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
136 JH/01 Avoid costly startup code when not strictly needed. This reduces time
137 for some exim process initialisations. It does mean that the logging
138 of TLS configuration problems is only done for the daemon startup.
140 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
142 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
143 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
146 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
147 DKIM signing is being done. Previously a transport_filter would always
148 disable CHUNKING, falling back to traditional DATA.
150 JH/05 Regard command-line receipients as tainted.
152 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
154 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
155 PAM library frees one of the arguments given to it, despite the
156 documentation. Therefore a plain malloc must be used.
158 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
159 on-stack buffers were used, resulting in a taint trap when DSN information
160 copied from a received message was written into the buffer.
162 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
163 the ordering of its ARC headers. This caused a crash.
165 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
166 a new record was being constructed with information from the peer, a trap
169 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
170 installation would get error messages from DMARC verify, when it hit the
171 nonexistent file indicated by the default. Distros wanting DMARC enabled
172 should both provide the file and set the option.
173 Also enforce no DMARC verification for command-line sourced messages.
175 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
176 could, depending on the platform, hang at the STARTTLS response.
178 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
179 message on a connection. Previously if one message had ARC headers and
180 the following one did not, a crash could result when adding an
181 Authentication-Results: header.
183 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
184 Exim main code and Exim-related utities. The introduction of taint
185 tracking also did many adjustments to string handling. Since then, eximon
186 frequently terminated with an assert failure.
188 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
189 check for 452 responses. This slightly helps the inefficieny of doing
190 a large alias-expansion into a recipient-limited target. The max_rcpt
191 transport option still applies (and at the current default, will override
192 the new feature). The check is done for either cause of synch, and forces
193 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
194 connection. The new facility is not tunable at this time.
196 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
197 library live data was being used, so the results became garbage. Make
198 copies while it is still usable.
200 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
201 on delivery deferred (==) and failed (**) lines (if a delivery was
202 attemtped). Previously it was only on completion (=>) lines.
204 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
205 for the expansion of the server_scram_iter and server_scram_salt options.
207 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
208 are now specifically given a NO_DATA response without hitting the system
209 resolver. The library goes on to do the now-standard TXT lookup.
210 Use of dnsdb lookups is not affected.
212 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
213 only retrieve the errormessage once. Previously two calls to dlerror()
214 were used, and the second one (for mainlog/paniclog) retrieved null
217 JH/20 Taint checking: disallow use of tainted data for
218 - the appendfile transport file and directory options
219 - the pipe transport command
220 - the autoreply transport file, log and once options
221 - file names used by the redirect router (including filter files)
223 - paths used by single-key lookups
224 Previously this was permitted.
226 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
227 adjusted the size of a major service buffer; this failed because the
228 buffer was in use at the time. Change to a compile-time increase in the
229 buffer size, when this authenticator is compiled into exim.
231 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
232 previous fast-mode was untenable in the face of glibs using mmap to
233 support larger malloc requests.
235 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
236 New values supported, if defined on system where compiled:
237 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
238 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
240 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
241 running a limited number of proceses in parallel, a benefit is gained. The
242 amount varies with the platform hardware and load. The use of the option
243 queue_run_in_order means we cannot do this, as ordering becomes
246 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
247 had introduced a string-copy (for ensuring NUL-termination) which was not
248 appropriate for that case, which can include embedded NUL bytes in the
249 block of data. Investigation showed the copy to actually be needless, the
250 data being length-specified.
252 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
253 done during a receiving connection, and both used TLS, global info was
254 used rather than per-connection info for tracking the state of data
255 queued for transmission. This could result in a connection hang.
257 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
258 Previously, when delivering serveral messages down a single connection
259 only the first would provide a SIZE. This was due to the size information
260 not being properly tracked.
262 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
263 TAI (at 37 seconds currently), pretend to be in UTC for time-related
264 expansion and logging. Previously, spurious values such as a future
265 minute could be seen.
267 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
268 it could crash from a null-deref. This could also affect the
269 ${addresses: } operator and ${readsock } item.
271 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
272 message following a mime one, the variable was not reset.
274 JH/30 When an pipelined-connect fails at the first response, assume incorrect
275 cached capability (perhaps the peer reneged?) and immediately retry in
278 JH/31 Fix spurious detection of timeout while writing to transport filter.
280 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
281 an attempt to copy the string was made before checking it.
283 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
284 taint of the lookup key was maintained; we now regard the presence in the
285 filesystem as sufficient validation.
287 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
288 argument is supplied.
290 JH/35 The dsearch lookup now requires that the directory is an absolute path.
291 Previously this was not checked, and nonempty relative paths made an
292 access under Exim's current working directory.
294 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
295 Previously no event was raised.
297 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
298 parameter supplied by the sender MAIL FROM command. Previously it was
299 ignored, and only the check_spool_space option value for the required
302 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
303 the size of the signing public-key. Previously it was instead giving
304 the size of the signature hash.
306 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
307 the default. See the (new) dkim_verify_min_keysizes option.
309 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
310 and an ACL use a lookup for checking either the local_part or domain,
311 stale data could be accessed. Ensure that variable references are
312 dropped between messages.
314 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
315 by the client was not checked as pointing within response data before
316 being used. A malicious client could thus cause an out-of-bounds read and
317 possibly gain authentication. Fix by adding the check.
319 JH/42 Internationalisation: change the default for downconversion in the smtp
320 transport to be "if needed". Previously it was "as previously set" for
321 the message, which usually meant "if needed" for message-submission but
322 "no" for everything else. However, MTAs have been seen using SMTPUTF8
323 even when the envelope addresses did not need it, resulting in forwarding
324 failures to non-supporting MTAs. A downconvert in such cases will be
325 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
326 transport. The change does mean that addresses needing conversion will
327 be converted when previously a delivery failure would occur.
329 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
330 response was received it would be used unchecked in a fail-DSN, violating
331 standards on line-length limits. Truncate if needed.
333 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
334 doesn't work. (Additionally add a new main config option to configure the
341 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
342 side, to decode a TLS 1.3 packet capture.
344 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
345 Previously the default library behaviour applied, sending two, each in
348 JH/03 Debug output for ACL now gives the config file name and line number for
351 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
353 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
355 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
356 buffer overrun for (non-chunking) other transports.
358 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
359 TLS1.3, means that a server rejecting a client certificate is not visible
360 to the client until the first read of encrypted data (typically the
361 response to EHLO). Add detection for that case and treat it as a failed
362 TLS connection attempt, so that the normal retry-in-clear can work (if
363 suitably configured).
365 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
366 and/or domain. Found and fixed by Jason Betts.
368 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
369 configuration). If a CNAME target was not a wellformed name pattern, a
372 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
373 the OS reports them interleaved with other addresses.
375 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
376 used both for input and for a verify callout, both encrypted, SMTP
377 responses being sent by the server could be lost. This resulted in
378 dropped connections and sometimes bounces generated by a peer sending
381 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
382 string. Previously it was possible to cause undefined behaviour in a
383 library routine (usually a crash). Found by "zerons".
385 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
388 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
389 API was removed, so update to use the newer ones.
391 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
392 any timeout set, is taking a long time. Previously we would hang on to a
393 rotated logfile "forever" if the input was arriving with long gaps
394 (a previous attempt to fix addressed lack, for a long time, of initial
397 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
398 shared (NFS) environment. The length of the tempfile name is now
399 4 + 16 ("hdr.$message_exim_id") which might break on file
400 systems which restrict the file name length to lower values.
403 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
404 shared (NFS) environment.
406 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
407 did for all versions <4.90). Notably -M, -m, --invert, -I may be
410 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
411 on some platforms for bit 31.
413 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
414 to changes apparently associated with TLS1.3 handling some of the APIs
415 previously used were either nonfunctional or inappropriate. Strings
416 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
417 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
418 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
419 This affects log line X= elements, the $tls_{in,out}_cipher variables,
420 and the use of specific cipher names in the encrypted= ACL condition.
422 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
424 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
425 verification result was not updated unless hosts_require_ocsp applied.
427 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
428 queue_list_requires_admin set to false, non-admin users were denied the
431 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
432 directory-of-certs mode. Previously they were advertised despite the
435 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
436 A single TCP connection by a client will now hold a TLS connection open
437 for multiple message deliveries, by default. Previously the default was to
440 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
441 default. If built with the facility, DANE will be used. The facility
442 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
444 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
445 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
446 must be defined and you must still, unless you define DISABLE_TLS, manage
447 the the include-dir and library-file requirements that go with that
448 choice. Non-TLS builds are still supported.
450 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
453 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
454 default. If the platform supports and has the facility enabled, it will
455 be requested on all coneections.
457 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
458 controlled by the build-time option SUPPORT_PIPE_CONNECT.
460 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
462 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
463 success-DSN messages. Previously the From: header was always the default
464 one for these; the option was ignored.
466 JH/28 Fix the timeout on smtp response to apply to the whole response.
467 Previously it was reset for every read, so a teergrubing peer sending
468 single bytes within the time limit could extend the connection for a
469 long time. Credit to Qualsys Security Advisory Team for the discovery.
471 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
472 delivery address, which leaked information of the results of local
473 forwarding. Change to the original envelope recipient address, per
476 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
477 requested. Previously not bounce was generated and a log entry of
478 error ignored was made.
480 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
482 JH/32 Introduce a general tainting mechanism for values read from the input
483 channel, and values derived from them. Refuse to expand any tainted
484 values, to catch one form of exploit.
486 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
487 was unused and the unexpanded text used for the test. Found and
488 fixed by Ruben Jenster.
490 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
491 an attempt to use a TLS library read routine dereffed a nul pointer,
494 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
495 them in our smtp response.
497 JH/36 Have the general router option retry_use_local_part default to true when
498 any of the restrictive preconditions are set (to anything). Previously it
499 was only for check_local user. The change removes one item of manual
500 configuration which is required for proper retries when a remote router
501 handles a subset of addresses for a domain.
503 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
504 link count into consideration.
506 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
507 caused the extension of big_buffer, the following lines were ignored.
509 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
510 accordance with RFC 2308. Previously there was no expiry, so a longlived
511 receive process (eg. due to ACL delays) versus a short SOA value could
514 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
516 JH/39 Promote DMARC support to mainline.
518 JH/40 Bug 2452: Add a References: header to DSNs.
520 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
521 parameters. The relevant library call is documented as "Deprecated: This
522 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
523 3.6.0, DH parameters are negotiated following RFC7919."
525 HS/06 Change the default of dnssec_request_domains to "*"
527 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
528 carried on and emitted a BDAT command, even when PIPELINING was not
531 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
532 buffer was used for the filename, resulting in a trap when tainted
533 arguments (eg. $domain) were used.
535 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
536 recommended to avoid a possible server-load attack. The feature can be
537 re-enabled via the openssl_options main cofiguration option.
539 JH/45 local_scan API: documented the current smtp_printf() call. This changed
540 for version 4.90 - adding a "more data" boolean to the arguments.
541 Bumped the ABI version number also, this having been missed previously;
542 release versions 4.90 to 4.92.3 inclusive were effectively broken in
543 respect of usage of smtp_printf() by either local_scan code or libraries
544 accessed via the ${dlfunc } expansion item. Both will need coding
545 adjustment for any calls to smtp_printf() to match the new function
546 signature; a FALSE value for the new argument is always safe.
548 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
549 the file-offset (which the Linux syscall does, and exim expects); this
550 resulted in an indefinite loop.
552 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
553 to do ARC verification. The Authentication-Results: header line added
554 by the configuration then had no ARC item.
560 JH/01 Remove code calling the customisable local_scan function, unless a new
561 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
563 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
564 non-signal-safe functions being used.
566 JH/03 Bug 2269: When presented with a received message having a stupidly large
567 number of DKIM-Signature headers, disable DKIM verification to avoid
568 a resource-consumption attack. The limit is set at twenty.
570 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
571 report of oldest_pass in ${authres } in consequence, and separate out
572 some descriptions of reasons for verification fail.
574 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
575 files in the spool were present and unlocked. A queue-runner could spot
576 them, resulting in a duplicate delivery. Fix that by doing the unlock
577 after the unlink. Investigation by Tim Stewart. Take the opportunity to
578 add more error-checking on spoolfile handling while that code is being
581 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
582 No known attacks, no CVE, this is defensive hardening.
584 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
585 a queue-runner could start a delivery while other operations were ongoing.
586 Cutthrough delivery was a common victim, resulting in duplicate delivery.
587 Found and investigated by Tim Stewart. Fix by using the open message data
588 file handle rather than opening another, and not locally closing it (which
589 releases a lock) for that case, while creating the temporary .eml format
590 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
592 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
593 $sender_verify_failure/$recipient_verify_failure to "random".
595 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
598 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
599 Previously this would segfault.
601 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
604 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
605 like zero, since the resolver should be doing this for us, But we need one
606 as a CNAME but no MX presence gets the CNAME returned; we need to check
607 that doesn't point to an MX to declare it "no MX returned" rather than
608 "error, loop". A new main option is added so the older capability of
609 following some limited number of chain links is maintained.
611 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
613 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
614 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
615 now-deprecated earlier definitions used only specified the range up to TLS
616 1.2 (in the older-version library docs).
618 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
620 JH/15 Rework TLS client-side context management. Stop using a global, and
621 explicitly pass a context around. This enables future use of TLS for
622 connections to service-daemons (eg. malware scanning) while a client smtp
623 connection is using TLS; with cutthrough connections this is quite likely.
625 JH/16 Fix ARC verification to do AS checks in reverse order.
627 JH/17 Support a "tls" option on the ${readsocket } expansion item.
629 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
630 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
631 Previously the "utf8" would be re-prepended for every additional message.
633 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
634 Previously thery were accepted, resulting in issues when attempting to
635 forward messages to a non-supporting MTA.
637 PP/02 Let -n work with printing macros too, not just options.
639 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
640 one parent address was copied, and bogus data was used at delivery-logging
641 time. Either a crash (after delivery) or bogus log data could result.
642 Discovery and analysis by Tim Stewart.
644 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
645 Previously if the string ended mid-character, we did not insert the
646 promised '?' replacement.
648 PP/04 Documentation: current string operators work on bytes, not codepoints.
650 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
651 should pack well giving a smaller memory footprint so better caching and
652 therefore performance. Group the declarations where this can't be done so
653 that the byte-sized flag variables are not interspersed among pointer
654 variables, giving a better chance of good packing by the compiler.
656 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
657 non-null, to avoid issues with sites running BATV. Previously reports were
658 sent with an empty envelope sender so looked like bounces.
660 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
661 The ignore_error flag wasn't being returned from the filter subprocess so
662 was not set for later routers. Investigation and fix by Matthias Kurz.
664 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
665 and a msg:complete for the whole, when a message is manually removed using
666 -Mrm. Developement by Matthias Kurz, hacked on by JH.
668 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
669 a "Gnu special" function, asprintf() in the DB utility binary builds; I
670 hope that is portable enough.
672 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
673 requiring a known-CA anchor certificate; make it now rely entirely on the
674 TLSA as an anchor. Checking the name on the leaf cert against the name
675 on the A-record for the host is still done for TA (but not for EE mode).
677 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
678 included in delivery lines for non-proxied connections, when compiled with
679 SUPPORT_SOCKS and running with proxy logging enabled.
681 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
682 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
683 move the existing event to fire before the normal logging of message
684 failure so that custom logging is bracketed by normal logging.
686 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
687 msg:fail:internal event. Developement by Matthias Kurz.
689 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
690 far too small for todays use of crypto signatures stored there. Go all
691 the way to the max DNS message size of 64kB, even though this might be
692 overmuch for IOT constrained device use.
694 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
695 copy a string over itself. The library routine is documented as not
696 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
698 JH/32 For main options check_spool_space and check_inode_space, where the
699 platform supports 64b integers, support more than the previous 2^31 kB
700 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
701 the previous G, M, k.
703 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
704 $authenticated_fail_id variable on authentication failure. Previously
707 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
708 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
709 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
710 GNUTLS_SEC_PARAM_MEDIUM.
712 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
713 side. Previously we would continue as if no SNI had been received.
715 JH/36 Harden the handling of string-lists. When a list consisted of a sole
716 "<" character, which should be a list-separator specification, we walked
717 off past the nul-terimation.
719 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
720 causes) even when the retry time is not yet met. Previously they were
721 not, meaning that when (say) an account was over-quota and temp-rejecting,
722 and multiple senders' messages were queued, only one sender would get
723 notified on each configured delay_warning cycle.
725 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
727 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
728 cipher-suites, an error can be left on the stack even for a succeeding
729 accept; this results in impossible error messages when a later operation
732 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
733 return error codes indicating retry. Under TLS1.3 this becomes required.
735 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
736 it only wrote the new authenticators, resulting in a lack of tracking of
737 peer changes of ESMTP extensions until the next cache flush.
739 JH/41 Fix the loop reading a message header line to check for integer overflow,
740 and more-often against header_maxsize. Previously a crafted message could
741 induce a crash of the recive process; now the message is cleanly rejected.
743 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
744 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
750 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
751 When redis_servers is set to a list of > 1 element, and the Redis servers
752 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
753 case of MOVED into a DEFER case instead, thus moving the query onto the
754 next server in the list. For a cluster of N elements, all N servers must
755 be defined in redis_servers.
757 GF/02 Catch and remove uninitialized value warning in exiqsumm
758 Check for existence of @ARGV before looking at $ARGV[0]
760 JH/01 Replace the store_release() internal interface with store_newblock(),
761 which internalises the check required to safely use the old one, plus
762 the allocate and data copy operations duplicated in both (!) of the
763 extant use locations.
765 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
766 modifier. This matches the restriction on the commandline.
768 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
769 Previously only the last row was returned.
771 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
772 we assumed that tags in the header were well-formed, and parsed the
773 element content after inspecting only the first char of the tag.
774 Assumptions at that stage could crash the receive process on malformed
777 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
778 While running the DKIM ACL we operate on the Permanent memory pool so that
779 variables created with "set" persist to the DATA ACL. Also (at any time)
780 DNS lookups that fail create cache records using the Permanent pool. But
781 expansions release any allocations made on the current pool - so a dnsdb
782 lookup expansion done in the DKIM ACL releases the memory used for the
783 DNS negative-cache, and bad things result. Solution is to switch to the
784 Main pool for expansions.
785 While we're in that code, add checks on the DNS cache during store_reset,
786 active in the testsuite.
787 Problem spotted, and debugging aided, by Wolfgang Breyha.
789 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
790 When none of the hosts presented to a transport match an already-open
791 connection, close it and proceed with the list. Previously we would
792 queue the message. Spotted by Lena with Yahoo, probably involving
795 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
796 Previously a spurious "250 OK id=" response was appended to the proper
799 JH/08 The "support for" informational output now, which built with Content
800 Scanning support, has a line for the malware scanner interfaces compiled
801 in. Interface can be individually included or not at build time.
803 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
804 by the template makefile "src/EDITME". The "STREAM" support for an older
805 ClamAV interface method is removed.
807 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
808 rows affected is given instead).
810 JH/11 The runtime Berkeley DB library version is now additionally output by
811 "exim -d -bV". Previously only the compile-time version was shown.
813 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
814 SMTP connection. Previously, when one had more receipients than the
815 first, an abortive onward connection was made. Move to full support for
816 multiple onward connections in sequence, handling cutthrough connection
817 for all multi-message initiating connections.
819 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
820 routers. Previously, a multi-recipient message would fail to match the
821 onward-connection opened for the first recipient, and cause its closure.
823 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
824 a timeout on read on a GnuTLS initiating connection, resulting in the
825 initiating connection being dropped. This mattered most when the callout
826 was marked defer_ok. Fix to keep the two timeout-detection methods
829 JH/15 Relax results from ACL control request to enable cutthrough, in
830 unsupported situations, from error to silently (except under debug)
831 ignoring. This covers use with PRDR, frozen messages, queue-only and
834 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
836 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
837 metadata, resulting in a crash in free().
839 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
840 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
841 Broken also in d185889f4, with init system revamp.
843 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
844 Previously we abruptly closed the connection after reading a malware-
845 found indication; now we go on to read the "scan ok" response line,
848 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
849 ACL. Previously, a crash would result.
851 JH/19 Speed up macro lookups during configuration file read, by skipping non-
852 macro text after a replacement (previously it was only once per line) and
853 by skipping builtin macros when searching for an uppercase lead character.
855 JH/20 DANE support moved from Experimental to mainline. The Makefile control
856 for the build is renamed.
858 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
859 was allocated for every new TLS startup, meaning one per message. Fix
860 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
862 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
863 reported the original. Fix to report (as far as possible) the ACL
864 result replacing the original.
866 JH/23 Fix memory leak during multi-message connections using STARTTLS under
867 OpenSSL. Certificate information is loaded for every new TLS startup,
868 and the resources needed to be freed.
870 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
872 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
875 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
876 DATA response info to the (existing) per-recipient response info for
877 the "C=" log element. It can have useful tracking info from the
878 destination system. Patch from Simon Arlott.
880 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
881 length value. Previously this would segfault.
883 HS/02 Support Avast multiline protoocol, this allows passing flags to
884 newer versions of the scanner.
886 JH/28 Ensure that variables possibly set during message acceptance are marked
887 dead before release of memory in the daemon loop. This stops complaints
888 about them when the debug_store option is enabled. Discovered specifically
889 for sender_rate_period, but applies to a whole set of variables.
890 Do the same for the queue-runner and queue-list loops, for variables set
891 from spool message files. Do the same for the SMTP per-message loop, for
892 certain variables indirectly set in ACL operations.
894 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
895 as a multi-recipient message from a mailinglist manager). The coding had
896 an arbitrary cutoff number of characters while checking for more input;
897 enforced by writing a NUL into the buffer. This corrupted long / fast
898 input. The problem was exposed more widely when more pipelineing of SMTP
899 responses was introduced, and one Exim system was feeding another.
900 The symptom is log complaints of SMTP syntax error (NUL chars) on the
901 receiving system, and refused recipients seen by the sending system
902 (propating to people being dropped from mailing lists).
903 Discovered and pinpointed by David Carter.
905 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
906 replaced by the ${authresults } expansion.
908 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
910 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
911 allows proper process termination in container environments.
913 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
914 Previously the "final dot" had a newline after it; ensure it is CR,LF.
916 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
917 and "err_perm", deprecated since 4.83 when the RFC-defined words
918 "temperror" and "permerror" were introduced.
920 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
921 transport-filters or DKIM-signing. The restriction was lost in the
922 consolidation of verify-callout and delivery SMTP handling.
923 Extend the restriction to also cover ARC-signing.
925 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
926 in defer=pass mode supply a 450 to the initiator. Previously the message
929 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
930 tls_require_ciphers is used as before.
932 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
933 "pass_unscanned". Only tmpfails from the scanner are written to
934 the paniclog, as they may require admin intervention (permission
935 denied, license issues). Other scanner errors (like decompression
936 bombs) do not cause a paniclog entry.
938 JH/36 Fix reinitialisation of DKIM logging variable between messages.
939 Previously it was possible to log spurious information in receive log
942 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
943 triggered odd behaviour from Outlook Express clients.
945 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
948 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
949 since the IETF WG has not yet settled on that versus the original
950 "bare" representation.
952 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
953 Previously the millisecond value corrupted the output.
954 Fix also for syslog_pid=no and log_selector +pid, for which the pid
955 corrupted the output.
961 JH/01 Rework error string handling in TLS interface so that the caller in
962 more cases is responsible for logging. This permits library-sourced
963 string to be attached to addresses during delivery, and collapses
964 pairs of long lines into single ones.
966 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
967 during configuration. Wildcards are allowed and expanded.
969 JH/02 Rework error string handling in DKIM to pass more info back to callers.
970 This permits better logging.
972 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
973 do not close it down and have the child transport start it up again on
974 the passed-on TCP connection. Instead, proxy the child (and any
975 subsequent ones) for TLS via a unix-domain socket channel. Logging is
976 affected: the continued delivery log lines do not have any DNSSEC, TLS
977 Certificate or OCSP information. TLS cipher information is still logged.
979 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
980 identical IP addresses on different listening ports. Will also affect
983 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
984 add noisy ifdef guards to special-case this sillyness.
985 Patch from Bernd Kuhls.
987 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
988 than 255 are no longer allowed.
990 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
991 Disable the session-cache too, which might reduce our load. Since we
992 currrectly use a new context for every connection, both as server and
993 client, there is no benefit for these.
994 GnuTLS appears to not support tickets server-side by default (we don't
995 call gnutls_session_ticket_enable_server()) but client side is enabled
996 by default on recent versions (3.1.3 +) unless the PFS priority string
999 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1000 <https://reproducible-builds.org/specs/source-date-epoch/>.
1002 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1003 the check for any unsuccessful recipients did not notice the limit, and
1004 erroneously found still-pending ones.
1006 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1007 MSG_MORE. Only in-clear (not on TLS connections).
1009 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1010 transport-filter is involved we can buffer the headers in memory for
1011 creating the signature, and read the spool data file once for the
1012 signature and again for transmission.
1014 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1015 4.77 as the kernel support then wasn't solid, having issues in 64bit
1016 mode. Now, it's been long enough. Add support for FreeBSD also.
1018 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1019 case where the routing stage had gathered several addresses to send to
1020 a host before calling the transport for the first, we previously failed
1021 to close down TLS in the old transport process before passing the TCP
1022 connection to the new process. The new one sent a STARTTLS command
1023 which naturally failed, giving a failed delivery and bloating the retry
1024 database. Investigation and fix prototype from Wolfgang Breyha.
1026 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1027 false-negatives in the check that the sender had not preempted a response
1028 or prompt from Exim (running as a server), due to that code's lack of
1029 awareness of the SMTP input buffering.
1031 PP/04 Add commandline_checks_require_admin option.
1032 Exim drops privileges sanely, various checks such as -be aren't a
1033 security problem, as long as you trust local users with access to their
1034 own account. When invoked by services which pass untrusted data to
1035 Exim, this might be an issue. Set this option in main configuration
1036 AND make fixes to the calling application, such as using `--` to stop
1039 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1040 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1041 responses to those, into a single TLS record each way (this usually means
1042 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1045 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1046 affects you only if you're dancing at the edge of the param size limits.
1047 If you are, and this message makes sense to you, then: raise the
1048 configured limit or use OpenSSL 1.1. Nothing we can do for older
1051 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1052 cmdline element to get the documented default one. Previously it was
1055 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1056 are unsuable for non-message-specific reasons - eg. connection timeout,
1057 banner-time rejection.
1059 JH/16 Fix logging of delivery remote port, when specified by router, under
1062 PP/06 Repair manualroute's ability to take options in any order, even if one
1063 is the name of a transport.
1066 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1068 JH/17 Change the list-building routines interface to use the expanding-string
1069 triplet model, for better allocation and copying behaviour.
1071 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1072 Previously it was constructed the first time a possibly-matching string
1073 was met in the configuration file input during startup; now it is done
1076 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1077 compatible one, to avoid the (poorly documented) possibility of a config
1078 file in the working directory redirecting the DB files, possibly correpting
1079 some existing file. CVE-2017-10140 assigned for BDB.
1081 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1082 cache-hot. Previously, although the result was properly cached, the
1083 initial verify call returned a defer.
1085 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1086 the main verify for receipient in uncached-mode.
1088 JH/22 Retire historical build files to an "unsupported" subdir. These are
1089 defined as "ones for which we have no current evidence of testing".
1091 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1092 if present. Previously it was ignored.
1094 JH/24 Start using specified-initialisers in C structure init coding. This is
1095 a C99 feature (it's 2017, so now considered safe).
1097 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1098 if was a fixed-sized field and bitmask ops via macros; it is now more
1101 PP/07 GitHub PR 56: Apply MariaDB build fix.
1102 Patch provided by Jaroslav Škarvada.
1104 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1105 during Coverity cleanups [4.87 JH/47]
1106 Diagnosis and fix provided by Michael Fischer v. Mollard.
1108 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1109 the right size to place the terminating semicolon on its own folded
1110 line, the header hash was calculated to an incorrect value thanks to
1111 the (relaxed) space the fold became.
1113 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1114 and confused the parent.
1116 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1117 which could crash as a result. This could lead to undeliverable messages.
1119 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1122 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1123 matters on fast-turnover and PID-randomising systems, which were getting
1124 out-of-order delivery.
1126 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1127 a possibly-overlapping copy. The symptom was that "Remote host closed
1128 connection in response to HELO" was logged instead of the actual 4xx
1131 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1132 Previously only that bufferd was discarded, resulting in SYMTP command
1135 JH/32 DKIM: when a message has multiple signatures matching an identity given
1136 in dkim_verify_signers, run the dkim acl once for each. Previously only
1137 one run was done. Bug 2189.
1139 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1140 "panic the current process" to "deliberately defer". The panic log is
1141 still written with the problem list name; the mail and reject logs now
1142 get a temp-reject line for the message that was being handled, saying
1143 something like "domains check lookup or other defer". The SMTP 451
1144 message is still "Temporary local problem".
1146 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1147 A crafted sequence of BDAT commands could result in in-use memory beeing
1148 freed. CVE-2017-16943.
1150 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1151 from SMTP input. Previously it was always done; now only done for DATA
1152 and not BDAT commands. CVE-2017-16944.
1154 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1155 to the message (such as an overlong header line). Previously this was
1156 not done and we did not exit BDAT mode. Followon from the previous item
1157 though a different problem.
1163 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1164 than -2003 did; needs libidn2 in addition to libidn.
1166 JH/02 The path option on a pipe transport is now expanded before use.
1168 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1169 Patch provided by "Björn", documentation fix added too.
1171 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1172 missing a wire-to-host endian conversion.
1174 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1175 close after a BDAT command line could be taken as a following command,
1176 giving a synch failure. Fix by only checking for synch immediately
1177 before acknowledging the chunk.
1179 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1180 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1181 macro. Patches provided by Josh Soref.
1183 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1184 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1185 by IANA as a service extension. However, John Klensin suggests that we
1188 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1189 the dkim code may be unix-mode line endings rather than smtp wire-format
1190 CRLF, so prepend a CR to any bare LF.
1192 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1193 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1195 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1196 the first were themselves being wrongly included in the feed into dkim
1197 processing; with most chunk sizes in use this resulted in an incorrect
1198 body hash calculated value.
1200 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1201 DKIM signature block, for verification. Although advised against by
1202 standards it is specifically not ruled illegal.
1204 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1206 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1207 missing a body hash (the bh= tag).
1209 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1210 It seems that HAProxy sends the Proxy Protocol information in clear and
1211 only then does a TLS startup, so do the same.
1213 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1214 TCP connections (such as for Spamd) unless the daemon successfully set
1215 Fast Open mode on its listening sockets. This fixes breakage seen on
1216 too-old kernels or those not configured for Fast Open, at the cost of
1217 requiring both directions being enabled for TFO, and TFO never being used
1218 by non-daemon-related Exim processes.
1220 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1221 endings, at least on the first header line. Try to canonify any that get
1222 past that check, despite the cost.
1224 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1225 now limited to an arbitrary five deep, while parsing addresses with the
1226 strip_excess_angle_brackets option enabled.
1228 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1229 instead leave the unprompted TLS handshake in socket buffer for the
1230 TLS library to consume.
1232 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1234 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1236 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1237 region in one operation, for speed, and this leaves assigned pointers
1238 dangling. Add checks run only under the testsuite which checks all
1239 variables at a store-reset and panics on a dangling pointer; add code
1240 explicitly nulling out all the variables discovered. Fixes one known
1241 bug: a transport crash, where a dangling pointer for $sending_ip_address
1242 originally assigned in a verify callout, is re-used.
1244 PP/06 Drop '.' from @INC in various Perl scripts.
1246 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1248 PP/08 Reduce a number of compilation warnings under clang; building with
1249 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1250 should be warning-free.
1252 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1254 HS/01 Fix portability problems introduced by PP/08 for platforms where
1255 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1257 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1258 chunk. This allows us to accept broken chunked messages. We need a more
1259 general solution here.
1261 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1262 already-broken messages in the queue.
1264 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1266 JH/19 Fix reference counting bug in routing-generated-address tracking.
1272 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1273 supports it and a size is available (ie. the sending peer gave us one).
1275 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1276 years of being deprecated). The replacements are the ACLs
1277 acl_smtp_mime and acl_not_smtp_mime.
1279 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1280 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1281 or even an in-clear connection were permitted. Now, if the host lookup
1282 was dnssec and dane was requested then the host is only used if the
1283 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1284 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1285 if one fails this test.
1286 This means that a poorly-configured remote DNS will make it incommunicado;
1287 but it protects against a DNS-interception attack on it.
1289 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1290 non-noisy when a race steals the message being considered.
1292 JH/05 If main configuration option tls_certificate is unset, generate a
1293 self-signed certificate for inbound TLS connections.
1295 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1296 in rewrites and routers.
1298 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1299 and logged a warning sing 4.83; now they are a configuration file error.
1301 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1302 (lacking @domain). Apply the same qualification processing as RCPT.
1304 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1306 JH/10 Support ${sha256:} applied to a string (as well as the previous
1309 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1310 a cutthrough deliver is pending, as we always want to make a connection.
1311 This also avoids re-routing the message when later placing the cutthrough
1312 connection after a verify cache hit.
1313 Do not update it with the verify result either.
1315 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1316 when routing results in more than one destination address.
1318 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1319 signing (which inhibits the cutthrough capability). Previously only
1320 the presence of an option was tested; now an expansion evaluating as
1321 empty is permissible (obviously it should depend only on data available
1322 when the cutthrough connection is made).
1324 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1325 the relevant preceding SMTP command did not note the pipelining mode.
1327 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1328 Previously they were not counted.
1330 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1331 as one having no matching records. Previously we deferred the message
1332 that needed the lookup.
1334 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1335 distinguished as "(=".
1337 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1338 for missing MX records. Previously it only worked for missing A records.
1340 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1342 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1343 after the data-go-ahead and data-ack. Patch from Jason Betts.
1345 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1346 even for a "none" policy. Patch from Tony Meyer.
1348 JH/22 Fix continued use of a connection for further deliveries. If a port was
1349 specified by a router, it must also match for the delivery to be
1352 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1353 When one of the recipients of a message was unsuitable for the connection
1354 (has no matching addresses), we lost track of needing to mark it
1355 deferred. As a result mail would be lost.
1357 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1359 JH/25 Decoding ACL controls is now done using a binary search; the source code
1360 takes up less space and should be simpler to maintain. Merge the ACL
1361 condition decode tables also, with similar effect.
1363 JH/26 Fix problem with one_time used on a redirect router which returned the
1364 parent address unchanged. A retry would see the parent address marked as
1365 delivered, so not attempt the (identical) child. As a result mail would
1368 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1369 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1370 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1373 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1374 Main config options check_{log,spool}_{inodes,space} are now
1375 100 inodes, 10MB unless set otherwise in the configuration.
1377 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1378 Previously it only applied to the main-section connection policy
1381 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1383 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1384 by me. Added RFC7919 DH primes as an alternative.
1386 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1387 are not in the system include path.
1389 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1390 GnuTLS, when a session startup failed (eg because the client disconnected)
1391 Exim did stdio operations after fclose. This was exposed by a recent
1392 change which nulled out the file handle after the fclose.
1394 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1395 signed directly by the cert-signing cert, rather than an intermediate
1396 OCSP-signing cert. This is the model used by LetsEncrypt.
1398 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1400 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1401 an incoming connection.
1403 HS/02 Bug 1802: Do not half-close the connection after sending a request
1406 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1407 fallback to "prime256v1".
1409 JH/34 SECURITY: Use proper copy of DATA command in error message.
1410 Could leak key material. Remotely exploitable. CVE-2016-9963.
1416 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1417 and 3.4.4 - once the server is enabled to respond to an OCSP request
1418 it does even when not requested, resulting in a stapling non-aware
1419 client dropping the TLS connection.
1421 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1422 support variable-length bit vectors. No functional change.
1424 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1425 The I= interface field on outgoing lines is now after the H= remote
1426 host field, same as incoming lines. There is a separate
1427 outgoing_interface log selector which allows you to disable the
1430 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1431 If not running log_selector +smtp_connection the mainlog would be held
1432 open indefinitely after a "too many connections" event, including to a
1433 deleted file after a log rotate. Leave the per net connection logging
1434 leaving it open for efficiency as that will be quickly detected by the
1435 check on the next write.
1437 HS/01 Bug 1671: Fix post transport crash.
1438 Processing the wait-<transport> messages could crash the delivery
1439 process if the message IDs didn't exist for some reason. When
1440 using 'split_spool_directory=yes' the construction of the spool
1441 file name failed already, exposing the same netto behaviour.
1443 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1444 mime_regex ACL conditions.
1446 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1447 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1448 to HELO, local diagnostic string.
1450 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1451 log line to debug. Even when configured with a tls authenticator many
1452 client connections are expected to not authenticate in this way, so
1453 an authenticate fail is not an error.
1455 HS/02 Add the Exim version string to the process info. This way exiwhat
1456 gives some more detail about the running daemon.
1458 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1459 matter for fast-change records such as DNSBLs.
1461 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1462 retry record, even if constant. There may be multiple transports with
1463 different interface settings and the retry behaviour needs to be kept
1466 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1468 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1470 JH/10 Bug 840: fix log_defer_output option of pipe transport
1472 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1473 hosts_randomize. This matters a lot when combined with mua_wrapper.
1475 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1476 ${quote_pgsql:<string>} operator.
1478 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1480 JH/14 Change header file name for freeradius-client. Relevant if compiling
1481 with Radius support; from the Gentoo tree and checked under Fedora.
1483 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1485 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1486 option transport_filter, meaning no filtering.
1488 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1490 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1491 defaults to "*" (all hosts). The variable is now available when not built
1492 with TLS, default unset, mainly to enable keeping the testsuite sane.
1493 If a server certificate is not supplied (via tls_certificate) an error is
1494 logged, and clients will find TLS connections fail on startup. Presumably
1495 they will retry in-clear.
1496 Packagers of Exim are strongly encouraged to create a server certificate
1497 at installation time.
1499 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1500 with the $config_file variable.
1502 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1503 in transport context, after the attempt, and per-recipient. The latter type
1504 is per host attempted. The event data is the error message, and the errno
1505 information encodes the lookup type (A vs. MX) used for the (first) host,
1506 and the trailing two digits of the smtp 4xx response.
1508 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1509 to write to mainlog (or rejectlog, paniclog) in the window between file
1510 creation and permissions/ownership being changed. Particularly affects
1511 installations where exicyclog is run as root, rather than exim user;
1512 result is that the running daemon panics and dies.
1514 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1516 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1517 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1518 "pri" and "weight". Note that the previous implicit priority given by the
1519 list order is no longer honoured.
1521 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1522 for DKIM processing.
1524 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1525 by defining SUPPORT_SOCKS.
1527 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1528 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1529 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1530 variables are renamed to proxy_{local,external}_{address,port}.
1532 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1533 for a build by defining SUPPORT_I18N
1535 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1536 of the query string, and make ${quote_redis:} do that quoting.
1538 JH/29 Move Events support from Experimental to mainline, enabled by default
1539 and removable for a build by defining DISABLE_EVENT.
1541 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1543 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1544 cached by the daemon.
1546 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1547 by defining LOOKUP_REDIS. The libhiredis library is required.
1549 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1550 keys are given for lookup.
1552 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1553 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1554 only supported when built with TLS support. The PolarSSL SHA routines
1555 are still used when the TLS library is too old for convenient support.
1557 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1558 openssl_options), for security. OpenSSL forces this from version 1.1.0
1559 server-side so match that on older versions.
1561 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1562 allocation for $value could be released as the expansion processing
1563 concluded, but leaving the global pointer active for it.
1565 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1566 and to use the domains and local_parts ACL conditions.
1568 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1569 incorrectly not doubled on cutthrough transmission, hence seen as a
1570 body-termination at the receiving system - resulting in truncated mails.
1571 Commonly the sender saw a TCP-level error, and retransmitted the message
1572 via the normal store-and-forward channel. This could result in duplicates
1573 received - but deduplicating mailstores were liable to retain only the
1574 initial truncated version.
1576 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1578 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1580 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1581 we're in there, support oversigning also; bug 1309.
1583 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1585 HS/04 Add support for keep_environment and add_environment options.
1587 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1588 either intentional arithmetic overflow during PRNG, or testing config-
1591 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1592 delivery resulted in actual delivery. Cancel cutthrough before DATA
1595 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1596 rejects a recipient: pass the reject to the originator.
1598 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1599 Many were false-positives and ignorable, but it's worth fixing the
1602 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1603 for the new environment-manipulation done at startup. Move the routines
1604 from being local to tls.c to being global via the os.c file.
1606 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1607 an extract embedded as result-arg for a map, the first arg for extract
1608 is unavailable so we cannot tell if this is a numbered or keyed
1609 extraction. Accept either.
1615 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1618 JH/02 The smtp transport option "multi_domain" is now expanded.
1620 JH/03 The smtp transport now requests PRDR by default, if the server offers
1623 JH/04 Certificate name checking on server certificates, when exim is a client,
1624 is now done by default. The transport option tls_verify_cert_hostnames
1625 can be used to disable this per-host. The build option
1626 EXPERIMENTAL_CERTNAMES is withdrawn.
1628 JH/05 The value of the tls_verify_certificates smtp transport and main options
1629 default to the word "system" to access the system default CA bundle.
1630 For GnuTLS, only version 3.0.20 or later.
1632 JH/06 Verification of the server certificate for a TLS connection is now tried
1633 (but not required) by default. The verification status is now logged by
1634 default, for both outbound TLS and client-certificate supplying inbound
1637 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1640 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1641 Status Notification (bounce) messages are now MIME format per RFC 3464.
1642 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1643 under the control of the dsn_advertise_hosts option, and routers may
1644 have a dsn_lasthop option.
1646 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1647 default, modifiable by a malware= option. The list separator for
1648 the options can now be changed in the usual way. Bug 68.
1650 JH/10 The smtp_receive_timeout main option is now expanded before use.
1652 JH/11 The incoming_interface log option now also enables logging of the
1653 local interface on delivery outgoing connections.
1655 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1656 if the interface and destination host and port all match.
1658 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1661 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1662 Patch from Andrew Lewis.
1664 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1665 now supports optional time-restrictions, weighting, and priority
1666 modifiers per server. Patch originally by <rommer@active.by>.
1668 JH/16 The spamd_address main option now supports a mixed list of local
1669 and remote servers. Remote servers can be IPv6 addresses, and
1670 specify a port-range.
1672 JH/17 Bug 68: The spamd_address main option now supports an optional
1673 timeout value per server.
1675 JH/18 Bug 1581: Router and transport options headers_add/remove can
1676 now have the list separator specified.
1678 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1681 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1684 JH/21 Support for the A6 type of dns record is withdrawn.
1686 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1687 rather than the verbs used.
1689 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1690 from 255 to 1024 chars.
1692 JH/24 Verification callouts now attempt to use TLS by default.
1694 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1695 are generic router options now. The defaults didn't change.
1697 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1698 Original patch from Alexander Shikoff, worked over by JH.
1700 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1701 files if ZCAT_COMMAND is not executable.
1703 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1705 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1707 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1708 Normally benign, it bites when the pair was led to by a CNAME;
1709 modern usage is to not canonicalize the domain to a CNAME target
1710 (and we were inconsistent anyway for A-only vs AAAA+A).
1712 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1714 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1715 when evaluating $sender_host_dnssec.
1717 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1718 $sender_helo_dnssec variable.
1720 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1722 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1724 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1726 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1727 documented as working, but never had. Support all but $spam_report.
1729 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1730 added for tls authenticator.
1732 HS/03 Add perl_taintmode main config option
1738 TL/01 When running the test suite, the README says that variables such as
1739 no_msglog_check are global and can be placed anywhere in a specific
1740 test's script, however it was observed that placement needed to be near
1741 the beginning for it to behave that way. Changed the runtest perl
1742 script to read through the entire script once to detect and set these
1743 variables, reset to the beginning of the script, and then run through
1744 the script parsing/test process like normal.
1746 TL/02 The BSD's have an arc4random API. One of the functions to induce
1747 adding randomness was arc4random_stir(), but it has been removed in
1748 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1749 function when detected.
1751 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1752 cause callback expansion.
1754 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1755 syntax errors in an expansion can be treated as a string instead of
1756 logging or causing an error, due to the internal use of bool_lax
1757 instead of bool when processing it.
1759 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1760 server certificates when making smtp deliveries.
1762 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1764 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1766 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1768 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1769 Merged patch from Sebastian Wiedenroth.
1771 JH/05 Fix results-pipe from transport process. Several recipients, combined
1772 with certificate use, exposed issues where response data items split
1773 over buffer boundaries were not parsed properly. This eventually
1774 resulted in duplicates being sent. This issue only became common enough
1775 to notice due to the introduction of connection certificate information,
1776 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1778 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1779 size buffer was used, resulting in syntax errors when an expansion
1782 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1783 version 3.3.6 or later.
1785 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1786 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1787 both become "event_action", the variables become $event_name, $event_data
1788 and $event_defer_errno. There is a new variable $verify_mode, usable in
1789 routers, transports and related events. The tls:cert event is now also
1790 raised for inbound connections, if the main configuration event_action
1793 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1794 early OCSP support, but no stapling (appears to be less than 1.0.0).
1796 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1797 server certificate names available under the smtp transport option
1798 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1801 JH/10 Time-related extraction expansions from certificates now use the main
1802 option "timezone" setting for output formatting, and are consistent
1803 between OpenSSL and GnuTLS compilations. Bug 1541.
1805 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1806 encoded parameter in the incoming message. Bug 1558.
1808 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1809 include certificate info, eximon was claiming there were spoolfile
1812 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1814 JH/14 Log delivery-related information more consistently, using the sequence
1815 "H=<name> [<ip>]" wherever possible.
1817 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1818 are problematic for Debian distribution, omit them from the release
1821 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1823 JH/16 Fix string representation of time values on 64bit time_t architectures.
1826 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1827 output list separator was used.
1832 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1833 checkers that were complaining about end of non-void function with no
1836 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1837 This was a regression introduced in 4.83 by another bugfix.
1839 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1841 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1842 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1848 TF/01 Correctly close the server side of TLS when forking for delivery.
1850 When a message was received over SMTP with TLS, Exim failed to clear up
1851 the incoming connection properly after forking off the child process to
1852 deliver the message. In some situations the subsequent outgoing
1853 delivery connection happened to have the same fd number as the incoming
1854 connection previously had. Exim would try to use TLS and fail, logging
1855 a "Bad file descriptor" error.
1857 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1858 utilities have not been installed.
1860 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1861 temporary space as the ACL may create new global variables.
1863 TL/01 LDAP support uses per connection or global context settings, depending
1864 upon the detected version of the libraries at build time.
1866 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1867 to extract and use the src ip:port in logging and expansions as if it
1868 were a direct connection from the outside internet. PPv2 support was
1869 updated based on HAProxy spec change in May 2014.
1871 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1873 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1874 Properly escape header and check for NULL return.
1876 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1879 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1881 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1882 characters in header names, implemented as a verify condition.
1883 Contributed by Michael Fischer v. Mollard.
1885 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1886 results permerror and temperror. Previous values are deprecated but
1887 still accepted. In a future release, err_perm and err_temp will be
1888 completely removed, which will be a backward incompatibility if the
1889 ACL tests for either of these two old results. Patch contributed by
1890 user bes-internal on the mailing list.
1892 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1894 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1895 selectors, in both main and reject logs.
1897 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1900 JH/07 Add malware type "sock" for talking to simple daemon.
1902 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1904 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1905 routers/transports under cutthrough routing.
1907 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1908 numbers. Touch up "bool" conditional to keep the same definition.
1910 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1912 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1914 JH/12 Expand items in router/transport headers_add or headers_remove lists
1915 individually rather than the list as a whole. Bug 1452.
1917 Required for reasonable handling of multiple headers_ options when
1918 they may be empty; requires that headers_remove items with embedded
1919 colons must have them doubled (or the list-separator changed).
1921 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1922 view the policy declared in the DMARC record. Currently, $dmarc_status
1923 is a combined value of both the record presence and the result of the
1926 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1928 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1929 dnslookup router and the smtp transport (applying to the forward
1932 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1933 of ldap servers used for a specific lookup. Patch provided by Heiko
1936 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1937 New variable $lookup_dnssec_authenticated for observability.
1939 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1940 Patch submitted by Lars Timman.
1942 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1944 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1945 Requires trusted mode and valid format message id, aborts otherwise.
1946 Patch contributed by Heiko Schlichting.
1948 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1949 certextract with support for various fields. Bug 1358.
1951 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1952 is requested by default, modifiable by smtp transport option
1955 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1956 operate on certificate variables to give certificate fingerprints
1957 Also new ${sha256:cert_variable}.
1959 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1961 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1962 Christian Aistleitner.
1964 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1966 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1967 file. Patch from Wolfgang Breyha.
1969 JH/25 Expand the coverage of the delivery $host and $host_address to
1970 client authenticators run in verify callout. Bug 1476.
1972 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1973 align with daemon_smtp_ports. Bug 72.
1975 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1976 support and error reporting did not work properly.
1978 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1979 and is readable. Patch from Andrew Colin Kissa.
1981 TL/14 Enhance documentation of ${run expansion and how it parses the
1982 commandline after expansion, particularly in the case when an
1983 unquoted variable expansion results in an empty value.
1985 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1987 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1988 by renaming away from it. Addresses GH issue 15, reported by
1991 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1993 TL/15 SECURITY: prevent double expansion in math comparison functions
1994 (can expand unsanitized data). Not remotely exploitable.
2001 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2003 PP/02 Make -n do something, by making it not do something.
2004 When combined with -bP, the name of an option is not output.
2006 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2009 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2010 $sender_host_name and config options to manage this, and basic check
2013 PP/05 DSCP support for outbound connections and control modifier for inbound.
2015 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2016 (Only plugin which currently uses this is kerberos4, which nobody should
2017 be using, but we should make it available and other future plugins might
2018 conceivably use it, even though it would break NAT; stuff *should* be
2019 using channel bindings instead).
2021 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2022 name; added for Sendmail compatibility; requires admin caller.
2023 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2024 just ignore it); requires trusted caller.
2025 Also parse but ignore: -Ac -Am -X<logfile>
2028 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2030 TL/02 Add +smtp_confirmation as a default logging option.
2032 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2033 Patch by Magnus Holmgren from 2007-02-20.
2035 TL/04 Bugzilla 1281 - Spec typo.
2036 Bugzilla 1283 - Spec typo.
2037 Bugzilla 1290 - Spec grammar fixes.
2039 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2041 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2043 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2044 mailing list by by Dmitry Isaikin.
2046 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2048 JH/02 Support "G" suffix to numbers in ${if comparisons.
2050 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2052 NM/01 Bugzilla 1197 - Spec typo
2053 Bugzilla 1196 - Spec examples corrections
2055 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2057 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2058 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2061 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2062 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2064 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2065 "acl {{name}{arg}...}", and optional args on acl condition
2068 JH/05 Permit multiple router/transport headers_add/remove lines.
2070 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2072 JH/07 Avoid using a waiting database for a single-message-only transport.
2073 Performance patch from Paul Fisher. Bugzilla 1262.
2075 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2078 JH/09 Add $headers_added variable, with content from use of ACL modifier
2079 add_header (but not yet added to the message). Bugzilla 199.
2081 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2082 Pulled from Bugzilla 817 by Wolfgang Breyha.
2084 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2086 (nb: this is the same fix as in Exim 4.80.1)
2088 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2091 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2094 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2096 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2097 advertise SMTP AUTH mechanism to us, instead of a generic
2098 protocol violation error. Also, make Exim more robust to bad
2099 data from the Dovecot auth socket.
2101 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2103 When a queue runner is handling a message, Exim first routes the
2104 recipient addresses, during which it prunes them based on the retry
2105 hints database. After that it attempts to deliver the message to
2106 any remaining recipients. It then updates the hints database using
2109 So if a recipient address works intermittently, it can get repeatedly
2110 deferred at routing time. The retry hints record remains fresh so the
2111 address never reaches the final cutoff time.
2113 This is a fairly common occurrence when a user is bumping up against
2114 their storage quota. Exim had some logic in its local delivery code
2115 to deal with this. However it did not apply to per-recipient defers
2116 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2118 This change adds a proper retry rule check during routing so that the
2119 final cutoff time is checked against the message's age. We only do
2120 this check if there is an address retry record and there is not a
2121 domain retry record; this implies that previous attempts to handle
2122 the address had the retry_use_local_parts option turned on. We use
2123 this as an approximation for the destination being like a local
2124 delivery, as in LMTP.
2126 I suspect this new check makes the old local delivery cutoff check
2127 redundant, but I have not verified this so I left the code in place.
2129 TF/02 Correct gecos expansion when From: is a prefix of the username.
2131 Test 0254 submits a message to Exim with the header
2135 When I ran the test suite under the user fanf2, Exim expanded
2136 the header to contain my full name, whereas it should have added
2137 a Resent-Sender: header. It erroneously treats any prefix of the
2138 username as equal to the username.
2140 This change corrects that bug.
2142 GF/01 DCC debug and logging tidyup
2143 Error conditions log to paniclog rather than rejectlog.
2144 Debug lines prefixed by "DCC: " to remove any ambiguity.
2146 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2148 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2149 Bug spotted by Jeremy Harris; was flawed since initial commit.
2150 Would have resulted in OCSP responses post-SNI triggering an Exim
2151 NULL dereference and crash.
2153 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2155 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2156 Bug detection, analysis and fix by Samuel Thibault.
2157 Bugzilla 1331, Debian bug #698092.
2159 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2161 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2162 Server implementation by Todd Lyons, client by JH.
2163 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2164 config variable "prdr_enable" controls whether the server
2165 advertises the facility. If the client requests PRDR a new
2166 acl_data_smtp_prdr ACL is called once for each recipient, after
2167 the body content is received and before the acl_smtp_data ACL.
2168 The client is controlled by both of: a hosts_try_prdr option
2169 on the smtp transport, and the server advertisement.
2170 Default client logging of deliveries and rejections involving
2171 PRDR are flagged with the string "PRDR".
2173 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2174 fclose(). Diagnosis by Todd Lyons.
2176 PP/17 Update configure.default to handle IPv6 localhost better.
2177 Patch by Alain Williams (plus minor tweaks).
2180 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2181 This is now consistent with GnuTLS, and is now documented: the
2182 previous undocumented portable approach to treating the option as
2183 unset was to force an expansion failure. That still works, and
2184 an empty string is now equivalent.
2186 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2187 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2188 not performing validation itself.
2190 PP/20 Added force_command boolean option to pipe transport.
2191 Patch from Nick Koston, of cPanel Inc.
2193 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2196 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2198 PP/21 Fix eximon continuous updating with timestamped log-files.
2199 Broken in a format-string cleanup in 4.80, missed when I repaired the
2200 other false fix of the same issue.
2201 Report and fix from Heiko Schlichting.
2204 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2205 Report from Prashanth Katuri.
2207 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2208 It's SecureTransport, so affects any MacOS clients which use the
2209 system-integrated TLS libraries, including email clients.
2211 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2212 using a MIME ACL for non-SMTP local injection.
2213 Report and assistance in diagnosis by Warren Baker.
2215 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2217 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2219 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2220 last id that failed so it may be referenced in subsequent ACL's.
2222 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2225 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2226 ldap library initialization, allowing self-signed CA's to be
2227 used. Also properly sets require_cert option later in code by
2228 using NULL (global ldap config) instead of ldap handle (per
2229 session). Bug diagnosis and testing by alxgomz.
2231 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2232 the src/util/ subdirectory.
2234 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2235 renamed to Transport Post Delivery Action by Jeremy Harris, as
2238 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2239 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2240 redis_servers = needs to be configured which will be used by the redis
2241 lookup. Patch from Warren Baker, of The Packet Hub.
2243 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2245 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2246 hostname or reverse DNS when processing a host list. Used suggestions
2247 from multiple comments on this bug.
2249 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2251 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2252 Missed a few lines, added it to make the runtest require no keyboard
2255 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2256 contains upper case chars. Make router use caseful_local_part.
2258 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2259 support when GnuTLS has been built with p11-kit.
2265 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2267 This, or similar/improved, will also be change PP/11 of 4.82.
2273 PP/01 Handle short writes when writing local log-files.
2274 In practice, only affects FreeBSD (8 onwards).
2275 Bugzilla 1053, with thanks to Dmitry Isaikin.
2277 NM/01 Bugzilla 949 - Documentation tweak
2279 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2282 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2284 PP/02 Implemented gsasl authenticator.
2286 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2288 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2289 `pkg-config foo` for cflags/libs.
2291 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2292 with rest of GSASL and with heimdal_gssapi.
2294 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2295 `pkg-config foo` for cflags/libs for the TLS implementation.
2297 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2298 properties get this fed in as external SSF. A number of robustness
2299 and debugging improvements to the cyrus_sasl authenticator.
2301 PP/08 cyrus_sasl server now expands the server_realm option.
2303 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2304 Patch by Jeremy Harris.
2306 PP/10 Added dbmjz lookup type.
2308 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2310 PP/12 MAIL args handles TAB as well as SP, for better interop with
2311 non-compliant senders.
2312 Analysis and variant patch by Todd Lyons.
2314 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2315 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2316 Patch from Dirk Mueller <dmueller@suse.com>
2318 PP/13 tls_peerdn now print-escaped for spool files.
2319 Observed some $tls_peerdn in wild which contained \n, which resulted
2320 in spool file corruption.
2322 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2323 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2324 or write after TLS renegotiation, which otherwise led to messages
2327 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2328 as a tracking header (ie: a signed header comes before the signature).
2329 Patch from Wolfgang Breyha.
2331 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2332 comma-sep list; embedded commas doubled.
2334 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2336 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2338 Report and patch from Dmitry Banschikov.
2340 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2341 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2342 failures. We appear to now support TLS1.1+ with Exim.
2344 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2345 lets Exim select keys and certificates based upon TLS SNI from client.
2346 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2347 before an outbound SMTP session. New log_selector, +tls_sni.
2349 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2350 NULL dereference. Report and patch from Alun Jones.
2352 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2353 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2354 Not seeing resolver debug output on NetBSD, but suspect this is a
2355 resolver implementation change.
2357 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2358 Left warnings. Added "eximon gdb" invocation mode.
2360 PP/21 Defaulting "accept_8bitmime" to true, not false.
2362 PP/22 Added -bw for inetd wait mode support.
2364 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2365 locate the relevant includes and libraries. Made this the default.
2367 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2368 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2370 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2371 This may cause build issues on older platforms.
2373 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2374 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2375 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2376 Added SNI support via GnuTLS too.
2377 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2379 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2381 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2382 Applied second patch from Janne, implementing suggestion to default
2383 multiple-strings-in-record handling to match SPF spec.
2385 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2387 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2388 read-only, out of scope).
2389 Patch from Wolfgang Breyha, report from Stuart Northfield.
2391 PP/29 Fix three issues highlighted by clang analyser static analysis.
2392 Only crash-plausible issue would require the Cambridge-specific
2393 iplookup router and a misconfiguration.
2394 Report from Marcin Mirosław.
2396 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2398 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2399 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2400 As part of this, removing so much warning spew let me fix some minor
2401 real issues in debug logging.
2403 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2404 assignment on my part. Fixed.
2406 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2407 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2408 Janne Snabb (who went above and beyond: thank you).
2410 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2411 string otherwise requires a connection and a bunch more work and it's
2412 relatively easy to get wrong. Should also expose TLS library linkage
2415 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2416 64-bit ${eval} (JH/03).
2418 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2419 GNU libc to support some of the 64-bit stuff, should not lead to
2420 conflicts. Defined before os.h is pulled in, so if a given platform
2421 needs to override this, it can.
2423 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2424 protection layer was required, which is not implemented.
2425 Bugzilla 1254, patch from Wolfgang Breyha.
2427 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2428 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2429 tls_dhparam take prime identifiers. Also unbreak combination of
2430 OpenSSL+DH_params+TLSSNI.
2432 PP/39 Disable SSLv2 by default in OpenSSL support.
2438 PP/01 Solaris build fix for Oracle's LDAP libraries.
2439 Bugzilla 1109, patch from Stephen Usher.
2441 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2443 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2446 TF/02 Fix a couple more cases where we did not log the error message
2447 when unlink() failed. See also change 4.74-TF/03.
2449 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2450 lock up or crash if it happened to be inside a call to libc when it
2451 got a SIGUSR1 from exiwhat.
2453 The SIGUSR1 handler appends the current process status to the process
2454 log which is later printed by exiwhat. It used to use the general
2455 purpose logging code to do this, but several functions it calls are
2456 not safe for signals.
2458 The new output code in the SIGUSR1 handler is specific to the process
2459 log, and simple enough that it's easy to inspect for signal safety.
2460 Removing some special cases also simplifies the general logging code.
2461 Removing the spurious timestamps from the process log simplifies
2464 TF/04 Improved ratelimit ACL condition.
2466 The /noupdate option has been deprecated in favour of /readonly which
2467 has clearer semantics. The /leaky, /strict, and /readonly update modes
2468 are mutually exclusive. The update mode is no longer included in the
2469 database key; it just determines when the database is updated. (This
2470 means that when you upgrade Exim will forget old rate measurements.)
2472 Exim now checks that the per_* options are used with an update mode that
2473 makes sense for the current ACL. For example, when Exim is processing a
2474 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2475 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2476 must specify per_mail/readonly. If you omit the update mode it defaults to
2477 /leaky where that makes sense (as before) or /readonly where required.
2479 The /noupdate option is now undocumented but still supported for
2480 backwards compatibility. It is equivalent to /readonly except that in
2481 ACLs where /readonly is required you may specify /leaky/noupdate or
2482 /strict/noupdate which are treated the same as /readonly.
2484 A useful new feature is the /count= option. This is a generalization
2485 of the per_byte option, so that you can measure the throughput of other
2486 aggregate values. For example, the per_byte option is now equivalent
2487 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2489 The per_rcpt option has been generalized using the /count= mechanism
2490 (though it's more complicated than the per_byte equivalence). When it is
2491 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2492 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2493 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2494 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2495 Note that using per_rcpt with a non-readonly update mode in more than
2496 one ACL will cause the recipients to be double-counted. (The per_mail
2497 and per_byte options don't have this problem.)
2499 The handling of very low rates has changed slightly. If the computed rate
2500 is less than the event's count (usually one) then this event is the first
2501 after a long gap. In this case the rate is set to the same as this event's
2502 count, so that the first message of a spam run is counted properly.
2504 The major new feature is a mechanism for counting the rate of unique
2505 events. The new per_addr option counts the number of different
2506 recipients that someone has sent messages to in the last time period. It
2507 behaves like per_rcpt if all the recipient addresses are different, but
2508 duplicate recipient addresses do not increase the measured rate. Like
2509 the /count= option this is a general mechanism, so the per_addr option
2510 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2511 example, measure the rate that a client uses different sender addresses
2512 with the options per_mail/unique=$sender_address. There are further
2513 details in the main documentation.
2515 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2517 TF/06 Removed a few PCRE remnants.
2519 TF/07 Automatically extract Exim's version number from tags in the git
2520 repository when doing development or release builds.
2522 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2523 Bugzilla 879. Patch from Paul Fisher.
2525 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2526 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2529 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2531 PP/05 Variable $av_failed, true if the AV scanner deferred.
2532 Bugzilla 1078. Patch from John Horne.
2534 PP/06 Stop make process more reliably on build failure.
2535 Bugzilla 1087. Patch from Heiko Schlittermann.
2537 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2538 Bugzilla 1089. Patch from Heiko Schlittermann.
2540 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2541 Bugzilla 1131. Patch from Holger Weiß.
2543 PP/09 Handle IPv6 addresses with SPF.
2544 Bugzilla 860. Patch from Wolfgang Breyha.
2546 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2548 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2551 PP/11 match_* no longer expand right-hand-side by default.
2552 New compile-time build option, EXPAND_LISTMATCH_RHS.
2553 New expansion conditions, "inlist", "inlisti".
2555 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2557 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2559 PP/14 fix log_write() format string regression from TF/03.
2560 Bugzilla 1152. Patch from Dmitry Isaikin.
2566 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2568 PP/02 Harmonised TLS library version reporting; only show if debugging.
2569 Layout now matches that introduced for other libraries in 4.74 PP/03.
2571 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2573 PP/04 New "dns_use_edns0" global option.
2575 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2578 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2579 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2581 TK/01 Updated PolarSSL code to 0.14.2.
2582 Bugzilla 1097. Patch from Andreas Metzler.
2584 PP/07 Catch divide-by-zero in ${eval:...}.
2585 Fixes bugzilla 1102.
2587 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2590 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2591 format-string attack -- SECURITY: remote arbitrary code execution.
2593 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2594 time unintentionally subject to list matching rules, letting the header
2595 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2596 arbitrary string expansion). This allowed for information disclosure.
2598 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2599 INT_MIN/-1 -- value coerced to INT_MAX.
2605 NM/01 Workaround for PCRE version dependency in version reporting
2608 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2609 This fixes portability to compilers other than gcc, notably
2610 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2612 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2613 makefiles for portability to HP-UX and POSIX correctness.
2615 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2616 Also via indented variable definition in the Makefile.
2617 (Debugging by Oliver Heesakkers).
2619 PP/02 Restore caching of spamd results with expanded spamd_address.
2620 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2622 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2623 Improves build reliability. Fix from: Frank Elsner
2625 NM/02 Fix wide character breakage in the rfc2047 coding
2626 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2628 NM/03 Allow underscore in dnslist lookups
2629 Fixes bug 1026. Patch from Graeme Fowler
2631 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2632 Code patches from Adam Ciarcinski of NetBSD.
2634 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2637 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2638 is logged, to avoid truncation. Patch from John Horne.
2640 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2641 Patch from Jakob Hirsch.
2643 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2644 SQL string expansion failure details.
2645 Patch from Andrey Oktyabrski.
2647 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2648 Patch from Simon Arlott.
2650 PP/09 New lookups functionality failed to compile on old gcc which rejects
2651 extern declarations in function scope.
2652 Patch from Oliver Fleischmann
2654 PP/10 Use sig_atomic_t for flags set from signal handlers.
2655 Check getgroups() return and improve debugging.
2656 Fixed developed for diagnosis in bug 927 (which turned out to be
2659 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2660 Patch from Mark Zealey.
2662 PP/12 Bugzilla 1056: Improved spamd server selection.
2663 Patch from Mark Zealey.
2665 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2666 Based on patch from Heiko Schlittermann.
2668 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2669 Patch from Uwe Doering, sign-off by Michael Haardt.
2671 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2672 variable declaration deep within a block. Bug and patch from
2675 PP/15 lookups-Makefile IRIX compatibility coercion.
2677 PP/16 Make DISABLE_DKIM build knob functional.
2679 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2680 Patch by Simon Arlott
2682 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2683 variable argument macros. Our copy now differs from upstream.
2689 TF/01 Failure to get a lock on a hints database can have serious
2690 consequences so log it to the panic log.
2692 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2693 controlled using the smtp_confirmation log selector.
2695 TF/03 Include the error message when we fail to unlink a spool file.
2697 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2698 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2699 for maintaining out-of-tree patches for some time.
2701 PP/01 Bugzilla 139: Documentation and portability issues.
2702 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2703 Handle per-OS dynamic-module compilation flags.
2705 PP/02 Let /dev/null have normal permissions.
2706 The 4.73 fixes were a little too stringent and complained about the
2707 permissions on /dev/null. Exempt it from some checks.
2708 Reported by Andreas M. Kirchwitz.
2710 PP/03 Report version information for many libraries, including
2711 Exim version information for dynamically loaded libraries. Created
2712 version.h, now support a version extension string for distributors
2713 who patch heavily. Dynamic module ABI change.
2715 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2716 privilege escalation vulnerability whereby the Exim run-time user
2717 can cause root to append content of the attacker's choosing to
2720 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2723 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2724 If dropping privileges for untrusted macros, we disabled normal logging
2725 on the basis that it would fail; for the Exim run-time user, this is not
2726 the case, and it resulted in successful deliveries going unlogged.
2727 Fixed. Reported by Andreas Metzler.
2733 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2734 only prepend for the Resent-* case. Fixes regression introduced in
2735 Exim 4.70 by NM/22 for Bugzilla 607.
2737 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2738 increasing numbers of administrators be bitten by this.
2740 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2742 PP/03 Bugzilla 994: added openssl_options main configuration option.
2744 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2746 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2748 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2750 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2751 without a peer certificate, leading to a segfault because of an
2752 assumption that peers always have certificates. Be a little more
2753 paranoid. Problem reported by Martin Tscholak.
2755 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2756 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2757 NB: ClamAV planning to remove STREAM in "middle of 2010".
2758 CL also introduces -bmalware, various -d+acl logging additions and
2759 more caution in buffer sizes.
2761 PP/09 Implemented reverse_ip expansion operator.
2763 PP/10 Bugzilla 937: provide a "debug" ACL control.
2765 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2767 PP/12 Bugzilla 973: Implement --version.
2769 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2771 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2773 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2775 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2776 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2777 ignore trailing whitespace.
2779 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2781 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2784 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2785 Notification from Dr Andrew Aitchison.
2787 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2788 ExtendedDetectionInfo response format.
2789 Notification from John Horne.
2791 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2794 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2795 XSL and documented dependency on system catalogs, with examples of how
2798 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2801 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2802 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2803 configuration file which is writeable by the Exim user or group.
2805 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2806 of configuration files to cover files specified with the -C option if
2807 they are going to be used with root privileges, not just the default
2810 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2811 option (effectively making it always true).
2813 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2814 files to be used while preserving root privileges.
2816 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2817 that rogue child processes cannot use them.
2819 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2820 run-time user, instead of root.
2822 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2823 Exim run-time user without dropping privileges.
2825 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2826 result string, instead of calling string_vformat() twice with the same
2829 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2830 for other users. Others should always drop root privileges if they use
2831 -C on the command line, even for a whitelisted configure file.
2833 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2835 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2841 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2842 $data_path, and $header_path variables; fixed documentation bugs and
2845 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2846 exipick to access non-standard spools, including the "frozen" queue
2849 NM/01 Bugzilla 965: Support mysql stored procedures.
2850 Patch from Alain Williams
2852 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2854 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2855 Patch from Andreas Metzler
2857 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2858 Patch from Kirill Miazine
2860 NM/05 Bugzilla 671: Added umask to procmail example.
2862 JJ/03 installed exipick 20100323.0, fixing doc bug
2864 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2865 directory. Notification and patch from Dan Rosenberg.
2867 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2869 TK/02 Improve log output when DKIM signing operation fails.
2871 MH/01 Treat the transport option dkim_domain as a colon separated
2872 list, not as a single string, and sign the message with each element,
2873 omitting multiple occurences of the same signer.
2875 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2876 Bugzilla 985, 986. Patch by Simon Arlott
2878 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2879 Patch by Simon Arlott
2881 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2882 MBX locking. Notification from Dan Rosenberg.
2888 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2890 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2892 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2894 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2896 NM/04 Bugzilla 915: Flag broken perl installation during build.
2902 TK/01 Added patch by Johannes Berg that expands the main option
2903 "spamd_address" if it starts with a dollar sign.
2905 TK/02 Write list of recipients to X-Envelope-Sender header when building
2906 the mbox-format spool file for content scanning (suggested by Jakob
2909 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2910 (http://www.dcc-servers.net/) support via dccifd. Activated by
2911 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2913 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2914 by Mark Daniel Reidel <mr@df.eu>.
2916 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2917 When building exim an external PCRE library is now needed -
2918 PCRE is a system library on the majority of modern systems.
2919 See entry on PCRE_LIBS in EDITME file.
2921 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2922 conversation. Added nologin parameter to request.
2923 Patch contributed by Kirill Miazine.
2925 TF/01 Do not log submission mode rewrites if they do not change the address.
2927 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2929 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2930 log files in place. Contributed by Roberto Lima.
2932 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2934 TF/03 Bugzilla 615: When checking the local_parts router precondition
2935 after a local_part_suffix or local_part_prefix option, Exim now
2936 does not use the address's named list lookup cache, since this
2937 contains cached lookups for the whole local part.
2939 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2940 Robert Millan. Documentation is in experimental-spec.txt.
2942 TF/04 Bugzilla 668: Fix parallel build (make -j).
2944 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2946 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2947 Patch provided by Jan Srzednicki.
2949 TF/05 Leading white space used to be stripped from $spam_report which
2950 wrecked the formatting. Now it is preserved.
2952 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2953 that they are available at delivery time.
2955 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2957 TF/08 TLS error reporting now respects the incoming_interface and
2958 incoming_port log selectors.
2960 TF/09 Produce a more useful error message if an SMTP transport's hosts
2961 setting expands to an empty string.
2963 NM/06 Bugzilla 744: EXPN did not work under TLS.
2964 Patch provided by Phil Pennock.
2966 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2967 Patch provided by Richard Godbee.
2969 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2970 acl_smtp_notquit, added index entry.
2972 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2973 Patch provided by Eugene Bujak.
2975 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2976 accept(). Patch provided by Maxim Dounin.
2978 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2979 Patch provided by Phil Pennock.
2981 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2983 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2984 Patch provided by Brad "anomie" Jorsch.
2986 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2987 Patch provided by Dean Brooks.
2989 TK/05 Add native DKIM support (does not depend on external libraries).
2991 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2992 Patch provided by Graeme Fowler.
2994 NM/16 Bugzilla 851: Documentation example syntax fix.
2996 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2998 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3001 NM/19 Bugzilla 745: TLS version reporting.
3002 Patch provided by Phil Pennock.
3004 NM/20 Bugzilla 167: bool: condition support.
3005 Patch provided by Phil Pennock.
3007 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3008 clients. Patch provided by Phil Pennock.
3010 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3011 Patch provided by Brad "anomie" Jorsch.
3013 NM/23 Bugzilla 687: Fix misparses in eximstats.
3014 Patch provided by Heiko Schlittermann.
3016 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3017 Patch provided by Heiko Schlittermann.
3019 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3020 plus update to original patch.
3022 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3024 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3025 Patch provided by David Brownlee.
3027 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3029 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3031 NM/30 Bugzilla 888: TLS documentation bugfixes.
3033 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3035 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3036 Unlike the original bugzilla I have changed all shell scripts in src tree.
3038 NM/33 Bugzilla 898: Transport filter timeout fix.
3039 Patch by Todd Rinaldo.
3041 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3042 Patch by Serge Demonchaux.
3044 NM/35 Bugzilla 39: Base64 decode bug fixes.
3045 Patch by Jakob Hirsch.
3047 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3049 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3051 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3053 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3059 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3060 ALT-N's libdkim that I have put here:
3061 http://duncanthrax.net/exim-experimental/
3063 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3064 were called 'true' and it seems that C99 defines that as a reserved
3065 keyword to be used with 'bool' variable types. That means you could
3066 not include C99-style headers which use bools without triggering
3067 build errors in sieve.c.
3069 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3070 as mailq or other aliases. Changed the --help handling significantly
3071 to do whats expected. exim_usage() emits usage/help information.
3073 SC/01 Added the -bylocaldomain option to eximstats.
3075 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3077 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3079 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3085 PH/01 Another patch from the Sieve maintainer.
3087 PH/02 When an IPv6 address is converted to a string for single-key lookup
3088 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3089 dots are used instead of colons so that keys in lsearch files need not
3090 contain colons. This was done some time before quoting was made available
3091 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3092 (notated using the quote facility) so as to distinguish them from IPv4
3093 keys. This meant that lookups for IP addresses in host lists did not work
3094 for iplsearch lookups.
3096 This has been fixed by arranging for IPv6 addresses to be expressed with
3097 colons if the lookup type is iplsearch. This is not incompatible, because
3098 previously such lookups could never work.
3100 The situation is now rather anomalous, since one *can* have colons in
3101 ordinary lsearch keys. However, making the change in all cases is
3102 incompatible and would probably break a number of configurations.
3104 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3107 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3108 conversion specification without a maximum field width, thereby enabling
3109 a rogue spamd server to cause a buffer overflow. While nobody in their
3110 right mind would setup Exim to query an untrusted spamd server, an
3111 attacker that gains access to a server running spamd could potentially
3112 exploit this vulnerability to run arbitrary code as the Exim user.
3114 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3115 $primary_hostname instead of what libspf2 thinks the hosts name is.
3117 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3118 a directory entry by the name of the lookup key. Previously, if a
3119 symlink pointed to a non-existing file or a file in a directory that
3120 Exim lacked permissions to read, a lookup for a key matching that
3121 symlink would fail. Now it is enough that a matching directory entry
3122 exists, symlink or not. (Bugzilla 503.)
3124 PH/03 The body_linecount and body_zerocount variables are now exported in the
3127 PH/04 Added the $dnslist_matched variable.
3129 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3130 This means they are set thereafter only if the connection becomes
3133 PH/06 Added the client_condition to authenticators so that some can be skipped
3134 by clients under certain conditions.
3136 PH/07 The error message for a badly-placed control=no_multiline_responses left
3137 "_responses" off the end of the name.
3139 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3141 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3142 (without spaces) instead of just copying the configuration text.
3144 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3146 PH/11 Added $max_received_linelength.
3148 PH/12 Added +ignore_defer and +include_defer to host lists.
3150 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3151 way in which PCRE > 7.0 is built.
3153 PH/14 Implemented queue_only_load_latch.
3155 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3156 MAIL command. The effect was to mangle the value on 64-bit systems.
3158 PH/16 Another patch from the Sieve maintainer.
3160 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3162 PH/18 If a system quota error occurred while trying to create the file for
3163 a maildir delivery, the message "Mailbox is full" was not appended to the
3164 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3165 only to a quota excession during the actual writing of the file.
3167 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3168 characters?) which causes problems in log lines. The DN values are now
3169 passed through string_printing() before being added to log lines.
3171 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3172 and InterBase are left for another time.)
3174 PH/21 Added message_body_newlines option.
3176 PH/22 Guard against possible overflow in moan_check_errorcopy().
3178 PH/23 POSIX allows open() to be a macro; guard against that.
3180 PH/24 If the recipient of an error message contained an @ in the local part
3181 (suitably quoted, of course), incorrect values were put in $domain and
3182 $local_part during the evaluation of errors_copy.
3188 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3189 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3192 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3193 issue a MAIL command.
3195 PH/02 In an ACL statement such as
3197 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3199 if a client was not listed at all, or was listed with a value other than
3200 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3201 the condition was not true (as it should be), so access was not denied.
3202 The bug was that the ! inversion was incorrectly passed on to the second
3203 item. This has been fixed.
3205 PH/03 Added additional dnslists conditions == and =& which are different from
3206 = and & when the dns lookup returns more than one IP address.
3208 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3209 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3211 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3212 FSYNC, which compiles an option called disable_fsync that allows for
3213 bypassing fsync(). The documentation is heavily laced with warnings.
3215 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3217 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3218 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3219 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3220 including adding "make clean"; (3) Added -fPIC when compiling the test
3221 dynamically loaded module, to get rid of a warning.
3223 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3224 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3225 = 0s. The bug is otherwise harmless.
3227 PH/07 There was a bug in the dovecot authenticator such that the value of
3228 $auth1 could be overwritten, and so not correctly preserved, after a
3229 successful authentication. This usually meant that the value preserved by
3230 the server_setid option was incorrect.
3232 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3234 PH/09 Installed PCRE release 7.0.
3236 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3237 run for batched SMTP input. It is now run at the start of every message
3238 in the batch. While fixing this I discovered that the process information
3239 (output by running exiwhat) was not always getting set for -bs and -bS
3240 input. This is fixed, and it now also says "batched" for BSMTP.
3242 PH/11 Added control=no_pipelining.
3244 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3245 patch, slightly modified), and move the expansion of helo_data till after
3246 the connection is made in the smtp transport (so it can use these
3249 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3251 PH/14 Added log_selector = +pid.
3253 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3255 PH/16 Add ${if forany and ${if forall.
3257 PH/17 Added dsn_from option to vary the From: line in DSNs.
3259 PH/18 Flush SMTP output before performing a callout, unless control =
3260 no_callout_flush is set.
3262 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3263 was true (the default) a successful delivery failed to delete the retry
3264 item, thus causing premature timeout of the address. The bug is now
3267 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3269 PH/21 Long custom messages for fakedefer and fakereject are now split up
3270 into multiline responses in the same way that messages for "deny" and
3271 other ACL rejections are.
3273 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3274 with slight modification.
3276 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3277 draft, changing the syntax and factoring some duplicate code".
3279 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3280 for deliveries of the second and subsequent messages over the same SMTP
3283 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3284 ${reduce, with only minor "tidies".
3286 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3288 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3289 expansion side effects.
3291 PH/27 When a message times out after an over-quota error from an Exim-imposed
3292 quota, the bounce message says "mailbox is full". This message was not
3293 being given when it was a system quota that was exceeded. It now should
3296 MH/03 Made $recipients available in local_scan(). local_scan() already has
3297 better access to the recipient list through recipients_list[], but
3298 $recipients can be useful in postmaster-provided expansion strings.
3300 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3301 in the case of a MAIL command with additional options following the
3302 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3303 were accidentally chopped off.
3305 PH/29 SMTP synchronization checks are implemented when a command is read -
3306 there is a check that no more input is waiting when there shouldn't be
3307 any. However, for some commands, a delay in an ACL can mean that it is
3308 some time before the response is written. In this time, more input might
3309 arrive, invalidly. So now there are extra checks after an ACL has run for
3310 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3311 pipelining has not been advertised.
3313 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3315 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3316 returned in the SMTP response when smtp_return_error_details was false.
3317 This has been fixed.
3319 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3320 instead of the C I/O that was originally supplied, because problems were
3321 reported on Solaris.
3323 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3324 Exim which did not show up earlier: it was assuming that a call to
3325 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3326 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3327 was a macro that became an assignment, so it seemed to work. This has
3328 changed to a proper function call with a void return, hence the compile
3329 error. Exim's code has been fixed.
3331 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3334 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3336 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3338 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3339 because it turns out to be too big or there is a timeout) there is no
3340 "Completed" line in the log. When some messages of this type were
3341 selected by exigrep, they were listed as "not completed". Others were
3342 picked up by some special patterns. I have improved the selection
3343 criteria to be more general.
3345 PH/38 The host_find_failed option in the manualroute router can now be set
3346 to "ignore", to completely ignore a host whose IP address cannot be
3347 found. If all hosts are ignored, the behaviour is controlled by the new
3348 host_all_ignored option.
3350 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3351 homing or because of multiple MX records with /mx) generated more than
3352 one IP address, and the following item turned out to be the local host,
3353 all the secondary addresses of the first item were incorrectly removed
3354 from the list, along with the local host and any following hosts (which
3355 is what is supposed to happen).
3357 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3358 whoever called Exim into the -H file. In the case of the daemon it was
3359 behaving confusingly. When first started, it used values for whoever
3360 started the daemon, but after a SIGHUP it used the Exim user (because it
3361 calls itself on a restart). I have changed the code so that it now always
3364 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3365 message are rejected with the same error (e.g. no authentication or bad
3366 sender address), and a DATA command is nevertheless sent (as can happen
3367 with PIPELINING or a stupid MUA), the error message that was given to the
3368 RCPT commands is included in the rejection of the DATA command. This is
3369 intended to be helpful for MUAs that show only the final error to their
3372 PH/42 Another patch from the Sieve maintainer.
3374 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3375 Eximstats - Fixed some broken HTML links and added missing column headers
3377 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3378 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3380 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3382 PH/43 Yet another patch from the Sieve maintainer.
3384 PH/44 I found a way to check for a TCP/IP connection going away before sending
3385 the response to the final '.' that terminates a message, but only in the
3386 case where the client has not sent further data following the '.'
3387 (unfortunately, this is allowed). However, in many cases there won't be
3388 any further data because there won't be any more messages to send. A call
3389 to select() can be used: if it shows that the input is "ready", there is
3390 either input waiting, or the socket has been closed. An attempt to read
3391 the next input character can distinguish the two cases. Previously, Exim
3392 would have sent an OK response which the client would never have see.
3393 This could lead to message repetition. This fix should cure that, at
3394 least in a lot of common cases.
3396 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3397 advertised in response to EHLO.
3403 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3404 fixed by 4.65/MH/01 (is this a record?) are fixed:
3406 (i) An empty string was always treated as zero by the numeric comparison
3407 operators. This behaviour has been restored.
3409 (ii) It is documented that the numeric comparison operators always treat
3410 their arguments as decimal numbers. This was broken in that numbers
3411 starting with 0 were being interpreted as octal.
3413 While fixing these problems I realized that there was another issue that
3414 hadn't been noticed. Values of message_size_limit (both the global option
3415 and the transport option) were treated as octal if they started with 0.
3416 The documentation was vague. These values are now always treated as
3417 decimal, and I will make that clear in the documentation.
3423 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3424 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3427 MH/01 Don't check that the operands of numeric comparison operators are
3428 integers when their expansion is in "skipping" mode (fixes bug
3429 introduced by 4.64-PH/07).
3431 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3432 child addresses, Exim now panics and dies. Previously, because the count
3433 is held in a short int, deliveries were likely to be lost. As such a
3434 large number of recipients for a single message is ridiculous
3435 (performance will be very, very poor), I have chosen to impose a limit
3436 rather than extend the field.
3442 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3443 leftover -K file (the existence of which was triggered by #402).
3444 While we were at it, introduced process PID as part of the -K
3445 filename. This should rule out race conditions when creating
3448 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3449 processing considerably. Previous code took too long for large mails,
3450 triggering a timeout which in turn triggers #401.
3452 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3453 in the DK code in transports.c. sendfile() is not really portable,
3454 hence the _LINUX specificness.
3456 TF/01 In the add_headers option to the mail command in an Exim filter,
3457 there was a bug that Exim would claim a syntax error in any
3458 header after the first one which had an odd number of characters
3461 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3462 callout verification, Exim cached a "reject" for the entire domain. This
3463 is correct for most verifications, but it is not correct for a recipient
3464 verification with use_sender or use_postmaster set, because in that case
3465 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3466 case of MAIL FROM:<> rejection from other early rejections (e.g.
3467 rejection of HELO). When verifying a recipient using a non-null MAIL
3468 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3469 Whatever the result of the callout, the value of the domain cache is
3470 left unchanged (for any other kind of callout, getting as far as trying
3471 RCPT means that the domain itself is ok).
3473 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3476 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3477 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3478 session, the daemon ignores ECONNECT errors and logs others; it now
3479 ignores EPIPE as well.
3481 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3482 (quoted-printable decoding).
3484 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3485 later the small subsequent patch to fix an introduced bug.
3487 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3489 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3491 PH/08 An error is now given if message_size_limit is specified negative.
3493 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3494 to be given (somewhat) arbitrary names.
3496 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3499 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3500 miscellaneous code fixes
3502 PH/10 Added the log_reject_target ACL modifier to specify where to log
3505 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3506 hostname. This is wrong, because it relates to the incoming message (and
3507 probably the interface on which it is arriving) and not to the outgoing
3508 callout (which could be using a different interface). This has been
3509 changed to use the value of the helo_data option from the smtp transport
3510 instead - this is what is used when a message is actually being sent. If
3511 there is no remote transport (possible with a router that sets up host
3512 addresses), $smtp_active_hostname is used.
3514 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3515 tweaks were necessary in order to get it to work (see also 21 below):
3516 (a) The code assumed that strncpy() returns a negative number on buffer
3517 overflow, which isn't the case. Replaced with Exim's string_format()
3519 (b) There were several signed/unsigned issues. I just did the minimum
3520 hacking in of casts. There is scope for a larger refactoring.
3521 (c) The code used strcasecmp() which is not a standard C function.
3522 Replaced with Exim's strcmpic() function.
3523 (d) The code set only $1; it now sets $auth1 as well.
3524 (e) A simple test gave the error "authentication client didn't specify
3525 service in request". It would seem that Dovecot has changed its
3526 interface. Fortunately there's a specification; I followed it and
3527 changed what the client sends and it appears to be working now.
3529 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3532 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3533 address A is aliased to B and C, where B exists and C does not. Without
3534 -v the output is "A verified" because verification stops after a
3535 successful redirection if more than one address is generated. However,
3536 with -v the child addresses are also verified. Exim was outputting "A
3537 failed to verify" and then showing the successful verification for C,
3538 with its parentage. It now outputs "B failed to verify", showing B's
3539 parentage before showing the successful verification of C.
3541 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3542 look up a TXT record in a specific list after matching in a combined
3545 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3546 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3547 they consult the DNS. I had assumed they would set it the way they
3548 wanted; and indeed my experiments on Linux seem to show that in some
3549 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3550 To be on the safe side, however, I have now made the interface to
3551 host_find_byname() similar to host_find_bydns(), with an argument
3552 containing the DNS resolver options. The host_find_byname() function now
3553 sets these options at its start, just as host_find_bydns() does. The smtp
3554 transport options dns_qualify_single and dns_search_parents are passed to
3555 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3556 of host_find_byname() use the default settings of RES_DEFNAMES
3557 (qualify_single) but not RES_DNSRCH (search_parents).
3559 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3560 spool_read_header() do less string testing, by means of a preliminary
3561 switch on the second character of optional "-foo" lines. (This is
3562 overdue, caused by the large number of possibilities that now exist.
3563 Originally there were few.) While I was there, I also converted the
3564 str(n)cmp tests so they don't re-test the leading "-" and the first
3565 character, in the hope this might squeeze out yet more improvement.
3567 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3568 flag allowing group syntax was set by the header_syntax check but not
3569 turned off, possible causing trouble later; (2) The flag was not being
3570 set at all for the header_verify test, causing "group"-style headers to
3571 be rejected. I have now set it in this case, and also caused header_
3572 verify to ignore an empty address taken from a group. While doing this, I
3573 came across some other cases where the code for allowing group syntax
3574 while scanning a header line wasn't quite right (mostly, not resetting
3575 the flag correctly in the right place). These bugs could have caused
3576 trouble for malformed header lines. I hope it is now all correct.
3578 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3579 with the "reply" argument non-NULL. The code, however (which originally
3580 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3581 but it didn't always do it. This confused somebody who was copying the
3582 code for some other use. I have removed all the tests.
3584 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3585 feature that was used to support insecure browsers during the U.S. crypto
3586 embargo. It requires special client support, and Exim is probably the
3587 only MTA that supported it -- and would never use it because real RSA is
3588 always available. This code has been removed, because it had the bad
3589 effect of slowing Exim down by computing (never used) parameters for the
3590 RSA_EXPORT functionality.
3592 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3593 authenticator to fail if there's a tab character in the incoming data
3594 (there should never be unless someone is messing about, as it's supposed
3595 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3596 if the connection is using TLS or if the remote IP is the same as the
3597 local IP, and the "valid-client-cert option" if a client certificate has
3600 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3601 authenticators. This can be used for authorization after authentication
3602 succeeds. (In the case of plaintext, it servers for both authentication
3605 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3606 if any retry times were supplied.
3608 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3609 connection, where there is no client IP address to check. In this
3610 situation, the verify now always succeeds.
3612 PH/25 Applied John Jetmore's -Mset patch.
3614 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3616 PH/27 In a string expansion for a processed (not raw) header when multiple
3617 headers of the same name were present, leading whitespace was being
3618 removed from all of them, but trailing whitespace was being removed only
3619 from the last one. Now trailing whitespace is removed from each header
3620 before concatenation. Completely empty headers in a concatenation (as
3621 before) are ignored.
3623 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3624 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3626 PH/29 [Removed. This was a change that I later backed out, and forgot to
3627 correct the ChangeLog entry (that I had efficiently created) before
3628 committing the later change.]
3630 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3631 address errors (4xx response to RCPT) over the same connection as other
3632 messages routed to the same hosts. Such deliveries are always "forced",
3633 so retry times are not inspected. This resulted in far too many retries
3634 for the affected addresses. The effect occurred only when there were more
3635 hosts than the hosts_max_try setting in the smtp transport when it had
3636 the 4xx errors. Those hosts that it had tried were not added to the list
3637 of hosts for which the message was waiting, so if all were tried, there
3638 was no problem. Two fixes have been applied:
3640 (i) If there are any address or message errors in an SMTP delivery, none
3641 of the hosts (tried or untried) are now added to the list of hosts
3642 for which the message is waiting, so the message should not be a
3643 candidate for sending over the same connection that was used for a
3644 successful delivery of some other message. This seems entirely
3645 reasonable: after all the message is NOT "waiting for some host".
3646 This is so "obvious" that I'm not sure why it wasn't done
3647 previously. Hope I haven't missed anything, but it can't do any
3648 harm, as the worst effect is to miss an optimization.
3650 (ii) If, despite (i), such a delivery is accidentally attempted, the
3651 routing retry time is respected, so at least it doesn't keep
3652 hammering the server.
3654 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3655 in ${readsocket because some servers need this prod.
3657 PH/32 Added some extra debug output when updating a wait-xxx database.
3659 PH/33 The hint "could be header name not terminated by colon", which has been
3660 given for certain expansion errors for a long time, was not being given
3661 for the ${if def:h_colon_omitted{... case.
3663 PH/34 The spec says: "With one important exception, whenever a domain list is
3664 being scanned, $domain contains the subject domain." There was at least
3665 one case where this was not true.
3667 PH/35 The error "getsockname() failed: connection reset by peer" was being
3668 written to the panic log as well as the main log, but it isn't really
3669 panic-worthy as it just means the connection died rather early on. I have
3670 removed the panic log writing for the ECONNRESET error when getsockname()
3673 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3674 runs only) independently of the message's sender address. This meant
3675 that, if the 4xx error was in fact related to the sender, a different
3676 message to the same recipient with a different sender could confuse
3677 things. In particular, this can happen when sending to a greylisting
3678 server, but other circumstances could also provoke similar problems.
3679 I have changed the default so that the retry time for these errors is now
3680 based a combination of the sender and recipient addresses. This change
3681 can be overridden by setting address_retry_include_sender=false in the
3684 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3685 remote server are returned as part of bounce messages. This was not
3686 happening for LMTP over a pipe (the lmtp transport), but now it is the
3687 same for both kinds of LMTP.
3689 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3690 in header lines that were in fact CNAMEs. This is no longer the case.
3692 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3693 and queue runs started by the daemon processed all messages. This has
3694 been fixed so that -R and -S can now usefully be given with -q<time>.
3696 PH/40 Import PCRE release 6.7 (fixes some bugs).
3698 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3700 PH/42 Give an error if -q is specified more than once.
3702 PH/43 Renamed the variables $interface_address and $interface_port as
3703 $received_ip_address and $received_port, to make it clear that these
3704 values apply to message reception, and not to the outgoing interface when
3705 a message is delivered. (The old names remain recognized, of course.)
3707 PH/44 There was no timeout on the connect() call when using a Unix domain
3708 socket in the ${readsocket expansion. There now is.
3710 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3711 be meaningful with "accept".
3713 SC/01 Eximstats V1.43
3714 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3716 SC/02 Eximstats V1.44
3717 Use a glob alias rather than an array ref in the generated
3718 parser. This improves both readability and performance.
3720 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3721 Collect SpamAssassin and rejection statistics.
3722 Don't display local sender or destination tables unless
3723 there is data to show.
3724 Added average volumes into the top table text output.
3726 SC/04 Eximstats V1.46
3727 Collect data on the number of addresses (recipients)
3728 as well as the number of messages.
3730 SC/05 Eximstats V1.47
3731 Added 'Message too big' to the list of mail rejection
3732 reasons (thanks to Marco Gaiarin).
3734 SC/06 Eximstats V1.48
3735 Mainlog lines which have GMT offsets and are too short to
3736 have a flag are now skipped.
3738 SC/07 Eximstats V1.49 (Alain Williams)
3739 Added the -emptyok flag.
3741 SC/08 Eximstats V1.50
3742 Fixes for obtaining the IP address from reject messages.
3744 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3745 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3746 whitespace changes from 4.64-PH/27
3748 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3751 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3752 are found, allow negative numbers in numeric criteria)
3754 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3756 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3757 to match changes made in 4.64-PH/43
3759 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3761 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3762 "domain" is actually an IP address, and force a failure. This locks out
3763 those revolvers/nameservers that support "A-for-A" lookups, in
3764 contravention of the specifications.
3766 PH/48 When a host name was looked up from an IP address, and the subsequent
3767 forward lookup of the name timed out, the host name was left in
3768 $sender_host_name, contrary to the specification.
3770 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3771 restricted to single-key lookups, Exim was not diagnosing an error if
3772 * or *@ was used with a query-style lookup.
3774 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3776 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3777 long ago, but noone interested enough thought of it. Let's just say that
3778 the "1.1" means that there are some new functions that weren't there at
3779 some point in the past.
3781 PH/51 Error processing for expansion failure of helo_data from an smtp
3782 transport during callout processing was broken.
3784 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3785 tested/used via the -bh/-bhc/-bs options.
3787 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3788 bug, fixed in subsequent PCRE releases).
3790 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3791 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3793 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3799 SC/01 Use a glob alias rather than an array ref in eximstats generated
3800 parser. This improves both readability and performance.
3802 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3803 Don't display local sender or destination tables in eximstats unless
3804 there is data to show.
3805 Added average volumes into the eximstats top table text output.
3807 SC/03 Collect data on the number of addresses (recipients) as well
3808 as the number of messages in eximstats.
3810 TF/01 Correct an error in the documentation for the redirect router. Exim
3811 does (usually) call initgroups() when daemonizing.
3813 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3814 with consistent privilege compared to when running as a daemon.
3816 TF/03 Note in the spec that $authenticated_id is not set for local
3817 submissions from trusted users.
3819 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3820 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3822 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3823 by adding some example configuration directives to the default
3824 configuration file. A little bit of work is required to uncomment the
3825 directives and define how usernames and passwords are checked, but
3826 there is now a framework to start from.
3828 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3829 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3830 without this. I don't know how relevant this is to other LDAP libraries.
3832 PH/02 Add the verb name to the "unknown ACL verb" error.
3834 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3836 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3838 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3839 directory not expanded when it should be if an expanded home directory
3840 was set for the address (which is overridden by the transport).
3842 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3845 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3846 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3847 because it is too late at that time, and has no effect.
3849 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3850 security issue with \' (bugzilla #107). I could not use the
3851 PQescapeStringConn() function, because it needs a PGconn value as one of
3854 PH/08 When testing addresses using -bt, indicate those final addresses that
3855 are duplicates that would not cause an additional delivery. At least one
3856 person was confused, thinking that -bt output corresponded to deliveries.
3857 (Suppressing duplicates isn't a good idea as you lose the information
3858 about possibly different redirections that led to the duplicates.)
3860 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3861 systems where poll() doesn't work, in particular OS X.
3863 PH/10 Added more information to debugging output for retry time not reached.
3865 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3866 operations in malware.c.
3868 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3871 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3872 syslog_duplication set false, log lines that would normally be written
3873 both the the main log and to the reject log were not written to syslog at
3876 PH/14 In the default configuration, change the use of "message" in ACL warn
3877 statements to "add_header".
3879 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3880 not followed by a command (e.g. "seen endif").
3882 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3883 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3886 PH/17 Added extra conditions to the default value of delay_warning_condition
3890 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3891 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3892 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3895 The Auto-Submitted: and various List- headers are standardised, whereas I
3896 don't think Precedence: ever was.
3898 PH/18 Refactored debugging code in route_finduser() to show more information,
3899 in particular, the error code if getpwnam() issues one.
3901 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3902 This is apparently needed in addition to the PH/07 change above to avoid
3903 any possible encoding problems.
3905 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3906 but not after initializing Perl.
3908 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3909 output them only if debugging. By default they are written stderr,
3910 apparently, which is not desirable.
3912 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3915 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3918 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3920 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3921 authenticated or an ident call has been made. Suppress the default
3922 values for $authenticated_id and $authenticated_sender (but permit -oMai
3923 and -oMas) when testing with -bh.
3925 PH/24 Re-jigged the order of the tests in the default configuration so that the
3926 tests for valid domains and recipients precede the DNS black list and CSA
3927 tests, on the grounds that those ones are more expensive.
3929 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3930 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3931 This bug exists in every version of Exim that I still have, right back to
3934 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3935 However, an attempt to turn on case-sensitivity in a regex key by
3936 including (?-i) didn't work because the subject string was already
3937 lowercased, and the effects were non-intuitive. It turns out that a
3938 one-line patch can be used to allow (?-i) to work as expected.
3944 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3945 other effects) broke the use of negated acl sub-conditions.
3947 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3950 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3951 "Deny" causes Exim to reject the incoming connection with a 554 error.
3952 Unfortunately, if there is a major crisis, such as a disk failure,
3953 tcp-wrappers gives "deny", whereas what one would like would be some
3954 kind of temporary error. A kludge has been added to help with this.
3955 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3956 554 error is used if errno is still zero or contains ENOENT (which occurs
3957 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3960 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3962 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3963 errors. Otherwise a message that provokes a temporary error (when other
3964 messages do not) can cause a whole host to time out.
3966 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3967 addresses were routed directly to files or pipes from a redirect router.
3968 File deliveries just didn't batch; pipe deliveries might have suffered
3971 PH/06 A failure to get a lock for a hints database would erroneously always say
3972 "Failed to get write lock", even when it was really a read lock.
3974 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3975 of 0600. This has been changed to use the value of the lockfile_mode
3976 option (which defaults to 0600).
3978 PH/08 Applied small patch from the Sieve maintainer.
3980 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3981 folder from quota calculations, a direct delivery into this folder messed
3982 up the contents of the maildirsize file. This was because the regex was
3983 used only to exclude .Trash (or whatever) when the size of the mailbox
3984 was calculated. There was no check that a delivery was happening into an
3985 excluded directory. This bug has been fixed by ignoring all quota
3986 processing for deliveries into excluded directories.
3988 PH/10 Added the maildirfolder_create_regex option to appendfile.
3994 PH/01 The code for finding all the local interface addresses on a FreeBSD
3995 system running IPv6 was broken. This may well have applied to all BSD
3996 systems, as well as to others that have similar system calls. The broken
3997 code found IPv4 interfaces correctly, but gave incorrect values for the
3998 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3999 that it would not match correctly against @[] and not recognize the IPv6
4002 PH/02 The ipliteral router was not recognizing addresses of the form user@
4003 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4005 PH/03 Added disable_ipv6.
4007 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4008 decoded form, because it is most often used to construct To: headers
4009 lines in autoreplies, and the decoded form may well be syntactically
4010 invalid. However, $reply_address has leading white space removed, and all
4011 newlines turned into spaces so that the autoreply transport does not
4014 PH/05 If group was specified without a user on a router, and no group or user
4015 was specified on a transport, the group from the router was ignored.
4017 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4018 for visible compile-time settings that can be used to change these
4019 numbers, for those that want even more. Backwards compatibility with old
4020 spool files has been maintained. However, going back to a previous Exim
4021 release will lost any variables that are in spool files.
4023 PH/07 Two small changes when running in the test harness: increase delay when
4024 passing a TCP/IP connection to a new process, in case the original
4025 process has to generate a bounce, and remove special handling of
4026 127.0.0.2 (sic), which is no longer necessary.
4028 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4029 be the same on different OS.
4031 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4034 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4035 whether --show-vars was specified or not
4037 JJ/02 exipick: Added support for new ACL variable spool format introduced
4040 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4041 syntactically invalid From: or Reply-to: line, and a filter used this to
4042 generate an autoreply, and therefore failed to obtain an address for the
4043 autoreply, Exim could try to deliver to a non-existent relative file
4044 name, causing unrelated and misleading errors. What now happens is that
4045 it logs this as a hard delivery error, but does not attempt to create a
4048 PH/11 The exinext utility has a -C option for testing purposes, but although
4049 the given file was scanned by exinext itself; it wasn't being passed on
4050 when Exim was called.
4052 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4053 an end-of-file indication when reading a command response.
4055 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4056 compiled. In many other places in Exim, IPv6 addresses are always
4057 recognized, so I have changed this. It also means that IPv4 domain
4058 literals of the form [IPV4:n.n.n.n] are now always recognized.
4060 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4061 used if the router is not running as root, for example, when verifying at
4062 ACL time, or when using -bh. The debugging output from this situation was
4063 non-existent - all you got was a failure to exec. I have made two
4066 (a) Failures to set uid/gid, the current directory, or a process leader
4067 in a subprocess such as that created by queryprogram now generate
4068 suitable debugging output when -d is set.
4070 (b) The queryprogram router detects when it is not running as root,
4071 outputs suitable debugging information if -d is set, and then runs
4072 the subprocess without attempting to change uid/gid.
4074 PH/15 Minor change to Makefile for building test_host (undocumented testing
4077 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4078 additional section of a DNS packet that returns MX or SRV records.
4079 Instead, it always explicitly searches for A/AAAA records. This avoids
4080 major problems that occur when a DNS server includes only records of one
4081 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4082 fixed another bug: if SRV records were looked up and the corresponding
4083 address records were *not* found in the additional section, the port
4084 values from the SRV records were lost.
4086 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4087 using the correct key (the original address) when searching the retry
4088 rules in order to find which one to use for generating the retry hint.
4090 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4091 adding the default one. Similarly, if it contains a Reply-To: header, the
4092 errors_reply_to option, if set, is not used.
4094 PH/19 When calculating a retry time, Exim used to measure the "time since
4095 failure" by looking at the "first failed" field in the retry record. Now
4096 it does not use this if it is later than than the arrival time of the
4097 message. Instead it uses the arrival time. This makes for better
4098 behaviour in cases where some deliveries succeed, thus re-setting the
4099 "first failed" field. An example is a quota failure for a huge message
4100 when small messages continue to be delivered. Without this change, the
4101 "time since failure" will always be short, possible causing more frequent
4102 delivery attempts for the huge message than are intended.
4103 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4105 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4106 $1, $2, $3) because the numerical variables can be reset during some
4107 expansion items (e.g. "match"), thereby losing the authentication data.
4109 PH/21 Make -bV show the size of off_t variables so that the test suite can
4110 decide whether to run tests for quotas > 2G.
4112 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4113 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4114 filecount value is greater than 2G or if a quota value is greater than 2G
4115 on a system where the size of off_t is not greater than 4, a panic error
4118 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4119 never match. The debug and -bh output now contains an explicit error
4120 message indicating a malformed IPv4 address or mask.
4122 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4123 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4124 PH/23 above applies.
4126 PH/25 Do not write to syslog when running in the test harness. The only
4127 occasion when this arises is a failure to open the main or panic logs
4128 (for which there is an explicit test).
4130 PH/26 Added the /no_tell option to "control=freeze".
4132 PH/27 If a host name lookup failed very early in a connection, for example, if
4133 the IP address matched host_lookup and the reverse lookup yielded a name
4134 that did not have a forward lookup, an error message of the form "no IP
4135 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4136 could be logged. Now it outputs the IP address instead of "NULL".
4138 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4139 allows the sender and the authenticated sender to be set when
4140 submitting a message from within Exim. Since child_open_exim() is
4141 documented for local_scan(), the new function should be too.
4143 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4144 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4145 results in an empty string is now treated as unset.
4147 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4149 PH/31 Added sender_verify_fail logging option.
4151 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4152 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4153 tidied the source and removed it altogether.
4155 PH/33 When a queue run was abandoned because the load average was too high, a
4156 log line was always written; now it is written only if the queue_run log
4157 selector is set. In addition, the log line for abandonment now contains
4158 information about the queue run such as the pid. This is always present
4159 in "start" and "stop" lines but was omitted from the "abandon" line.
4161 PH/34 Omit spaces between a header name and the colon in the error message that
4162 is given when verify = headers_syntax fails (if there are lots of them,
4163 the message gets confusing).
4165 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4166 names, as there are now some PTR records that contain slashes. This check
4167 is only to protect against broken name servers that fall over on strange
4168 characters, so the fact that it applies to all lookups doesn't matter.
4170 PH/36 Now that the new test suite is complete, we can remove some of the
4171 special code in Exim that was needed for the old test suite. For example,
4172 sorting DNS records because real resolvers return them in an arbitrary
4173 order. The new test suite's fake resolver always returns records in the
4176 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4177 bounces) except when queue_only is set, to avoid logging races between
4178 the different processes.
4180 PH/38 Panic-die if .include specifies a non-absolute path.
4182 PH/39 A tweak to the "H" retry rule from its user.
4184 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4185 a label. They prevented compilation on older perls.
4187 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4188 a warning to be raised on newish perls.
4190 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4191 on queue. Changes to match documented behaviour of showing count of
4192 messages matching specified criteria.
4194 PH/40 Changed the default ident timeout from 30s to 5s.
4196 PH/41 Added support for the use of login_cap features, on those BSD systems
4197 that have them, for controlling the resources used by pipe deliveries.
4199 PH/42 The content-scanning code uses fopen() to create files in which to put
4200 message data. Previously it was not paying any attention to the mode of
4201 the files. Exim runs with umask(0) because the rest of the code creates
4202 files with open(), and sets the required mode explicitly. Thus, these
4203 files were ending up world-writeable. This was not a big issue, because,
4204 being within the spool directory, they were not world-accessible. I have
4205 created a function called modefopen, which takes an additional mode
4206 argument. It sets umask(777), creates the file, chmods it to the required
4207 mode, then resets the umask. All the relevant calls to fopen() in the
4208 content scanning code have been changed to use this function.
4210 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4211 to 24 hours. This avoids potential overflow problems when processing G
4212 and H retry rules. I suspect nobody ever tinkers with this value.
4214 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4216 PH/45 When the plaintext authenticator is running as a client, the server's
4217 challenges are checked to ensure they are valid base64 strings. By
4218 default, the authentication attempt is cancelled if an invalid string is
4219 received. Setting client_ignore_invalid_base64 true ignores these errors.
4220 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4221 they are received. Thus, the responses can be made to depend on the
4222 challenges. If an invalid string is ignored, an empty string is placed in
4225 PH/46 Messages that are created by the autoreply transport now contains a
4226 References: header, in accordance with RFCs 2822 and 3834.
4228 PH/47 Added authenticated_sender_force to the smtp transport.
4230 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4232 PH/49 Installed latest patch from the Sieve maintainer.
4234 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4235 was also set, the appendfile transport was unnecessarily scanning a
4236 directory of message files (e.g. for maildir delivery) to find the count
4237 of files (along with the size), even though it did not need this
4238 information. It now does the scan only if it needs to find either the
4239 size of the count of files.
4241 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4243 PH/52 Two bugs concerned with error handling when the smtp transport is
4246 (i) Exim was not creating retry information for temporary errors given
4247 for individual recipients after the DATA command when the smtp transport
4248 was used in LMTP mode. This meant that they could be retried too
4249 frequently, and not timed out correctly.
4251 (ii) Exim was setting the flag that allows error details to be returned
4252 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4253 recipients that were returned after the DATA command.
4255 PH/53 This is related to PH/52, but is more general: for any failing address,
4256 when detailed error information was permitted to be returned to the
4257 sender, but the error was temporary, then after the final timeout, only
4258 "retry timeout exceeded" was returned. Now it returns the full error as
4259 well as "retry timeout exceeded".
4261 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4262 do this, and (what is worse) MTAs that accept it.
4264 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4265 will now be deprecated.
4267 PH/56 New os.c-cygwin from the Cygwin maintainer.
4269 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4270 formats (previously only available in exim formats via -bpr, -bpru,
4271 and -bpra. Now also available in native and exiqgrep formats)
4273 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4274 with very large, slow to parse queues
4276 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4278 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4280 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4281 responses to authentication challenges, though it was showing the
4282 challenges; (ii) I've removed the CR characters from the debug output for
4285 PH/58 Allow for the insertion of a newline as well as a space when a string
4286 is turned into more than one encoded-word during RFC 2047 encoding. The
4287 Sieve code now uses this.
4289 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4290 data_4xx, lost_connection, tls_required.
4292 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4293 message was being sent as an SMTP response.
4295 PH/61 Add -l and -k options to exicyclog.
4297 PH/62 When verifying, if an address was redirected to one new address, so that
4298 verification continued, and the new address failed or deferred after
4299 having set something in $address_data, the value of $address_data was not
4300 passed back to the ACL. This was different to the case when no
4301 redirection occurred. The value is now passed back in both cases.
4303 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4304 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4305 use login_cap.h, so on its own it isn't the distinguishing feature. The
4306 new name refers directly to the setclassresources() function.
4308 PH/65 Added configuration files for NetBSD3.
4310 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4312 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4313 is preferred over IPv4.
4315 PH/68 The bounce_return_message and bounce_return_body options were not being
4316 honoured for bounces generated during the reception of non-SMTP messages.
4317 In particular, this applied to messages rejected by the ACL. This bug has
4318 been fixed. However, if bounce_return_message is true and bounce_return_
4319 body is false, the headers that are returned for a non-SMTP message
4320 include only those that have been read before the error was detected.
4321 (In the case of an ACL rejection, they have all been read.)
4323 PH/69 The HTML version of the specification is now built in a directory called
4324 spec_html instead of spec.html, because the latter looks like a path with
4325 a MIME-type, and this confuses some software.
4327 PH/70 Catch two compiler warnings in sieve.c.
4329 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4330 function verify_get_ident() calls ip_connect() to connect a socket, but
4331 if the "connect()" function timed out, ip_connect() used to close the
4332 socket. However, verify_get_ident() also closes the socket later, and in
4333 between Exim writes to the log, which may get opened at this point. When
4334 the socket was closed in ip_connect(), the log could get the same file
4335 descriptor number as the socket. This naturally causes chaos. The fix is
4336 not to close the socket in ip_connect(); the socket should be closed by
4337 the function that creates it. There was only one place in the code where
4338 this was missing, in the iplookup router, which I don't think anybody now
4339 uses, but I've fixed it anyway.
4341 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4342 well as to direct DNS lookups. Otherwise the handling of names in host
4343 lists is inconsistent and therefore confusing.
4349 PH/01 Two changes to the default runtime configuration:
4351 (1) Move the checks for relay_from_hosts and authenticated clients from
4352 after to before the (commented out) DNS black list checks.
4354 (2) Add control=submission to the relay_from_hosts and authenticated
4355 clients checks, on the grounds that messages accepted by these
4356 statements are most likely to be submissions.
4358 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4360 (1) Generate an error if the third argument for the ${prvs expansion is
4363 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4366 (3) Reset the variables that are obtained from the first argument of
4367 ${prvscheck and used in the second argument before leaving the code,
4368 because their memory is reclaimed, so using them afterwards may do
4371 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4372 one (it's much easier than Tom thought :-).
4374 (5) Because of (4), we can now allow for the use of $prvscheck_result
4375 inside the third argument.
4377 PH/03 For some reason, the default setting of PATH when running a command from
4378 a pipe transport was just "/usr/bin". I have changed it to
4381 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4382 anything to be listed in the output from -bV.
4384 PH/05 When a filter generated an autoreply, the entire To: header line was
4385 quoted in the delivery log line, like this:
4387 => >A.N.Other <ano@some.domain> <original@ddress> ...
4389 This has been changed so that it extracts the operative address. There
4390 may be more than one such address. If so, they are comma-separated, like
4393 => >ano@some.domain,ona@other.domain <original@ddress> ...
4395 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4396 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4397 not being looked up in the rDNS to get a host name, Exim was showing the
4398 IP address twice in Received: lines, even though the IP addresses were
4399 identical. For example:
4401 Received: from [1.2.3.4] (helo=[1.2.3.4])
4403 However, if the real host name was known, it was omitting the HELO data
4404 if it matched the actual IP address. This has been tidied up so that it
4405 doesn't show the same IP address twice.
4407 PH/07 When both +timestamp and +memory debugging was on, the value given by
4408 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4409 called by the debug printing, thereby overwriting the timestamp buffer.
4410 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4412 PH/08 When the original message was included in an autoreply transport, it
4413 always said "this is a copy of the message, including all the headers",
4414 even if body_only or headers_only was set. It now gives an appropriate
4417 PH/09 Applied a patch from the Sieve maintainer which:
4419 o fixes some comments
4420 o adds the (disabled) notify extension core
4421 o adds some debug output for the result of if/elsif tests
4422 o points to the current vacation draft in the documentation
4423 and documents the missing references header update
4427 o fixes a bug in processing the envelope test (when testing
4428 multiple envelope elements, the last element determined the
4431 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4432 Electronic Mail") by including:
4434 Auto-submitted: auto-generated
4436 in the messages that it generates (bounce messages and others, such as
4437 warnings). In the case of bounce messages for non-SMTP messages, there was
4438 also a typo: it was using "Auto_submitted" (underscore instead of
4439 hyphen). Since every message generated by Exim is necessarily in response
4440 to another message, thes have all been changed to:
4442 Auto-Submitted: auto-replied
4444 in accordance with these statements in the RFC:
4446 The auto-replied keyword:
4448 - SHOULD be used on messages sent in direct response to another
4449 message by an automatic process,
4451 - MUST NOT be used on manually-generated messages,
4453 - MAY be used on Delivery Status Notifications (DSNs) and Message
4454 Disposition Notifications (MDNs),
4456 - MUST NOT be used on messages generated by automatic or periodic
4457 processes, except for messages which are automatic responses to
4460 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4461 to the default Received: header definition.
4463 PH/12 Added log selector acl_warn_skipped (default on).
4465 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4466 variables because (a) they are in the wrong storage pool and (b) even if
4467 they were copied, it wouldn't work properly because of the caching.
4469 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4470 checking when decoding. Apparently there are clients that generate
4471 overlong encoded strings. Why am I not surprised?
4473 PH/15 If the first argument of "${if match_address" was not empty, but did not
4474 contain an "@" character, Exim crashed. Now it writes a panic log message
4475 and treats the condition as false.
4477 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4479 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4480 extension "envelope-auth". The code is finished and in agreement with
4481 other implementations, but there is no documentation so far and in fact,
4482 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4483 not changing the active code.
4485 Print executed "if" and "elsif" statements when debugging is used. This
4486 helps a great deal to understand what a filter does.
4488 Document more things not specified clearly in RFC3028. I had all this
4489 sorted out, when out of a sudden new issues came to my mind. Oops."
4491 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4494 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4495 canonical form (as documented). However, after a host name lookup from
4496 the IP address, check_host() was doing a simple string comparison with
4497 addresses acquired from the DNS when checking that the found name did
4498 have the original IP as one of its addresses. Since any found IPv6
4499 addresses are likely to be in abbreviated form, the comparison could
4500 fail. Luckily, there already exists a function for doing the comparison
4501 by converting both addresses to binary, so now that is used instead of
4502 the text comparison.
4504 PH/20 There was another similar case to PH/19, when a complete host name was
4505 given in a host list; looking up its IP address could give an abbreviated
4506 form, whereas the current host's name might or might not be abbreviated.
4507 The same fix has been applied.
4513 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4514 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4517 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4518 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4520 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4522 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4523 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4524 Also, for "Auto-Submitted", treat anything other than "no" as
4525 non-personal, in accordance with RFC 3834. (Previously it treated
4526 anything starting "auto-" as non-personal.)
4528 TF/01 The control=submission/name=... option had a problem with syntax
4529 errors if the name included a slash character. The /name= option
4530 now slurps the rest of the string, so it can include any characters
4531 but it must come last in the list of options (after /sender_retain
4534 PH/05 Some modifications to the interface to the fake nameserver for the new
4542 TK/01 Added the "success_on_redirect" address verification option. See
4543 NewStuff for rationale and an example.
4545 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4547 PH/02 Patch to exigrep to allow it to work on syslog lines.
4549 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4550 fread() to skip over the body file's header line, because in Cygwin the
4551 header line is locked and is inaccessible.
4553 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4554 co-exist for some time) to make it clear that it is the Exim ID that is
4555 referenced, not the Message-ID: header line.
4557 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4558 string_format() function, because snprintf() does not exist on all
4561 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4562 ${stat: expansion item.
4564 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4565 protocol synchronization error", to keep the pedants happy.
4567 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4568 well as for IRIX systems, when gcc is being used. See the host.c source
4571 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4573 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4576 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4577 transport and to the smtp transport in LMTP mode.
4579 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4581 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4582 run a filter in a subprocess. This could lead to confusion in subsequent
4583 lookups in the parent process. There should also be a search_tidyup() at
4584 the end of the subprocess.
4586 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4587 only if the host matched helo_try_verify_hosts, which caused the
4588 verification to occur when the EHLO/HELO command was issued. The ACL just
4589 tested the remembered result. Now, if a previous verification attempt has
4590 not happened, "verify = helo" does it there and then.
4592 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4594 TK/03 Fix log output including CR from clamd.
4596 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4597 exist provoked a memory error which could cause a segfault.
4599 PH/15 Installed PCRE 6.2
4601 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4603 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4604 of the problem. Specifically, suggested +O2 rather than +O1 for the
4607 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4609 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4610 record, the port was not correctly logged when the outgoing_port log
4611 selector was set (it logged the transort's default port).
4613 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4614 fallback_hosts, and "hosts" in the smtp transport.
4616 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4617 host errors such as "Connection refused".
4619 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4620 authentication with radiusclient 0.4.9:
4622 - Error returned from rc_read_config was caught wrongly
4623 - Username/password not passed on to radius server due to wrong length.
4625 The presumption is that some radiusclient API changes for 4.51/PH/17
4626 were not taken care of correctly. The code is still untested by me (my
4627 Linux distribution still has 0.3.2 of radiusclient), but it was
4628 contributed by a Radius user.
4630 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4631 expanding the "port" option of the smtp transport.
4633 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4634 while reading a MIME header. Thanks to Tom Hughes for a patch.
4636 PH/24 Include config.h inside local_scan.h so that configuration settings are
4639 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4640 that in an ACL for RCPT (for example), you can examine exactly what was
4643 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4644 commands, but it was not correctly comparing the address with the actual
4645 client host address. Thus, it would show the EHLO address in Received:
4646 header lines when this was not necessary.
4648 PH/27 Added the % operator to ${eval:}.
4650 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4651 it should be ignoring failures (because with -C, for example, it has lost
4652 privilege). It wasn't ignoring creation failures other than "already
4655 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4658 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4659 because an input line was too long, either on its own, or by virtue of
4660 too many continuations, the temporary file was not being removed, and the
4661 return code was incorrect.
4663 PH/31 Missing "BOOL" in function definition in filtertest.c.
4665 PH/32 Applied Sieve patches from the maintainer.
4667 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4669 PH/33 Added "verify = not_blind".
4671 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4672 Local/Makefile (with some defaults set). These are used in built scripts
4673 such as exicyclog, but they have never been used in the exim_install
4674 script (though there are many overriding facilities there). I have
4675 arranged that the exim_install script now takes note of these two
4678 PH/35 Installed configuration files for Dragonfly.
4680 PH/36 When a locally submitted message by a trusted user did not contain a
4681 From: header, and the sender address was obtained from -f or from an SMTP
4682 MAIL command, and the trusted user did not use -F to supply a sender
4683 name, $originator_name was incorrectly used when constructing a From:
4684 header. Furthermore, $originator_name was used for submission mode
4685 messages from external hosts without From: headers in a similar way,
4686 which is clearly wrong.
4688 PH/37 Added control=suppress_local_fixups.
4690 PH/38 When log_selector = +received_sender was set, and the addition of the
4691 sender made the log line's construction buffer exactly full, or one byte
4692 less than full, an overflow happened when the terminating "\n" was
4695 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4696 when the result of a list match is failure because a DNS lookup failed.
4698 PH/40 RM_COMMAND is now used in the building process.
4700 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4701 the "build-* directories that it finds.
4703 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4704 address in a domain literal was a prefix of an interface address.
4706 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4707 when verifying a sender address, unless rewrite_headers is false.
4709 PH/44 Wrote a long comment about why errors_to addresses are verified as
4710 recipients, not senders.
4712 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4713 the ratelimit ACL was added.
4715 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4717 PH/46 Added extra information about PostgreSQL errors to the error string.
4719 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4720 avoiding the need to install special zones in a real server. This is
4721 backwards compatible; if it can't find the fake resolver, it drops back.
4722 Thus, both old and new test suites can be run.
4724 TF/02 Added util/ratelimit.pl
4726 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4727 clock is set back in time.
4729 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4730 Candler <B.Candler@pobox.com>.
4732 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4733 OK for addresses that are the result of redirections.
4735 PH/48 A number of further additions for the benefit of the new test suite,
4736 including a fake gethostbyname() that interfaces to the fake DNS resolver
4739 TF/06 The fix for widen_domains has also been applied to qualify_single and
4740 search_parents which are the other dnslookup options that can cause
4743 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4746 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4748 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4749 service" if it encountered an MX record with an empty target hostname.
4750 The message is now "an MX or SRV record indicated no SMTP service".
4752 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4753 if the DNS of the sending site is misconfigured. This is quite a
4754 common situation. This change restores the behaviour of treating a
4755 helo verification defer as a failure.
4757 PH/51 If self=fail was set on a router, the bounce message did not include the
4758 actual error message.
4764 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4766 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4767 command itself ended in error, the underlying message about the transport
4768 filter timeout was being overwritten with the pipe command error. Now the
4769 underlying error message should be appended to the second error message.
4771 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4773 PH/02 Reduce the amount of output that "make" produces by default. Full output
4774 can still be requested.
4776 PH/03 The warning log line about a condition test deferring for a "warn" verb
4777 was being output only once per connection, rather than after each
4778 occurrence (because it was using the same function as for successful
4779 "warn" verbs). This seems wrong, so I have changed it.
4781 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4782 it should not have, which might have caused a crash in the right
4783 circumstances, but probably never did.
4785 PH/04 Installed a modified version of Tony Finch's patch to make submission
4786 mode fix the return path as well as the Sender: header line, and to
4787 add a /name= option so that you can make the user's friendly name appear
4790 TF/03 Added the control = fakedefer ACL modifier.
4792 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4793 Mark Lowes for thorough testing.
4795 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4797 TK/03 Merged latest SRS patch from Miles Wilton.
4799 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4800 with the definition in sysexits.h (which is #included earlier).
4801 Fortunately, Exim does not actually use EX_OK. The code used to try to
4802 preserve the sysexits.h value, by assuming that macro definitions were
4803 scanned for macro replacements. I have been disabused of this notion,
4804 so now the code just undefines EX_OK before #including unistd.h.
4806 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4807 in the smtp transport. When a block could not be written in a single
4808 write() function, the timeout was being re-applied to each part-write.
4809 This seems wrong - if the receiver was accepting one byte at a time it
4810 would take for ever. The timeout is now adjusted when this happens. It
4811 doesn't have to be particularly precise.
4813 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4814 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4816 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4817 without a domain if the check to <postmaster@domain> fails.
4819 SC/01 Eximstats: added -xls and the ability to specify output files
4820 (patch written by Frank Heydlauf).
4822 SC/02 Eximstats: use FileHandles for outputting results.
4824 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4826 SC/04 Eximstats: fixed display of large numbers with -nvr option
4828 SC/05 Eximstats: fixed merging of reports with empty tables.
4830 SC/06 Eximstats: added the -include_original_destination flag
4832 SC/07 Eximstats: removed tabs and trailing whitespace.
4834 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4836 TK/06 MBOX spool code: Add real "From " MBOX separator line
4837 so the .eml file is really in mbox format (even though
4838 most programs do not really care). Patch from Alex Miller.
4840 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4841 The latter is generated from $received_to and is only set if the
4842 message has one envelope recipient. SA can use these headers,
4843 obviously out-of-the-box. Patch from Alex Miller.
4845 PH/08 The ${def test on a variable was returning false if the variable's
4846 value was "0", contrary to what the specification has always said!
4847 The result should be true unless the variable is empty.
4849 PH/09 The syntax error of a character other than { following "${if
4850 def:variable_name" (after optional whitespace) was not being diagnosed.
4851 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4852 accidental colon was present, for example, could give incorrect results.
4854 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4855 result is used (not including appendfile, where other changes are about
4858 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4859 This involved changing a lot of size variables from int to off_t. It
4860 should work with maildirs and everything.
4862 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4863 spamd dying while we are connected to it.
4865 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4868 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4869 function for BDB 4.3.
4871 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4873 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4874 the sources for PCRE have been re-arranged and more files are now
4877 PH/15 The code I had for printing potentially long long variables in PH/11
4878 above was not the best (it lost precision). The length of off_t variables
4879 is now inspected at build time, and an appropriate printing format (%ld
4880 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4881 to be "long long int" or "long int". This is needed for the internal
4882 formatting function string_vformat().
4884 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4885 the configuration file to be ":syslog", then the script "guesses" where
4886 the logs files are, rather than using the compiled in default. In our
4887 case the guess is not the same as the compiled default, so the script
4888 suddenly stopped working when I started to use syslog. The patch checks
4889 to see if log_file_path is "". If so, it attempts to read it from exim
4890 with no configuration file to get the compiled in version, before it
4891 falls back to the previous guessing code."
4893 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4894 implementing BATV in an Exim configuration. See NewStuff for the gory
4897 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4898 Makefile that are specific to HP-UX.
4900 PH/18 If the "use_postmaster" option was set for a recipient callout together
4901 with the "random" option, the postmaster address was used as the MAIL
4902 FROM address for the random test, but not for the subsequent recipient
4903 test. It is now used for both.
4905 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4906 patch removes a few documentation additions to RFC 3028, because the
4907 latest draft now contains them. It adds the new en;ascii-case comparator
4908 and a new error check for 8bit text in MIME parts. Comparator and
4909 require names are now matched exactly. I enabled the subaddress
4910 extension, but it is not well tested yet (read: it works for me)."
4912 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4913 rework some of the code of TK/09 above to avoid the hardwired use of
4914 "%lld" and "long long". Replaced the call to snprintf() with a call to
4917 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4918 records point to non-existent hosts", "retry timeout exceeded", and
4919 "retry time not reached for any host after a long failure period".
4921 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4922 experimental DomainKeys support:
4924 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4925 (2) On an error such as an illegally used "control", the wrong name for
4926 the control was given.
4928 These problems did NOT occur unless DomainKeys support was compiled.
4930 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4932 PH/24 Added ${if match_ip condition.
4934 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4935 databases so that it will be absolutely obvious if a crash occurs in the
4936 DB library. This is a regular occurrence (often caused by mis-matched
4939 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4940 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4941 picked up on a user's system that detects such things. There doesn't seem
4942 to be a gcc warning option for this - only an attribute that has to be
4943 put on the function's prototype. It seems that in Fedora Core 4 they have
4944 set this on a number of new functions. No doubt there will be more in due
4947 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4948 specify a transport. However, if an address that was verified by such a
4949 router was the subject of a callout, Exim crashed because it tried to
4950 read the rcpt_include_affixes from the non-existent transport. Now it
4951 just assumes that the setting of that option is false. This bug was
4952 introduced by 4.51/PH/31.
4954 PH/28 Changed -d+all to exclude +memory, because that information is very
4955 rarely of interest, but it makes the output a lot bigger. People tend to
4956 do -d+all out of habit.
4958 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4959 code in os-type was giving problems when libc.so lives in lib64, like on
4962 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4963 aren't the modern standard, and it seems that some systems' include files
4964 don't always have them. Exim was already checking for some of the newer
4965 ones like T_AAAA, and defining it itself. I've added checks for all the
4966 record types that Exim uses.
4968 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4969 not automatically generating a new one, as it is supposed to. This
4970 prevented TLS from working. If the file did exist, but contained invalid
4971 data, a new version was generated, as expected. It was only the case of a
4972 non-existent file that was broken.
4974 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4975 with a change in libdomainkeys > 0.64.
4977 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4978 from DNS. If the selector record carries the flag, it now has
4979 precedence over the domain-wide flag.
4981 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4983 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4984 the use of a port name that isn't defined in /etc/services) occurred, the
4985 message was deferred as in a normal delivery, and thus remained on the
4986 spool, instead of being failed because of the mua_wrapper setting. This
4987 is now fixed, and I tidied up some of the mua_wrapper messages at the
4990 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4991 the messages in a hash of arrays rather than using individual hashes.
4992 This is a bit cleaner and results in dramatic memory savings, albeit
4993 at a slight CPU cost.
4995 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4996 as requested by Marc Sherman.
4998 SC/10 Eximstats: added histograms for user specified patterns as requested
5001 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5003 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5004 fopen() in the content-scanning modules that did not already have it.
5010 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5011 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5013 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5015 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5017 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5018 to test Sieve filters that use "vacation".
5020 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5021 that changes the way the GnuTLS parameters are stored in the cache file.
5022 The new format can be generated externally. For backward compatibility,
5023 if the data in the cache doesn't make sense, Exim assumes it has read an
5024 old-format file, and it generates new data and writes a new file. This
5025 means that you can't go back to an older release without removing the
5028 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5029 work if there are any delivery delays because "one_time" forces the
5030 parent to be marked "delivered", so its unseen clone is never tried
5031 again. For this reason, Exim now forbids the simultaneous setting of
5034 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5035 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5036 entry if you want to know the details. The fix, however, neglected to
5037 consider the case when local delivery batching is involved. The test for
5038 "previously delivered" was not happening when checking to see if an
5039 address could be batched with a previous (undelivered) one; under
5040 certain circumstances this could lead to multiple deliveries to the same
5043 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5044 in its include files, and this causes problems building Exim.
5046 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5047 header_syntax) but Exim was just ignoring anything given after a slash.
5048 In particular, this caused confusion with an attempt to use "verify =
5049 reverse_host_lookup/defer_ok". An error is now given when options are
5050 supplied for verify items that do not have them. (Maybe reverse_host_
5051 lookup should have a defer_ok option, but that's a different point.)
5053 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5054 defined by RFC 821) to 2048, because there were problems with some AUTH
5055 commands, and RFC 1869 says the size should be increased for extended
5056 SMTP commands that take arguments.
5058 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5061 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5062 "unknown" error; now it says that the functionality isn't in the binary.
5064 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5065 an address' error message when a string expansion fails (syntax or
5066 whatever). Otherwise the password may appear in the log. Following change
5067 PH/42 below, there is no longer a chance of it appearing in a bounce
5070 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5072 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5073 of its addresses was ever tried. (Bugzilla bug #2.)
5075 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5076 the result incorrectly in the debug output. (It correctly added a newline
5077 to what was transported.)
5079 TF/01 Added $received_time.
5081 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5082 commented out examples of how to interface to a virus scanner and to
5083 SpamAssassin. Also added commented examples of av_scanner and
5084 spamd_address settings.
5086 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5087 and controls are allowed in which ACLs. There were a couple of minor
5088 errors. Some of the entries in the conditions table (which is a table of
5089 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5090 negation of where the condition IS allowed.
5092 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5094 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5095 header file does not have a version number, so I've had to invent a new
5096 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5097 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5098 radiusclient), but it was contributed by a Radius user.
5100 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5101 files or directories, for OpenSSL.
5103 PH/19 When an Exim process that is running as root has to create an Exim log
5104 file, it does so in a subprocess that runs as exim:exim so as to get the
5105 ownership right at creation (otherwise, other Exim processes might see
5106 the file with the wrong ownership). There was no test for failure of this
5107 fork() call, which would lead to the process getting stuck as it waited
5108 for a non-existent subprocess. Forks do occasionally fail when resources
5109 run out. I reviewed all the other calls to fork(); they all seem to check
5112 PH/20 When checking for unexpected SMTP input at connect time (before writing
5113 the banner), Exim was not dealing correctly with a non-positive return
5114 from the read() function. If the client had disconnected by this time,
5115 the result was a log entry for a synchronization error with an empty
5116 string after "input=" when read() returned zero. If read() returned -1
5117 (an event I could not check), uninitialized data bytes were printed.
5118 There were reports of junk text (parts of files, etc) appearing after
5121 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5123 PH/22 Added support for macro redefinition, and (re)definition in between
5124 driver and ACL definitions.
5126 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5127 forgetting to use the resulting value; it was using the unexpanded value.
5129 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5130 hadn't been configured. The fix is from Juergen Kreileder, who
5131 understands it better than I do:
5133 "Here's what I see happening with three configured cyrus_sasl
5134 authenticators configured (plain, login, cram-md5):
5136 On startup auth_cyrus_sasl_init() gets called for each of these.
5137 This means three calls to sasl_listmech() without a specified mech_list.
5138 => SASL tests which mechs of all available mechs actually work
5139 => three warnings about OTP not working
5140 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5142 With the patch, sasl_listmech() also gets called three times. But now
5143 SASL's mech_list option is set to the server_mech specified in the the
5144 authenticator. Or in other words, the answer from sasl_listmech()
5145 gets limited to just the mech you're testing for (which is different
5147 => the return list contains just 'plain' or 'login', 'cram-md5' or
5148 nothing depending on the value of ob->server_mech.
5150 I've just tested the patch: Authentication still works fine,
5151 unavailable mechs specified in the exim configuration are still
5152 caught, and the auth.log warnings about OTP are gone."
5154 PH/25 When debugging is enabled, the contents of the command line are added
5155 to the debugging output, even when log_selector=+arguments is not
5158 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5159 answer is "GNU", and only if the return is "GNU/something" is the answer
5162 PH/27 $acl_verify_message is now set immediately after the failure of a
5163 verification in an ACL, and so is available in subsequent modifiers. In
5164 particular, the message can be preserved by coding like this:
5166 warn !verify = sender
5167 set acl_m0 = $acl_verify_message
5169 Previously, $acl_verify_message was set only while expanding "message"
5170 and "log_message" when a very denied access.
5172 PH/28 Modified OS/os.c-Linux with
5174 -#ifndef OS_LOAD_AVERAGE
5175 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5177 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5178 nomenclature these days.)
5180 PH/29 Installed patch from the Sieve maintainer that adds the options
5181 sieve_useraddress and sieve_subaddress to the redirect router.
5183 PH/30 In these circumstances:
5184 . Two addresses routed to the same list of hosts;
5185 . First host does not offer TLS;
5186 . First host accepts first address;
5187 . First host gives temporary error to second address;
5188 . Second host offers TLS and a TLS session is established;
5189 . Second host accepts second address.
5190 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5191 and peerdn, if requested) that were in fact used only for the second
5194 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5195 attention to any local part prefix or suffix that was matched by the
5196 router that accepted the address. It now behaves in the same way as it
5197 does for delivery: the affixes are removed from the local part unless
5198 rcpt_include_affixes is set on the transport.
5200 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5201 timeout during the DATA phase of an incoming message.
5203 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5204 applied a patch sanctioned by the Sieve maintainer.
5206 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5207 the uid or gid is negative. A case of a negative gid caused this to be
5208 noticed. The fix allows for either to be negative.
5210 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5211 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5214 PH/36 The change PH/12 above was broken. Fixed it.
5216 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5217 the grounds that routing the same address twice would always produce the
5218 same answer. This might have been true once, but it is certainly no
5219 longer true now. Routing a child address may depend on the previous
5220 routing that produced that child. Some complicated redirection strategies
5221 went wrong when messages had multiple recipients, and made Exim's
5222 behaviour dependent on the order in which the addresses were given.
5224 I have moved the duplicate checking until after the routing is complete.
5225 Exim scans the addresses that are assigned to local and remote
5226 transports, and removes any duplicates. This means that more work will be
5227 done, as duplicates will always all be routed, but duplicates are
5228 presumably rare, so I don't expect this is of any significance.
5230 For deliveries to pipes, files, and autoreplies, the duplicate checking
5231 still happens during the routing process, since they are not going to be
5234 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5235 It corrects a timeout issue with spamd. This is Ian's comment: "The
5236 background is that sometimes spamd either never reads data from a
5237 connection it has accepted, or it never writes response data. The exiscan
5238 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5239 blindly assumes that writes won't block so it may never time out."
5241 PH/39 Allow G after quota size as well as K and M.
5243 PH/40 The value set for $authenticated_id in an authenticator may not contain
5244 binary zeroes or newlines because the value is written to log lines and
5245 to spool files. There was no check on this. Now the value is run through
5246 the string_printing() function so that such characters are converted to
5247 printable escape sequences.
5249 PH/41 $message_linecount is a new variable that contains the total number of
5250 lines in the message. Compare $body_linecount, which is the count for the
5253 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5254 bounce and delay warning messages, except in certain special cases, which
5257 (a) An SMTP error message from a remote host;
5258 (b) A message specified in a :fail: redirection;
5259 (c) A message specified in a "fail" command in a system filter;
5260 (d) A message specified in a FAIL return from the queryprogram router;
5261 (e) A message specified by the cannot_route_message router option.
5263 In these cases only, Exim does include the error details in bounce and
5264 warning messages. There are also a few cases where bland messages such
5265 as "unrouteable address" or "local delivery error" are given.
5267 PH/43 $value is now also set for the "else" part of a ${run expansion.
5269 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5270 being worked on, but at least Exim now implements the latest version to
5273 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5274 process to complete was treated as a delivery failure, a timeout while
5275 writing the message to the pipe was logged, but erroneously treated as a
5276 successful delivery. Such timeouts include transport filter timeouts. For
5277 consistency with the overall process timeout, these timeouts are now
5278 treated as errors, giving rise to delivery failures by default. However,
5279 there is now a new Boolean option for the pipe transport called
5280 timeout_defer, which, if set TRUE, converts the failures into defers for
5281 both kinds of timeout. A transport filter timeout is now identified in
5284 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5285 systems where "make" and "gmake" are different, calling "gmake" at top
5286 level broke things. I've arranged for the value of $(MAKE) to be passed
5287 from the Makefile to this script so that it can call the same version of
5291 A note about Exim versions 4.44 and 4.50
5292 ----------------------------------------
5294 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5295 changes of various kinds. As a consequence, a big documentation update was
5296 needed. This delayed the release for rather longer than seemed good, especially
5297 in the light of a couple of (minor) security issues. Therefore, the changes
5298 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5299 release. So 4.44 and 4.50 are in effect two different branches that both start
5302 I have left the 4.50 change log unchanged; it contains all the changes since
5303 4.43. The change log for 4.44 is below; many of its items are identical to
5304 those for 4.50. This seems to be the most sensible way to preserve the
5305 historical information.
5311 1. Minor wording change to the doc/README.SIEVE file.
5313 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5314 computation of the current number of files was incorrect.
5316 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5317 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5320 4. Give more explanation in the error message when the command for a transport
5321 filter fails to execute.
5323 5. There are several places where Exim runs a non-Exim command in a
5324 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5325 was being done only for the command run by the queryprogram router. It is
5326 now done for all such subprocesses. The other cases are: ${run, transport
5327 filters, and the commands run by the lmtp and pipe transports.
5329 6. Added CONFIGURE_GROUP build-time option.
5331 7. Some older OS have a limit of 256 on the maximum number of file
5332 descriptors. Exim was using setrlimit() to set 1000 as a large value
5333 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5334 systems. I've change it so that if it can't get 1000, it tries for 256.
5336 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5337 was an oversight, and furthermore, ever since the addition of extra
5338 controls (e.g. 4.43/32), the checks on when to allow different forms of
5339 "control" were broken. There should now be diagnostics for all cases when a
5340 control that does not make sense is encountered.
5342 9. Added the /retain_sender option to "control=submission".
5344 10. $recipients is now available in the predata ACL (oversight).
5346 11. Tidy the search cache before the fork to do a delivery from a message
5347 received from the command line. Otherwise the child will trigger a lookup
5348 failure and thereby defer the delivery if it tries to use (for example) a
5349 cached ldap connection that the parent has called unbind on.
5351 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5352 of $address_data from the recipient verification was clobbered by the
5353 sender verification.
5355 13. The value of address_data from a sender verification is now available in
5356 $sender_address_data in subsequent conditions in the ACL statement.
5358 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5360 15. Added a new option "connect=<time>" to callout options, to set a different
5363 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5364 was its contents. (It was OK if the option was not defined at all.)
5366 17. A "Completed" log line is now written for messages that are removed from
5367 the spool by the -Mrm option.
5369 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5370 information about exactly what failed.
5372 19. Added -dd to debug only the daemon process.
5374 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5375 handles timeouts, both on the server side and network timeouts. Renamed the
5376 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5378 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5379 It is now set to "smtps".
5381 22. $host_address is now set to the target address during the checking of
5382 ignore_target_hosts.
5384 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5385 being passed; this would have caused $sender_host_name to have been used if
5386 matching the list had actually called for a host name (not very likely,
5387 since this list is usually IP addresses). A host name is now passed as
5390 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5391 code that specifies a non-restarting handler (typically sigaction() in
5392 modern systems) in an attempt to fix a rare and obscure crash bug.
5394 25. Narrowed the window for a race in the daemon that could cause it to ignore
5395 SIGCHLD signals. This is not a major problem, because they are used only to
5396 wake it up if nothing else does.
5398 26. A malformed maildirsize file could cause Exim to calculate negative values
5399 for the mailbox size or file count. Odd effects could occur as a result.
5400 The maildirsize information is now recalculated if the size or filecount
5403 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5404 support for a long time. Removed HAVE_SYS_VFS_H.
5406 28. Installed the latest version of exipick from John Jetmore.
5408 29. In an address list, if the pattern was not a regular expression, an empty
5409 subject address (from a bounce message) matched only if the pattern was an
5410 empty string. Non-empty patterns were not even tested. This was the wrong
5411 because it is perfectly reasonable to use an empty address as part of a
5412 database query. An empty address is now tested by patterns that are
5413 lookups. However, all the other forms of pattern expect the subject to
5414 contain a local part and a domain, and therefore, for them, an empty
5415 address still always fails if the pattern is not itself empty.
5417 30. Exim went into a mad DNS loop when attempting to do a callout where the
5418 host was specified on an smtp transport, and looking it up yielded more
5419 than one IP address.
5421 31. Re-factored the code for checking spool and log partition space into a
5422 function that finds that data and another that does the check. The former
5423 is then used to implement four new variables: $spool_space, $log_space,
5424 $spool_inodes, and $log_inodes.
5426 32. The RFC2047 encoding function was originally intended for short strings
5427 such as real names; it was not keeping to the 75-character limit for
5428 encoded words that the RFC imposes. It now respects the limit, and
5429 generates multiple encoded words if necessary. To be on the safe side, I
5430 have increased the buffer size for the ${rfc2047: expansion operator from
5433 33. It is now permitted to omit both strings after an "if" condition; if the
5434 condition is true, the result is "true". As before, when the second string
5435 is omitted, a false condition yields an empty string. This makes it less
5436 cumbersome to write custom ACL and router conditions.
5438 34. Failure to deliver a bounce message always caused it to be frozen, even if
5439 there was an errors_to setting on the router. The errors_to setting is now
5442 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5443 canonical form (fully expanded) before being placed in
5444 $sender_host_address.
5446 36. The table in the code that translates DNS record types into text (T_A to
5447 "A" for instance) was missing entries for NS and CNAME. It is just possible
5448 that this could have caused confusion if both these types were looked up
5449 for the same domain, because the text type is used as part of Exim's
5450 per-process caching. But the chance of anyone hitting this buglet seems
5453 37. The dnsdb lookup has been extended in a number of ways.
5455 (1) There is a new type, "zns", which walks up the domain tree until it
5456 finds some nameserver records. It should be used with care.
5458 (2) There is a new type, "mxh", which is like "mx" except that it returns
5459 just the host names, not the priorities.
5461 (3) It is now possible to give a list of domains (or IP addresses) to be
5462 looked up. The behaviour when one of the lookups defers can be
5463 controlled by a keyword.
5465 (4) It is now possible to specify the separator character for use when
5466 multiple records are returned.
5468 38. The dnslists ACL condition has been extended: it is now possible to supply
5469 a list of IP addresses and/or domains to be looked up in a particular DNS
5472 39. Added log_selector=+queue_time_overall.
5474 40. When running the queue in the test harness, wait just a tad after forking a
5475 delivery process, to get repeatability of debugging output.
5477 41. Include certificate and key file names in error message when GnuTLS fails
5478 to set them up, because the GnuTLS error message doesn't include the name
5479 of the failing file when there is a problem reading it.
5481 42. Allow both -bf and -bF in the same test run.
5483 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5485 44. The "Exiscan patch" is now merged into the mainline Exim source.
5487 45. Sometimes the final signoff response after QUIT could fail to get
5488 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5489 < 0 before doing a fflush(). This bug looks as though it goes back to the
5490 introduction of TLS in release 3.20, but "sometimes" must have been rare
5491 because the tests only now provoked it.
5493 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5494 (this can affect the format of dates).
5496 47. exim_tidydb, when checking for the continued existence of a message for
5497 which it has found a message-specific retry record, was not finding
5498 messages that were in split spool directories. Consequently, it was
5499 deleting retry records that should have stayed in existence.
5501 48. Steve fixed some bugs in eximstats.
5503 49. The SPA authentication driver was not abandoning authentication and moving
5504 on to the next authenticator when an expansion was forced to fail,
5505 contradicting the general specification for all authenticators. Instead it
5506 was generating a temporary error. It now behaves as specified.
5508 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5509 (the order specifies the preference for clients). The order is now AES256,
5510 AES128, 3DES, ARCFOUR128.
5512 51. Small patch to Sieve code - explicitly set From: when generating an
5515 52. Exim crashed if a remote delivery caused a very long error message to be
5516 recorded - for instance if somebody sent an entire SpamAssassin report back
5517 as a large number of 550 error lines. This bug was coincidentally fixed by
5518 increasing the size of one of Exim's internal buffers (big_buffer) that
5519 happened as part of the Exiscan merge. However, to be on the safe side, I
5520 have made the code more robust (and fixed the comments that describe what
5523 53. Now that there can be additional text after "Completed" in log lines (if
5524 the queue_time_overall log selector is set), a one-byte patch to exigrep
5525 was needed to allow it to recognize "Completed" as not the last thing in
5528 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5529 patch that reportedly fixes this has been added. I am not expert enough to
5530 create a test for it. This is what the patch creator wrote:
5532 "I found a little strange behaviour of ldap code when working with
5533 Windows 2003 AD Domain, where users was placed in more than one
5534 Organization Units. When I tried to give exim partial DN, the exit code
5535 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5536 But simultaneously result of request was absolutely normal ldap result,
5537 so I produce this patch..."
5539 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5540 REFERENCE, so I have modified the code to exclude the patch when that macro
5543 55. Some experimental protocols are using DNS PTR records for new purposes. The
5544 keys for these records are domain names, not reversed IP addresses. The
5545 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5546 leaves it alone. Component reversal etc. now happens only for IP addresses.
5549 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5551 57. Double the size of the debug message buffer (to 2048) so that more of very
5552 long debug lines gets shown.
5554 58. The exicyclog utility now does better if the number of log files to keep
5555 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5557 59. Two changes related to the smtp_active_hostname option:
5559 (1) $smtp_active_hostname is now available as a variable.
5560 (2) The default for smtp_banner uses $smtp_active_hostname instead
5561 of $primary_hostname.
5563 60. The host_aton() function is supposed to be passed a string that is known
5564 to be a valid IP address. However, in the case of IPv6 addresses, it was
5565 not checking this. This is a hostage to fortune. Exim now panics and dies
5566 if the condition is not met. A case was found where this could be provoked
5567 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5568 components; fortuitously, this particular loophole had already been fixed
5569 by change 4.50/55 above.
5571 If there are any other similar loopholes, the new check in host_aton()
5572 itself should stop them being exploited. The report I received stated that
5573 data on the command line could provoke the exploit when Exim was running as
5574 exim, but did not say which command line option was involved. All I could
5575 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5576 running as the user.
5579 61. There was a buffer overflow vulnerability in the SPA authentication code
5580 (which came originally from the Samba project). I have added a test to the
5581 spa_base64_to_bits() function which I hope fixes it.
5584 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5585 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5587 63. The daemon start-up calls getloadavg() while still root for those OS that
5588 need the first call to be done as root, but it missed one case: when
5589 deliver_queue_load_max is set with deliver_drop_privilege. This is
5590 necessary for the benefit of the queue runner, because there is no re-exec
5591 when deliver_drop_privilege is set.
5593 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5594 This has been fixed.
5596 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5597 was in use, was not putting the data itself into the right store pool;
5598 consequently, it could be overwritten for a subsequent message in the same
5599 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5602 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5604 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5605 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5606 the function were treating the return as a boolean value, which happened to
5607 work because 0=false and not-0=true, but is not correct code.
5609 68. The host_aton() function was not handling scoped IPv6 addresses (those
5610 with, for example, "%eth0" on the end) correctly.
5612 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5613 negated items (that is, ~something) in unsigned ints. Some compilers
5614 apparently mutter when there is no cast.
5616 70. If an address verification called from an ACL failed, and did not produce a
5617 user-specific message (i.e. there was only a "system" message), nothing was
5618 put in $acl_verify_message. In this situation, it now puts the system
5621 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5622 session; change 4.31/43 added the unwanted input to the log line - except
5623 that it did not do this in the start of session case. It now does.
5625 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5626 This is wrong and can cause the other end to generate a synchronization
5627 error if it is another Exim or anything else that does the synchronization
5628 check. A QUIT command is no longer sent after a timeout.
5630 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5631 during host lookups.
5633 74. The defer_ok option of callout verification was not working if it was used
5634 when verifying addresses in header lines, that is, for this case:
5636 verify = header_sender/callout=defer_ok
5638 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5639 those file descriptors could be used for SMTP connections. If anything
5640 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5641 could be sent to the SMTP client, causing chaos. The daemon now opens
5642 stdin, stdout, and stderr to /dev/null when it puts itself into the
5645 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5646 log by default. The user can override this with suitable Perl magic.
5648 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5649 the log message when discard triggers, was not working for the DATA ACL or
5650 for the non-SMTP ACL.
5652 78. Error message wording change in sieve.c.
5654 79. If smtp_accept_max_per_host was set, the number of connections could be
5655 restricted to fewer than expected, because the daemon was trying to set up
5656 a new connection before checking whether the processes handling previous
5657 connections had finished. The check for completed processes is now done
5658 earlier. On busy systems, this bug wouldn't be noticed because something
5659 else would have woken the daemon, and it would have reaped the completed
5662 80. If a message was submitted locally by a user whose login name contained one
5663 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5664 It caused a spool format error. I have fixed the spool reading code. A
5665 related problem was that the "from" clause in the Received: line became
5666 illegal because of the space(s). It is now covered by ${quote_local_part.
5668 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5671 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5673 83. Patch from Sieve maintainer to fix unterminated string problem in
5674 "vacation" handling.
5676 84. Some minor changes to the Linux configuration files to help with other
5677 OS variants using glibc.
5679 85. One more patch for Sieve to update vacation handling to latest spec.
5682 ----------------------------------------------------
5683 See the note above about the 4.44 and 4.50 releases.
5684 ----------------------------------------------------
5690 1. Change 4.43/35 introduced a bug that caused file counts to be
5691 incorrectly computed when quota_filecount was set in an appendfile
5694 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5695 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5698 3. Give more explanation in the error message when the command for a transport
5699 filter fails to execute.
5701 4. There are several places where Exim runs a non-Exim command in a
5702 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5703 was being done only for the command run by the queryprogram router. It is
5704 now done for all such subprocesses. The other cases are: ${run, transport
5705 filters, and the commands run by the lmtp and pipe transports.
5707 5. Some older OS have a limit of 256 on the maximum number of file
5708 descriptors. Exim was using setrlimit() to set 1000 as a large value
5709 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5710 systems. I've change it so that if it can't get 1000, it tries for 256.
5712 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5713 was an oversight, and furthermore, ever since the addition of extra
5714 controls (e.g. 4.43/32), the checks on when to allow different forms of
5715 "control" were broken. There should now be diagnostics for all cases when a
5716 control that does not make sense is encountered.
5718 7. $recipients is now available in the predata ACL (oversight).
5720 8. Tidy the search cache before the fork to do a delivery from a message
5721 received from the command line. Otherwise the child will trigger a lookup
5722 failure and thereby defer the delivery if it tries to use (for example) a
5723 cached ldap connection that the parent has called unbind on.
5725 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5726 of $address_data from the recipient verification was clobbered by the
5727 sender verification.
5729 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5730 was its contents. (It was OK if the option was not defined at all.)
5732 11. A "Completed" log line is now written for messages that are removed from
5733 the spool by the -Mrm option.
5735 12. $host_address is now set to the target address during the checking of
5736 ignore_target_hosts.
5738 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5739 being passed; this would have caused $sender_host_name to have been used if
5740 matching the list had actually called for a host name (not very likely,
5741 since this list is usually IP addresses). A host name is now passed as
5744 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5745 code that specifies a non-restarting handler (typically sigaction() in
5746 modern systems) in an attempt to fix a rare and obscure crash bug.
5748 15. Narrowed the window for a race in the daemon that could cause it to ignore
5749 SIGCHLD signals. This is not a major problem, because they are used only to
5750 wake it up if nothing else does.
5752 16. A malformed maildirsize file could cause Exim to calculate negative values
5753 for the mailbox size or file count. Odd effects could occur as a result.
5754 The maildirsize information is now recalculated if the size or filecount
5757 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5758 support for a long time. Removed HAVE_SYS_VFS_H.
5760 18. Updated exipick to current release from John Jetmore.
5762 19. Allow an empty sender to be matched against a lookup in an address list.
5763 Previously the only cases considered were a regular expression, or an
5766 20. Exim went into a mad DNS lookup loop when doing a callout where the
5767 host was specified on the transport, if the DNS lookup yielded more than
5770 21. The RFC2047 encoding function was originally intended for short strings
5771 such as real names; it was not keeping to the 75-character limit for
5772 encoded words that the RFC imposes. It now respects the limit, and
5773 generates multiple encoded words if necessary. To be on the safe side, I
5774 have increased the buffer size for the ${rfc2047: expansion operator from
5777 22. Failure to deliver a bounce message always caused it to be frozen, even if
5778 there was an errors_to setting on the router. The errors_to setting is now
5781 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5782 canonical form (fully expanded) before being placed in
5783 $sender_host_address.
5785 24. Updated eximstats to version 1.33
5787 25. Include certificate and key file names in error message when GnuTLS fails
5788 to set them up, because the GnuTLS error message doesn't include the name
5789 of the failing file when there is a problem reading it.
5791 26. Expand error message when OpenSSL has problems setting up cert/key files.
5794 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5795 (this can affect the format of dates).
5797 28. exim_tidydb, when checking for the continued existence of a message for
5798 which it has found a message-specific retry record, was not finding
5799 messages that were in split spool directories. Consequently, it was
5800 deleting retry records that should have stayed in existence.
5802 29. eximstats updated to version 1.35
5803 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5804 1.35 - bugfix such that pie charts by volume are generated correctly
5806 30. The SPA authentication driver was not abandoning authentication and moving
5807 on to the next authenticator when an expansion was forced to fail,
5808 contradicting the general specification for all authenticators. Instead it
5809 was generating a temporary error. It now behaves as specified.
5811 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5812 (the order specifies the preference for clients). The order is now AES256,
5813 AES128, 3DES, ARCFOUR128.
5815 31. Small patch to Sieve code - explicitly set From: when generating an
5818 32. Exim crashed if a remote delivery caused a very long error message to be
5819 recorded - for instance if somebody sent an entire SpamAssassin report back
5820 as a large number of 550 error lines. This bug was coincidentally fixed by
5821 increasing the size of one of Exim's internal buffers (big_buffer) that
5822 happened as part of the Exiscan merge. However, to be on the safe side, I
5823 have made the code more robust (and fixed the comments that describe what
5826 33. Some experimental protocols are using DNS PTR records for new purposes. The
5827 keys for these records are domain names, not reversed IP addresses. The
5828 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5829 leaves it alone. Component reversal etc. now happens only for IP addresses.
5832 34. The host_aton() function is supposed to be passed a string that is known
5833 to be a valid IP address. However, in the case of IPv6 addresses, it was
5834 not checking this. This is a hostage to fortune. Exim now panics and dies
5835 if the condition is not met. A case was found where this could be provoked
5836 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5837 components; fortuitously, this particular loophole had already been fixed
5838 by change 4.50/55 or 4.44/33 above.
5840 If there are any other similar loopholes, the new check in host_aton()
5841 itself should stop them being exploited. The report I received stated that
5842 data on the command line could provoke the exploit when Exim was running as
5843 exim, but did not say which command line option was involved. All I could
5844 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5845 running as the user.
5848 35. There was a buffer overflow vulnerability in the SPA authentication code
5849 (which came originally from the Samba project). I have added a test to the
5850 spa_base64_to_bits() function which I hope fixes it.
5853 36. The daemon start-up calls getloadavg() while still root for those OS that
5854 need the first call to be done as root, but it missed one case: when
5855 deliver_queue_load_max is set with deliver_drop_privilege. This is
5856 necessary for the benefit of the queue runner, because there is no re-exec
5857 when deliver_drop_privilege is set.
5859 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5860 was in use, was not putting the data itself into the right store pool;
5861 consequently, it could be overwritten for a subsequent message in the same
5862 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5865 38. Sometimes the final signoff response after QUIT could fail to get
5866 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5867 < 0 before doing a fflush(). This bug looks as though it goes back to the
5868 introduction of TLS in release 3.20, but "sometimes" must have been rare
5869 because the tests only now provoked it.
5875 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5876 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5877 Now it returns an integer. A number of calls were still expecting a T/F
5878 return. Fortuitously, in all cases, the tests worked in OK situations,
5879 which is the norm. However, things would have gone wrong on any write
5880 failures on the smtp file descriptor. This function is used when sending
5881 messages over SMTP and also when doing verify callouts.
5883 2. When Exim is called to do synchronous delivery of a locally submitted
5884 message (the -odf or -odi options), it no longer closes stderr before doing
5887 3. Implemented the mua_wrapper option.
5889 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5891 5. Implemented the functions header_remove(), header_testname(),
5892 header_add_at_position(), and receive_remove_recipient(), and exported them
5895 6. If an ACL "warn" statement specified the addition of headers, Exim already
5896 inserted X-ACL-Warn: at the start if there was no header name. However, it
5897 was not making this test for the second and subsequent header lines if
5898 there were newlines in the string. This meant that an invalid header could
5899 be inserted if Exim was badly configured.
5901 7. Allow an ACL "warn" statement to add header lines at the start or after all
5902 the Received: headers, as well as at the end.
5904 8. Added the rcpt_4xx retry error code.
5906 9. Added postmaster_mailfrom=xxx to callout verification option.
5908 10. Added mailfrom=xxxx to the callout verification option, for verify=
5911 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5912 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5914 12. Inserted some casts to stop certain compilers warning when using pointer
5915 differences as field lengths or precisions in printf-type calls (mostly
5916 affecting debugging statements).
5918 13. Added optional readline() support for -be (dynamically loaded).
5920 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5921 same clock tick as a message's arrival, so that its received time was the
5922 same as the "first fail" time on the retry record, and that message
5923 remained on the queue past the ultimate address timeout, every queue runner
5924 would try a delivery (because it was past the ultimate address timeout) but
5925 after another failure, the ultimate address timeout, which should have then
5926 bounced the address, did not kick in. This was a "< instead of <=" error;
5927 in most cases the first failure would have been in the next clock tick
5928 after the received time, and all would be well.
5930 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5931 being recognized when the domain list was tested by the match_domain
5932 condition in an expansion string.
5934 16. Added the ${str2b64: operator.
5936 17. Exim was always calling setrlimit() to set a large limit for the number of
5937 processes, without checking whether the existing limit was already
5938 adequate. (It did check for the limit on file descriptors.) Furthermore,
5939 errors from getrlimit() and setrlimit() were being ignored. Now they are
5940 logged to the main and panic logs, but Exim does carry on, to try to do its
5941 job under whatever limits there are.
5943 18. Imported PCRE 5.0.
5945 19. Trivial typo in log message " temporarily refused connection" (the leading
5948 20. If the log selector return_path_on_delivery was set and an address was
5949 redirected to /dev/null, the delivery process crashed because it assumed
5950 that a return path would always be set for a "successful" delivery. In this
5951 case, the whole delivery is bypassed as an optimization, and therefore no
5954 21. Internal re-arrangement: the function for sending a challenge and reading
5955 a response while authentication was assuming a zero-terminated challenge
5956 string. It's now changed to take a pointer and a length, to allow for
5957 binary data in such strings.
5959 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5961 23. Exim was not respecting finduser_retries when seeking the login of the
5962 uid under which it was called; it was always trying 10 times. (The default
5963 setting of finduser_retries is zero.) Also, it was sleeping after the final
5964 failure, which is pointless.
5966 24. Implemented tls_on_connect_ports.
5968 25. Implemented acl_smtp_predata.
5970 26. If the domain in control=submission is set empty, Exim assumes that the
5971 authenticated id is a complete email address when it generates From: or
5972 Sender: header lines.
5974 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5975 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5976 chown and chgrp in /bin and hostname in /usr/bin.
5978 28. Exim was keeping the "process log" file open after each use, just as it
5979 does for the main log. This opens the possibility of it remaining open for
5980 long periods when the USR1 signal hits a daemon. Occasional processlog
5981 errors were reported, that could have been caused by this. Anyway, it seems
5982 much more sensible not to leave this file open at all, so that is what now
5985 29. The long-running daemon process does not normally write to the log once it
5986 has entered its main loop, and it closes the log before doing so. This is
5987 so that log files can straightforwardly be renamed and moved. However,
5988 there are a couple of unusual error situations where the daemon does write
5989 log entries, and I had neglected to close the log afterwards.
5991 30. The text of an SMTP error response that was received during a remote
5992 delivery was being truncated at 512 bytes. This is too short for some of
5993 the long messages that one sometimes sees. I've increased the limit to
5996 31. It is now possible to make retry rules that apply only when a message has a
5997 specific sender, in particular, an empty sender.
5999 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6000 it possible to be selective about when SMTP synchronization is enforced.
6002 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6004 32. Implemented hosts_connection_nolog.
6006 33. Added an ACL for QUIT.
6008 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6011 35. Added mailbox_size and mailbox_filecount to appendfile.
6013 36. Added control = no_multiline_responses to ACLs.
6015 37. There was a bug in the logic of the code that waits for the clock to tick
6016 in the case where the clock went backwards by a substantial amount such
6017 that the microsecond fraction of "now" was more than the microsecond
6018 fraction of "then" (but the whole seconds number was less).
6020 38. Added support for the libradius Radius client library this is found on
6021 FreeBSD (previously only the radiusclient library was supported).
6027 1. When certain lookups returned multiple values in the form name=value, the
6028 quoting of the values was not always being done properly. Specifically:
6029 (a) If the value started with a double quote, but contained no whitespace,
6031 (b) If the value contained whitespace other than a space character (i.e.
6032 tabs or newlines or carriage returns) it was not quoted.
6033 This fix has been applied to the mysql and pgsql lookups by writing a
6034 separate quoting function and calling it from the lookup code. The fix
6035 should probably also be applied to nisplus, ibase and oracle lookups, but
6036 since I cannot test any of those, I have not disturbed their existing code.
6038 2. A hit in the callout cache for a specific address caused a log line with no
6039 reason for rejecting RCPT. Now it says "Previous (cached) callout
6040 verification failure".
6042 3. There was an off-by-one bug in the queryprogram router. An over-long
6043 return line was truncated at 256 instead of 255 characters, thereby
6044 overflowing its buffer with the terminating zero. As well as fixing this, I
6045 have increased the buffer size to 1024 (and made a note to document this).
6047 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6048 when Exim is waiting for an SMTP response from a remote server, Exim
6049 restarts its select() call on the socket, thereby resetting its timeout.
6050 This is not a problem when such interrupts are rare. Somebody set up a cron
6051 job to run exiwhat every 2 minutes, which is less than the normal select()
6052 timeout (5 or 10 minutes). This meant that the select() timeout never
6053 kicked in because it was always reset. I have fixed this by comparing the
6054 time when an interrupt arrives with the time at the start of the first call
6055 to select(). If more time than the timeout has elapsed, the interrupt is
6056 treated as a timeout.
6058 5. Some internal re-factoring in preparation for the addition of Sieve
6059 extensions (by MH). In particular, the "personal" test is moved to a
6060 separate function, and given an option for scanning Cc: and Bcc: (which is
6061 not set for Exim filters).
6063 6. When Exim created an email address using the login of the caller as the
6064 local part (e.g. when creating a From: or Sender: header line), it was not
6065 quoting the local part when it contained special characters such as @.
6067 7. Installed new OpenBSD configuration files.
6069 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6070 try to make them clearer.
6072 9. Callout options, other than the timeout value, were being ignored when
6073 verifying sender addresses in header lines. For example, when using
6075 verify = header_sender/callout=no_cache
6077 the cache was (incorrectly) being used.
6079 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6080 only the Cygwin environment.
6082 11. When return_path_on_delivery was set as a log selector, if different remote
6083 addresses in the same message used different return paths and parallel
6084 remote delivery occurred, the wrong values would sometimes be logged.
6085 (Whenever a remote delivery process finished, the return path value from
6086 the most recently started remote delivery process was logged.)
6088 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6089 lines when AUTH and/or TLS are in use. This is the "received protocol"
6090 field. Exim used to use "asmtp" for authenticated SMTP, without any
6091 indication (in the protocol name) for TLS use. Now it follows the RFC and
6092 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6093 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6094 names appear in log lines as well as in Received: header lines.
6096 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6097 extensions, and comparison tests, and to fix some bugs.
6099 14. Changes to the "personal" filter test:
6101 (1) The test was buggy in that it was just doing the equivalent of
6102 "contains" tests on header lines. For example, if a user's address was
6103 anne@some.where, the "personal" test would incorrectly be true for
6105 To: susanne@some.where
6107 This test is now done by extracting each address from the header in turn,
6108 and checking the entire address. Other tests that are part of "personal"
6109 are now done using regular expressions (for example, to check local parts
6110 of addresses in From: header lines).
6112 (2) The list of non-personal local parts in From: addresses has been
6113 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6114 taken from the Sieve specification recommendations.
6116 (3) If the message contains any header line starting with "List-" it is
6117 treated as non-personal.
6119 (4) The test for "circular" in the Subject: header line has been removed
6120 because it now seems ill-conceived.
6122 15. Minor typos in src/EDITME comments corrected.
6124 16. Installed latest exipick from John Jetmore.
6126 17. If headers_add on a router specified a text string that was too long for
6127 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6128 of string_sprintf() is now avoided.
6130 18. $message_body_size was not set (it was always zero) when running the DATA
6131 ACL and the local_scan() function.
6133 19. For the "mail" command in an Exim filter, no default was being set for
6134 the once_repeat time, causing a random time value to be used if "once" was
6135 specified. (If the value happened to be <= 0, no repeat happened.) The
6136 default is now 0s, meaning "never repeat". The "vacation" command was OK
6137 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6138 (I found it when inspecting the code).
6140 20. There is now an overall timeout for performing a callout verification. It
6141 defaults to 4 times the callout timeout, which applies to individual SMTP
6142 commands during the callout. The overall timeout applies when there is more
6143 than one host that can be tried. The timeout is checked before trying the
6144 next host. This prevents very long delays if there are a large number of
6145 hosts and all are timing out (e.g. when the network connections are timing
6146 out). The value of the overall timeout can be changed by specifying an
6147 additional sub-option for "callout", called "maxwait". For example:
6149 verify = sender/callout=5s,maxwait=20s
6151 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6152 to the end before writing, but this should make it even safer).
6154 22. Exim was forgetting that it had advertised PIPELINING for the second and
6155 subsequent messages on an SMTP connection. It was also not resetting its
6156 memory on STARTTLS and an internal HELO.
6158 23. When Exim logs an SMTP synchronization error within a session, it now
6159 records whether PIPELINING has been advertised or not.
6161 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6162 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6163 rather than long int.
6165 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6167 26. Added the never_mail option to autoreply.
6173 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6174 crash if the getsockname() call failed; this can happen if a connection is
6175 closed very soon after it is established. The problem was simply in the
6176 order in which certain operations were done, causing Exim to try to write
6177 to the SMTP stream before it had set up the file descriptor. The bug has
6178 been fixed by making things happen in the correct order.
6184 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6185 before the connection was closed, thus losing the rejection response.
6187 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6188 some early Solaris releases, but causes trouble in current releases where
6189 socklen_t is defined.
6191 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6194 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6197 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6198 that releasing the top of it at the end releases what was used for sub-
6199 expansions (unless the block got too big). However, discard this block if
6200 the first thing is a variable or header, so that we can use its block when
6201 it is dynamic (useful for very large $message_headers, for example).
6203 6. Lookups now cache *every* query, not just the most recent. A new, separate
6204 store pool is used for this. It can be recovered when all lookup caches are
6205 flushed. Lookups now release memory at the end of their result strings.
6206 This has involved some general refactoring of the lookup sources.
6208 7. Some code has been added to the store_xxx() functions to reduce the amount
6209 of flapping under certain conditions.
6211 8. log_incoming_interface used to affect only the <= reception log lines. Now
6212 it causes the local interface and port to be added to several more SMTP log
6213 lines, for example "SMTP connection from", and rejection lines.
6215 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6217 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6219 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6220 could be overwritten at the end of the current message (or the start of a
6221 new message if it was set in a HELO ACL). The value is now preserved for
6222 the duration of the SMTP connection.
6224 12. If a transport had a headers_rewrite setting, and a matching header line
6225 contained an unqualified address, that address was qualified, even if it
6226 did not match any rewriting rules. The underlying bug was that the values
6227 of the flags that permit the existence of unqualified sender and recipient
6228 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6229 non-local messages, and by -bnq for local messages) were not being
6230 preserved with the message after it was received.
6232 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6233 "next input=" as part of the text comprising the host identity instead of
6234 the correct text. The code was using the same buffer for two different
6235 strings. However, depending on which order the printing function evaluated
6236 its arguments, the bug did not always show up. Under Linux, for example, my
6237 test suite worked just fine.
6239 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6240 doesn't work with some older versions of Perl. It has been changed to "my",
6241 which in any case is probably the better facility to use.
6243 15. A really picky compiler found some instances of statements for creating
6244 error messages that either had too many or two few arguments for the format
6247 16. The size of the buffer for calls to the DNS resolver has been increased
6248 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6249 for addresses that have a lot of PTR records. This alleviates a problem; it
6250 does not fully solve it.
6252 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6253 buffer now truncates the list and logs the incident, which is the same
6254 action as happens when Exim is looking up a host name and its aliases.
6255 Previously in this situation something unpredictable would happen;
6256 sometimes it was "internal error: store_reset failed".
6258 18. If a server dropped the connection unexpectedly when an Exim client was
6259 using GnuTLS and trying to read a response, the client delivery process
6260 crashed while trying to generate an error log message.
6262 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6263 string, for example:
6265 warn message = H1: something\nH2: something
6267 the text was added as a single header line from Exim's point of view
6268 though it ended up OK in the delivered message. However, searching for the
6269 second and subsequent header lines using $h_h2: did not work. This has been
6270 fixed. Similarly, if a system filter added multiple headers in this way,
6271 the routers could not see them.
6273 20. Expanded the error message when iplsearch is called with an invalid key to
6274 suggest using net-iplsearch in a host list.
6276 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6277 ACLs are no longer actually imposed (and a message to that effect is
6280 22. If a "gecos" field in a passwd entry contained escaped characters, in
6281 particular, if it contained a \" sequence, Exim got it wrong when building
6282 a From: or a Sender: header from that name. A second bug also caused
6283 incorrect handling when an unquoted " was present following a character
6284 that needed quoting.
6286 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6287 was not being matched caselessly.
6289 24. Arranged for all hyphens in the exim.8 source to be escaped with
6292 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6293 properly. Recipient callout cache records were still being keyed to include
6294 the sender, even when use_sender was set false. This led to far more
6295 callouts that were necessary. The sender is no longer included in the key
6296 when use_sender is false.
6298 26. Added "control = submission" modifier to ACLs.
6300 27. Added the ${base62d: operator to decode base 62 numbers.
6302 28. dnsdb lookups can now access SRV records.
6304 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6305 the configuration file.
6307 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6308 (-v) mode. This makes the output for a verbose queue run more intelligible.
6310 31. Added a use_postmaster feature to recipient callouts.
6312 32. Added the $body_zerocount variable, containing the number of binary zero
6313 bytes in the message body.
6315 33. The time of last modification of the "new" subdirectory is now used as the
6316 "mailbox time last read" when there is a quota error for a maildir
6319 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6321 35. Added +ignore_unknown as a special item in host lists.
6323 36. Code for decoding IPv6 addresses in host lists is now included, even if
6324 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6325 address was recognized as an IP address, but was then not correctly decoded
6326 into binary, causing unexpected and incorrect effects when compared with
6333 1. Very minor rewording of debugging text in manualroute to say "list of
6334 hosts" instead of "hostlist".
6336 2. If verify=header_syntax was set, and a header line with an unqualified
6337 address (no domain) and a large number of spaces between the end of the
6338 name and the colon was received, the reception process suffered a buffer
6339 overflow, and (when I tested it) crashed. This was caused by some obsolete
6340 code that should have been removed. The fix is to remove it!
6342 3. When running in the test harness, delay a bit after writing a bounce
6343 message to get a bit more predictability in the log output.
6345 4. Added a call to search_tidyup() just before forking a reception process. In
6346 theory, someone could use a lookup in the expansion of smtp_accept_max_
6347 per_host which, without the tidyup, could leave open a database connection.
6349 5. Added the variables $recipient_data and $sender_data which get set from a
6350 lookup success in an ACL "recipients" or "senders" condition, or a router
6351 "senders" option, similar to $domain_data and $local_part_data.
6353 6. Moved the writing of debug_print from before to after the "senders" test
6356 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6357 problems for message scanning, either using a data ACL, or using
6358 local_scan() because the Received: header was not generated till after they
6359 were called (in order to set the time as the time of reception completion).
6360 I have revised the way this works. The header is now generated after the
6361 body is received, but before the ACL or local_scan() are called. After they
6362 are run, the timestamp in the header is updated.
6368 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6369 before starting a queue runner without re-exec. This happened only when
6370 deliver_drop_privilege was set or when the Exim user was set to root. The
6371 effect of the bug was that timeouts during subsequent deliveries caused
6372 crashes instead of being properly handled. The handler is now left at its
6373 default (and expected) setting.
6375 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6376 message, again when deliver_drop_privilege is set or Exim is run as root.
6377 The bug described in (1) was not present in this case, but the tidying up
6378 of the other signals was missing. I have made the two cases consistent.
6380 3. The ignore_target_hosts setting on a manualroute router was being ignored
6381 for hosts that were looked up using the /MX notation.
6383 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6386 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6387 operated on the sender address. After changing the $sender_address to <>
6388 for the sender address verify, Exim was re-instated it as the original
6389 (before rewriting) address, but remembering that it had rewritten it, so it
6390 wasn't rewriting it again. This bug also had the effect of breaking the
6391 sender address verification caching when the sender address was rewritten.
6393 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6394 This has been changed so that if the ip literal address matches
6395 ignore_target_hosts, the router declines.
6397 7. Added expansion conditions match_domain, match_address, and match_local_
6398 part (NOT match_host).
6400 8. The placeholder for the Received: header didn't have a length field set.
6402 9. Added code to Exim itself and to exim_lock to test for a specific race
6403 condition that could lead to file corruption when using MBX delivery. The
6404 issue is with the lockfile that is created in /tmp. If this file is removed
6405 after a process has opened it but before that process has acquired a lock,
6406 there is the potential for a second process to recreate the file and also
6407 acquire a lock. This could lead to two Exim processes writing to the file
6408 at the same time. The added code performs the same test as UW imapd; it
6409 checks after acquiring the lock that its file descriptor still refers to
6410 the same named file.
6412 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6413 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6416 11. Added the smtp_active_hostname option. If used, this will typically be made
6417 to depend on the incoming interface address. Because $interface_address is
6418 not set up until the daemon has forked a reception process, error responses
6419 that can happen earlier (such as "too many connections") no longer contain
6422 12. If an expansion in a condition on a "warn" statement fails because a lookup
6423 defers, the "warn" statement is abandoned, and the next ACL statement is
6424 processed. Previously this caused the whole ACL to be aborted.
6426 13. Added the iplsearch lookup type.
6428 14. Added ident_timeout as a log selector.
6430 15. Added tls_certificate_verified as a log selector.
6432 16. Added a global option tls_require_ciphers (compare the smtp transport
6433 option of the same name). This controls incoming TLS connections.
6435 17. I finally figured out how to make tls_require_ciphers do a similar thing
6436 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6437 before starting the TLS session.
6439 18. Tabs are now shown as \t in -bP output.
6441 19. If the log selector return_path_on_delivery was set, Exim crashed when
6442 bouncing a message because it had too many Received: header lines.
6444 20. If two routers both had headers_remove settings, and the first one included
6445 a superfluous trailing colon, the final name in the first list and the
6446 first name in the second list were incorrectly joined into one item (with a
6447 colon in the middle).
6453 1. Added -C and -D options to the exinext utility, mainly to make it easier
6454 to include in the automated testing, but these could be helpful when
6455 multiple configurations are in use.
6457 2. The exinext utility was not formatting the output nicely when there was
6458 an alternate port involved in the retry record key, nor when there was a
6459 message id as well (for retries that were specific to a specific message
6460 and a specific host). It was also confused by IPv6 addresses, because of
6461 the additional colons they contain. I have fixed the IPv4 problem, and
6462 patched it up to do a reasonable job for IPv6.
6464 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6465 delivery, the log line now contains "pipelined" if PIPELINING was used.
6467 4. An SMTP transport process used to panic and die if the bind() call to set
6468 an explicit outgoing interface failed. This has been changed; it is now
6469 treated in the same way as a connect() failure.
6471 5. A reference to $sender_host_name in the part of a conditional expansion
6472 that was being skipped was still causing a DNS lookup. This no longer
6475 6. The def: expansion condition was not recognizing references to header lines
6476 that used bh_ and bheader_.
6478 7. Added the _cache feature to named lists.
6480 8. The code for checking quota_filecount in the appendfile transport was
6481 allowing one more file than it should have been.
6483 9. For compatibility with Sendmail, the command line option
6491 and sets the incoming protocol and host name (for trusted callers). The
6492 host name and its colon can be omitted when only the protocol is to be set.
6493 Note the Exim already has two private options, -pd and -ps, that refer to
6494 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6495 "s", but I don't think that's a major issue.
6497 10. A number of refactoring changes to the code, none of which should affect
6500 (a) The number of logging options was getting close to filling up the
6501 32-bit word that was used as a bit map. I have split them into two classes:
6502 those that are passed in the argument to log_write(), and those that are
6503 only ever tested independently outside of that function. These are now in
6504 separate 32-bit words, so there is plenty of room for expansion again.
6505 There is no change in the user interface or the logging behaviour.
6507 (b) When building, for example, log lines, the code previously used a
6508 macro that called string_cat() twice, in order to add two strings. This is
6509 not really sufficiently general. Furthermore, there was one instance where
6510 it was actually wrong because one of the argument was used twice, and in
6511 one call a function was used. (As it happened, calling the function twice
6512 did not affect the overall behaviour.) The macro has been replaced by a
6513 function that can join an arbitrary number of extra strings onto a growing
6516 (c) The code for expansion conditions now uses a table and a binary chop
6517 instead of a serial search (which was left over from when there were very
6518 few conditions). Also, it now recognizes conditions like "pam" even when
6519 the relevant support is not compiled in: a suitably worded error message is
6520 given if an attempt is made to use such a condition.
6522 11. Added ${time_interval:xxxxx}.
6524 12. A bug was causing one of the ddress fields not to be passed back correctly
6525 from remote delivery subprocesses. The field in question was not being
6526 subsequently used, so this caused to problems in practice.
6528 13. Added new log selectors queue_time and deliver_time.
6530 14. Might have fixed a bug in maildirsizefile handling that threw up
6531 "unexpected character" debug warnings, and recalculated the data
6532 unnecessarily. In any case, I expanded the warning message to give more
6535 15. Added the message "Restricted characters in address" to the statements in
6536 the default ACL that block characters like @ and % in local parts.
6538 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6539 Three changes have been made:
6541 (a) There was a serious bug; a negative response to MAIL caused the whole
6542 recipient domain to be cached as invalid, thereby blocking all messages
6543 to all local parts at the same domain, from all senders. This bug has
6544 been fixed. The domain is no longer cached after a negative response to
6545 MAIL if the sender used is not empty.
6547 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6550 (c) A new callout option, "use_sender" has been added for people who want
6551 the modified behaviour.
6557 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6560 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6561 indeed breaks things for older releases.
6563 3. Added additional logging to the case where there is a problem reading data
6564 from a filter that is running in a subprocess using a pipe, in order to
6565 try to track down a specific problem.
6567 4. Testing facility fudge: when running in the test harness and attempting
6568 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6569 getting "No route to host". Convert this to a timeout.
6571 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6574 6. Some OS don't have socklen_t but use size_t instead. This affects the
6575 fifth argument of getsockopt() amongst other things. This is now
6576 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6577 can be set for individual OS. I have set it for SunOS5, OSF1, and
6578 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6579 some earlier ones do not.
6581 7. Change 4.30/15 was not doing the test caselessly.
6583 8. The standard form for an IPv6 address literal was being rejected by address
6584 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6585 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6586 this, as well as the form without the "IPv6" on the front (but only when
6587 address literals are enabled, of course).
6589 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6591 10. Exim crashed if a message with an empty sender address specified by -f
6592 encountered a router with an errors_to setting. This could be provoked only
6593 by a command such as
6597 where an empty string was supplied; "<>" did not hit this bug.
6599 11. Installed PCRE release 4.5.
6601 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6602 remained set. It is now erased.
6604 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6605 times from message ids (which are base 36 rather than the normal 62).
6607 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6608 were being counted as actual protocol errors, and logged if the log
6609 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6610 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6611 and DATA following a set of rejected RCPTs do not count as protocol errors.
6612 In other words, Exim assumes they were pipelined, though this may not
6613 actually be the case. Of course, in all cases the client gets an
6614 appropriate error code.
6616 15. If a lookup fails in an ACL condition, a message about the failure may
6617 be available; it is used if testing the ACL cannot continue, because most
6618 such messages specify what the cause of the deferral is. However, some
6619 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6620 that caused an old message to be retained and used if a later statement
6621 caused a defer, replacing the real cause of the deferral.
6623 16. If an IP address had so many PTR records that the DNS lookup buffer
6624 was not large enough to hold them, Exim could crash while trying to process
6625 the truncated data. It now detects and logs this case.
6627 17. Further to 4.21/58, another change has been made: if (and only if) the
6628 first line of a message (the first header line) ends with CRLF, a bare LF
6629 in a subsequent header line has a space inserted after it, so as not to
6630 terminate the header.
6632 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6633 unnecessarily, used atoi() instead of strtol(), and didn't check the
6634 termination when getting file sizes from file names by regex.
6636 19. Completely re-implemented the support for maildirsize files, in the light
6637 of a number of problems with the previous contributed implementation
6638 (4.30/29). In particular:
6640 . If the quota is zero, the maildirsize file is maintained, but no quota is
6643 . If the maildir directory does not exist, it is created before any attempt
6644 to write a maildirsize file.
6646 . The quota value in the file is just a cache; if the quota is changed in
6647 the transport, the new value overrides.
6649 . A regular expression is available for excluding directories from the
6652 20. The autoreply transport checks the characters in options that define the
6653 message's headers; it allows continued headers, but it was checking with
6654 isspace() after an embedded newline instead of explicitly looking for a
6657 21. If all the "regular" hosts to which an address was routed had passed their
6658 expiry times, and had not reached their retry times, the address was
6659 bounced, even if fallback hosts were defined. Now Exim should go on to try
6662 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6663 equivalent code in the SMTP transport. Some hosts send humungous responses
6664 to HELO/EHLO, more than 1024 it seems.
6666 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6667 gives compiler warnings in some environments. I've now done it "properly",
6670 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6671 (because of problems with the built-in one) was declared to return uschar *
6672 instead of char *, causing compiler failure.
6674 25. Fixed a file descriptor leak when processing alias/forward files.
6676 26. Fixed a minor format string issue in dbfn.c.
6678 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6680 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6681 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6684 29. When a sender address is verified, it is cached, to save repeating the test
6685 when there is more than one recipient in a message. However, when the
6686 verification involves a callout, it is possible for different callout
6687 options to be set for different recipients. It is too complicated to keep
6688 track of this in the cache, so now Exim always runs a verification when a
6689 callout is required, relying on the callout cache for the optimization.
6690 The overhead is duplication of the address routing, but this should not be
6693 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6694 to be verified with callout=postmaster, and the main callout worked but the
6695 postmaster check failed, the verification correctly failed. However, if a
6696 subsequent RCPT command asked for sender verification *without* the
6697 postmaster check, incorrect caching caused this verification also to fail,
6700 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6701 it was not caching the DNS options (qualify_single, search_parents) that
6702 were used when the lookup failed. A subsequent lookup with different
6703 options therefore always gave the same answer, though there were cases
6704 where it should not have. (Example: a "domains = !$mx_any" option on a
6705 dnslookup router: the "domains" option is always processed without any
6706 widening, but the router might have qualify_single set.) Now Exim uses the
6707 cached value only when the same options are set.
6709 32. Added John Jetmore's "exipick" utility to the distribution.
6711 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6712 than a timeout (e.g. a certificate is required, and is not provided), an
6713 Exim server now closes the connection immediately. Previously it waited for
6714 the client to close - but if the client is SSL, it seems that they each
6715 wait for each other, leading to a delay before one of them times out.
6717 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6718 maintained 0.8.x compatibility because I don't think many are using it, and
6719 it is clearly obsolete.
6721 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6724 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6725 client certificate was expired. A simple patch fixes this, though I don't
6726 understand the full logic of why the verify callback is called multiple
6729 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6730 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6731 which causes problems with some clients (such as the Certicom SSL Plus
6732 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6733 disables the coutermeasure allowing Eudora to connect."
6735 38. Exim was not checking that a write() to a log file succeeded. This could
6736 lead to Bad Things if a log got too big, in particular if it hit a file
6737 size limit. Exim now panics and dies if it cannot write to a log file, just
6738 as it does if it cannot open a log file.
6740 39. Modified OS/Makefile-Linux so that it now contains
6742 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6744 The two -D definitions ensure that Exim is compiled with large file
6745 support, which makes it possible to handle log files that are bigger than
6748 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6749 instance) a domain was checked against a named list that involved a lookup,
6750 causing $domain_data to be set, then another domain was checked against the
6751 same list, then the first domain was re-checked, the value of $domain_data
6752 after the final check could be wrong. In particular, if the second check
6753 failed, it could be set empty. This bug probably also applied to
6756 41. The strip_trailing_dot option was not being applied to the address given
6757 with the -f command-line option.
6759 42. The code for reading a message's header from the spool was incrementing
6760 $received_count, but never initializing it. This meant that the value was
6761 incorrect (doubled) while delivering a message in the same process in which
6762 it was received. In the most common configuration of Exim, this never
6763 happens - a fresh exec is done - but it can happen when
6764 deliver_drop_privilege is set.
6766 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6767 it now includes up to 150 characters of the unexpected data in the log
6770 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6771 and building data strings. The size of both of these buffers was 10 000
6772 bytes - far larger than anybody would *ever* want, thought I. Needless to
6773 say, somebody hit the limit. I have increased the maximum line length to
6774 20 000 and the maximum data length of concatenated lines to 100 000. I have
6775 also fixed two bugs, because there was no checking on these buffers. Tsk,
6776 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6777 buffer is too small.
6779 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6780 lsearch lookups. Now it does.
6782 46. When parsing a route_list item in a manualroute router, a fixed-length
6783 buffer was used for the list of hosts. I made this 1024 bytes long,
6784 thinking that nobody would ever have a list of hosts that long. Wrong.
6785 Somebody had a whole pile of complicated expansion conditions, and the
6786 string was silently truncated, leading to an expansion error. It turns out
6787 that it is easier to change to an unlimited length (owing to other changes
6788 that have happened since this code was originally written) than to build
6789 structure for giving a limitation error. The length of the item that
6790 expands into the list of hosts is now unlimited.
6792 47. The lsearch lookup could not handle data where the length of text line was
6793 more than 4095 characters. Such lines were truncated, leading to shortened
6794 data being returned. It should now handle lines of any length.
6796 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6797 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6800 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6801 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6802 no longer happens in comment lines. A list of replacements is now placed
6803 at the head of all of the source files, except those whose only change is
6804 to replace PERL_COMMAND in the very first #! line.
6806 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6807 messages on the queue, with a bottom-up merge sort, using code contributed
6808 by Michael Haardt. This should make operations like -bp somewhat faster on
6809 large queues. It won't affect queue runners, except when queue_run_in_order
6812 51. Installed eximstats 1.31 in the distribution.
6814 52. Added support for SRV lookups to the dnslookup router.
6816 53. If an ACL referred to $message_body or $message_body_end, the value was not
6817 reset for any messages that followed in the same SMTP session.
6819 54. The store-handling optimization for building very long strings was not
6820 differentiating between the different store pools. I don't think this
6821 actually made any difference in practice, but I've tidied it.
6823 55. While running the routers to verify a sender address, $sender_address
6824 was still set to the sender address. This is wrong, because when routing to
6825 send a bounce to the sender, it would be empty. Therefore, I have changed
6826 it so that, while verifying a sender address, $sender_address is set to <>.
6827 (There is no change to what happens when verifying a recipient address.)
6829 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6830 target A or AAAA records (if not already returned) without resetting the
6831 qualify_single or search_parents options of the DNS resolver. These are
6832 inappropriate in this case because the targets of MX and SRV records must
6833 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6834 target that, when qualified, matched something in the local domain. These
6835 two options are now turned off when doing these lookups.
6837 57. It seems that at least some releases of Reiserfs (which does not have the
6838 concept of a fixed number of inodes) returns zero and not -1 for the
6839 number of available inodes. This interacted badly with check_spool_inodes,
6840 which assumed that -1 was the "no such thing" setting. What I have done is
6841 to check that the total number of inodes is greater than zero before doing
6842 the test of how many are available.
6844 58. When a "warn" ACL statement has a log_message modifier, the message is
6845 remembered, and not repeated. This is to avoid a lot of repetition when a
6846 message has many recipients that cause the same warning to be written.
6847 However, Exim was preserving the list of already written lines for an
6848 entire SMTP session, which doesn't seem right. The memory is now reset if a
6849 new message is started.
6851 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6852 debugging output unless log_rewrite was also set.
6854 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6855 of (int)(handle) when we know that handle contains (void *)(-1).
6857 61. The Exim daemon panic-logs an error return when it closes the incoming
6858 connection. However "connection reset by peer" seems to be common, and
6859 isn't really an error worthy of noting specially, so that particular error
6862 62. When Exim is trying to find all the local interfaces, it used to panic and
6863 die if the ioctl to get the interface flags failed. However, it seems that
6864 on at least one OS (Solaris 9) it is possible to have an interface that is
6865 included in the list of interfaces, but for which you get a failure error
6866 for this call. This happens when the interface is not "plumbed" into a
6867 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6868 failure of the "get flags" call assumes that the interface is down.
6870 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6871 makes life easier for people who are doing arithmetic on fields extracted
6872 from dates, where you often get leading zeros that should not be
6873 interpreted as octal.
6875 64. Added qualify_domain to the redirect router, to override the global
6878 65. If a pathologically long header line contained very many addresses (the
6879 report of this problem mentioned 10 000) and each of them was rewritten,
6880 Exim could use up a very large amount of memory. (It kept on making new
6881 copies of the header line as it rewrote, and never released the old ones.)
6882 At the expense of a bit more processing, the header rewriting function has
6883 been changed so that it no longer eats memory in this way.
6885 66. The generation of the Received: header has been moved from the time that a
6886 message starts to be received, to the time that it finishes. The timestamp
6887 in the Received: header should now be very close to that of the <= log
6888 line. There are two side-effects of this change:
6890 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6891 logged header lines no longer include the local Received: line, because
6892 it has not yet been created. The same applies to a copy of the message
6893 that is returned to a non-SMTP sender when a message is rejected.
6895 (b) When a filter file is tested using -bf, no additional Received: header
6896 is added to the test message. After some thought, I decided that this
6899 This change does not affect the value of $received_for. It is still set
6900 after address rewriting, but before local_scan() is called.
6902 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6904 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6905 gave an unhelpful panic error message, and a defer error. I have managed to
6906 change this behaviour so that it now rejects any supplied certificate,
6907 which seems right, as the list of acceptable certificates is empty.
6909 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6910 gave an unhelpful defer error. I have not managed to make this reject any
6911 supplied certificates, but the error message it gives is "no certificate
6912 supplied", which is not helpful.
6914 70. exigrep's output now also includes lines that are not associated with any
6915 message, but which match the given pattern. Implemented by a patch from
6916 Martin Sluka, which also tidied up the Perl a bit.
6918 71. Recipient callout verification, like sender verification, was using <> in
6919 the MAIL FROM command. This isn't really the right thing, since the actual
6920 sender may affect whether the remote host accepts the recipient or not. I
6921 have changed it to use the actual sender in the callout; this means that
6922 the cache record is now keyed on a recipient/sender pair, not just the
6923 recipient address. There doesn't seem to be a real danger of callout loops,
6924 since a callout by the remote host to check the sender would use <>.
6925 [SEE ABOVE: changed after hitting problems.]
6927 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6928 temporary errors. However, in the case of such a code being given after
6929 the end of a data transmission (i.e. after ".") Exim was failing to write
6930 a retry record for the message. (Yes, there was some broken host that was
6931 actually sending 8xx at this point.)
6933 73. An unknown lookup type in a host list could cause Exim to panic-die when
6934 the list was checked. (An example that provoked this was putting <; in the
6935 middle of a list instead of at the start.) If this happened during a DATA
6936 ACL check, a -D file could be left lying around. This kind of configuration
6937 error no longer causes Exim to die; instead it causes a defer error. The
6938 incident is still logged to the main and panic logs.
6940 74. Buglet left over from Exim 3 conversion. The message "too many messages
6941 in one connection" was written to the rejectlog but not the mainlog, except
6942 when address rewriting (yes!) was being logged.
6944 75. Added write_rejectlog option.
6946 76. When a system filter was run not as root (that is, when system_filter_user
6947 was set), the values of the $n variables were not being returned to the
6948 main process; thus, they were not subsequently available in the $sn
6951 77. Added +return_path_on_delivery log selector.
6953 78. A connection timeout was being treated differently from recipients deferred
6954 when testing hosts_max_try with a message that was older than the host's
6955 retry timeout. (The host should not be counted, thus allowing all hosts to
6956 be tried at least once before bouncing.) This may have been the cause of an
6957 occasionally reported bug whereby a message would remain on the queue
6958 longer than the retry timeout, but would be bounced if a delivery was
6959 forced. I say "may" because I never totally pinned down the problem;
6960 setting up timeout/retry tests is difficult. See also the next item.
6962 79. The ultimate address timeout was not being applied to errors that involved
6963 a combination of host plus message (for example, a timeout on a MAIL
6964 command). When an address resolved to a number of possible hosts, and they
6965 were not all tried for each delivery (e.g. because of hosts_max_try), a
6966 message could remain on the queue longer than the retry timeout.
6968 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6971 81. Fixed an obscure SMTP outgoing bug which required at least the following
6972 conditions: (a) there was another message waiting for the same server;
6973 (b) the server returned 5xx to all RCPT commands in the first message so
6974 that the message was not completed; (c) the server dropped the connection
6975 or gave a negative response to the RSET that Exim sends to abort the
6976 transaction. The observed case was a dropped connection after DATA that had
6977 been sent in pipelining mode. That is, the server had advertised PIPELINING
6978 but was not implementing it correctly. The effect of the bug was incorrect
6979 behaviour, such as trying another host, and this could lead to a crash.
6985 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6986 and daemon.c were passed as pointers to ints; they should have been
6987 pointers to socklen_t variables (which are typically unsigned ints).
6989 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6992 3. Fixed a really odd bug that affected only the testing scheme; patching a
6993 certain fixed string in the binary changed the value of another string that
6994 happened to be identical to the end of the original first string.
6996 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6997 name", it returns that address as the IP address. On some operating
6998 systems (e.g. Solaris), it also passes back the IP address string as the
6999 "host name". However, on others (e.g. Linux), it passes back an empty
7000 string. Exim wasn't checking for this, and was changing the host name to an
7001 empty string, assuming it had been canonicalized.
7003 5. Although rare, it is permitted to have more than one PTR record for a given
7004 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7005 all the names associated with an address, because they do in Solaris.
7006 However, it seems that they do not in Linux for data that comes from the
7007 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7008 I found this out when I moved to a new Linux workstation and tried to run
7009 the Exim test suite.
7011 To get round this problem I have changed the code so that it now does its
7012 own call to the DNS to look up PTR records when searching for a host name.
7013 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7014 addresses that are only in /etc/hosts are still found.
7016 This behaviour is, however, controlled by an option called host_lookup_
7017 order, which defaults to "bydns:byaddr". If people want to use the other
7018 order, or indeed, just use one or the other means of lookup, they can
7019 specify it in this variable.
7021 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7022 some operating systems, this comes back from gethostbyaddr() as an empty
7023 string, and this is what Exim used to test for. However, it seems that in
7024 other systems, "." is yielded. Exim now tests for this case too.
7026 7. The values of check_spool_space and check_log_space are now held internally
7027 as a number of kilobytes instead of an absolute number of bytes. If a
7028 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7029 kilobyte. This means that much larger values can be stored.
7031 8. Exim monitor: an attempt to get the action menu when not actually pointing
7032 at a message produces an empty menu entitled "No message selected". This
7033 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7034 no entries in it ("Shell widget menu has zero width and/or height"). So I
7035 have added a single, blank menu entry in this case.
7037 9. Added ${quote_local_part.
7039 10. MIME decoding is now applied to the contents of Subject: header lines when
7042 11. Now that a reference to $sender_host_address automatically causes a reverse
7043 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7044 host lookup before query-style lookups in lists that might use this
7045 variable. This has therefore been abolished, and the "net-" prefix is no
7046 longer necessary for query-style lookups.
7048 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7049 have been a typo for LFLAGS, so it has been changed.
7051 13. The install script calls Exim with "-C /dev/null" in order to find the
7052 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7053 to be output. However, since Exim outputs its version number before the
7054 error, it didn't break the script. It just looked ugly. I fixed this by
7055 always allowing "-C /dev/null" if the caller is root.
7057 14. Ignore overlarge ACL variable number when reading spool file - insurance
7058 against a later release with more variables having written the file.
7060 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7061 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7062 this, as well as the form without the "IPv6" on the front.
7064 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7065 OS/Makefile-Darwin file.
7067 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7068 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7070 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7071 variable. However, some OS use ints for inodes. I've added cast to long int
7072 to get rid of the compiler warning.
7074 19. I had forgotten to lock out "/../" in configuration file names when
7075 ALT_CONFIG_PREFIX was set.
7077 20. Routers used for verification do not need to specify transports. However,
7078 if such a router generated a host list, and callout was configured, Exim
7079 crashed, because it could not find a port number from the (non-existent)
7080 transport. It now assumes port 25 in this circumstance.
7082 21. Added the -t option to exigrep.
7084 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7085 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7086 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7087 them is set, LOOKUP_LSEARCH is forced.
7089 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7091 24. Added sender and host information to the "rejected by local_scan()" log
7092 line; previously there was no indication of these.
7094 25. Added .include_if_exists.
7096 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7097 Linux. It turns out that not all file systems support this. Apparently some
7098 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7099 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7100 not supported on the file descriptor, is now ignored when Exim is trying to
7101 sync a directory. This applies only to Linux.
7103 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7105 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7106 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7109 29. Added support for maildirsize files from supplied patch (modified a bit).
7111 30. The use of :fail: followed by an empty string could lead Exim to respond to
7112 sender verification failures with (e.g.):
7114 550 Verification failed for <xxx>
7115 550 Sender verify failed
7117 where the first response line was missing the '-' that indicates it is not
7118 the final line of the response.
7120 31. The loop for finding the name of the user that called Exim had a hardwired
7121 limit of 10; it now uses the value of finduser_retries, which is used for
7122 all other user lookups.
7124 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7127 33. Exim was neglecting to zero errno before one call of strtol() when
7128 expanding a string and expecting an integer value. On some systems this
7129 resulted in spurious "integer overflow" errors. Also, it was casting the
7130 result into an int without checking.
7132 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7133 did not work. The code looks as if it has *never* worked, though it appears
7134 to have been documented since at least release 1.62. I have made it work.
7136 35. The "timeout_DNS" error in retry rules, also documented since at least
7137 1.62, also never worked. As it isn't clear exactly what this means, and
7138 clearly it isn't a major issue, I have abolished the feature by treating it
7139 as "timeout", and writing a warning to the main and panic logs.
7141 36. The display of retry rules for -brt wasn't always showing the error code
7144 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7145 timeout_connect_A, timeout_connect_MX.
7147 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7148 to the empty sender.
7150 39. The daemon was not analysing the content of -oX till after it had closed
7151 stderr and disconnected from the controlling terminal. This meant that any
7152 syntax errors were only noted on the panic log, and the return code from
7153 the command was 0. By re-arranging the code a little, I've made the
7154 decoding happen first, so such errors now appear on stderr, and the return
7155 code is 1. However, the actual setting up of the sockets still happens in
7156 the disconnected process, so errors there are still only recorded on the
7159 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7160 connections (as happens on some IP stacks) was logged at start up time as
7161 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7162 it from "IPv6 and IPv4", which means that two separate sockets are being
7165 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7166 says whether AF_INET or AF_INET6 was passed as an argument.
7168 42. Exiwhat output was messed up when time zones were included in log
7171 43. Exiwhat now gives more information about the daemon's listening ports,
7172 and whether -tls-on-connect was used.
7174 44. The "port" option of the smtp transport is now expanded.
7176 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7177 silently ignored. Now an error message is written to the main and panic
7180 46. There's a new ACL modifier called "logwrite" which writes to a log file
7181 as soon as it is encountered.
7183 47. Added $local_user_uid and $local_user_gid at routing time.
7185 48. Exim crashed when trying to verify a sender address that was being
7188 49. Exim was recognizing only a space character after ".include". It now also
7189 recognizes a tab character.
7191 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7192 extracting the relevant information from the specification. The man page no
7193 longer contains scrambled data for the -d option, and I've added a section
7194 at the front about calling Exim under different names.
7196 51. Added "extra_headers" argument to the "mail" command in filter files.
7198 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7201 53. Installed eximstats 1.29.
7203 54. Added transport_filter_timeout as a generic transport option.
7205 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7206 Cc: header lines. This was required by RFC 822, but it not required by RFC
7209 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7210 incoming messages that did not have them. Now it does so only if the
7211 message originates locally, that is, if there is no associated remote host
7212 address. When Resent- header lines are present, this applies to the Resent-
7213 lines rather than the non-Resent- lines.
7215 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7216 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7218 58. Messages for configuration errors now include the name of the main
7219 configuration file - useful now that there may be more than one file in a
7220 list (.included file names were always shown).
7222 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7223 for those rare installations that do not start the daemon as root or run it
7224 setuid root. I've cut out the call to initgroups() if the daemon is not
7227 60. The Exim user and group can now be bound into the binary as text strings
7228 that are looked up at the start of Exim's processing.
7230 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7232 62. Added $mailstore_basename variable.
7234 63. Installed patch to sieve.c from Michael Haardt.
7236 64. When Exim failed to open the panic log after failing to open the main log,
7237 the original message it was trying to log was written to stderr and debug
7238 output, but if they were not available (the usual case in production), it
7239 was lost. Now it is written to syslog before the two lines that record the
7240 failures to open the logs.
7242 65. Users' Exim filters run in subprocesses under the user's uid. It is
7243 possible for a "deliver" command or an alias in a "personal" command to
7244 provoke an address rewrite. If logging of address rewriting is configured,
7245 this fails because the process is not running as root or exim. There may be
7246 a better way of dealing with this, but for the moment (because 4.30 needs
7247 to be released), I have disabled address rewrite logging when running a
7248 filter in a non-root, non-exim process.
7254 1. The buildconfig auxiliary program wasn't quoting the value set for
7255 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7256 not defined. This bug was masked in 4.22 by the effect that was fixed in
7259 2. Some messages that were rejected after a message id was allocated were
7260 shown as "incomplete" by exigrep. It no longer does this for messages that
7261 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7263 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7264 have allow_domain_literals set, the ID did not get logged in the <= line.
7265 Domain literals are now always recognized in Message-ID: header lines.
7267 4. The first argument for a ${extract expansion item is the key name or field
7268 number. Leading and trailing spaces in this item were not being ignored,
7269 causing some misleading effects.
7271 5. When deliver_drop_privilege was set, single queue runner processes started
7272 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7273 same command in the process it spins off) were not dropping privilege.
7275 6. When the daemon running as "exim" started a queue runner, it always
7276 re-executed Exim in the spun-off process. This is a waste of effort when
7277 deliver_drop_privilege is set. The new process now just calls the
7278 queue-runner function directly.
7284 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7287 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7288 ignored. Though the use of -oP was forcing the writing of a pid file, it
7289 was always written to the default place.
7291 3. If the message "no IP address found for host xxxx" is generated during
7292 incoming verification, it is now followed by identification of the incoming
7293 connection (so you can more easily find what provoked it).
7295 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7297 5. Added some features to "harden" Exim a bit more against certain attacks:
7299 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7300 be put in Local/Makefile. This is like the never_users runtime option,
7301 but it cannot be overridden. The default setting is "root".
7303 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7304 prefix string with which any file named in a -C command line option
7307 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7308 is retained for -C and -D only if the caller of Exim is root. Without
7309 it, the exim user may also use -C and -D and retain privilege.
7311 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7312 command line option is disabled.
7314 6. Macro names set by the -D option must start with an upper case letter, just
7315 like macro names defined in the configuration file.
7317 7. Added "dereference=" facility to LDAP.
7319 8. Two instances of the typo "uknown" in the source files are fixed.
7321 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7322 the Configure-Makefile script screwed up while processing it.
7324 10. Incorporated PCRE 4.4.
7326 11. The SMTP synchronization check was not operating right at the start of an
7327 SMTP session. For example, it could not catch a HELO sent before the client
7328 waited for the greeting. There is now a check for outstanding input at the
7329 point when the greeting is written. Because of the duplex, asynchronous
7330 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7331 way, but not yet received, when the check is performed.
7333 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7334 on TCP/IP sockets, because this apparently causes some broken clients to
7337 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7338 unchanged) from the Cygwin maintainer.
7340 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7341 was supported instead of testing for mbx. Effectively a typo.
7343 15. The spa authenticator server code was not checking that the input it
7344 received was valid base64.
7346 16. The debug output line for the "set" modifier in ACLs was not showing the
7347 name of the variable that was being set.
7349 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7351 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7352 lookup on its own. Something else was needed to trigger the lookup. For
7353 example, a match in host_lookup or the need for a host name in a host list.
7354 Now, if $sender_host_name is referenced and the host name has not yet been
7355 looked up, a lookup is performed. If the lookup fails, the variable remains
7356 empty, and $host_lookup_failed is set to "1".
7358 19. Added "eqi" as a case-independent comparison operator.
7360 20. The saslauthd authentication condition could segfault if neither service
7361 nor realm was specified.
7363 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7364 error message that was logged was misleading, and incoming SMTP
7365 connections were dropped. The message is now more accurate, and temporary
7366 errors are given to SMTP connections.
7368 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7369 (except RSET) with a 421 error, until QUIT is received. However, it was
7370 failing to send a response to QUIT.
7372 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7373 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7374 ACL is now run after the helo_try_verify_hosts code.
7376 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7377 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7378 case-independent, so other case variants are also recognized). Apparently
7379 some systems use these upper case variants.
7381 25. If more than two messages were waiting for the same host, and a transport
7382 filter was specified for the transport, Exim sent two messages over the
7383 same TCP/IP connection, and then failed with "socket operation on non-
7384 socket" when it tried to send the third.
7386 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7388 27. The extern definition of crypt16() in expand.c was not being excluded when
7389 the OS had its own crypt16() function.
7391 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7392 as a preferred synonym for return_size_limit, both as an option and as an
7395 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7397 30. Changed the default configuration ACL to relax the local part checking rule
7398 for addresses that are not in any local domains. For these addresses,
7399 slashes and pipe symbols are allowed within local parts, but the sequence
7400 /../ is explicitly forbidden.
7402 31. SPA server authentication was not clearing the challenge buffer before
7405 32. log_message in a "warn" ACL statement was writing to the reject log as
7406 well as to the main log, which contradicts the documentation and doesn't
7407 seem right (because no rejection is happening). So I have stopped it.
7409 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7410 However, I am unable to do any testing of this.
7412 34. Fixed an infelicity in the appendfile transport. When checking directories
7413 for a mailbox, to see if any needed to be created, it was accidentally
7414 using path names with one or more superfluous leading slashes; tracing
7415 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7417 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7418 contents are added to the log line that is written for every discarded
7419 recipient. (Previously a log_message setting was ignored.)
7421 36. The ${quote: operator now quotes the string if it is empty.
7423 37. The install script runs exim in order to find its version number. If for
7424 some reason other than non-existence or emptiness, which it checks, it
7425 could not run './exim', it was installing it with an empty version number,
7426 i.e. as "exim-". This error state is now caught, and the installation is
7429 38. An argument was missing from the function that creates an error message
7430 when Exim fails to connect to the socket for saslauthd authentication.
7431 This could cause Exim to crash, or give a corrupted message.
7433 39. Added isip, isip4, and isip6 to ${if conditions.
7435 40. The ACL variables $acl_xx are now saved with the message, and can be
7436 accessed later in routers, transports, and filters.
7438 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7439 strings in the file are not string-expanded.
7441 42. If a MAIL command specified a SIZE value that was too large to fit into an
7442 int variable, the check against message_size_limit failed. Such values are
7443 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7444 day this will have to be increased, but I don't think I want to be around
7445 when emails are that large.
7452 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7453 iconv() is not standard in FreeBSD.
7455 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7456 IPv6 enabled. The observed symptom was a segmentation fault on return from
7457 the function os_common_find_running_interfaces() in src/os.c.
7459 3. In the check_special_case() function in daemon.c I had used "errno" as an
7460 argument name, which causes warnings on some systems. This was basically a
7461 typo, since it was named "eno" in the comments!
7463 4. The code that waits for the clock to tick (at a resolution of some fraction
7464 of a second) so as to ensure message-id uniqueness was always waiting for
7465 at least one whole tick, when it could have waited for less. [This is
7466 almost certainly not relevant at current processor speeds, where it is
7467 unlikely to ever wait at all. But we try to future-proof.]
7469 5. The function that sleeps for a time interval that includes fractions of a
7470 second contained a race. It did not block SIGALRM between setting the
7471 timer, and suspending (a couple of lines later). If the interval was short
7472 and the sigsuspend() was delayed until after it had expired, the suspension
7473 never ended. On busy systems this could lead to processes getting stuck for
7476 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7477 process, before it forks any delivery processes. The open lookup caching
7478 mechanism meant that the open file or database connection was passed into
7479 the delivery process. The problem was that delivery processes always tidy
7480 up cached lookup data. This could cause a problem for the next delivery
7481 process started by the queue runner, because the external queue runner
7482 process does not know about the closure. So the next delivery process
7483 still has data in the lookup cache. In the case of a file lookup, there was
7484 no problem because closing a file descriptor in a subprocess doesn't affect
7485 the parent. However, if the lookup was caching a connection to a database,
7486 the connection was closed, and the second delivery process was likely to
7487 see errors such as "PGSQL: query failed: server closed the connection
7488 unexpectedly". The problem has been fixed by closing all cached lookups
7489 in a queue runner before running a delivery process.
7491 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7492 seem to have the "const" qualifier which it has on other OS. I've
7495 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7496 *of the same type* (client or server) with the same public name that an
7497 error should be diagnosed.
7499 9. When Exim looked up a host name for an IP address, but failed to find the
7500 original IP address when looking up the host name (a safety check), it
7501 output the message "<ip address> does not match any IP for NULL", which was
7502 confusing, to say the least. The bug was that the host name should have
7503 appeared instead of "NULL".
7505 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7506 user that is built into the binary, and the -C or -D options is used, root
7507 privilege is dropped before the configuration file is read. In addition,
7508 logging is switched to stderr instead of the normal log files. If the
7509 configuration then re-defines the Exim user, the unprivileged environment
7510 is probably not what is expected, so Exim logs a panic warning message (but
7513 However, if deliver_drop_privilege is set, the unprivileged state may well
7514 be exactly what is intended, so the warning has been cut out in that case,
7515 and Exim is allowed to try to write to its normal log files.
7521 1. smtp_return_error_details was not giving details for temporary sender
7522 or receiver verification errors.
7524 2. Diagnose a configuration error if two authenticators have the same public
7527 3. Exim used not to create the message log file for a message until the first
7528 delivery attempt. This could be confusing when incoming messages were held
7529 for policy or load reasons. The message log file is now created at the time
7530 the message is received, and an initial "Received" line is written to it.
7532 4. The automatically generated man page for command line options had a minor
7533 bug that caused no ill effects; however, a more serious problem was that
7534 the procedure for building the man page automatically didn't always
7535 operate. Consequently, release 4.20 contains an out-of-date version. This
7536 shouldn't happen again.
7538 5. When building Exim with embedded Perl support, the script that builds the
7539 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7540 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7542 6. The freeze_tell option was not being used for messages that were frozen on
7543 arrival, either by an ACL or by local_scan().
7545 7. Added the smtp_incomplete_transaction log selector.
7547 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7548 was accepting AUTH without a new EHLO.
7550 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7551 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7554 10. Logging of TCP/IP connections (when configured) now happens in the main
7555 daemon process instead of the child process, so that the TCP/IP connection
7556 count is more accurate (but it can never be perfect).
7558 11. The use of "drop" in a nested ACL was not being handled correctly in the
7559 outer ACL. Now, if condition failure induced by the nested "drop" causes
7560 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7561 or "require"), the connection is dropped.
7563 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7564 that yield "discard" can now be used with an "accept" or a "discard" verb,
7565 but an error is generated for any others (because I can't see a useful way
7566 to define what should happen).
7568 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7569 are now processed in the same way as lines in the Exim configuration file.
7570 In particular, continuation lines are supported.
7572 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7574 15. Added -ti meaning -t -i.
7576 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7577 domains, and warn by logging if others are found.
7579 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7580 returned from GIFCONF when Exim is trying to find the list of interfaces on
7581 a host. The code in os.c has been modified to copy each ifreq to an aligned
7582 structure in all cases.
7584 Also, in some cases, the returned ifreq's were being copied to a 'struct
7585 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7586 means the last couple of bytes of an IPv6 address could be chopped if the
7587 ifreq contained only a normal sockaddr (14 bytes storage).
7589 18. Named domain lists were not supported in the hosts_treat_as_local option.
7590 An entry such as +xxxx was not recognized, and was treated as a literal
7593 19. Ensure that header lines added by a DATA ACL are included in the reject log
7594 if the ACL subsequently rejects the message.
7596 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7597 MD5 (which is deprecated).
7599 21. When testing a filter file using -bf, Exim was writing a message when it
7600 took the sender from a "From " line in the message, but it was not doing so
7601 when it took $return_path from a Return-Path: header line. It now does.
7603 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7604 with a valid header line field name (a series of printing characters
7605 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7607 23. Changed "disc" in the source to "disk" to conform to the documentation and
7608 the book and for uniformity.
7610 24. Ignore Sendmail's -Ooption=value command line item.
7612 25. When execve() failed while trying to run a command in a pipe transport,
7613 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7614 could be confused with a return value of 69 from the command itself. This
7615 has been changed to 127, the value the shell returns if it is asked to run
7616 a non-existent command. The wording for the related log line suggests a
7617 non-existent command as the problem.
7619 26. If received_header_text expands to an empty string, do not add a Received:
7620 header line to the message. (Well, it adds a token one on the spool, but
7621 marks it "old" so that it doesn't get used or transmitted.)
7623 27. Installed eximstats 1.28 (addition of -nt option).
7625 28. There was no check for failure on the call to getsockname() in the daemon
7626 code. This can fail if there is a shortage of resources on the system, with
7627 ENOMEM, for example. A temporary error is now given on failure.
7629 29. Contrary to the C standard, it seems that in some environments, the
7630 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7631 program. Exim now does this explicitly; it affects the formatting of
7632 timestamps using strftime().
7634 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7635 complaints. I've now initialized all the variables, to avoid this.
7637 32. Header lines added by a system filter were not being "seen" during
7638 transport-time rewrites.
7640 33. The info_callback() function passed to OpenSSL is set up with type void
7641 (*)(SSL *, int, int), as described somewhere. However, when calling the
7642 function (actually a macro) that sets it up, the type void(*)() is
7643 expected. I've put in a cast to prevent warnings from picky compilers.
7645 34. If a DNS black list lookup found a CNAME record, but there were no A
7646 records associated with the domain it pointed at, Exim crashed.
7648 35. If a DNS black list lookup returned more than one A record, Exim ignored
7649 all but the first. It now scans all returned addresses if a particular IP
7650 value is being sought. In this situation, the contents of the
7651 $dnslist_value variable are a list of all the addresses, separated by a
7654 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7655 to accept a host name and all its aliases if the forward lookup for any of
7656 them yielded the IP address of the incoming connection. Now it accepts only
7657 those names whose forward lookup yields the correct IP address. Any other
7658 names are discarded. This closes a loophole whereby a rogue DNS
7659 administrator could create reverse DNS records to break through a
7660 wildcarded host restriction in an ACL.
7662 37. If a user filter or a system filter that ran in a subprocess used any of
7663 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7664 the wrong values were passed to the pipe command ($thisaddress had the
7665 value of $0, $0 had the value of $1, etc). This bug was introduced by
7666 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7668 38. Improved the line breaking for long SMTP error messages from ACLs.
7669 Previously, if there was no break point between 40 and 75 characters, Exim
7670 left the rest of the message alone. Two changes have been made: (a) I've
7671 reduced the minimum length to 35 characters; (b) if it can't find a break
7672 point between 35 and 75 characters, it looks ahead and uses the first one
7673 that it finds. This may give the occasional overlong line, but at least the
7674 remaining text gets split now.
7676 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7677 file descriptors might be low, and that setting 1000 would always raise it.
7678 It turns out that in some environments, the limit is already over 1000 and
7679 that lowering it causes trouble. So now Exim takes care not to decrease it.
7681 40. When delivering a message, the value of $return_path is set to $sender_
7682 address at the start of routing (routers may change the value). By an
7683 oversight, this default was not being set up when an address was tested by
7684 -bt or -bv, which affected the outcome if any router or filter referred to
7687 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7688 log only when "message" or "log_message" are set. However, if one of the
7689 conditions was an address verification, or a call to a nested ACL, the
7690 messages generated by the underlying test were being passed through. This
7691 no longer happens. The underlying message is available in $acl_verify_
7692 message for both "message" and "log_message" expansions, so it can be
7693 passed through if needed.
7695 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7696 new expansion $bh_ to give the encoded byte string without charset
7697 translation. Translation happens only if iconv() is available; HAVE_ICONV
7698 indicates this at build time. HEADERS_CHARSET gives the charset to
7699 translate to; headers_charset can change it in the configuration, and
7700 "headers charset" can change it in an individual filter file.
7702 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7703 that creates RFC 2047 encoded "words" labels them as that charset instead
7704 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7705 expansion operator; (ii) when Exim creates a From: line for a local
7706 message; (iii) when a header line is rewritten to include a "phrase" part.
7708 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7709 buggy, causing it to skip the first lines of messages whose message ID
7710 ended in 'D'. This would not have bitten before Exim release 4.14, because
7711 message IDs were unlikely to end in 'D' before then. The effect was to have
7712 incorrect size information for certain domains.
7714 45. #include "config.h" was missing at the start of the crypt16.c module. This
7715 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7718 46. If there was a timeout during a "random" callout check, Exim treated it as
7719 a failure of the random address, and carried on sending RSET and the real
7720 address. If the delay was just some slowness somewhere, the response to the
7721 original RCPT would be taken as a response to RSET and so on, causing
7722 mayhem of various kinds.
7724 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7725 when I implemented it. It didn't allow for the fact that some option values
7726 may legitimately be negative (e.g. size_addition), and it didn't even do
7727 the right test for positive values.
7729 48. Domain names in DNS records are case-independent. Exim always looks them up
7730 in lower case. Some resolvers return domain names in exactly the case they
7731 appear in the zone file, that is, they may contain uppercase letters. Not
7732 all resolvers do this - some return always lower case. Exim was treating a
7733 change of case by a resolver as a change of domain, similar to a widening
7734 of a domain abbreviation. This triggered its re-routing code and so it was
7735 trying to route what was effectively the same domain again. This normally
7736 caused routing to fail (because the router wouldn't handle the domain
7737 twice). Now Exim checks for this case specially, and just changes the
7738 casing of the domain that it ultimately uses when it transmits the message
7741 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7744 50. If a filter generated a file delivery with a non-absolute name (possible if
7745 no home directory exists for the router), the forbid_file option was not
7748 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7749 the existing equality matching.
7751 52. Exim was using ints instead of ino_t variables in some places where it was
7752 dealing with inode numbers.
7754 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7755 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7756 called TMPDIR, and if it finds it is different, it changes its value.
7758 54. The smtp_printf() function is now made available to local_scan() so
7759 additional output lines can be written before returning. There is also an
7760 smtp_fflush() function to enable the detection of a dropped connection.
7761 The variables smtp_input and smtp_batched_input are exported to
7764 55. Changed the default runtime configuration: the message "Unknown user"
7765 has been removed from the ACL, and instead placed on the localuser router,
7766 using the cannot_route_message feature. This means that any verification
7767 failures that generate their own messages won't get overridden. Similarly,
7768 the "Unrouteable address" message that was in the ACL for unverifiable
7769 relay addresses has also been removed.
7771 56. Added hosts_avoid_esmtp to the smtp transport.
7773 57. The exicyclog script was not checking for the esoteric option
7774 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7775 will work only if exicyclog is run under the appropriate euid.
7777 58. Following a discussion on the list, the rules by which Exim recognises line
7778 endings on incoming messages have been changed. The -dropcr and drop_cr
7779 options are now no-ops, retained only for backwards compatibility. The
7780 following line terminators are recognized: LF CRLF CR. However, special
7781 processing applies to CR:
7783 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7784 nor a local message in the state where . is a terminator.
7786 (ii) If a bare CR is encountered in a header line, an extra space is added
7787 after the line terminator so as not to end the header. The reasoning
7788 behind this is that bare CRs in header lines are most likely either
7789 to be mistakes, or people trying to play silly games.
7791 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7792 was being incorrectly given as 18 bytes larger than it should have been.
7793 This is a VOB (very old bug).
7795 60. This may never have affected anything current, but just in case it has:
7796 When the local host is found other than at the start of a list of hosts,
7797 the local host, those with the same MX, and any that follow, are discarded.
7798 When the list in question was part of a longer list of hosts, the following
7799 hosts (not currently being processed) were also being discarded. This no
7800 longer happens. I'm not sure if this situation could ever has previously
7803 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7806 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7807 header. This is something that is recommended in a new Internet Draft, and
7808 is something that is documented as being done by Sendmail. There are two
7809 possible values. For messages generated by the autoreply transport, Exim
7812 Auto-Submitted: auto-replied
7814 whereas for all other generated messages (e.g. bounces) it adds
7816 Auto-Submitted: auto-generated
7818 63. The "personal" condition in filters now includes a test for the
7819 Auto-Submitted: header. If it contains the string "auto-" the message it
7820 not considered personal.
7822 64. Added rcpt_include_affixes as a generic transport option.
7824 65. Added queue_only_override (default true).
7826 66. Added the syslog_duplication option.
7828 67. If what should have been the first header line of a message consisted of
7829 a space followed by a colon, Exim was mis-interpreting it as a header line.
7830 It isn't of course - it is syntactically invalid and should therefore be
7831 treated as the start of the message body. The misbehaviour could have
7832 caused a number of strange effects, including loss of data in subsequent
7833 header lines, and spool format errors.
7835 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7836 client host had authenticated. This control can now be exercised by an ACL
7837 for more flexibility.
7839 69. By default, callouts do not happen when testing with -bh. There is now a
7840 variant, -bhc, which does actually run the callout code, including
7841 consulting and updating the callout cache.
7843 70. Added support for saslauthd authentication, courtesy of Alexander
7846 71. If statvfs() failed on the spool or log directories while checking their
7847 size for availability, Exim confusingly gave the error "space shortage".
7848 Furthermore, in debugging mode it crashed with a floating point exception.
7849 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7850 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7851 serious problem, Exim now writes to the main and panic logs when this
7852 happens, with details of the failure. It then refuses to accept the
7853 incoming message, giving the message "spool directory problem" or "log
7854 directory problem" with a 421 code for SMTP messages.
7856 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7857 0, 1, and 2 exist, because some OS complain for execs without them (see
7858 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7859 descriptors. However, the code omitted to check that the open succeeded,
7860 causing mysterious errors if for some reason the permissions on /dev/null
7861 got screwed. Now Exim writes a message to the main and panic logs, and
7862 bombs out if it can't open /dev/null.
7864 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7865 interact so that it is all more flexible. It is supposed to remain
7866 backwards compatible. Also added extra_local_interfaces.
7868 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7869 to bomb out with an assertion failure - to the client this appears as a
7870 connection drop. This problem occurs in the part of the code that was taken
7871 from the Samba project. Fortunately, the assertion is in a very simple
7872 function, so I have fixed this by reproducing the function inline in the
7873 one place where it is called, and arranging for authentication to fail
7874 instead of killing the process with assert().
7876 75. The SPA client code was not working when the server requested OEM rather
7877 than Unicode encoding.
7879 76. Added code to make require_files with a specific uid setting more usable in
7880 the case where statting the file as root fails - usually a non-root-mounted
7881 NFS file system. When this happens and the failure is EACCES, Exim now
7882 forks a subprocess and does the per-uid checking as the relevant uid.
7884 77. Added process_log_path.
7886 78. If log_file_path was not explicitly set, a setting of check_log_space or
7887 check_log_inodes was ignored.
7889 79. If a space check for the spool or log partitions fails, the incident is now
7890 logged. Of course, in the latter case the data may get lost...
7892 80. Added the %p formatting code to string_format() so that it can be used to
7893 print addresses in debug_print(). Adjusted all the address printing in the
7894 debugging in store.c to use %p rather than %d.
7896 81. There was a concern that a line of code in smtp_in.c could overflow a
7897 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7898 initially expressed, the concern was not well-founded, because trailing
7899 spaces are removed early. However, if the trailing spaces were followed by
7900 a NULL, they did not get removed, so the overflow was possible. Two fixes
7903 (a) I re-wrote the offending code in a cleaner fashion.
7904 (b) If an incoming SMTP command contains a NULL character, it is rejected
7907 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7908 runs initgroups(), so that if the Exim user is in any additional groups,
7909 they will be used during message reception.
7915 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob