didn't check the
6647 termination when getting file sizes from file names by regex.
6649 19. Completely re-implemented the support for maildirsize files, in the light
6650 of a number of problems with the previous contributed implementation
6651 (4.30/29). In particular:
6653 . If the quota is zero, the maildirsize file is maintained, but no quota is
6656 . If the maildir directory does not exist, it is created before any attempt
6657 to write a maildirsize file.
6659 . The quota value in the file is just a cache; if the quota is changed in
6660 the transport, the new value overrides.
6662 . A regular expression is available for excluding directories from the
6665 20. The autoreply transport checks the characters in options that define the
6666 message's headers; it allows continued headers, but it was checking with
6667 isspace() after an embedded newline instead of explicitly looking for a
6670 21. If all the "regular" hosts to which an address was routed had passed their
6671 expiry times, and had not reached their retry times, the address was
6672 bounced, even if fallback hosts were defined. Now Exim should go on to try
6675 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6676 equivalent code in the SMTP transport. Some hosts send humungous responses
6677 to HELO/EHLO, more than 1024 it seems.
6679 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6680 gives compiler warnings in some environments. I've now done it "properly",
6683 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6684 (because of problems with the built-in one) was declared to return uschar *
6685 instead of char *, causing compiler failure.
6687 25. Fixed a file descriptor leak when processing alias/forward files.
6689 26. Fixed a minor format string issue in dbfn.c.
6691 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6693 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6694 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6697 29. When a sender address is verified, it is cached, to save repeating the test
6698 when there is more than one recipient in a message. However, when the
6699 verification involves a callout, it is possible for different callout
6700 options to be set for different recipients. It is too complicated to keep
6701 track of this in the cache, so now Exim always runs a verification when a
6702 callout is required, relying on the callout cache for the optimization.
6703 The overhead is duplication of the address routing, but this should not be
6706 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6707 to be verified with callout=postmaster, and the main callout worked but the
6708 postmaster check failed, the verification correctly failed. However, if a
6709 subsequent RCPT command asked for sender verification *without* the
6710 postmaster check, incorrect caching caused this verification also to fail,
6713 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6714 it was not caching the DNS options (qualify_single, search_parents) that
6715 were used when the lookup failed. A subsequent lookup with different
6716 options therefore always gave the same answer, though there were cases
6717 where it should not have. (Example: a "domains = !$mx_any" option on a
6718 dnslookup router: the "domains" option is always processed without any
6719 widening, but the router might have qualify_single set.) Now Exim uses the
6720 cached value only when the same options are set.
6722 32. Added John Jetmore's "exipick" utility to the distribution.
6724 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6725 than a timeout (e.g. a certificate is required, and is not provided), an
6726 Exim server now closes the connection immediately. Previously it waited for
6727 the client to close - but if the client is SSL, it seems that they each
6728 wait for each other, leading to a delay before one of them times out.
6730 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6731 maintained 0.8.x compatibility because I don't think many are using it, and
6732 it is clearly obsolete.
6734 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6737 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6738 client certificate was expired. A simple patch fixes this, though I don't
6739 understand the full logic of why the verify callback is called multiple
6742 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6743 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6744 which causes problems with some clients (such as the Certicom SSL Plus
6745 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6746 disables the coutermeasure allowing Eudora to connect."
6748 38. Exim was not checking that a write() to a log file succeeded. This could
6749 lead to Bad Things if a log got too big, in particular if it hit a file
6750 size limit. Exim now panics and dies if it cannot write to a log file, just
6751 as it does if it cannot open a log file.
6753 39. Modified OS/Makefile-Linux so that it now contains
6755 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6757 The two -D definitions ensure that Exim is compiled with large file
6758 support, which makes it possible to handle log files that are bigger than
6761 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6762 instance) a domain was checked against a named list that involved a lookup,
6763 causing $domain_data to be set, then another domain was checked against the
6764 same list, then the first domain was re-checked, the value of $domain_data
6765 after the final check could be wrong. In particular, if the second check
6766 failed, it could be set empty. This bug probably also applied to
6769 41. The strip_trailing_dot option was not being applied to the address given
6770 with the -f command-line option.
6772 42. The code for reading a message's header from the spool was incrementing
6773 $received_count, but never initializing it. This meant that the value was
6774 incorrect (doubled) while delivering a message in the same process in which
6775 it was received. In the most common configuration of Exim, this never
6776 happens - a fresh exec is done - but it can happen when
6777 deliver_drop_privilege is set.
6779 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6780 it now includes up to 150 characters of the unexpected data in the log
6783 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6784 and building data strings. The size of both of these buffers was 10 000
6785 bytes - far larger than anybody would *ever* want, thought I. Needless to
6786 say, somebody hit the limit. I have increased the maximum line length to
6787 20 000 and the maximum data length of concatenated lines to 100 000. I have
6788 also fixed two bugs, because there was no checking on these buffers. Tsk,
6789 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6790 buffer is too small.
6792 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6793 lsearch lookups. Now it does.
6795 46. When parsing a route_list item in a manualroute router, a fixed-length
6796 buffer was used for the list of hosts. I made this 1024 bytes long,
6797 thinking that nobody would ever have a list of hosts that long. Wrong.
6798 Somebody had a whole pile of complicated expansion conditions, and the
6799 string was silently truncated, leading to an expansion error. It turns out
6800 that it is easier to change to an unlimited length (owing to other changes
6801 that have happened since this code was originally written) than to build
6802 structure for giving a limitation error. The length of the item that
6803 expands into the list of hosts is now unlimited.
6805 47. The lsearch lookup could not handle data where the length of text line was
6806 more than 4095 characters. Such lines were truncated, leading to shortened
6807 data being returned. It should now handle lines of any length.
6809 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6810 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6813 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6814 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6815 no longer happens in comment lines. A list of replacements is now placed
6816 at the head of all of the source files, except those whose only change is
6817 to replace PERL_COMMAND in the very first #! line.
6819 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6820 messages on the queue, with a bottom-up merge sort, using code contributed
6821 by Michael Haardt. This should make operations like -bp somewhat faster on
6822 large queues. It won't affect queue runners, except when queue_run_in_order
6825 51. Installed eximstats 1.31 in the distribution.
6827 52. Added support for SRV lookups to the dnslookup router.
6829 53. If an ACL referred to $message_body or $message_body_end, the value was not
6830 reset for any messages that followed in the same SMTP session.
6832 54. The store-handling optimization for building very long strings was not
6833 differentiating between the different store pools. I don't think this
6834 actually made any difference in practice, but I've tidied it.
6836 55. While running the routers to verify a sender address, $sender_address
6837 was still set to the sender address. This is wrong, because when routing to
6838 send a bounce to the sender, it would be empty. Therefore, I have changed
6839 it so that, while verifying a sender address, $sender_address is set to <>.
6840 (There is no change to what happens when verifying a recipient address.)
6842 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6843 target A or AAAA records (if not already returned) without resetting the
6844 qualify_single or search_parents options of the DNS resolver. These are
6845 inappropriate in this case because the targets of MX and SRV records must
6846 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6847 target that, when qualified, matched something in the local domain. These
6848 two options are now turned off when doing these lookups.
6850 57. It seems that at least some releases of Reiserfs (which does not have the
6851 concept of a fixed number of inodes) returns zero and not -1 for the
6852 number of available inodes. This interacted badly with check_spool_inodes,
6853 which assumed that -1 was the "no such thing" setting. What I have done is
6854 to check that the total number of inodes is greater than zero before doing
6855 the test of how many are available.
6857 58. When a "warn" ACL statement has a log_message modifier, the message is
6858 remembered, and not repeated. This is to avoid a lot of repetition when a
6859 message has many recipients that cause the same warning to be written.
6860 However, Exim was preserving the list of already written lines for an
6861 entire SMTP session, which doesn't seem right. The memory is now reset if a
6862 new message is started.
6864 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6865 debugging output unless log_rewrite was also set.
6867 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6868 of (int)(handle) when we know that handle contains (void *)(-1).
6870 61. The Exim daemon panic-logs an error return when it closes the incoming
6871 connection. However "connection reset by peer" seems to be common, and
6872 isn't really an error worthy of noting specially, so that particular error
6875 62. When Exim is trying to find all the local interfaces, it used to panic and
6876 die if the ioctl to get the interface flags failed. However, it seems that
6877 on at least one OS (Solaris 9) it is possible to have an interface that is
6878 included in the list of interfaces, but for which you get a failure error
6879 for this call. This happens when the interface is not "plumbed" into a
6880 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6881 failure of the "get flags" call assumes that the interface is down.
6883 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6884 makes life easier for people who are doing arithmetic on fields extracted
6885 from dates, where you often get leading zeros that should not be
6886 interpreted as octal.
6888 64. Added qualify_domain to the redirect router, to override the global
6891 65. If a pathologically long header line contained very many addresses (the
6892 report of this problem mentioned 10 000) and each of them was rewritten,
6893 Exim could use up a very large amount of memory. (It kept on making new
6894 copies of the header line as it rewrote, and never released the old ones.)
6895 At the expense of a bit more processing, the header rewriting function has
6896 been changed so that it no longer eats memory in this way.
6898 66. The generation of the Received: header has been moved from the time that a
6899 message starts to be received, to the time that it finishes. The timestamp
6900 in the Received: header should now be very close to that of the <= log
6901 line. There are two side-effects of this change:
6903 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6904 logged header lines no longer include the local Received: line, because
6905 it has not yet been created. The same applies to a copy of the message
6906 that is returned to a non-SMTP sender when a message is rejected.
6908 (b) When a filter file is tested using -bf, no additional Received: header
6909 is added to the test message. After some thought, I decided that this
6912 This change does not affect the value of $received_for. It is still set
6913 after address rewriting, but before local_scan() is called.
6915 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6917 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6918 gave an unhelpful panic error message, and a defer error. I have managed to
6919 change this behaviour so that it now rejects any supplied certificate,
6920 which seems right, as the list of acceptable certificates is empty.
6922 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6923 gave an unhelpful defer error. I have not managed to make this reject any
6924 supplied certificates, but the error message it gives is "no certificate
6925 supplied", which is not helpful.
6927 70. exigrep's output now also includes lines that are not associated with any
6928 message, but which match the given pattern. Implemented by a patch from
6929 Martin Sluka, which also tidied up the Perl a bit.
6931 71. Recipient callout verification, like sender verification, was using <> in
6932 the MAIL FROM command. This isn't really the right thing, since the actual
6933 sender may affect whether the remote host accepts the recipient or not. I
6934 have changed it to use the actual sender in the callout; this means that
6935 the cache record is now keyed on a recipient/sender pair, not just the
6936 recipient address. There doesn't seem to be a real danger of callout loops,
6937 since a callout by the remote host to check the sender would use <>.
6938 [SEE ABOVE: changed after hitting problems.]
6940 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6941 temporary errors. However, in the case of such a code being given after
6942 the end of a data transmission (i.e. after ".") Exim was failing to write
6943 a retry record for the message. (Yes, there was some broken host that was
6944 actually sending 8xx at this point.)
6946 73. An unknown lookup type in a host list could cause Exim to panic-die when
6947 the list was checked. (An example that provoked this was putting <; in the
6948 middle of a list instead of at the start.) If this happened during a DATA
6949 ACL check, a -D file could be left lying around. This kind of configuration
6950 error no longer causes Exim to die; instead it causes a defer error. The
6951 incident is still logged to the main and panic logs.
6953 74. Buglet left over from Exim 3 conversion. The message "too many messages
6954 in one connection" was written to the rejectlog but not the mainlog, except
6955 when address rewriting (yes!) was being logged.
6957 75. Added write_rejectlog option.
6959 76. When a system filter was run not as root (that is, when system_filter_user
6960 was set), the values of the $n variables were not being returned to the
6961 main process; thus, they were not subsequently available in the $sn
6964 77. Added +return_path_on_delivery log selector.
6966 78. A connection timeout was being treated differently from recipients deferred
6967 when testing hosts_max_try with a message that was older than the host's
6968 retry timeout. (The host should not be counted, thus allowing all hosts to
6969 be tried at least once before bouncing.) This may have been the cause of an
6970 occasionally reported bug whereby a message would remain on the queue
6971 longer than the retry timeout, but would be bounced if a delivery was
6972 forced. I say "may" because I never totally pinned down the problem;
6973 setting up timeout/retry tests is difficult. See also the next item.
6975 79. The ultimate address timeout was not being applied to errors that involved
6976 a combination of host plus message (for example, a timeout on a MAIL
6977 command). When an address resolved to a number of possible hosts, and they
6978 were not all tried for each delivery (e.g. because of hosts_max_try), a
6979 message could remain on the queue longer than the retry timeout.
6981 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6984 81. Fixed an obscure SMTP outgoing bug which required at least the following
6985 conditions: (a) there was another message waiting for the same server;
6986 (b) the server returned 5xx to all RCPT commands in the first message so
6987 that the message was not completed; (c) the server dropped the connection
6988 or gave a negative response to the RSET that Exim sends to abort the
6989 transaction. The observed case was a dropped connection after DATA that had
6990 been sent in pipelining mode. That is, the server had advertised PIPELINING
6991 but was not implementing it correctly. The effect of the bug was incorrect
6992 behaviour, such as trying another host, and this could lead to a crash.
6998 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6999 and daemon.c were passed as pointers to ints; they should have been
7000 pointers to socklen_t variables (which are typically unsigned ints).
7002 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7005 3. Fixed a really odd bug that affected only the testing scheme; patching a
7006 certain fixed string in the binary changed the value of another string that
7007 happened to be identical to the end of the original first string.
7009 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7010 name", it returns that address as the IP address. On some operating
7011 systems (e.g. Solaris), it also passes back the IP address string as the
7012 "host name". However, on others (e.g. Linux), it passes back an empty
7013 string. Exim wasn't checking for this, and was changing the host name to an
7014 empty string, assuming it had been canonicalized.
7016 5. Although rare, it is permitted to have more than one PTR record for a given
7017 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7018 all the names associated with an address, because they do in Solaris.
7019 However, it seems that they do not in Linux for data that comes from the
7020 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7021 I found this out when I moved to a new Linux workstation and tried to run
7022 the Exim test suite.
7024 To get round this problem I have changed the code so that it now does its
7025 own call to the DNS to look up PTR records when searching for a host name.
7026 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7027 addresses that are only in /etc/hosts are still found.
7029 This behaviour is, however, controlled by an option called host_lookup_
7030 order, which defaults to "bydns:byaddr". If people want to use the other
7031 order, or indeed, just use one or the other means of lookup, they can
7032 specify it in this variable.
7034 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7035 some operating systems, this comes back from gethostbyaddr() as an empty
7036 string, and this is what Exim used to test for. However, it seems that in
7037 other systems, "." is yielded. Exim now tests for this case too.
7039 7. The values of check_spool_space and check_log_space are now held internally
7040 as a number of kilobytes instead of an absolute number of bytes. If a
7041 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7042 kilobyte. This means that much larger values can be stored.
7044 8. Exim monitor: an attempt to get the action menu when not actually pointing
7045 at a message produces an empty menu entitled "No message selected". This
7046 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7047 no entries in it ("Shell widget menu has zero width and/or height"). So I
7048 have added a single, blank menu entry in this case.
7050 9. Added ${quote_local_part.
7052 10. MIME decoding is now applied to the contents of Subject: header lines when
7055 11. Now that a reference to $sender_host_address automatically causes a reverse
7056 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7057 host lookup before query-style lookups in lists that might use this
7058 variable. This has therefore been abolished, and the "net-" prefix is no
7059 longer necessary for query-style lookups.
7061 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7062 have been a typo for LFLAGS, so it has been changed.
7064 13. The install script calls Exim with "-C /dev/null" in order to find the
7065 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7066 to be output. However, since Exim outputs its version number before the
7067 error, it didn't break the script. It just looked ugly. I fixed this by
7068 always allowing "-C /dev/null" if the caller is root.
7070 14. Ignore overlarge ACL variable number when reading spool file - insurance
7071 against a later release with more variables having written the file.
7073 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7074 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7075 this, as well as the form without the "IPv6" on the front.
7077 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7078 OS/Makefile-Darwin file.
7080 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7081 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7083 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7084 variable. However, some OS use ints for inodes. I've added cast to long int
7085 to get rid of the compiler warning.
7087 19. I had forgotten to lock out "/../" in configuration file names when
7088 ALT_CONFIG_PREFIX was set.
7090 20. Routers used for verification do not need to specify transports. However,
7091 if such a router generated a host list, and callout was configured, Exim
7092 crashed, because it could not find a port number from the (non-existent)
7093 transport. It now assumes port 25 in this circumstance.
7095 21. Added the -t option to exigrep.
7097 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7098 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7099 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7100 them is set, LOOKUP_LSEARCH is forced.
7102 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7104 24. Added sender and host information to the "rejected by local_scan()" log
7105 line; previously there was no indication of these.
7107 25. Added .include_if_exists.
7109 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7110 Linux. It turns out that not all file systems support this. Apparently some
7111 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7112 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7113 not supported on the file descriptor, is now ignored when Exim is trying to
7114 sync a directory. This applies only to Linux.
7116 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7118 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7119 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7122 29. Added support for maildirsize files from supplied patch (modified a bit).
7124 30. The use of :fail: followed by an empty string could lead Exim to respond to
7125 sender verification failures with (e.g.):
7127 550 Verification failed for <xxx>
7128 550 Sender verify failed
7130 where the first response line was missing the '-' that indicates it is not
7131 the final line of the response.
7133 31. The loop for finding the name of the user that called Exim had a hardwired
7134 limit of 10; it now uses the value of finduser_retries, which is used for
7135 all other user lookups.
7137 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7140 33. Exim was neglecting to zero errno before one call of strtol() when
7141 expanding a string and expecting an integer value. On some systems this
7142 resulted in spurious "integer overflow" errors. Also, it was casting the
7143 result into an int without checking.
7145 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7146 did not work. The code looks as if it has *never* worked, though it appears
7147 to have been documented since at least release 1.62. I have made it work.
7149 35. The "timeout_DNS" error in retry rules, also documented since at least
7150 1.62, also never worked. As it isn't clear exactly what this means, and
7151 clearly it isn't a major issue, I have abolished the feature by treating it
7152 as "timeout", and writing a warning to the main and panic logs.
7154 36. The display of retry rules for -brt wasn't always showing the error code
7157 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7158 timeout_connect_A, timeout_connect_MX.
7160 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7161 to the empty sender.
7163 39. The daemon was not analysing the content of -oX till after it had closed
7164 stderr and disconnected from the controlling terminal. This meant that any
7165 syntax errors were only noted on the panic log, and the return code from
7166 the command was 0. By re-arranging the code a little, I've made the
7167 decoding happen first, so such errors now appear on stderr, and the return
7168 code is 1. However, the actual setting up of the sockets still happens in
7169 the disconnected process, so errors there are still only recorded on the
7172 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7173 connections (as happens on some IP stacks) was logged at start up time as
7174 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7175 it from "IPv6 and IPv4", which means that two separate sockets are being
7178 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7179 says whether AF_INET or AF_INET6 was passed as an argument.
7181 42. Exiwhat output was messed up when time zones were included in log
7184 43. Exiwhat now gives more information about the daemon's listening ports,
7185 and whether -tls-on-connect was used.
7187 44. The "port" option of the smtp transport is now expanded.
7189 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7190 silently ignored. Now an error message is written to the main and panic
7193 46. There's a new ACL modifier called "logwrite" which writes to a log file
7194 as soon as it is encountered.
7196 47. Added $local_user_uid and $local_user_gid at routing time.
7198 48. Exim crashed when trying to verify a sender address that was being
7201 49. Exim was recognizing only a space character after ".include". It now also
7202 recognizes a tab character.
7204 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7205 extracting the relevant information from the specification. The man page no
7206 longer contains scrambled data for the -d option, and I've added a section
7207 at the front about calling Exim under different names.
7209 51. Added "extra_headers" argument to the "mail" command in filter files.
7211 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7214 53. Installed eximstats 1.29.
7216 54. Added transport_filter_timeout as a generic transport option.
7218 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7219 Cc: header lines. This was required by RFC 822, but it not required by RFC
7222 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7223 incoming messages that did not have them. Now it does so only if the
7224 message originates locally, that is, if there is no associated remote host
7225 address. When Resent- header lines are present, this applies to the Resent-
7226 lines rather than the non-Resent- lines.
7228 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7229 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7231 58. Messages for configuration errors now include the name of the main
7232 configuration file - useful now that there may be more than one file in a
7233 list (.included file names were always shown).
7235 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7236 for those rare installations that do not start the daemon as root or run it
7237 setuid root. I've cut out the call to initgroups() if the daemon is not
7240 60. The Exim user and group can now be bound into the binary as text strings
7241 that are looked up at the start of Exim's processing.
7243 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7245 62. Added $mailstore_basename variable.
7247 63. Installed patch to sieve.c from Michael Haardt.
7249 64. When Exim failed to open the panic log after failing to open the main log,
7250 the original message it was trying to log was written to stderr and debug
7251 output, but if they were not available (the usual case in production), it
7252 was lost. Now it is written to syslog before the two lines that record the
7253 failures to open the logs.
7255 65. Users' Exim filters run in subprocesses under the user's uid. It is
7256 possible for a "deliver" command or an alias in a "personal" command to
7257 provoke an address rewrite. If logging of address rewriting is configured,
7258 this fails because the process is not running as root or exim. There may be
7259 a better way of dealing with this, but for the moment (because 4.30 needs
7260 to be released), I have disabled address rewrite logging when running a
7261 filter in a non-root, non-exim process.
7267 1. The buildconfig auxiliary program wasn't quoting the value set for
7268 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7269 not defined. This bug was masked in 4.22 by the effect that was fixed in
7272 2. Some messages that were rejected after a message id was allocated were
7273 shown as "incomplete" by exigrep. It no longer does this for messages that
7274 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7276 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7277 have allow_domain_literals set, the ID did not get logged in the <= line.
7278 Domain literals are now always recognized in Message-ID: header lines.
7280 4. The first argument for a ${extract expansion item is the key name or field
7281 number. Leading and trailing spaces in this item were not being ignored,
7282 causing some misleading effects.
7284 5. When deliver_drop_privilege was set, single queue runner processes started
7285 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7286 same command in the process it spins off) were not dropping privilege.
7288 6. When the daemon running as "exim" started a queue runner, it always
7289 re-executed Exim in the spun-off process. This is a waste of effort when
7290 deliver_drop_privilege is set. The new process now just calls the
7291 queue-runner function directly.
7297 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7300 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7301 ignored. Though the use of -oP was forcing the writing of a pid file, it
7302 was always written to the default place.
7304 3. If the message "no IP address found for host xxxx" is generated during
7305 incoming verification, it is now followed by identification of the incoming
7306 connection (so you can more easily find what provoked it).
7308 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7310 5. Added some features to "harden" Exim a bit more against certain attacks:
7312 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7313 be put in Local/Makefile. This is like the never_users runtime option,
7314 but it cannot be overridden. The default setting is "root".
7316 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7317 prefix string with which any file named in a -C command line option
7320 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7321 is retained for -C and -D only if the caller of Exim is root. Without
7322 it, the exim user may also use -C and -D and retain privilege.
7324 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7325 command line option is disabled.
7327 6. Macro names set by the -D option must start with an upper case letter, just
7328 like macro names defined in the configuration file.
7330 7. Added "dereference=" facility to LDAP.
7332 8. Two instances of the typo "uknown" in the source files are fixed.
7334 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7335 the Configure-Makefile script screwed up while processing it.
7337 10. Incorporated PCRE 4.4.
7339 11. The SMTP synchronization check was not operating right at the start of an
7340 SMTP session. For example, it could not catch a HELO sent before the client
7341 waited for the greeting. There is now a check for outstanding input at the
7342 point when the greeting is written. Because of the duplex, asynchronous
7343 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7344 way, but not yet received, when the check is performed.
7346 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7347 on TCP/IP sockets, because this apparently causes some broken clients to
7350 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7351 unchanged) from the Cygwin maintainer.
7353 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7354 was supported instead of testing for mbx. Effectively a typo.
7356 15. The spa authenticator server code was not checking that the input it
7357 received was valid base64.
7359 16. The debug output line for the "set" modifier in ACLs was not showing the
7360 name of the variable that was being set.
7362 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7364 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7365 lookup on its own. Something else was needed to trigger the lookup. For
7366 example, a match in host_lookup or the need for a host name in a host list.
7367 Now, if $sender_host_name is referenced and the host name has not yet been
7368 looked up, a lookup is performed. If the lookup fails, the variable remains
7369 empty, and $host_lookup_failed is set to "1".
7371 19. Added "eqi" as a case-independent comparison operator.
7373 20. The saslauthd authentication condition could segfault if neither service
7374 nor realm was specified.
7376 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7377 error message that was logged was misleading, and incoming SMTP
7378 connections were dropped. The message is now more accurate, and temporary
7379 errors are given to SMTP connections.
7381 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7382 (except RSET) with a 421 error, until QUIT is received. However, it was
7383 failing to send a response to QUIT.
7385 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7386 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7387 ACL is now run after the helo_try_verify_hosts code.
7389 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7390 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7391 case-independent, so other case variants are also recognized). Apparently
7392 some systems use these upper case variants.
7394 25. If more than two messages were waiting for the same host, and a transport
7395 filter was specified for the transport, Exim sent two messages over the
7396 same TCP/IP connection, and then failed with "socket operation on non-
7397 socket" when it tried to send the third.
7399 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7401 27. The extern definition of crypt16() in expand.c was not being excluded when
7402 the OS had its own crypt16() function.
7404 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7405 as a preferred synonym for return_size_limit, both as an option and as an
7408 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7410 30. Changed the default configuration ACL to relax the local part checking rule
7411 for addresses that are not in any local domains. For these addresses,
7412 slashes and pipe symbols are allowed within local parts, but the sequence
7413 /../ is explicitly forbidden.
7415 31. SPA server authentication was not clearing the challenge buffer before
7418 32. log_message in a "warn" ACL statement was writing to the reject log as
7419 well as to the main log, which contradicts the documentation and doesn't
7420 seem right (because no rejection is happening). So I have stopped it.
7422 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7423 However, I am unable to do any testing of this.
7425 34. Fixed an infelicity in the appendfile transport. When checking directories
7426 for a mailbox, to see if any needed to be created, it was accidentally
7427 using path names with one or more superfluous leading slashes; tracing
7428 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7430 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7431 contents are added to the log line that is written for every discarded
7432 recipient. (Previously a log_message setting was ignored.)
7434 36. The ${quote: operator now quotes the string if it is empty.
7436 37. The install script runs exim in order to find its version number. If for
7437 some reason other than non-existence or emptiness, which it checks, it
7438 could not run './exim', it was installing it with an empty version number,
7439 i.e. as "exim-". This error state is now caught, and the installation is
7442 38. An argument was missing from the function that creates an error message
7443 when Exim fails to connect to the socket for saslauthd authentication.
7444 This could cause Exim to crash, or give a corrupted message.
7446 39. Added isip, isip4, and isip6 to ${if conditions.
7448 40. The ACL variables $acl_xx are now saved with the message, and can be
7449 accessed later in routers, transports, and filters.
7451 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7452 strings in the file are not string-expanded.
7454 42. If a MAIL command specified a SIZE value that was too large to fit into an
7455 int variable, the check against message_size_limit failed. Such values are
7456 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7457 day this will have to be increased, but I don't think I want to be around
7458 when emails are that large.
7465 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7466 iconv() is not standard in FreeBSD.
7468 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7469 IPv6 enabled. The observed symptom was a segmentation fault on return from
7470 the function os_common_find_running_interfaces() in src/os.c.
7472 3. In the check_special_case() function in daemon.c I had used "errno" as an
7473 argument name, which causes warnings on some systems. This was basically a
7474 typo, since it was named "eno" in the comments!
7476 4. The code that waits for the clock to tick (at a resolution of some fraction
7477 of a second) so as to ensure message-id uniqueness was always waiting for
7478 at least one whole tick, when it could have waited for less. [This is
7479 almost certainly not relevant at current processor speeds, where it is
7480 unlikely to ever wait at all. But we try to future-proof.]
7482 5. The function that sleeps for a time interval that includes fractions of a
7483 second contained a race. It did not block SIGALRM between setting the
7484 timer, and suspending (a couple of lines later). If the interval was short
7485 and the sigsuspend() was delayed until after it had expired, the suspension
7486 never ended. On busy systems this could lead to processes getting stuck for
7489 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7490 process, before it forks any delivery processes. The open lookup caching
7491 mechanism meant that the open file or database connection was passed into
7492 the delivery process. The problem was that delivery processes always tidy
7493 up cached lookup data. This could cause a problem for the next delivery
7494 process started by the queue runner, because the external queue runner
7495 process does not know about the closure. So the next delivery process
7496 still has data in the lookup cache. In the case of a file lookup, there was
7497 no problem because closing a file descriptor in a subprocess doesn't affect
7498 the parent. However, if the lookup was caching a connection to a database,
7499 the connection was closed, and the second delivery process was likely to
7500 see errors such as "PGSQL: query failed: server closed the connection
7501 unexpectedly". The problem has been fixed by closing all cached lookups
7502 in a queue runner before running a delivery process.
7504 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7505 seem to have the "const" qualifier which it has on other OS. I've
7508 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7509 *of the same type* (client or server) with the same public name that an
7510 error should be diagnosed.
7512 9. When Exim looked up a host name for an IP address, but failed to find the
7513 original IP address when looking up the host name (a safety check), it
7514 output the message "<ip address> does not match any IP for NULL", which was
7515 confusing, to say the least. The bug was that the host name should have
7516 appeared instead of "NULL".
7518 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7519 user that is built into the binary, and the -C or -D options is used, root
7520 privilege is dropped before the configuration file is read. In addition,
7521 logging is switched to stderr instead of the normal log files. If the
7522 configuration then re-defines the Exim user, the unprivileged environment
7523 is probably not what is expected, so Exim logs a panic warning message (but
7526 However, if deliver_drop_privilege is set, the unprivileged state may well
7527 be exactly what is intended, so the warning has been cut out in that case,
7528 and Exim is allowed to try to write to its normal log files.
7534 1. smtp_return_error_details was not giving details for temporary sender
7535 or receiver verification errors.
7537 2. Diagnose a configuration error if two authenticators have the same public
7540 3. Exim used not to create the message log file for a message until the first
7541 delivery attempt. This could be confusing when incoming messages were held
7542 for policy or load reasons. The message log file is now created at the time
7543 the message is received, and an initial "Received" line is written to it.
7545 4. The automatically generated man page for command line options had a minor
7546 bug that caused no ill effects; however, a more serious problem was that
7547 the procedure for building the man page automatically didn't always
7548 operate. Consequently, release 4.20 contains an out-of-date version. This
7549 shouldn't happen again.
7551 5. When building Exim with embedded Perl support, the script that builds the
7552 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7553 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7555 6. The freeze_tell option was not being used for messages that were frozen on
7556 arrival, either by an ACL or by local_scan().
7558 7. Added the smtp_incomplete_transaction log selector.
7560 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7561 was accepting AUTH without a new EHLO.
7563 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7564 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7567 10. Logging of TCP/IP connections (when configured) now happens in the main
7568 daemon process instead of the child process, so that the TCP/IP connection
7569 count is more accurate (but it can never be perfect).
7571 11. The use of "drop" in a nested ACL was not being handled correctly in the
7572 outer ACL. Now, if condition failure induced by the nested "drop" causes
7573 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7574 or "require"), the connection is dropped.
7576 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7577 that yield "discard" can now be used with an "accept" or a "discard" verb,
7578 but an error is generated for any others (because I can't see a useful way
7579 to define what should happen).
7581 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7582 are now processed in the same way as lines in the Exim configuration file.
7583 In particular, continuation lines are supported.
7585 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7587 15. Added -ti meaning -t -i.
7589 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7590 domains, and warn by logging if others are found.
7592 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7593 returned from GIFCONF when Exim is trying to find the list of interfaces on
7594 a host. The code in os.c has been modified to copy each ifreq to an aligned
7595 structure in all cases.
7597 Also, in some cases, the returned ifreq's were being copied to a 'struct
7598 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7599 means the last couple of bytes of an IPv6 address could be chopped if the
7600 ifreq contained only a normal sockaddr (14 bytes storage).
7602 18. Named domain lists were not supported in the hosts_treat_as_local option.
7603 An entry such as +xxxx was not recognized, and was treated as a literal
7606 19. Ensure that header lines added by a DATA ACL are included in the reject log
7607 if the ACL subsequently rejects the message.
7609 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7610 MD5 (which is deprecated).
7612 21. When testing a filter file using -bf, Exim was writing a message when it
7613 took the sender from a "From " line in the message, but it was not doing so
7614 when it took $return_path from a Return-Path: header line. It now does.
7616 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7617 with a valid header line field name (a series of printing characters
7618 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7620 23. Changed "disc" in the source to "disk" to conform to the documentation and
7621 the book and for uniformity.
7623 24. Ignore Sendmail's -Ooption=value command line item.
7625 25. When execve() failed while trying to run a command in a pipe transport,
7626 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7627 could be confused with a return value of 69 from the command itself. This
7628 has been changed to 127, the value the shell returns if it is asked to run
7629 a non-existent command. The wording for the related log line suggests a
7630 non-existent command as the problem.
7632 26. If received_header_text expands to an empty string, do not add a Received:
7633 header line to the message. (Well, it adds a token one on the spool, but
7634 marks it "old" so that it doesn't get used or transmitted.)
7636 27. Installed eximstats 1.28 (addition of -nt option).
7638 28. There was no check for failure on the call to getsockname() in the daemon
7639 code. This can fail if there is a shortage of resources on the system, with
7640 ENOMEM, for example. A temporary error is now given on failure.
7642 29. Contrary to the C standard, it seems that in some environments, the
7643 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7644 program. Exim now does this explicitly; it affects the formatting of
7645 timestamps using strftime().
7647 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7648 complaints. I've now initialized all the variables, to avoid this.
7650 32. Header lines added by a system filter were not being "seen" during
7651 transport-time rewrites.
7653 33. The info_callback() function passed to OpenSSL is set up with type void
7654 (*)(SSL *, int, int), as described somewhere. However, when calling the
7655 function (actually a macro) that sets it up, the type void(*)() is
7656 expected. I've put in a cast to prevent warnings from picky compilers.
7658 34. If a DNS black list lookup found a CNAME record, but there were no A
7659 records associated with the domain it pointed at, Exim crashed.
7661 35. If a DNS black list lookup returned more than one A record, Exim ignored
7662 all but the first. It now scans all returned addresses if a particular IP
7663 value is being sought. In this situation, the contents of the
7664 $dnslist_value variable are a list of all the addresses, separated by a
7667 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7668 to accept a host name and all its aliases if the forward lookup for any of
7669 them yielded the IP address of the incoming connection. Now it accepts only
7670 those names whose forward lookup yields the correct IP address. Any other
7671 names are discarded. This closes a loophole whereby a rogue DNS
7672 administrator could create reverse DNS records to break through a
7673 wildcarded host restriction in an ACL.
7675 37. If a user filter or a system filter that ran in a subprocess used any of
7676 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7677 the wrong values were passed to the pipe command ($thisaddress had the
7678 value of $0, $0 had the value of $1, etc). This bug was introduced by
7679 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7681 38. Improved the line breaking for long SMTP error messages from ACLs.
7682 Previously, if there was no break point between 40 and 75 characters, Exim
7683 left the rest of the message alone. Two changes have been made: (a) I've
7684 reduced the minimum length to 35 characters; (b) if it can't find a break
7685 point between 35 and 75 characters, it looks ahead and uses the first one
7686 that it finds. This may give the occasional overlong line, but at least the
7687 remaining text gets split now.
7689 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7690 file descriptors might be low, and that setting 1000 would always raise it.
7691 It turns out that in some environments, the limit is already over 1000 and
7692 that lowering it causes trouble. So now Exim takes care not to decrease it.
7694 40. When delivering a message, the value of $return_path is set to $sender_
7695 address at the start of routing (routers may change the value). By an
7696 oversight, this default was not being set up when an address was tested by
7697 -bt or -bv, which affected the outcome if any router or filter referred to
7700 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7701 log only when "message" or "log_message" are set. However, if one of the
7702 conditions was an address verification, or a call to a nested ACL, the
7703 messages generated by the underlying test were being passed through. This
7704 no longer happens. The underlying message is available in $acl_verify_
7705 message for both "message" and "log_message" expansions, so it can be
7706 passed through if needed.
7708 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7709 new expansion $bh_ to give the encoded byte string without charset
7710 translation. Translation happens only if iconv() is available; HAVE_ICONV
7711 indicates this at build time. HEADERS_CHARSET gives the charset to
7712 translate to; headers_charset can change it in the configuration, and
7713 "headers charset" can change it in an individual filter file.
7715 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7716 that creates RFC 2047 encoded "words" labels them as that charset instead
7717 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7718 expansion operator; (ii) when Exim creates a From: line for a local
7719 message; (iii) when a header line is rewritten to include a "phrase" part.
7721 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7722 buggy, causing it to skip the first lines of messages whose message ID
7723 ended in 'D'. This would not have bitten before Exim release 4.14, because
7724 message IDs were unlikely to end in 'D' before then. The effect was to have
7725 incorrect size information for certain domains.
7727 45. #include "config.h" was missing at the start of the crypt16.c module. This
7728 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7731 46. If there was a timeout during a "random" callout check, Exim treated it as
7732 a failure of the random address, and carried on sending RSET and the real
7733 address. If the delay was just some slowness somewhere, the response to the
7734 original RCPT would be taken as a response to RSET and so on, causing
7735 mayhem of various kinds.
7737 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7738 when I implemented it. It didn't allow for the fact that some option values
7739 may legitimately be negative (e.g. size_addition), and it didn't even do
7740 the right test for positive values.
7742 48. Domain names in DNS records are case-independent. Exim always looks them up
7743 in lower case. Some resolvers return domain names in exactly the case they
7744 appear in the zone file, that is, they may contain uppercase letters. Not
7745 all resolvers do this - some return always lower case. Exim was treating a
7746 change of case by a resolver as a change of domain, similar to a widening
7747 of a domain abbreviation. This triggered its re-routing code and so it was
7748 trying to route what was effectively the same domain again. This normally
7749 caused routing to fail (because the router wouldn't handle the domain
7750 twice). Now Exim checks for this case specially, and just changes the
7751 casing of the domain that it ultimately uses when it transmits the message
7754 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7757 50. If a filter generated a file delivery with a non-absolute name (possible if
7758 no home directory exists for the router), the forbid_file option was not
7761 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7762 the existing equality matching.
7764 52. Exim was using ints instead of ino_t variables in some places where it was
7765 dealing with inode numbers.
7767 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7768 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7769 called TMPDIR, and if it finds it is different, it changes its value.
7771 54. The smtp_printf() function is now made available to local_scan() so
7772 additional output lines can be written before returning. There is also an
7773 smtp_fflush() function to enable the detection of a dropped connection.
7774 The variables smtp_input and smtp_batched_input are exported to
7777 55. Changed the default runtime configuration: the message "Unknown user"
7778 has been removed from the ACL, and instead placed on the localuser router,
7779 using the cannot_route_message feature. This means that any verification
7780 failures that generate their own messages won't get overridden. Similarly,
7781 the "Unrouteable address" message that was in the ACL for unverifiable
7782 relay addresses has also been removed.
7784 56. Added hosts_avoid_esmtp to the smtp transport.
7786 57. The exicyclog script was not checking for the esoteric option
7787 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7788 will work only if exicyclog is run under the appropriate euid.
7790 58. Following a discussion on the list, the rules by which Exim recognises line
7791 endings on incoming messages have been changed. The -dropcr and drop_cr
7792 options are now no-ops, retained only for backwards compatibility. The
7793 following line terminators are recognized: LF CRLF CR. However, special
7794 processing applies to CR:
7796 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7797 nor a local message in the state where . is a terminator.
7799 (ii) If a bare CR is encountered in a header line, an extra space is added
7800 after the line terminator so as not to end the header. The reasoning
7801 behind this is that bare CRs in header lines are most likely either
7802 to be mistakes, or people trying to play silly games.
7804 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7805 was being incorrectly given as 18 bytes larger than it should have been.
7806 This is a VOB (very old bug).
7808 60. This may never have affected anything current, but just in case it has:
7809 When the local host is found other than at the start of a list of hosts,
7810 the local host, those with the same MX, and any that follow, are discarded.
7811 When the list in question was part of a longer list of hosts, the following
7812 hosts (not currently being processed) were also being discarded. This no
7813 longer happens. I'm not sure if this situation could ever has previously
7816 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7819 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7820 header. This is something that is recommended in a new Internet Draft, and
7821 is something that is documented as being done by Sendmail. There are two
7822 possible values. For messages generated by the autoreply transport, Exim
7825 Auto-Submitted: auto-replied
7827 whereas for all other generated messages (e.g. bounces) it adds
7829 Auto-Submitted: auto-generated
7831 63. The "personal" condition in filters now includes a test for the
7832 Auto-Submitted: header. If it contains the string "auto-" the message it
7833 not considered personal.
7835 64. Added rcpt_include_affixes as a generic transport option.
7837 65. Added queue_only_override (default true).
7839 66. Added the syslog_duplication option.
7841 67. If what should have been the first header line of a message consisted of
7842 a space followed by a colon, Exim was mis-interpreting it as a header line.
7843 It isn't of course - it is syntactically invalid and should therefore be
7844 treated as the start of the message body. The misbehaviour could have
7845 caused a number of strange effects, including loss of data in subsequent
7846 header lines, and spool format errors.
7848 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7849 client host had authenticated. This control can now be exercised by an ACL
7850 for more flexibility.
7852 69. By default, callouts do not happen when testing with -bh. There is now a
7853 variant, -bhc, which does actually run the callout code, including
7854 consulting and updating the callout cache.
7856 70. Added support for saslauthd authentication, courtesy of Alexander
7859 71. If statvfs() failed on the spool or log directories while checking their
7860 size for availability, Exim confusingly gave the error "space shortage".
7861 Furthermore, in debugging mode it crashed with a floating point exception.
7862 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7863 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7864 serious problem, Exim now writes to the main and panic logs when this
7865 happens, with details of the failure. It then refuses to accept the
7866 incoming message, giving the message "spool directory problem" or "log
7867 directory problem" with a 421 code for SMTP messages.
7869 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7870 0, 1, and 2 exist, because some OS complain for execs without them (see
7871 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7872 descriptors. However, the code omitted to check that the open succeeded,
7873 causing mysterious errors if for some reason the permissions on /dev/null
7874 got screwed. Now Exim writes a message to the main and panic logs, and
7875 bombs out if it can't open /dev/null.
7877 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7878 interact so that it is all more flexible. It is supposed to remain
7879 backwards compatible. Also added extra_local_interfaces.
7881 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7882 to bomb out with an assertion failure - to the client this appears as a
7883 connection drop. This problem occurs in the part of the code that was taken
7884 from the Samba project. Fortunately, the assertion is in a very simple
7885 function, so I have fixed this by reproducing the function inline in the
7886 one place where it is called, and arranging for authentication to fail
7887 instead of killing the process with assert().
7889 75. The SPA client code was not working when the server requested OEM rather
7890 than Unicode encoding.
7892 76. Added code to make require_files with a specific uid setting more usable in
7893 the case where statting the file as root fails - usually a non-root-mounted
7894 NFS file system. When this happens and the failure is EACCES, Exim now
7895 forks a subprocess and does the per-uid checking as the relevant uid.
7897 77. Added process_log_path.
7899 78. If log_file_path was not explicitly set, a setting of check_log_space or
7900 check_log_inodes was ignored.
7902 79. If a space check for the spool or log partitions fails, the incident is now
7903 logged. Of course, in the latter case the data may get lost...
7905 80. Added the %p formatting code to string_format() so that it can be used to
7906 print addresses in debug_print(). Adjusted all the address printing in the
7907 debugging in store.c to use %p rather than %d.
7909 81. There was a concern that a line of code in smtp_in.c could overflow a
7910 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7911 initially expressed, the concern was not well-founded, because trailing
7912 spaces are removed early. However, if the trailing spaces were followed by
7913 a NULL, they did not get removed, so the overflow was possible. Two fixes
7916 (a) I re-wrote the offending code in a cleaner fashion.
7917 (b) If an incoming SMTP command contains a NULL character, it is rejected
7920 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7921 runs initgroups(), so that if the Exim user is in any additional groups,
7922 they will be used during message reception.
7928 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob