1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
6 Since Exim version 4.94
7 -----------------------
9 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
10 as arguments, so an implementation trying to copy these into a local
11 buffer was taking a taint-enforcement trap. Fix by using dynamically
12 created buffers. Similar fix for radius expansion condition.
14 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
15 reasonable, eg. to count headers. Fix by using dynamically created
16 buffers rather than a local. Do similar fixes for ACL actions "dcc",
17 "log_reject_target", "malware" and "spam"; the arguments are expanded
18 so could be handling tainted values.
20 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
21 broken the (no-op) support for this sendmail command. Restore it
22 to doing nothing, silently, and returning good status.
24 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
25 record path was given (or the default used) without a leading directory
26 path, an error occurred on trying to open it. Use the transport's working
29 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
30 transport. Previously it was the name on the DNS A-record; use instead
31 the head of the CNAME chain leading there (if there is one). This seems
32 to align better with RFC 6125.
34 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
35 smtp_accept_max_per_host allocated resources which were not released
36 when the limit was exceeded. This eventually crashed the daemon. Fix
37 by adding a relase action in that path.
39 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
40 expanded; previously using tainted values was rejected. Fix by using
41 dynamically-created buffers.
43 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
44 Previously a macro used one argument twice; when called with the
45 argument as an expression having side-effects, incorrect operation
46 resulted. Use an inlineable function.
48 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
49 held open for a verify callout. Previously this wan not accounted for
50 and a corrupt onward SMTP conversation resulted.
52 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
53 excluded, not matching the documentation.
55 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
56 was given for the sqlite_dbfile a trap resulted.
58 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
59 "name" argument resulted in a trap. There is no reason to disallow such;
60 this was a coding error.
62 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
63 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
64 spent suspended, ignoring the Posix definition. Previously we assumed
65 it did and a constant offset from real time could be used as a correction.
66 Change to using the same clock source for the start-of-message and the
67 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
68 exists, just to get a clock slightly more aligned to reality.
70 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
71 RFC says it is optional some validators care. The missing char was not
72 intended but triggered by a line-wrap alignement. Discovery and fix by
73 Guillaume Outters, hacked on by JH.
75 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
76 name being quoted was tainted a trap would be taken. Fix by using
77 dynamicaly created buffers. The routine could have been called by a
78 rewrite with the "h" flag, by using the "-F" command-line option, or
79 by using a "name=" option on a control=submission ACL modifier.
81 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
82 Previously when a whitespace character was specified it was not inserted
83 after removing the newline.
85 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
86 is_tainted() had an off-by-one error in the overenthusiastic direction.
87 Find and fix by Gavan. Although NetBSD is not a supported platform for
88 4.94 this bug could affect other platforms.
90 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
91 is_tainted() had an off-by-one error in the overenthusiastic direction.
92 Find and fix by Gavan. Although NetBSD is not a supported platform for
93 4.94 this bug could affect other platforms.
94 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
95 Previously when a whitespace character was specified it was not inserted
96 after removing the newline.
98 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
99 the domain part of the recipient address. This overrides any tls_sni
100 option set, which was previously used.
102 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
105 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
106 than one server was defined and depending on the platform memory layout
107 details, an internal consistency trap could be hit while walking the list
110 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
111 files. The value is used by the authresults expansion item. Previously
112 if this was used in a router or transport, a crash could result.
114 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
115 applied. This resulted, if any header-line rewrite rules were configured,
116 in a panic-log trigerrable by sending a message with a long address in
117 a header. Fix by increaing the arbitrary limit to larger than a single
118 (dewrapped) 5322 header line maximum size.
120 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
121 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
122 RFC 6648 which deprecates X- options in protocols as a general practice.
123 Changeover between the implementations is handled by the mechanisms
126 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
127 a remote which uses callout/hold. Previously the local was queued.
129 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
132 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
134 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
135 authenticator client_send option. Previously the next char, after a pair
136 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
137 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
138 documentation. There is still no way to get a leading ^ immediately
139 after a NUL (ie. for the password of a PLAIN method authenticator.
141 JH/39 Bug 2691: fix $local_part_data. When the matching list element
142 referred to a file, bad data was returned. This likely also affected
145 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
146 left undeleted; the attempt to re-create it then failed - resulting in
147 the usual "SIGHUP tp have daemon reload configuration" to not work.
148 This affected any platform not supporting "abstract" Unix-domain
149 sockets (i.e. not Linux).
151 JH/42 Bug 2692: Harden against a peer which reneges on a 452 "too many
152 recipients" response to RCPT in a later response, with a 250. The
153 previous coding assumed this would not happen, and under PIPELINING
154 would result in both lost and duplicate recipients for a message.
162 JH/01 Avoid costly startup code when not strictly needed. This reduces time
163 for some exim process initialisations. It does mean that the logging
164 of TLS configuration problems is only done for the daemon startup.
166 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
168 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
169 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
172 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
173 DKIM signing is being done. Previously a transport_filter would always
174 disable CHUNKING, falling back to traditional DATA.
176 JH/05 Regard command-line receipients as tainted.
178 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
180 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
181 PAM library frees one of the arguments given to it, despite the
182 documentation. Therefore a plain malloc must be used.
184 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
185 on-stack buffers were used, resulting in a taint trap when DSN information
186 copied from a received message was written into the buffer.
188 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
189 the ordering of its ARC headers. This caused a crash.
191 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
192 a new record was being constructed with information from the peer, a trap
195 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
196 installation would get error messages from DMARC verify, when it hit the
197 nonexistent file indicated by the default. Distros wanting DMARC enabled
198 should both provide the file and set the option.
199 Also enforce no DMARC verification for command-line sourced messages.
201 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
202 could, depending on the platform, hang at the STARTTLS response.
204 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
205 message on a connection. Previously if one message had ARC headers and
206 the following one did not, a crash could result when adding an
207 Authentication-Results: header.
209 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
210 Exim main code and Exim-related utities. The introduction of taint
211 tracking also did many adjustments to string handling. Since then, eximon
212 frequently terminated with an assert failure.
214 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
215 check for 452 responses. This slightly helps the inefficieny of doing
216 a large alias-expansion into a recipient-limited target. The max_rcpt
217 transport option still applies (and at the current default, will override
218 the new feature). The check is done for either cause of synch, and forces
219 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
220 connection. The new facility is not tunable at this time.
222 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
223 library live data was being used, so the results became garbage. Make
224 copies while it is still usable.
226 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
227 on delivery deferred (==) and failed (**) lines (if a delivery was
228 attemtped). Previously it was only on completion (=>) lines.
230 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
231 for the expansion of the server_scram_iter and server_scram_salt options.
233 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
234 are now specifically given a NO_DATA response without hitting the system
235 resolver. The library goes on to do the now-standard TXT lookup.
236 Use of dnsdb lookups is not affected.
238 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
239 only retrieve the errormessage once. Previously two calls to dlerror()
240 were used, and the second one (for mainlog/paniclog) retrieved null
243 JH/20 Taint checking: disallow use of tainted data for
244 - the appendfile transport file and directory options
245 - the pipe transport command
246 - the autoreply transport file, log and once options
247 - file names used by the redirect router (including filter files)
249 - paths used by single-key lookups
250 Previously this was permitted.
252 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
253 adjusted the size of a major service buffer; this failed because the
254 buffer was in use at the time. Change to a compile-time increase in the
255 buffer size, when this authenticator is compiled into exim.
257 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
258 previous fast-mode was untenable in the face of glibs using mmap to
259 support larger malloc requests.
261 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
262 New values supported, if defined on system where compiled:
263 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
264 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
266 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
267 running a limited number of proceses in parallel, a benefit is gained. The
268 amount varies with the platform hardware and load. The use of the option
269 queue_run_in_order means we cannot do this, as ordering becomes
272 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
273 had introduced a string-copy (for ensuring NUL-termination) which was not
274 appropriate for that case, which can include embedded NUL bytes in the
275 block of data. Investigation showed the copy to actually be needless, the
276 data being length-specified.
278 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
279 done during a receiving connection, and both used TLS, global info was
280 used rather than per-connection info for tracking the state of data
281 queued for transmission. This could result in a connection hang.
283 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
284 Previously, when delivering serveral messages down a single connection
285 only the first would provide a SIZE. This was due to the size information
286 not being properly tracked.
288 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
289 TAI (at 37 seconds currently), pretend to be in UTC for time-related
290 expansion and logging. Previously, spurious values such as a future
291 minute could be seen.
293 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
294 it could crash from a null-deref. This could also affect the
295 ${addresses: } operator and ${readsock } item.
297 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
298 message following a mime one, the variable was not reset.
300 JH/30 When an pipelined-connect fails at the first response, assume incorrect
301 cached capability (perhaps the peer reneged?) and immediately retry in
304 JH/31 Fix spurious detection of timeout while writing to transport filter.
306 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
307 an attempt to copy the string was made before checking it.
309 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
310 taint of the lookup key was maintained; we now regard the presence in the
311 filesystem as sufficient validation.
313 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
314 argument is supplied.
316 JH/35 The dsearch lookup now requires that the directory is an absolute path.
317 Previously this was not checked, and nonempty relative paths made an
318 access under Exim's current working directory.
320 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
321 Previously no event was raised.
323 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
324 parameter supplied by the sender MAIL FROM command. Previously it was
325 ignored, and only the check_spool_space option value for the required
328 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
329 the size of the signing public-key. Previously it was instead giving
330 the size of the signature hash.
332 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
333 the default. See the (new) dkim_verify_min_keysizes option.
335 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
336 and an ACL use a lookup for checking either the local_part or domain,
337 stale data could be accessed. Ensure that variable references are
338 dropped between messages.
340 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
341 by the client was not checked as pointing within response data before
342 being used. A malicious client could thus cause an out-of-bounds read and
343 possibly gain authentication. Fix by adding the check.
345 JH/42 Internationalisation: change the default for downconversion in the smtp
346 transport to be "if needed". Previously it was "as previously set" for
347 the message, which usually meant "if needed" for message-submission but
348 "no" for everything else. However, MTAs have been seen using SMTPUTF8
349 even when the envelope addresses did not need it, resulting in forwarding
350 failures to non-supporting MTAs. A downconvert in such cases will be
351 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
352 transport. The change does mean that addresses needing conversion will
353 be converted when previously a delivery failure would occur.
355 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
356 response was received it would be used unchecked in a fail-DSN, violating
357 standards on line-length limits. Truncate if needed.
359 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
360 doesn't work. (Additionally add a new main config option to configure the
367 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
368 side, to decode a TLS 1.3 packet capture.
370 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
371 Previously the default library behaviour applied, sending two, each in
374 JH/03 Debug output for ACL now gives the config file name and line number for
377 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
379 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
381 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
382 buffer overrun for (non-chunking) other transports.
384 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
385 TLS1.3, means that a server rejecting a client certificate is not visible
386 to the client until the first read of encrypted data (typically the
387 response to EHLO). Add detection for that case and treat it as a failed
388 TLS connection attempt, so that the normal retry-in-clear can work (if
389 suitably configured).
391 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
392 and/or domain. Found and fixed by Jason Betts.
394 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
395 configuration). If a CNAME target was not a wellformed name pattern, a
398 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
399 the OS reports them interleaved with other addresses.
401 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
402 used both for input and for a verify callout, both encrypted, SMTP
403 responses being sent by the server could be lost. This resulted in
404 dropped connections and sometimes bounces generated by a peer sending
407 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
408 string. Previously it was possible to cause undefined behaviour in a
409 library routine (usually a crash). Found by "zerons".
411 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
414 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
415 API was removed, so update to use the newer ones.
417 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
418 any timeout set, is taking a long time. Previously we would hang on to a
419 rotated logfile "forever" if the input was arriving with long gaps
420 (a previous attempt to fix addressed lack, for a long time, of initial
423 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
424 shared (NFS) environment. The length of the tempfile name is now
425 4 + 16 ("hdr.$message_exim_id") which might break on file
426 systems which restrict the file name length to lower values.
429 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
430 shared (NFS) environment.
432 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
433 did for all versions <4.90). Notably -M, -m, --invert, -I may be
436 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
437 on some platforms for bit 31.
439 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
440 to changes apparently associated with TLS1.3 handling some of the APIs
441 previously used were either nonfunctional or inappropriate. Strings
442 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
443 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
444 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
445 This affects log line X= elements, the $tls_{in,out}_cipher variables,
446 and the use of specific cipher names in the encrypted= ACL condition.
448 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
450 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
451 verification result was not updated unless hosts_require_ocsp applied.
453 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
454 queue_list_requires_admin set to false, non-admin users were denied the
457 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
458 directory-of-certs mode. Previously they were advertised despite the
461 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
462 A single TCP connection by a client will now hold a TLS connection open
463 for multiple message deliveries, by default. Previously the default was to
466 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
467 default. If built with the facility, DANE will be used. The facility
468 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
470 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
471 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
472 must be defined and you must still, unless you define DISABLE_TLS, manage
473 the the include-dir and library-file requirements that go with that
474 choice. Non-TLS builds are still supported.
476 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
479 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
480 default. If the platform supports and has the facility enabled, it will
481 be requested on all coneections.
483 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
484 controlled by the build-time option SUPPORT_PIPE_CONNECT.
486 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
488 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
489 success-DSN messages. Previously the From: header was always the default
490 one for these; the option was ignored.
492 JH/28 Fix the timeout on smtp response to apply to the whole response.
493 Previously it was reset for every read, so a teergrubing peer sending
494 single bytes within the time limit could extend the connection for a
495 long time. Credit to Qualsys Security Advisory Team for the discovery.
497 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
498 delivery address, which leaked information of the results of local
499 forwarding. Change to the original envelope recipient address, per
502 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
503 requested. Previously not bounce was generated and a log entry of
504 error ignored was made.
506 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
508 JH/32 Introduce a general tainting mechanism for values read from the input
509 channel, and values derived from them. Refuse to expand any tainted
510 values, to catch one form of exploit.
512 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
513 was unused and the unexpanded text used for the test. Found and
514 fixed by Ruben Jenster.
516 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
517 an attempt to use a TLS library read routine dereffed a nul pointer,
520 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
521 them in our smtp response.
523 JH/36 Have the general router option retry_use_local_part default to true when
524 any of the restrictive preconditions are set (to anything). Previously it
525 was only for check_local user. The change removes one item of manual
526 configuration which is required for proper retries when a remote router
527 handles a subset of addresses for a domain.
529 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
530 link count into consideration.
532 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
533 caused the extension of big_buffer, the following lines were ignored.
535 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
536 accordance with RFC 2308. Previously there was no expiry, so a longlived
537 receive process (eg. due to ACL delays) versus a short SOA value could
540 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
542 JH/39 Promote DMARC support to mainline.
544 JH/40 Bug 2452: Add a References: header to DSNs.
546 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
547 parameters. The relevant library call is documented as "Deprecated: This
548 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
549 3.6.0, DH parameters are negotiated following RFC7919."
551 HS/06 Change the default of dnssec_request_domains to "*"
553 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
554 carried on and emitted a BDAT command, even when PIPELINING was not
557 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
558 buffer was used for the filename, resulting in a trap when tainted
559 arguments (eg. $domain) were used.
561 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
562 recommended to avoid a possible server-load attack. The feature can be
563 re-enabled via the openssl_options main cofiguration option.
565 JH/45 local_scan API: documented the current smtp_printf() call. This changed
566 for version 4.90 - adding a "more data" boolean to the arguments.
567 Bumped the ABI version number also, this having been missed previously;
568 release versions 4.90 to 4.92.3 inclusive were effectively broken in
569 respect of usage of smtp_printf() by either local_scan code or libraries
570 accessed via the ${dlfunc } expansion item. Both will need coding
571 adjustment for any calls to smtp_printf() to match the new function
572 signature; a FALSE value for the new argument is always safe.
574 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
575 the file-offset (which the Linux syscall does, and exim expects); this
576 resulted in an indefinite loop.
578 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
579 to do ARC verification. The Authentication-Results: header line added
580 by the configuration then had no ARC item.
586 JH/01 Remove code calling the customisable local_scan function, unless a new
587 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
589 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
590 non-signal-safe functions being used.
592 JH/03 Bug 2269: When presented with a received message having a stupidly large
593 number of DKIM-Signature headers, disable DKIM verification to avoid
594 a resource-consumption attack. The limit is set at twenty.
596 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
597 report of oldest_pass in ${authres } in consequence, and separate out
598 some descriptions of reasons for verification fail.
600 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
601 files in the spool were present and unlocked. A queue-runner could spot
602 them, resulting in a duplicate delivery. Fix that by doing the unlock
603 after the unlink. Investigation by Tim Stewart. Take the opportunity to
604 add more error-checking on spoolfile handling while that code is being
607 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
608 No known attacks, no CVE, this is defensive hardening.
610 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
611 a queue-runner could start a delivery while other operations were ongoing.
612 Cutthrough delivery was a common victim, resulting in duplicate delivery.
613 Found and investigated by Tim Stewart. Fix by using the open message data
614 file handle rather than opening another, and not locally closing it (which
615 releases a lock) for that case, while creating the temporary .eml format
616 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
618 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
619 $sender_verify_failure/$recipient_verify_failure to "random".
621 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
624 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
625 Previously this would segfault.
627 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
630 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
631 like zero, since the resolver should be doing this for us, But we need one
632 as a CNAME but no MX presence gets the CNAME returned; we need to check
633 that doesn't point to an MX to declare it "no MX returned" rather than
634 "error, loop". A new main option is added so the older capability of
635 following some limited number of chain links is maintained.
637 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
639 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
640 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
641 now-deprecated earlier definitions used only specified the range up to TLS
642 1.2 (in the older-version library docs).
644 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
646 JH/15 Rework TLS client-side context management. Stop using a global, and
647 explicitly pass a context around. This enables future use of TLS for
648 connections to service-daemons (eg. malware scanning) while a client smtp
649 connection is using TLS; with cutthrough connections this is quite likely.
651 JH/16 Fix ARC verification to do AS checks in reverse order.
653 JH/17 Support a "tls" option on the ${readsocket } expansion item.
655 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
656 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
657 Previously the "utf8" would be re-prepended for every additional message.
659 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
660 Previously thery were accepted, resulting in issues when attempting to
661 forward messages to a non-supporting MTA.
663 PP/02 Let -n work with printing macros too, not just options.
665 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
666 one parent address was copied, and bogus data was used at delivery-logging
667 time. Either a crash (after delivery) or bogus log data could result.
668 Discovery and analysis by Tim Stewart.
670 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
671 Previously if the string ended mid-character, we did not insert the
672 promised '?' replacement.
674 PP/04 Documentation: current string operators work on bytes, not codepoints.
676 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
677 should pack well giving a smaller memory footprint so better caching and
678 therefore performance. Group the declarations where this can't be done so
679 that the byte-sized flag variables are not interspersed among pointer
680 variables, giving a better chance of good packing by the compiler.
682 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
683 non-null, to avoid issues with sites running BATV. Previously reports were
684 sent with an empty envelope sender so looked like bounces.
686 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
687 The ignore_error flag wasn't being returned from the filter subprocess so
688 was not set for later routers. Investigation and fix by Matthias Kurz.
690 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
691 and a msg:complete for the whole, when a message is manually removed using
692 -Mrm. Developement by Matthias Kurz, hacked on by JH.
694 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
695 a "Gnu special" function, asprintf() in the DB utility binary builds; I
696 hope that is portable enough.
698 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
699 requiring a known-CA anchor certificate; make it now rely entirely on the
700 TLSA as an anchor. Checking the name on the leaf cert against the name
701 on the A-record for the host is still done for TA (but not for EE mode).
703 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
704 included in delivery lines for non-proxied connections, when compiled with
705 SUPPORT_SOCKS and running with proxy logging enabled.
707 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
708 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
709 move the existing event to fire before the normal logging of message
710 failure so that custom logging is bracketed by normal logging.
712 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
713 msg:fail:internal event. Developement by Matthias Kurz.
715 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
716 far too small for todays use of crypto signatures stored there. Go all
717 the way to the max DNS message size of 64kB, even though this might be
718 overmuch for IOT constrained device use.
720 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
721 copy a string over itself. The library routine is documented as not
722 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
724 JH/32 For main options check_spool_space and check_inode_space, where the
725 platform supports 64b integers, support more than the previous 2^31 kB
726 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
727 the previous G, M, k.
729 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
730 $authenticated_fail_id variable on authentication failure. Previously
733 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
734 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
735 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
736 GNUTLS_SEC_PARAM_MEDIUM.
738 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
739 side. Previously we would continue as if no SNI had been received.
741 JH/36 Harden the handling of string-lists. When a list consisted of a sole
742 "<" character, which should be a list-separator specification, we walked
743 off past the nul-terimation.
745 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
746 causes) even when the retry time is not yet met. Previously they were
747 not, meaning that when (say) an account was over-quota and temp-rejecting,
748 and multiple senders' messages were queued, only one sender would get
749 notified on each configured delay_warning cycle.
751 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
753 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
754 cipher-suites, an error can be left on the stack even for a succeeding
755 accept; this results in impossible error messages when a later operation
758 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
759 return error codes indicating retry. Under TLS1.3 this becomes required.
761 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
762 it only wrote the new authenticators, resulting in a lack of tracking of
763 peer changes of ESMTP extensions until the next cache flush.
765 JH/41 Fix the loop reading a message header line to check for integer overflow,
766 and more-often against header_maxsize. Previously a crafted message could
767 induce a crash of the recive process; now the message is cleanly rejected.
769 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
770 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
776 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
777 When redis_servers is set to a list of > 1 element, and the Redis servers
778 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
779 case of MOVED into a DEFER case instead, thus moving the query onto the
780 next server in the list. For a cluster of N elements, all N servers must
781 be defined in redis_servers.
783 GF/02 Catch and remove uninitialized value warning in exiqsumm
784 Check for existence of @ARGV before looking at $ARGV[0]
786 JH/01 Replace the store_release() internal interface with store_newblock(),
787 which internalises the check required to safely use the old one, plus
788 the allocate and data copy operations duplicated in both (!) of the
789 extant use locations.
791 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
792 modifier. This matches the restriction on the commandline.
794 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
795 Previously only the last row was returned.
797 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
798 we assumed that tags in the header were well-formed, and parsed the
799 element content after inspecting only the first char of the tag.
800 Assumptions at that stage could crash the receive process on malformed
803 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
804 While running the DKIM ACL we operate on the Permanent memory pool so that
805 variables created with "set" persist to the DATA ACL. Also (at any time)
806 DNS lookups that fail create cache records using the Permanent pool. But
807 expansions release any allocations made on the current pool - so a dnsdb
808 lookup expansion done in the DKIM ACL releases the memory used for the
809 DNS negative-cache, and bad things result. Solution is to switch to the
810 Main pool for expansions.
811 While we're in that code, add checks on the DNS cache during store_reset,
812 active in the testsuite.
813 Problem spotted, and debugging aided, by Wolfgang Breyha.
815 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
816 When none of the hosts presented to a transport match an already-open
817 connection, close it and proceed with the list. Previously we would
818 queue the message. Spotted by Lena with Yahoo, probably involving
821 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
822 Previously a spurious "250 OK id=" response was appended to the proper
825 JH/08 The "support for" informational output now, which built with Content
826 Scanning support, has a line for the malware scanner interfaces compiled
827 in. Interface can be individually included or not at build time.
829 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
830 by the template makefile "src/EDITME". The "STREAM" support for an older
831 ClamAV interface method is removed.
833 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
834 rows affected is given instead).
836 JH/11 The runtime Berkeley DB library version is now additionally output by
837 "exim -d -bV". Previously only the compile-time version was shown.
839 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
840 SMTP connection. Previously, when one had more receipients than the
841 first, an abortive onward connection was made. Move to full support for
842 multiple onward connections in sequence, handling cutthrough connection
843 for all multi-message initiating connections.
845 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
846 routers. Previously, a multi-recipient message would fail to match the
847 onward-connection opened for the first recipient, and cause its closure.
849 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
850 a timeout on read on a GnuTLS initiating connection, resulting in the
851 initiating connection being dropped. This mattered most when the callout
852 was marked defer_ok. Fix to keep the two timeout-detection methods
855 JH/15 Relax results from ACL control request to enable cutthrough, in
856 unsupported situations, from error to silently (except under debug)
857 ignoring. This covers use with PRDR, frozen messages, queue-only and
860 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
862 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
863 metadata, resulting in a crash in free().
865 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
866 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
867 Broken also in d185889f4, with init system revamp.
869 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
870 Previously we abruptly closed the connection after reading a malware-
871 found indication; now we go on to read the "scan ok" response line,
874 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
875 ACL. Previously, a crash would result.
877 JH/19 Speed up macro lookups during configuration file read, by skipping non-
878 macro text after a replacement (previously it was only once per line) and
879 by skipping builtin macros when searching for an uppercase lead character.
881 JH/20 DANE support moved from Experimental to mainline. The Makefile control
882 for the build is renamed.
884 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
885 was allocated for every new TLS startup, meaning one per message. Fix
886 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
888 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
889 reported the original. Fix to report (as far as possible) the ACL
890 result replacing the original.
892 JH/23 Fix memory leak during multi-message connections using STARTTLS under
893 OpenSSL. Certificate information is loaded for every new TLS startup,
894 and the resources needed to be freed.
896 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
898 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
901 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
902 DATA response info to the (existing) per-recipient response info for
903 the "C=" log element. It can have useful tracking info from the
904 destination system. Patch from Simon Arlott.
906 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
907 length value. Previously this would segfault.
909 HS/02 Support Avast multiline protoocol, this allows passing flags to
910 newer versions of the scanner.
912 JH/28 Ensure that variables possibly set during message acceptance are marked
913 dead before release of memory in the daemon loop. This stops complaints
914 about them when the debug_store option is enabled. Discovered specifically
915 for sender_rate_period, but applies to a whole set of variables.
916 Do the same for the queue-runner and queue-list loops, for variables set
917 from spool message files. Do the same for the SMTP per-message loop, for
918 certain variables indirectly set in ACL operations.
920 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
921 as a multi-recipient message from a mailinglist manager). The coding had
922 an arbitrary cutoff number of characters while checking for more input;
923 enforced by writing a NUL into the buffer. This corrupted long / fast
924 input. The problem was exposed more widely when more pipelineing of SMTP
925 responses was introduced, and one Exim system was feeding another.
926 The symptom is log complaints of SMTP syntax error (NUL chars) on the
927 receiving system, and refused recipients seen by the sending system
928 (propating to people being dropped from mailing lists).
929 Discovered and pinpointed by David Carter.
931 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
932 replaced by the ${authresults } expansion.
934 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
936 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
937 allows proper process termination in container environments.
939 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
940 Previously the "final dot" had a newline after it; ensure it is CR,LF.
942 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
943 and "err_perm", deprecated since 4.83 when the RFC-defined words
944 "temperror" and "permerror" were introduced.
946 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
947 transport-filters or DKIM-signing. The restriction was lost in the
948 consolidation of verify-callout and delivery SMTP handling.
949 Extend the restriction to also cover ARC-signing.
951 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
952 in defer=pass mode supply a 450 to the initiator. Previously the message
955 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
956 tls_require_ciphers is used as before.
958 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
959 "pass_unscanned". Only tmpfails from the scanner are written to
960 the paniclog, as they may require admin intervention (permission
961 denied, license issues). Other scanner errors (like decompression
962 bombs) do not cause a paniclog entry.
964 JH/36 Fix reinitialisation of DKIM logging variable between messages.
965 Previously it was possible to log spurious information in receive log
968 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
969 triggered odd behaviour from Outlook Express clients.
971 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
974 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
975 since the IETF WG has not yet settled on that versus the original
976 "bare" representation.
978 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
979 Previously the millisecond value corrupted the output.
980 Fix also for syslog_pid=no and log_selector +pid, for which the pid
981 corrupted the output.
987 JH/01 Rework error string handling in TLS interface so that the caller in
988 more cases is responsible for logging. This permits library-sourced
989 string to be attached to addresses during delivery, and collapses
990 pairs of long lines into single ones.
992 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
993 during configuration. Wildcards are allowed and expanded.
995 JH/02 Rework error string handling in DKIM to pass more info back to callers.
996 This permits better logging.
998 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
999 do not close it down and have the child transport start it up again on
1000 the passed-on TCP connection. Instead, proxy the child (and any
1001 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1002 affected: the continued delivery log lines do not have any DNSSEC, TLS
1003 Certificate or OCSP information. TLS cipher information is still logged.
1005 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1006 identical IP addresses on different listening ports. Will also affect
1009 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1010 add noisy ifdef guards to special-case this sillyness.
1011 Patch from Bernd Kuhls.
1013 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1014 than 255 are no longer allowed.
1016 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1017 Disable the session-cache too, which might reduce our load. Since we
1018 currrectly use a new context for every connection, both as server and
1019 client, there is no benefit for these.
1020 GnuTLS appears to not support tickets server-side by default (we don't
1021 call gnutls_session_ticket_enable_server()) but client side is enabled
1022 by default on recent versions (3.1.3 +) unless the PFS priority string
1025 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1026 <https://reproducible-builds.org/specs/source-date-epoch/>.
1028 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1029 the check for any unsuccessful recipients did not notice the limit, and
1030 erroneously found still-pending ones.
1032 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1033 MSG_MORE. Only in-clear (not on TLS connections).
1035 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1036 transport-filter is involved we can buffer the headers in memory for
1037 creating the signature, and read the spool data file once for the
1038 signature and again for transmission.
1040 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1041 4.77 as the kernel support then wasn't solid, having issues in 64bit
1042 mode. Now, it's been long enough. Add support for FreeBSD also.
1044 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1045 case where the routing stage had gathered several addresses to send to
1046 a host before calling the transport for the first, we previously failed
1047 to close down TLS in the old transport process before passing the TCP
1048 connection to the new process. The new one sent a STARTTLS command
1049 which naturally failed, giving a failed delivery and bloating the retry
1050 database. Investigation and fix prototype from Wolfgang Breyha.
1052 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1053 false-negatives in the check that the sender had not preempted a response
1054 or prompt from Exim (running as a server), due to that code's lack of
1055 awareness of the SMTP input buffering.
1057 PP/04 Add commandline_checks_require_admin option.
1058 Exim drops privileges sanely, various checks such as -be aren't a
1059 security problem, as long as you trust local users with access to their
1060 own account. When invoked by services which pass untrusted data to
1061 Exim, this might be an issue. Set this option in main configuration
1062 AND make fixes to the calling application, such as using `--` to stop
1065 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1066 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1067 responses to those, into a single TLS record each way (this usually means
1068 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1071 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1072 affects you only if you're dancing at the edge of the param size limits.
1073 If you are, and this message makes sense to you, then: raise the
1074 configured limit or use OpenSSL 1.1. Nothing we can do for older
1077 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1078 cmdline element to get the documented default one. Previously it was
1081 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1082 are unsuable for non-message-specific reasons - eg. connection timeout,
1083 banner-time rejection.
1085 JH/16 Fix logging of delivery remote port, when specified by router, under
1088 PP/06 Repair manualroute's ability to take options in any order, even if one
1089 is the name of a transport.
1092 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1094 JH/17 Change the list-building routines interface to use the expanding-string
1095 triplet model, for better allocation and copying behaviour.
1097 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1098 Previously it was constructed the first time a possibly-matching string
1099 was met in the configuration file input during startup; now it is done
1102 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1103 compatible one, to avoid the (poorly documented) possibility of a config
1104 file in the working directory redirecting the DB files, possibly correpting
1105 some existing file. CVE-2017-10140 assigned for BDB.
1107 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1108 cache-hot. Previously, although the result was properly cached, the
1109 initial verify call returned a defer.
1111 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1112 the main verify for receipient in uncached-mode.
1114 JH/22 Retire historical build files to an "unsupported" subdir. These are
1115 defined as "ones for which we have no current evidence of testing".
1117 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1118 if present. Previously it was ignored.
1120 JH/24 Start using specified-initialisers in C structure init coding. This is
1121 a C99 feature (it's 2017, so now considered safe).
1123 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1124 if was a fixed-sized field and bitmask ops via macros; it is now more
1127 PP/07 GitHub PR 56: Apply MariaDB build fix.
1128 Patch provided by Jaroslav Škarvada.
1130 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1131 during Coverity cleanups [4.87 JH/47]
1132 Diagnosis and fix provided by Michael Fischer v. Mollard.
1134 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1135 the right size to place the terminating semicolon on its own folded
1136 line, the header hash was calculated to an incorrect value thanks to
1137 the (relaxed) space the fold became.
1139 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1140 and confused the parent.
1142 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1143 which could crash as a result. This could lead to undeliverable messages.
1145 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1148 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1149 matters on fast-turnover and PID-randomising systems, which were getting
1150 out-of-order delivery.
1152 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1153 a possibly-overlapping copy. The symptom was that "Remote host closed
1154 connection in response to HELO" was logged instead of the actual 4xx
1157 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1158 Previously only that bufferd was discarded, resulting in SYMTP command
1161 JH/32 DKIM: when a message has multiple signatures matching an identity given
1162 in dkim_verify_signers, run the dkim acl once for each. Previously only
1163 one run was done. Bug 2189.
1165 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1166 "panic the current process" to "deliberately defer". The panic log is
1167 still written with the problem list name; the mail and reject logs now
1168 get a temp-reject line for the message that was being handled, saying
1169 something like "domains check lookup or other defer". The SMTP 451
1170 message is still "Temporary local problem".
1172 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1173 A crafted sequence of BDAT commands could result in in-use memory beeing
1174 freed. CVE-2017-16943.
1176 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1177 from SMTP input. Previously it was always done; now only done for DATA
1178 and not BDAT commands. CVE-2017-16944.
1180 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1181 to the message (such as an overlong header line). Previously this was
1182 not done and we did not exit BDAT mode. Followon from the previous item
1183 though a different problem.
1189 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1190 than -2003 did; needs libidn2 in addition to libidn.
1192 JH/02 The path option on a pipe transport is now expanded before use.
1194 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1195 Patch provided by "Björn", documentation fix added too.
1197 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1198 missing a wire-to-host endian conversion.
1200 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1201 close after a BDAT command line could be taken as a following command,
1202 giving a synch failure. Fix by only checking for synch immediately
1203 before acknowledging the chunk.
1205 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1206 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1207 macro. Patches provided by Josh Soref.
1209 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1210 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1211 by IANA as a service extension. However, John Klensin suggests that we
1214 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1215 the dkim code may be unix-mode line endings rather than smtp wire-format
1216 CRLF, so prepend a CR to any bare LF.
1218 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1219 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1221 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1222 the first were themselves being wrongly included in the feed into dkim
1223 processing; with most chunk sizes in use this resulted in an incorrect
1224 body hash calculated value.
1226 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1227 DKIM signature block, for verification. Although advised against by
1228 standards it is specifically not ruled illegal.
1230 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1232 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1233 missing a body hash (the bh= tag).
1235 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1236 It seems that HAProxy sends the Proxy Protocol information in clear and
1237 only then does a TLS startup, so do the same.
1239 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1240 TCP connections (such as for Spamd) unless the daemon successfully set
1241 Fast Open mode on its listening sockets. This fixes breakage seen on
1242 too-old kernels or those not configured for Fast Open, at the cost of
1243 requiring both directions being enabled for TFO, and TFO never being used
1244 by non-daemon-related Exim processes.
1246 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1247 endings, at least on the first header line. Try to canonify any that get
1248 past that check, despite the cost.
1250 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1251 now limited to an arbitrary five deep, while parsing addresses with the
1252 strip_excess_angle_brackets option enabled.
1254 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1255 instead leave the unprompted TLS handshake in socket buffer for the
1256 TLS library to consume.
1258 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1260 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1262 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1263 region in one operation, for speed, and this leaves assigned pointers
1264 dangling. Add checks run only under the testsuite which checks all
1265 variables at a store-reset and panics on a dangling pointer; add code
1266 explicitly nulling out all the variables discovered. Fixes one known
1267 bug: a transport crash, where a dangling pointer for $sending_ip_address
1268 originally assigned in a verify callout, is re-used.
1270 PP/06 Drop '.' from @INC in various Perl scripts.
1272 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1274 PP/08 Reduce a number of compilation warnings under clang; building with
1275 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1276 should be warning-free.
1278 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1280 HS/01 Fix portability problems introduced by PP/08 for platforms where
1281 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1283 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1284 chunk. This allows us to accept broken chunked messages. We need a more
1285 general solution here.
1287 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1288 already-broken messages in the queue.
1290 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1292 JH/19 Fix reference counting bug in routing-generated-address tracking.
1298 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1299 supports it and a size is available (ie. the sending peer gave us one).
1301 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1302 years of being deprecated). The replacements are the ACLs
1303 acl_smtp_mime and acl_not_smtp_mime.
1305 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1306 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1307 or even an in-clear connection were permitted. Now, if the host lookup
1308 was dnssec and dane was requested then the host is only used if the
1309 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1310 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1311 if one fails this test.
1312 This means that a poorly-configured remote DNS will make it incommunicado;
1313 but it protects against a DNS-interception attack on it.
1315 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1316 non-noisy when a race steals the message being considered.
1318 JH/05 If main configuration option tls_certificate is unset, generate a
1319 self-signed certificate for inbound TLS connections.
1321 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1322 in rewrites and routers.
1324 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1325 and logged a warning sing 4.83; now they are a configuration file error.
1327 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1328 (lacking @domain). Apply the same qualification processing as RCPT.
1330 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1332 JH/10 Support ${sha256:} applied to a string (as well as the previous
1335 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1336 a cutthrough deliver is pending, as we always want to make a connection.
1337 This also avoids re-routing the message when later placing the cutthrough
1338 connection after a verify cache hit.
1339 Do not update it with the verify result either.
1341 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1342 when routing results in more than one destination address.
1344 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1345 signing (which inhibits the cutthrough capability). Previously only
1346 the presence of an option was tested; now an expansion evaluating as
1347 empty is permissible (obviously it should depend only on data available
1348 when the cutthrough connection is made).
1350 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1351 the relevant preceding SMTP command did not note the pipelining mode.
1353 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1354 Previously they were not counted.
1356 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1357 as one having no matching records. Previously we deferred the message
1358 that needed the lookup.
1360 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1361 distinguished as "(=".
1363 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1364 for missing MX records. Previously it only worked for missing A records.
1366 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1368 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1369 after the data-go-ahead and data-ack. Patch from Jason Betts.
1371 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1372 even for a "none" policy. Patch from Tony Meyer.
1374 JH/22 Fix continued use of a connection for further deliveries. If a port was
1375 specified by a router, it must also match for the delivery to be
1378 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1379 When one of the recipients of a message was unsuitable for the connection
1380 (has no matching addresses), we lost track of needing to mark it
1381 deferred. As a result mail would be lost.
1383 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1385 JH/25 Decoding ACL controls is now done using a binary search; the source code
1386 takes up less space and should be simpler to maintain. Merge the ACL
1387 condition decode tables also, with similar effect.
1389 JH/26 Fix problem with one_time used on a redirect router which returned the
1390 parent address unchanged. A retry would see the parent address marked as
1391 delivered, so not attempt the (identical) child. As a result mail would
1394 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1395 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1396 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1399 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1400 Main config options check_{log,spool}_{inodes,space} are now
1401 100 inodes, 10MB unless set otherwise in the configuration.
1403 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1404 Previously it only applied to the main-section connection policy
1407 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1409 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1410 by me. Added RFC7919 DH primes as an alternative.
1412 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1413 are not in the system include path.
1415 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1416 GnuTLS, when a session startup failed (eg because the client disconnected)
1417 Exim did stdio operations after fclose. This was exposed by a recent
1418 change which nulled out the file handle after the fclose.
1420 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1421 signed directly by the cert-signing cert, rather than an intermediate
1422 OCSP-signing cert. This is the model used by LetsEncrypt.
1424 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1426 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1427 an incoming connection.
1429 HS/02 Bug 1802: Do not half-close the connection after sending a request
1432 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1433 fallback to "prime256v1".
1435 JH/34 SECURITY: Use proper copy of DATA command in error message.
1436 Could leak key material. Remotely exploitable. CVE-2016-9963.
1442 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1443 and 3.4.4 - once the server is enabled to respond to an OCSP request
1444 it does even when not requested, resulting in a stapling non-aware
1445 client dropping the TLS connection.
1447 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1448 support variable-length bit vectors. No functional change.
1450 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1451 The I= interface field on outgoing lines is now after the H= remote
1452 host field, same as incoming lines. There is a separate
1453 outgoing_interface log selector which allows you to disable the
1456 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1457 If not running log_selector +smtp_connection the mainlog would be held
1458 open indefinitely after a "too many connections" event, including to a
1459 deleted file after a log rotate. Leave the per net connection logging
1460 leaving it open for efficiency as that will be quickly detected by the
1461 check on the next write.
1463 HS/01 Bug 1671: Fix post transport crash.
1464 Processing the wait-<transport> messages could crash the delivery
1465 process if the message IDs didn't exist for some reason. When
1466 using 'split_spool_directory=yes' the construction of the spool
1467 file name failed already, exposing the same netto behaviour.
1469 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1470 mime_regex ACL conditions.
1472 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1473 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1474 to HELO, local diagnostic string.
1476 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1477 log line to debug. Even when configured with a tls authenticator many
1478 client connections are expected to not authenticate in this way, so
1479 an authenticate fail is not an error.
1481 HS/02 Add the Exim version string to the process info. This way exiwhat
1482 gives some more detail about the running daemon.
1484 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1485 matter for fast-change records such as DNSBLs.
1487 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1488 retry record, even if constant. There may be multiple transports with
1489 different interface settings and the retry behaviour needs to be kept
1492 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1494 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1496 JH/10 Bug 840: fix log_defer_output option of pipe transport
1498 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1499 hosts_randomize. This matters a lot when combined with mua_wrapper.
1501 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1502 ${quote_pgsql:<string>} operator.
1504 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1506 JH/14 Change header file name for freeradius-client. Relevant if compiling
1507 with Radius support; from the Gentoo tree and checked under Fedora.
1509 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1511 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1512 option transport_filter, meaning no filtering.
1514 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1516 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1517 defaults to "*" (all hosts). The variable is now available when not built
1518 with TLS, default unset, mainly to enable keeping the testsuite sane.
1519 If a server certificate is not supplied (via tls_certificate) an error is
1520 logged, and clients will find TLS connections fail on startup. Presumably
1521 they will retry in-clear.
1522 Packagers of Exim are strongly encouraged to create a server certificate
1523 at installation time.
1525 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1526 with the $config_file variable.
1528 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1529 in transport context, after the attempt, and per-recipient. The latter type
1530 is per host attempted. The event data is the error message, and the errno
1531 information encodes the lookup type (A vs. MX) used for the (first) host,
1532 and the trailing two digits of the smtp 4xx response.
1534 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1535 to write to mainlog (or rejectlog, paniclog) in the window between file
1536 creation and permissions/ownership being changed. Particularly affects
1537 installations where exicyclog is run as root, rather than exim user;
1538 result is that the running daemon panics and dies.
1540 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1542 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1543 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1544 "pri" and "weight". Note that the previous implicit priority given by the
1545 list order is no longer honoured.
1547 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1548 for DKIM processing.
1550 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1551 by defining SUPPORT_SOCKS.
1553 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1554 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1555 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1556 variables are renamed to proxy_{local,external}_{address,port}.
1558 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1559 for a build by defining SUPPORT_I18N
1561 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1562 of the query string, and make ${quote_redis:} do that quoting.
1564 JH/29 Move Events support from Experimental to mainline, enabled by default
1565 and removable for a build by defining DISABLE_EVENT.
1567 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1569 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1570 cached by the daemon.
1572 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1573 by defining LOOKUP_REDIS. The libhiredis library is required.
1575 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1576 keys are given for lookup.
1578 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1579 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1580 only supported when built with TLS support. The PolarSSL SHA routines
1581 are still used when the TLS library is too old for convenient support.
1583 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1584 openssl_options), for security. OpenSSL forces this from version 1.1.0
1585 server-side so match that on older versions.
1587 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1588 allocation for $value could be released as the expansion processing
1589 concluded, but leaving the global pointer active for it.
1591 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1592 and to use the domains and local_parts ACL conditions.
1594 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1595 incorrectly not doubled on cutthrough transmission, hence seen as a
1596 body-termination at the receiving system - resulting in truncated mails.
1597 Commonly the sender saw a TCP-level error, and retransmitted the message
1598 via the normal store-and-forward channel. This could result in duplicates
1599 received - but deduplicating mailstores were liable to retain only the
1600 initial truncated version.
1602 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1604 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1606 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1607 we're in there, support oversigning also; bug 1309.
1609 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1611 HS/04 Add support for keep_environment and add_environment options.
1613 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1614 either intentional arithmetic overflow during PRNG, or testing config-
1617 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1618 delivery resulted in actual delivery. Cancel cutthrough before DATA
1621 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1622 rejects a recipient: pass the reject to the originator.
1624 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1625 Many were false-positives and ignorable, but it's worth fixing the
1628 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1629 for the new environment-manipulation done at startup. Move the routines
1630 from being local to tls.c to being global via the os.c file.
1632 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1633 an extract embedded as result-arg for a map, the first arg for extract
1634 is unavailable so we cannot tell if this is a numbered or keyed
1635 extraction. Accept either.
1641 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1644 JH/02 The smtp transport option "multi_domain" is now expanded.
1646 JH/03 The smtp transport now requests PRDR by default, if the server offers
1649 JH/04 Certificate name checking on server certificates, when exim is a client,
1650 is now done by default. The transport option tls_verify_cert_hostnames
1651 can be used to disable this per-host. The build option
1652 EXPERIMENTAL_CERTNAMES is withdrawn.
1654 JH/05 The value of the tls_verify_certificates smtp transport and main options
1655 default to the word "system" to access the system default CA bundle.
1656 For GnuTLS, only version 3.0.20 or later.
1658 JH/06 Verification of the server certificate for a TLS connection is now tried
1659 (but not required) by default. The verification status is now logged by
1660 default, for both outbound TLS and client-certificate supplying inbound
1663 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1666 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1667 Status Notification (bounce) messages are now MIME format per RFC 3464.
1668 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1669 under the control of the dsn_advertise_hosts option, and routers may
1670 have a dsn_lasthop option.
1672 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1673 default, modifiable by a malware= option. The list separator for
1674 the options can now be changed in the usual way. Bug 68.
1676 JH/10 The smtp_receive_timeout main option is now expanded before use.
1678 JH/11 The incoming_interface log option now also enables logging of the
1679 local interface on delivery outgoing connections.
1681 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1682 if the interface and destination host and port all match.
1684 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1687 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1688 Patch from Andrew Lewis.
1690 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1691 now supports optional time-restrictions, weighting, and priority
1692 modifiers per server. Patch originally by <rommer@active.by>.
1694 JH/16 The spamd_address main option now supports a mixed list of local
1695 and remote servers. Remote servers can be IPv6 addresses, and
1696 specify a port-range.
1698 JH/17 Bug 68: The spamd_address main option now supports an optional
1699 timeout value per server.
1701 JH/18 Bug 1581: Router and transport options headers_add/remove can
1702 now have the list separator specified.
1704 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1707 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1710 JH/21 Support for the A6 type of dns record is withdrawn.
1712 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1713 rather than the verbs used.
1715 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1716 from 255 to 1024 chars.
1718 JH/24 Verification callouts now attempt to use TLS by default.
1720 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1721 are generic router options now. The defaults didn't change.
1723 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1724 Original patch from Alexander Shikoff, worked over by JH.
1726 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1727 files if ZCAT_COMMAND is not executable.
1729 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1731 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1733 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1734 Normally benign, it bites when the pair was led to by a CNAME;
1735 modern usage is to not canonicalize the domain to a CNAME target
1736 (and we were inconsistent anyway for A-only vs AAAA+A).
1738 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1740 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1741 when evaluating $sender_host_dnssec.
1743 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1744 $sender_helo_dnssec variable.
1746 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1748 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1750 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1752 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1753 documented as working, but never had. Support all but $spam_report.
1755 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1756 added for tls authenticator.
1758 HS/03 Add perl_taintmode main config option
1764 TL/01 When running the test suite, the README says that variables such as
1765 no_msglog_check are global and can be placed anywhere in a specific
1766 test's script, however it was observed that placement needed to be near
1767 the beginning for it to behave that way. Changed the runtest perl
1768 script to read through the entire script once to detect and set these
1769 variables, reset to the beginning of the script, and then run through
1770 the script parsing/test process like normal.
1772 TL/02 The BSD's have an arc4random API. One of the functions to induce
1773 adding randomness was arc4random_stir(), but it has been removed in
1774 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1775 function when detected.
1777 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1778 cause callback expansion.
1780 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1781 syntax errors in an expansion can be treated as a string instead of
1782 logging or causing an error, due to the internal use of bool_lax
1783 instead of bool when processing it.
1785 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1786 server certificates when making smtp deliveries.
1788 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1790 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1792 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1794 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1795 Merged patch from Sebastian Wiedenroth.
1797 JH/05 Fix results-pipe from transport process. Several recipients, combined
1798 with certificate use, exposed issues where response data items split
1799 over buffer boundaries were not parsed properly. This eventually
1800 resulted in duplicates being sent. This issue only became common enough
1801 to notice due to the introduction of connection certificate information,
1802 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1804 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1805 size buffer was used, resulting in syntax errors when an expansion
1808 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1809 version 3.3.6 or later.
1811 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1812 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1813 both become "event_action", the variables become $event_name, $event_data
1814 and $event_defer_errno. There is a new variable $verify_mode, usable in
1815 routers, transports and related events. The tls:cert event is now also
1816 raised for inbound connections, if the main configuration event_action
1819 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1820 early OCSP support, but no stapling (appears to be less than 1.0.0).
1822 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1823 server certificate names available under the smtp transport option
1824 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1827 JH/10 Time-related extraction expansions from certificates now use the main
1828 option "timezone" setting for output formatting, and are consistent
1829 between OpenSSL and GnuTLS compilations. Bug 1541.
1831 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1832 encoded parameter in the incoming message. Bug 1558.
1834 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1835 include certificate info, eximon was claiming there were spoolfile
1838 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1840 JH/14 Log delivery-related information more consistently, using the sequence
1841 "H=<name> [<ip>]" wherever possible.
1843 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1844 are problematic for Debian distribution, omit them from the release
1847 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1849 JH/16 Fix string representation of time values on 64bit time_t architectures.
1852 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1853 output list separator was used.
1858 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1859 checkers that were complaining about end of non-void function with no
1862 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1863 This was a regression introduced in 4.83 by another bugfix.
1865 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1867 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1868 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1874 TF/01 Correctly close the server side of TLS when forking for delivery.
1876 When a message was received over SMTP with TLS, Exim failed to clear up
1877 the incoming connection properly after forking off the child process to
1878 deliver the message. In some situations the subsequent outgoing
1879 delivery connection happened to have the same fd number as the incoming
1880 connection previously had. Exim would try to use TLS and fail, logging
1881 a "Bad file descriptor" error.
1883 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1884 utilities have not been installed.
1886 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1887 temporary space as the ACL may create new global variables.
1889 TL/01 LDAP support uses per connection or global context settings, depending
1890 upon the detected version of the libraries at build time.
1892 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1893 to extract and use the src ip:port in logging and expansions as if it
1894 were a direct connection from the outside internet. PPv2 support was
1895 updated based on HAProxy spec change in May 2014.
1897 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1899 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1900 Properly escape header and check for NULL return.
1902 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1905 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1907 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1908 characters in header names, implemented as a verify condition.
1909 Contributed by Michael Fischer v. Mollard.
1911 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1912 results permerror and temperror. Previous values are deprecated but
1913 still accepted. In a future release, err_perm and err_temp will be
1914 completely removed, which will be a backward incompatibility if the
1915 ACL tests for either of these two old results. Patch contributed by
1916 user bes-internal on the mailing list.
1918 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1920 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1921 selectors, in both main and reject logs.
1923 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1926 JH/07 Add malware type "sock" for talking to simple daemon.
1928 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1930 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1931 routers/transports under cutthrough routing.
1933 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1934 numbers. Touch up "bool" conditional to keep the same definition.
1936 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1938 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1940 JH/12 Expand items in router/transport headers_add or headers_remove lists
1941 individually rather than the list as a whole. Bug 1452.
1943 Required for reasonable handling of multiple headers_ options when
1944 they may be empty; requires that headers_remove items with embedded
1945 colons must have them doubled (or the list-separator changed).
1947 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1948 view the policy declared in the DMARC record. Currently, $dmarc_status
1949 is a combined value of both the record presence and the result of the
1952 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1954 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1955 dnslookup router and the smtp transport (applying to the forward
1958 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1959 of ldap servers used for a specific lookup. Patch provided by Heiko
1962 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1963 New variable $lookup_dnssec_authenticated for observability.
1965 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1966 Patch submitted by Lars Timman.
1968 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1970 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1971 Requires trusted mode and valid format message id, aborts otherwise.
1972 Patch contributed by Heiko Schlichting.
1974 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1975 certextract with support for various fields. Bug 1358.
1977 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1978 is requested by default, modifiable by smtp transport option
1981 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1982 operate on certificate variables to give certificate fingerprints
1983 Also new ${sha256:cert_variable}.
1985 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1987 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1988 Christian Aistleitner.
1990 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1992 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1993 file. Patch from Wolfgang Breyha.
1995 JH/25 Expand the coverage of the delivery $host and $host_address to
1996 client authenticators run in verify callout. Bug 1476.
1998 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1999 align with daemon_smtp_ports. Bug 72.
2001 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2002 support and error reporting did not work properly.
2004 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2005 and is readable. Patch from Andrew Colin Kissa.
2007 TL/14 Enhance documentation of ${run expansion and how it parses the
2008 commandline after expansion, particularly in the case when an
2009 unquoted variable expansion results in an empty value.
2011 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2013 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2014 by renaming away from it. Addresses GH issue 15, reported by
2017 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2019 TL/15 SECURITY: prevent double expansion in math comparison functions
2020 (can expand unsanitized data). Not remotely exploitable.
2027 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2029 PP/02 Make -n do something, by making it not do something.
2030 When combined with -bP, the name of an option is not output.
2032 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2035 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2036 $sender_host_name and config options to manage this, and basic check
2039 PP/05 DSCP support for outbound connections and control modifier for inbound.
2041 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2042 (Only plugin which currently uses this is kerberos4, which nobody should
2043 be using, but we should make it available and other future plugins might
2044 conceivably use it, even though it would break NAT; stuff *should* be
2045 using channel bindings instead).
2047 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2048 name; added for Sendmail compatibility; requires admin caller.
2049 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2050 just ignore it); requires trusted caller.
2051 Also parse but ignore: -Ac -Am -X<logfile>
2054 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2056 TL/02 Add +smtp_confirmation as a default logging option.
2058 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2059 Patch by Magnus Holmgren from 2007-02-20.
2061 TL/04 Bugzilla 1281 - Spec typo.
2062 Bugzilla 1283 - Spec typo.
2063 Bugzilla 1290 - Spec grammar fixes.
2065 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2067 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2069 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2070 mailing list by by Dmitry Isaikin.
2072 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2074 JH/02 Support "G" suffix to numbers in ${if comparisons.
2076 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2078 NM/01 Bugzilla 1197 - Spec typo
2079 Bugzilla 1196 - Spec examples corrections
2081 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2083 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2084 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2087 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2088 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2090 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2091 "acl {{name}{arg}...}", and optional args on acl condition
2094 JH/05 Permit multiple router/transport headers_add/remove lines.
2096 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2098 JH/07 Avoid using a waiting database for a single-message-only transport.
2099 Performance patch from Paul Fisher. Bugzilla 1262.
2101 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2104 JH/09 Add $headers_added variable, with content from use of ACL modifier
2105 add_header (but not yet added to the message). Bugzilla 199.
2107 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2108 Pulled from Bugzilla 817 by Wolfgang Breyha.
2110 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2112 (nb: this is the same fix as in Exim 4.80.1)
2114 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2117 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2120 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2122 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2123 advertise SMTP AUTH mechanism to us, instead of a generic
2124 protocol violation error. Also, make Exim more robust to bad
2125 data from the Dovecot auth socket.
2127 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2129 When a queue runner is handling a message, Exim first routes the
2130 recipient addresses, during which it prunes them based on the retry
2131 hints database. After that it attempts to deliver the message to
2132 any remaining recipients. It then updates the hints database using
2135 So if a recipient address works intermittently, it can get repeatedly
2136 deferred at routing time. The retry hints record remains fresh so the
2137 address never reaches the final cutoff time.
2139 This is a fairly common occurrence when a user is bumping up against
2140 their storage quota. Exim had some logic in its local delivery code
2141 to deal with this. However it did not apply to per-recipient defers
2142 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2144 This change adds a proper retry rule check during routing so that the
2145 final cutoff time is checked against the message's age. We only do
2146 this check if there is an address retry record and there is not a
2147 domain retry record; this implies that previous attempts to handle
2148 the address had the retry_use_local_parts option turned on. We use
2149 this as an approximation for the destination being like a local
2150 delivery, as in LMTP.
2152 I suspect this new check makes the old local delivery cutoff check
2153 redundant, but I have not verified this so I left the code in place.
2155 TF/02 Correct gecos expansion when From: is a prefix of the username.
2157 Test 0254 submits a message to Exim with the header
2161 When I ran the test suite under the user fanf2, Exim expanded
2162 the header to contain my full name, whereas it should have added
2163 a Resent-Sender: header. It erroneously treats any prefix of the
2164 username as equal to the username.
2166 This change corrects that bug.
2168 GF/01 DCC debug and logging tidyup
2169 Error conditions log to paniclog rather than rejectlog.
2170 Debug lines prefixed by "DCC: " to remove any ambiguity.
2172 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2174 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2175 Bug spotted by Jeremy Harris; was flawed since initial commit.
2176 Would have resulted in OCSP responses post-SNI triggering an Exim
2177 NULL dereference and crash.
2179 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2181 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2182 Bug detection, analysis and fix by Samuel Thibault.
2183 Bugzilla 1331, Debian bug #698092.
2185 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2187 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2188 Server implementation by Todd Lyons, client by JH.
2189 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2190 config variable "prdr_enable" controls whether the server
2191 advertises the facility. If the client requests PRDR a new
2192 acl_data_smtp_prdr ACL is called once for each recipient, after
2193 the body content is received and before the acl_smtp_data ACL.
2194 The client is controlled by both of: a hosts_try_prdr option
2195 on the smtp transport, and the server advertisement.
2196 Default client logging of deliveries and rejections involving
2197 PRDR are flagged with the string "PRDR".
2199 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2200 fclose(). Diagnosis by Todd Lyons.
2202 PP/17 Update configure.default to handle IPv6 localhost better.
2203 Patch by Alain Williams (plus minor tweaks).
2206 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2207 This is now consistent with GnuTLS, and is now documented: the
2208 previous undocumented portable approach to treating the option as
2209 unset was to force an expansion failure. That still works, and
2210 an empty string is now equivalent.
2212 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2213 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2214 not performing validation itself.
2216 PP/20 Added force_command boolean option to pipe transport.
2217 Patch from Nick Koston, of cPanel Inc.
2219 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2222 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2224 PP/21 Fix eximon continuous updating with timestamped log-files.
2225 Broken in a format-string cleanup in 4.80, missed when I repaired the
2226 other false fix of the same issue.
2227 Report and fix from Heiko Schlichting.
2230 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2231 Report from Prashanth Katuri.
2233 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2234 It's SecureTransport, so affects any MacOS clients which use the
2235 system-integrated TLS libraries, including email clients.
2237 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2238 using a MIME ACL for non-SMTP local injection.
2239 Report and assistance in diagnosis by Warren Baker.
2241 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2243 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2245 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2246 last id that failed so it may be referenced in subsequent ACL's.
2248 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2251 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2252 ldap library initialization, allowing self-signed CA's to be
2253 used. Also properly sets require_cert option later in code by
2254 using NULL (global ldap config) instead of ldap handle (per
2255 session). Bug diagnosis and testing by alxgomz.
2257 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2258 the src/util/ subdirectory.
2260 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2261 renamed to Transport Post Delivery Action by Jeremy Harris, as
2264 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2265 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2266 redis_servers = needs to be configured which will be used by the redis
2267 lookup. Patch from Warren Baker, of The Packet Hub.
2269 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2271 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2272 hostname or reverse DNS when processing a host list. Used suggestions
2273 from multiple comments on this bug.
2275 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2277 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2278 Missed a few lines, added it to make the runtest require no keyboard
2281 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2282 contains upper case chars. Make router use caseful_local_part.
2284 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2285 support when GnuTLS has been built with p11-kit.
2291 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2293 This, or similar/improved, will also be change PP/11 of 4.82.
2299 PP/01 Handle short writes when writing local log-files.
2300 In practice, only affects FreeBSD (8 onwards).
2301 Bugzilla 1053, with thanks to Dmitry Isaikin.
2303 NM/01 Bugzilla 949 - Documentation tweak
2305 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2308 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2310 PP/02 Implemented gsasl authenticator.
2312 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2314 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2315 `pkg-config foo` for cflags/libs.
2317 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2318 with rest of GSASL and with heimdal_gssapi.
2320 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2321 `pkg-config foo` for cflags/libs for the TLS implementation.
2323 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2324 properties get this fed in as external SSF. A number of robustness
2325 and debugging improvements to the cyrus_sasl authenticator.
2327 PP/08 cyrus_sasl server now expands the server_realm option.
2329 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2330 Patch by Jeremy Harris.
2332 PP/10 Added dbmjz lookup type.
2334 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2336 PP/12 MAIL args handles TAB as well as SP, for better interop with
2337 non-compliant senders.
2338 Analysis and variant patch by Todd Lyons.
2340 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2341 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2342 Patch from Dirk Mueller <dmueller@suse.com>
2344 PP/13 tls_peerdn now print-escaped for spool files.
2345 Observed some $tls_peerdn in wild which contained \n, which resulted
2346 in spool file corruption.
2348 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2349 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2350 or write after TLS renegotiation, which otherwise led to messages
2353 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2354 as a tracking header (ie: a signed header comes before the signature).
2355 Patch from Wolfgang Breyha.
2357 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2358 comma-sep list; embedded commas doubled.
2360 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2362 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2364 Report and patch from Dmitry Banschikov.
2366 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2367 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2368 failures. We appear to now support TLS1.1+ with Exim.
2370 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2371 lets Exim select keys and certificates based upon TLS SNI from client.
2372 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2373 before an outbound SMTP session. New log_selector, +tls_sni.
2375 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2376 NULL dereference. Report and patch from Alun Jones.
2378 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2379 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2380 Not seeing resolver debug output on NetBSD, but suspect this is a
2381 resolver implementation change.
2383 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2384 Left warnings. Added "eximon gdb" invocation mode.
2386 PP/21 Defaulting "accept_8bitmime" to true, not false.
2388 PP/22 Added -bw for inetd wait mode support.
2390 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2391 locate the relevant includes and libraries. Made this the default.
2393 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2394 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2396 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2397 This may cause build issues on older platforms.
2399 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2400 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2401 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2402 Added SNI support via GnuTLS too.
2403 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2405 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2407 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2408 Applied second patch from Janne, implementing suggestion to default
2409 multiple-strings-in-record handling to match SPF spec.
2411 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2413 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2414 read-only, out of scope).
2415 Patch from Wolfgang Breyha, report from Stuart Northfield.
2417 PP/29 Fix three issues highlighted by clang analyser static analysis.
2418 Only crash-plausible issue would require the Cambridge-specific
2419 iplookup router and a misconfiguration.
2420 Report from Marcin Mirosław.
2422 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2424 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2425 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2426 As part of this, removing so much warning spew let me fix some minor
2427 real issues in debug logging.
2429 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2430 assignment on my part. Fixed.
2432 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2433 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2434 Janne Snabb (who went above and beyond: thank you).
2436 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2437 string otherwise requires a connection and a bunch more work and it's
2438 relatively easy to get wrong. Should also expose TLS library linkage
2441 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2442 64-bit ${eval} (JH/03).
2444 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2445 GNU libc to support some of the 64-bit stuff, should not lead to
2446 conflicts. Defined before os.h is pulled in, so if a given platform
2447 needs to override this, it can.
2449 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2450 protection layer was required, which is not implemented.
2451 Bugzilla 1254, patch from Wolfgang Breyha.
2453 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2454 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2455 tls_dhparam take prime identifiers. Also unbreak combination of
2456 OpenSSL+DH_params+TLSSNI.
2458 PP/39 Disable SSLv2 by default in OpenSSL support.
2464 PP/01 Solaris build fix for Oracle's LDAP libraries.
2465 Bugzilla 1109, patch from Stephen Usher.
2467 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2469 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2472 TF/02 Fix a couple more cases where we did not log the error message
2473 when unlink() failed. See also change 4.74-TF/03.
2475 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2476 lock up or crash if it happened to be inside a call to libc when it
2477 got a SIGUSR1 from exiwhat.
2479 The SIGUSR1 handler appends the current process status to the process
2480 log which is later printed by exiwhat. It used to use the general
2481 purpose logging code to do this, but several functions it calls are
2482 not safe for signals.
2484 The new output code in the SIGUSR1 handler is specific to the process
2485 log, and simple enough that it's easy to inspect for signal safety.
2486 Removing some special cases also simplifies the general logging code.
2487 Removing the spurious timestamps from the process log simplifies
2490 TF/04 Improved ratelimit ACL condition.
2492 The /noupdate option has been deprecated in favour of /readonly which
2493 has clearer semantics. The /leaky, /strict, and /readonly update modes
2494 are mutually exclusive. The update mode is no longer included in the
2495 database key; it just determines when the database is updated. (This
2496 means that when you upgrade Exim will forget old rate measurements.)
2498 Exim now checks that the per_* options are used with an update mode that
2499 makes sense for the current ACL. For example, when Exim is processing a
2500 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2501 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2502 must specify per_mail/readonly. If you omit the update mode it defaults to
2503 /leaky where that makes sense (as before) or /readonly where required.
2505 The /noupdate option is now undocumented but still supported for
2506 backwards compatibility. It is equivalent to /readonly except that in
2507 ACLs where /readonly is required you may specify /leaky/noupdate or
2508 /strict/noupdate which are treated the same as /readonly.
2510 A useful new feature is the /count= option. This is a generalization
2511 of the per_byte option, so that you can measure the throughput of other
2512 aggregate values. For example, the per_byte option is now equivalent
2513 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2515 The per_rcpt option has been generalized using the /count= mechanism
2516 (though it's more complicated than the per_byte equivalence). When it is
2517 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2518 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2519 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2520 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2521 Note that using per_rcpt with a non-readonly update mode in more than
2522 one ACL will cause the recipients to be double-counted. (The per_mail
2523 and per_byte options don't have this problem.)
2525 The handling of very low rates has changed slightly. If the computed rate
2526 is less than the event's count (usually one) then this event is the first
2527 after a long gap. In this case the rate is set to the same as this event's
2528 count, so that the first message of a spam run is counted properly.
2530 The major new feature is a mechanism for counting the rate of unique
2531 events. The new per_addr option counts the number of different
2532 recipients that someone has sent messages to in the last time period. It
2533 behaves like per_rcpt if all the recipient addresses are different, but
2534 duplicate recipient addresses do not increase the measured rate. Like
2535 the /count= option this is a general mechanism, so the per_addr option
2536 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2537 example, measure the rate that a client uses different sender addresses
2538 with the options per_mail/unique=$sender_address. There are further
2539 details in the main documentation.
2541 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2543 TF/06 Removed a few PCRE remnants.
2545 TF/07 Automatically extract Exim's version number from tags in the git
2546 repository when doing development or release builds.
2548 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2549 Bugzilla 879. Patch from Paul Fisher.
2551 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2552 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2555 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2557 PP/05 Variable $av_failed, true if the AV scanner deferred.
2558 Bugzilla 1078. Patch from John Horne.
2560 PP/06 Stop make process more reliably on build failure.
2561 Bugzilla 1087. Patch from Heiko Schlittermann.
2563 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2564 Bugzilla 1089. Patch from Heiko Schlittermann.
2566 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2567 Bugzilla 1131. Patch from Holger Weiß.
2569 PP/09 Handle IPv6 addresses with SPF.
2570 Bugzilla 860. Patch from Wolfgang Breyha.
2572 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2574 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2577 PP/11 match_* no longer expand right-hand-side by default.
2578 New compile-time build option, EXPAND_LISTMATCH_RHS.
2579 New expansion conditions, "inlist", "inlisti".
2581 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2583 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2585 PP/14 fix log_write() format string regression from TF/03.
2586 Bugzilla 1152. Patch from Dmitry Isaikin.
2592 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2594 PP/02 Harmonised TLS library version reporting; only show if debugging.
2595 Layout now matches that introduced for other libraries in 4.74 PP/03.
2597 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2599 PP/04 New "dns_use_edns0" global option.
2601 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2604 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2605 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2607 TK/01 Updated PolarSSL code to 0.14.2.
2608 Bugzilla 1097. Patch from Andreas Metzler.
2610 PP/07 Catch divide-by-zero in ${eval:...}.
2611 Fixes bugzilla 1102.
2613 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2616 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2617 format-string attack -- SECURITY: remote arbitrary code execution.
2619 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2620 time unintentionally subject to list matching rules, letting the header
2621 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2622 arbitrary string expansion). This allowed for information disclosure.
2624 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2625 INT_MIN/-1 -- value coerced to INT_MAX.
2631 NM/01 Workaround for PCRE version dependency in version reporting
2634 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2635 This fixes portability to compilers other than gcc, notably
2636 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2638 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2639 makefiles for portability to HP-UX and POSIX correctness.
2641 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2642 Also via indented variable definition in the Makefile.
2643 (Debugging by Oliver Heesakkers).
2645 PP/02 Restore caching of spamd results with expanded spamd_address.
2646 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2648 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2649 Improves build reliability. Fix from: Frank Elsner
2651 NM/02 Fix wide character breakage in the rfc2047 coding
2652 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2654 NM/03 Allow underscore in dnslist lookups
2655 Fixes bug 1026. Patch from Graeme Fowler
2657 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2658 Code patches from Adam Ciarcinski of NetBSD.
2660 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2663 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2664 is logged, to avoid truncation. Patch from John Horne.
2666 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2667 Patch from Jakob Hirsch.
2669 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2670 SQL string expansion failure details.
2671 Patch from Andrey Oktyabrski.
2673 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2674 Patch from Simon Arlott.
2676 PP/09 New lookups functionality failed to compile on old gcc which rejects
2677 extern declarations in function scope.
2678 Patch from Oliver Fleischmann
2680 PP/10 Use sig_atomic_t for flags set from signal handlers.
2681 Check getgroups() return and improve debugging.
2682 Fixed developed for diagnosis in bug 927 (which turned out to be
2685 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2686 Patch from Mark Zealey.
2688 PP/12 Bugzilla 1056: Improved spamd server selection.
2689 Patch from Mark Zealey.
2691 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2692 Based on patch from Heiko Schlittermann.
2694 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2695 Patch from Uwe Doering, sign-off by Michael Haardt.
2697 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2698 variable declaration deep within a block. Bug and patch from
2701 PP/15 lookups-Makefile IRIX compatibility coercion.
2703 PP/16 Make DISABLE_DKIM build knob functional.
2705 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2706 Patch by Simon Arlott
2708 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2709 variable argument macros. Our copy now differs from upstream.
2715 TF/01 Failure to get a lock on a hints database can have serious
2716 consequences so log it to the panic log.
2718 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2719 controlled using the smtp_confirmation log selector.
2721 TF/03 Include the error message when we fail to unlink a spool file.
2723 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2724 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2725 for maintaining out-of-tree patches for some time.
2727 PP/01 Bugzilla 139: Documentation and portability issues.
2728 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2729 Handle per-OS dynamic-module compilation flags.
2731 PP/02 Let /dev/null have normal permissions.
2732 The 4.73 fixes were a little too stringent and complained about the
2733 permissions on /dev/null. Exempt it from some checks.
2734 Reported by Andreas M. Kirchwitz.
2736 PP/03 Report version information for many libraries, including
2737 Exim version information for dynamically loaded libraries. Created
2738 version.h, now support a version extension string for distributors
2739 who patch heavily. Dynamic module ABI change.
2741 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2742 privilege escalation vulnerability whereby the Exim run-time user
2743 can cause root to append content of the attacker's choosing to
2746 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2749 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2750 If dropping privileges for untrusted macros, we disabled normal logging
2751 on the basis that it would fail; for the Exim run-time user, this is not
2752 the case, and it resulted in successful deliveries going unlogged.
2753 Fixed. Reported by Andreas Metzler.
2759 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2760 only prepend for the Resent-* case. Fixes regression introduced in
2761 Exim 4.70 by NM/22 for Bugzilla 607.
2763 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2764 increasing numbers of administrators be bitten by this.
2766 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2768 PP/03 Bugzilla 994: added openssl_options main configuration option.
2770 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2772 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2774 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2776 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2777 without a peer certificate, leading to a segfault because of an
2778 assumption that peers always have certificates. Be a little more
2779 paranoid. Problem reported by Martin Tscholak.
2781 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2782 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2783 NB: ClamAV planning to remove STREAM in "middle of 2010".
2784 CL also introduces -bmalware, various -d+acl logging additions and
2785 more caution in buffer sizes.
2787 PP/09 Implemented reverse_ip expansion operator.
2789 PP/10 Bugzilla 937: provide a "debug" ACL control.
2791 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2793 PP/12 Bugzilla 973: Implement --version.
2795 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2797 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2799 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2801 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2802 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2803 ignore trailing whitespace.
2805 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2807 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2810 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2811 Notification from Dr Andrew Aitchison.
2813 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2814 ExtendedDetectionInfo response format.
2815 Notification from John Horne.
2817 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2820 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2821 XSL and documented dependency on system catalogs, with examples of how
2824 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2827 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2828 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2829 configuration file which is writeable by the Exim user or group.
2831 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2832 of configuration files to cover files specified with the -C option if
2833 they are going to be used with root privileges, not just the default
2836 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2837 option (effectively making it always true).
2839 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2840 files to be used while preserving root privileges.
2842 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2843 that rogue child processes cannot use them.
2845 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2846 run-time user, instead of root.
2848 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2849 Exim run-time user without dropping privileges.
2851 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2852 result string, instead of calling string_vformat() twice with the same
2855 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2856 for other users. Others should always drop root privileges if they use
2857 -C on the command line, even for a whitelisted configure file.
2859 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2861 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2867 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2868 $data_path, and $header_path variables; fixed documentation bugs and
2871 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2872 exipick to access non-standard spools, including the "frozen" queue
2875 NM/01 Bugzilla 965: Support mysql stored procedures.
2876 Patch from Alain Williams
2878 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2880 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2881 Patch from Andreas Metzler
2883 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2884 Patch from Kirill Miazine
2886 NM/05 Bugzilla 671: Added umask to procmail example.
2888 JJ/03 installed exipick 20100323.0, fixing doc bug
2890 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2891 directory. Notification and patch from Dan Rosenberg.
2893 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2895 TK/02 Improve log output when DKIM signing operation fails.
2897 MH/01 Treat the transport option dkim_domain as a colon separated
2898 list, not as a single string, and sign the message with each element,
2899 omitting multiple occurences of the same signer.
2901 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2902 Bugzilla 985, 986. Patch by Simon Arlott
2904 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2905 Patch by Simon Arlott
2907 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2908 MBX locking. Notification from Dan Rosenberg.
2914 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2916 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2918 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2920 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2922 NM/04 Bugzilla 915: Flag broken perl installation during build.
2928 TK/01 Added patch by Johannes Berg that expands the main option
2929 "spamd_address" if it starts with a dollar sign.
2931 TK/02 Write list of recipients to X-Envelope-Sender header when building
2932 the mbox-format spool file for content scanning (suggested by Jakob
2935 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2936 (http://www.dcc-servers.net/) support via dccifd. Activated by
2937 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2939 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2940 by Mark Daniel Reidel <mr@df.eu>.
2942 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2943 When building exim an external PCRE library is now needed -
2944 PCRE is a system library on the majority of modern systems.
2945 See entry on PCRE_LIBS in EDITME file.
2947 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2948 conversation. Added nologin parameter to request.
2949 Patch contributed by Kirill Miazine.
2951 TF/01 Do not log submission mode rewrites if they do not change the address.
2953 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2955 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2956 log files in place. Contributed by Roberto Lima.
2958 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2960 TF/03 Bugzilla 615: When checking the local_parts router precondition
2961 after a local_part_suffix or local_part_prefix option, Exim now
2962 does not use the address's named list lookup cache, since this
2963 contains cached lookups for the whole local part.
2965 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2966 Robert Millan. Documentation is in experimental-spec.txt.
2968 TF/04 Bugzilla 668: Fix parallel build (make -j).
2970 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2972 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2973 Patch provided by Jan Srzednicki.
2975 TF/05 Leading white space used to be stripped from $spam_report which
2976 wrecked the formatting. Now it is preserved.
2978 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2979 that they are available at delivery time.
2981 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2983 TF/08 TLS error reporting now respects the incoming_interface and
2984 incoming_port log selectors.
2986 TF/09 Produce a more useful error message if an SMTP transport's hosts
2987 setting expands to an empty string.
2989 NM/06 Bugzilla 744: EXPN did not work under TLS.
2990 Patch provided by Phil Pennock.
2992 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2993 Patch provided by Richard Godbee.
2995 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2996 acl_smtp_notquit, added index entry.
2998 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2999 Patch provided by Eugene Bujak.
3001 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3002 accept(). Patch provided by Maxim Dounin.
3004 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3005 Patch provided by Phil Pennock.
3007 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3009 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3010 Patch provided by Brad "anomie" Jorsch.
3012 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3013 Patch provided by Dean Brooks.
3015 TK/05 Add native DKIM support (does not depend on external libraries).
3017 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3018 Patch provided by Graeme Fowler.
3020 NM/16 Bugzilla 851: Documentation example syntax fix.
3022 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3024 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3027 NM/19 Bugzilla 745: TLS version reporting.
3028 Patch provided by Phil Pennock.
3030 NM/20 Bugzilla 167: bool: condition support.
3031 Patch provided by Phil Pennock.
3033 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3034 clients. Patch provided by Phil Pennock.
3036 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3037 Patch provided by Brad "anomie" Jorsch.
3039 NM/23 Bugzilla 687: Fix misparses in eximstats.
3040 Patch provided by Heiko Schlittermann.
3042 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3043 Patch provided by Heiko Schlittermann.
3045 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3046 plus update to original patch.
3048 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3050 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3051 Patch provided by David Brownlee.
3053 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3055 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3057 NM/30 Bugzilla 888: TLS documentation bugfixes.
3059 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3061 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3062 Unlike the original bugzilla I have changed all shell scripts in src tree.
3064 NM/33 Bugzilla 898: Transport filter timeout fix.
3065 Patch by Todd Rinaldo.
3067 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3068 Patch by Serge Demonchaux.
3070 NM/35 Bugzilla 39: Base64 decode bug fixes.
3071 Patch by Jakob Hirsch.
3073 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3075 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3077 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3079 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3085 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3086 ALT-N's libdkim that I have put here:
3087 http://duncanthrax.net/exim-experimental/
3089 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3090 were called 'true' and it seems that C99 defines that as a reserved
3091 keyword to be used with 'bool' variable types. That means you could
3092 not include C99-style headers which use bools without triggering
3093 build errors in sieve.c.
3095 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3096 as mailq or other aliases. Changed the --help handling significantly
3097 to do whats expected. exim_usage() emits usage/help information.
3099 SC/01 Added the -bylocaldomain option to eximstats.
3101 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3103 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3105 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3111 PH/01 Another patch from the Sieve maintainer.
3113 PH/02 When an IPv6 address is converted to a string for single-key lookup
3114 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3115 dots are used instead of colons so that keys in lsearch files need not
3116 contain colons. This was done some time before quoting was made available
3117 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3118 (notated using the quote facility) so as to distinguish them from IPv4
3119 keys. This meant that lookups for IP addresses in host lists did not work
3120 for iplsearch lookups.
3122 This has been fixed by arranging for IPv6 addresses to be expressed with
3123 colons if the lookup type is iplsearch. This is not incompatible, because
3124 previously such lookups could never work.
3126 The situation is now rather anomalous, since one *can* have colons in
3127 ordinary lsearch keys. However, making the change in all cases is
3128 incompatible and would probably break a number of configurations.
3130 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3133 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3134 conversion specification without a maximum field width, thereby enabling
3135 a rogue spamd server to cause a buffer overflow. While nobody in their
3136 right mind would setup Exim to query an untrusted spamd server, an
3137 attacker that gains access to a server running spamd could potentially
3138 exploit this vulnerability to run arbitrary code as the Exim user.
3140 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3141 $primary_hostname instead of what libspf2 thinks the hosts name is.
3143 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3144 a directory entry by the name of the lookup key. Previously, if a
3145 symlink pointed to a non-existing file or a file in a directory that
3146 Exim lacked permissions to read, a lookup for a key matching that
3147 symlink would fail. Now it is enough that a matching directory entry
3148 exists, symlink or not. (Bugzilla 503.)
3150 PH/03 The body_linecount and body_zerocount variables are now exported in the
3153 PH/04 Added the $dnslist_matched variable.
3155 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3156 This means they are set thereafter only if the connection becomes
3159 PH/06 Added the client_condition to authenticators so that some can be skipped
3160 by clients under certain conditions.
3162 PH/07 The error message for a badly-placed control=no_multiline_responses left
3163 "_responses" off the end of the name.
3165 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3167 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3168 (without spaces) instead of just copying the configuration text.
3170 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3172 PH/11 Added $max_received_linelength.
3174 PH/12 Added +ignore_defer and +include_defer to host lists.
3176 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3177 way in which PCRE > 7.0 is built.
3179 PH/14 Implemented queue_only_load_latch.
3181 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3182 MAIL command. The effect was to mangle the value on 64-bit systems.
3184 PH/16 Another patch from the Sieve maintainer.
3186 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3188 PH/18 If a system quota error occurred while trying to create the file for
3189 a maildir delivery, the message "Mailbox is full" was not appended to the
3190 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3191 only to a quota excession during the actual writing of the file.
3193 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3194 characters?) which causes problems in log lines. The DN values are now
3195 passed through string_printing() before being added to log lines.
3197 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3198 and InterBase are left for another time.)
3200 PH/21 Added message_body_newlines option.
3202 PH/22 Guard against possible overflow in moan_check_errorcopy().
3204 PH/23 POSIX allows open() to be a macro; guard against that.
3206 PH/24 If the recipient of an error message contained an @ in the local part
3207 (suitably quoted, of course), incorrect values were put in $domain and
3208 $local_part during the evaluation of errors_copy.
3214 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3215 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3218 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3219 issue a MAIL command.
3221 PH/02 In an ACL statement such as
3223 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3225 if a client was not listed at all, or was listed with a value other than
3226 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3227 the condition was not true (as it should be), so access was not denied.
3228 The bug was that the ! inversion was incorrectly passed on to the second
3229 item. This has been fixed.
3231 PH/03 Added additional dnslists conditions == and =& which are different from
3232 = and & when the dns lookup returns more than one IP address.
3234 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3235 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3237 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3238 FSYNC, which compiles an option called disable_fsync that allows for
3239 bypassing fsync(). The documentation is heavily laced with warnings.
3241 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3243 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3244 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3245 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3246 including adding "make clean"; (3) Added -fPIC when compiling the test
3247 dynamically loaded module, to get rid of a warning.
3249 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3250 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3251 = 0s. The bug is otherwise harmless.
3253 PH/07 There was a bug in the dovecot authenticator such that the value of
3254 $auth1 could be overwritten, and so not correctly preserved, after a
3255 successful authentication. This usually meant that the value preserved by
3256 the server_setid option was incorrect.
3258 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3260 PH/09 Installed PCRE release 7.0.
3262 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3263 run for batched SMTP input. It is now run at the start of every message
3264 in the batch. While fixing this I discovered that the process information
3265 (output by running exiwhat) was not always getting set for -bs and -bS
3266 input. This is fixed, and it now also says "batched" for BSMTP.
3268 PH/11 Added control=no_pipelining.
3270 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3271 patch, slightly modified), and move the expansion of helo_data till after
3272 the connection is made in the smtp transport (so it can use these
3275 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3277 PH/14 Added log_selector = +pid.
3279 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3281 PH/16 Add ${if forany and ${if forall.
3283 PH/17 Added dsn_from option to vary the From: line in DSNs.
3285 PH/18 Flush SMTP output before performing a callout, unless control =
3286 no_callout_flush is set.
3288 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3289 was true (the default) a successful delivery failed to delete the retry
3290 item, thus causing premature timeout of the address. The bug is now
3293 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3295 PH/21 Long custom messages for fakedefer and fakereject are now split up
3296 into multiline responses in the same way that messages for "deny" and
3297 other ACL rejections are.
3299 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3300 with slight modification.
3302 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3303 draft, changing the syntax and factoring some duplicate code".
3305 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3306 for deliveries of the second and subsequent messages over the same SMTP
3309 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3310 ${reduce, with only minor "tidies".
3312 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3314 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3315 expansion side effects.
3317 PH/27 When a message times out after an over-quota error from an Exim-imposed
3318 quota, the bounce message says "mailbox is full". This message was not
3319 being given when it was a system quota that was exceeded. It now should
3322 MH/03 Made $recipients available in local_scan(). local_scan() already has
3323 better access to the recipient list through recipients_list[], but
3324 $recipients can be useful in postmaster-provided expansion strings.
3326 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3327 in the case of a MAIL command with additional options following the
3328 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3329 were accidentally chopped off.
3331 PH/29 SMTP synchronization checks are implemented when a command is read -
3332 there is a check that no more input is waiting when there shouldn't be
3333 any. However, for some commands, a delay in an ACL can mean that it is
3334 some time before the response is written. In this time, more input might
3335 arrive, invalidly. So now there are extra checks after an ACL has run for
3336 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3337 pipelining has not been advertised.
3339 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3341 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3342 returned in the SMTP response when smtp_return_error_details was false.
3343 This has been fixed.
3345 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3346 instead of the C I/O that was originally supplied, because problems were
3347 reported on Solaris.
3349 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3350 Exim which did not show up earlier: it was assuming that a call to
3351 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3352 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3353 was a macro that became an assignment, so it seemed to work. This has
3354 changed to a proper function call with a void return, hence the compile
3355 error. Exim's code has been fixed.
3357 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3360 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3362 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3364 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3365 because it turns out to be too big or there is a timeout) there is no
3366 "Completed" line in the log. When some messages of this type were
3367 selected by exigrep, they were listed as "not completed". Others were
3368 picked up by some special patterns. I have improved the selection
3369 criteria to be more general.
3371 PH/38 The host_find_failed option in the manualroute router can now be set
3372 to "ignore", to completely ignore a host whose IP address cannot be
3373 found. If all hosts are ignored, the behaviour is controlled by the new
3374 host_all_ignored option.
3376 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3377 homing or because of multiple MX records with /mx) generated more than
3378 one IP address, and the following item turned out to be the local host,
3379 all the secondary addresses of the first item were incorrectly removed
3380 from the list, along with the local host and any following hosts (which
3381 is what is supposed to happen).
3383 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3384 whoever called Exim into the -H file. In the case of the daemon it was
3385 behaving confusingly. When first started, it used values for whoever
3386 started the daemon, but after a SIGHUP it used the Exim user (because it
3387 calls itself on a restart). I have changed the code so that it now always
3390 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3391 message are rejected with the same error (e.g. no authentication or bad
3392 sender address), and a DATA command is nevertheless sent (as can happen
3393 with PIPELINING or a stupid MUA), the error message that was given to the
3394 RCPT commands is included in the rejection of the DATA command. This is
3395 intended to be helpful for MUAs that show only the final error to their
3398 PH/42 Another patch from the Sieve maintainer.
3400 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3401 Eximstats - Fixed some broken HTML links and added missing column headers
3403 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3404 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3406 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3408 PH/43 Yet another patch from the Sieve maintainer.
3410 PH/44 I found a way to check for a TCP/IP connection going away before sending
3411 the response to the final '.' that terminates a message, but only in the
3412 case where the client has not sent further data following the '.'
3413 (unfortunately, this is allowed). However, in many cases there won't be
3414 any further data because there won't be any more messages to send. A call
3415 to select() can be used: if it shows that the input is "ready", there is
3416 either input waiting, or the socket has been closed. An attempt to read
3417 the next input character can distinguish the two cases. Previously, Exim
3418 would have sent an OK response which the client would never have see.
3419 This could lead to message repetition. This fix should cure that, at
3420 least in a lot of common cases.
3422 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3423 advertised in response to EHLO.
3429 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3430 fixed by 4.65/MH/01 (is this a record?) are fixed:
3432 (i) An empty string was always treated as zero by the numeric comparison
3433 operators. This behaviour has been restored.
3435 (ii) It is documented that the numeric comparison operators always treat
3436 their arguments as decimal numbers. This was broken in that numbers
3437 starting with 0 were being interpreted as octal.
3439 While fixing these problems I realized that there was another issue that
3440 hadn't been noticed. Values of message_size_limit (both the global option
3441 and the transport option) were treated as octal if they started with 0.
3442 The documentation was vague. These values are now always treated as
3443 decimal, and I will make that clear in the documentation.
3449 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3450 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3453 MH/01 Don't check that the operands of numeric comparison operators are
3454 integers when their expansion is in "skipping" mode (fixes bug
3455 introduced by 4.64-PH/07).
3457 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3458 child addresses, Exim now panics and dies. Previously, because the count
3459 is held in a short int, deliveries were likely to be lost. As such a
3460 large number of recipients for a single message is ridiculous
3461 (performance will be very, very poor), I have chosen to impose a limit
3462 rather than extend the field.
3468 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3469 leftover -K file (the existence of which was triggered by #402).
3470 While we were at it, introduced process PID as part of the -K
3471 filename. This should rule out race conditions when creating
3474 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3475 processing considerably. Previous code took too long for large mails,
3476 triggering a timeout which in turn triggers #401.
3478 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3479 in the DK code in transports.c. sendfile() is not really portable,
3480 hence the _LINUX specificness.
3482 TF/01 In the add_headers option to the mail command in an Exim filter,
3483 there was a bug that Exim would claim a syntax error in any
3484 header after the first one which had an odd number of characters
3487 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3488 callout verification, Exim cached a "reject" for the entire domain. This
3489 is correct for most verifications, but it is not correct for a recipient
3490 verification with use_sender or use_postmaster set, because in that case
3491 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3492 case of MAIL FROM:<> rejection from other early rejections (e.g.
3493 rejection of HELO). When verifying a recipient using a non-null MAIL
3494 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3495 Whatever the result of the callout, the value of the domain cache is
3496 left unchanged (for any other kind of callout, getting as far as trying
3497 RCPT means that the domain itself is ok).
3499 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3502 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3503 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3504 session, the daemon ignores ECONNECT errors and logs others; it now
3505 ignores EPIPE as well.
3507 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3508 (quoted-printable decoding).
3510 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3511 later the small subsequent patch to fix an introduced bug.
3513 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3515 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3517 PH/08 An error is now given if message_size_limit is specified negative.
3519 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3520 to be given (somewhat) arbitrary names.
3522 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3525 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3526 miscellaneous code fixes
3528 PH/10 Added the log_reject_target ACL modifier to specify where to log
3531 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3532 hostname. This is wrong, because it relates to the incoming message (and
3533 probably the interface on which it is arriving) and not to the outgoing
3534 callout (which could be using a different interface). This has been
3535 changed to use the value of the helo_data option from the smtp transport
3536 instead - this is what is used when a message is actually being sent. If
3537 there is no remote transport (possible with a router that sets up host
3538 addresses), $smtp_active_hostname is used.
3540 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3541 tweaks were necessary in order to get it to work (see also 21 below):
3542 (a) The code assumed that strncpy() returns a negative number on buffer
3543 overflow, which isn't the case. Replaced with Exim's string_format()
3545 (b) There were several signed/unsigned issues. I just did the minimum
3546 hacking in of casts. There is scope for a larger refactoring.
3547 (c) The code used strcasecmp() which is not a standard C function.
3548 Replaced with Exim's strcmpic() function.
3549 (d) The code set only $1; it now sets $auth1 as well.
3550 (e) A simple test gave the error "authentication client didn't specify
3551 service in request". It would seem that Dovecot has changed its
3552 interface. Fortunately there's a specification; I followed it and
3553 changed what the client sends and it appears to be working now.
3555 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3558 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3559 address A is aliased to B and C, where B exists and C does not. Without
3560 -v the output is "A verified" because verification stops after a
3561 successful redirection if more than one address is generated. However,
3562 with -v the child addresses are also verified. Exim was outputting "A
3563 failed to verify" and then showing the successful verification for C,
3564 with its parentage. It now outputs "B failed to verify", showing B's
3565 parentage before showing the successful verification of C.
3567 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3568 look up a TXT record in a specific list after matching in a combined
3571 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3572 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3573 they consult the DNS. I had assumed they would set it the way they
3574 wanted; and indeed my experiments on Linux seem to show that in some
3575 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3576 To be on the safe side, however, I have now made the interface to
3577 host_find_byname() similar to host_find_bydns(), with an argument
3578 containing the DNS resolver options. The host_find_byname() function now
3579 sets these options at its start, just as host_find_bydns() does. The smtp
3580 transport options dns_qualify_single and dns_search_parents are passed to
3581 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3582 of host_find_byname() use the default settings of RES_DEFNAMES
3583 (qualify_single) but not RES_DNSRCH (search_parents).
3585 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3586 spool_read_header() do less string testing, by means of a preliminary
3587 switch on the second character of optional "-foo" lines. (This is
3588 overdue, caused by the large number of possibilities that now exist.
3589 Originally there were few.) While I was there, I also converted the
3590 str(n)cmp tests so they don't re-test the leading "-" and the first
3591 character, in the hope this might squeeze out yet more improvement.
3593 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3594 flag allowing group syntax was set by the header_syntax check but not
3595 turned off, possible causing trouble later; (2) The flag was not being
3596 set at all for the header_verify test, causing "group"-style headers to
3597 be rejected. I have now set it in this case, and also caused header_
3598 verify to ignore an empty address taken from a group. While doing this, I
3599 came across some other cases where the code for allowing group syntax
3600 while scanning a header line wasn't quite right (mostly, not resetting
3601 the flag correctly in the right place). These bugs could have caused
3602 trouble for malformed header lines. I hope it is now all correct.
3604 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3605 with the "reply" argument non-NULL. The code, however (which originally
3606 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3607 but it didn't always do it. This confused somebody who was copying the
3608 code for some other use. I have removed all the tests.
3610 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3611 feature that was used to support insecure browsers during the U.S. crypto
3612 embargo. It requires special client support, and Exim is probably the
3613 only MTA that supported it -- and would never use it because real RSA is
3614 always available. This code has been removed, because it had the bad
3615 effect of slowing Exim down by computing (never used) parameters for the
3616 RSA_EXPORT functionality.
3618 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3619 authenticator to fail if there's a tab character in the incoming data
3620 (there should never be unless someone is messing about, as it's supposed
3621 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3622 if the connection is using TLS or if the remote IP is the same as the
3623 local IP, and the "valid-client-cert option" if a client certificate has
3626 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3627 authenticators. This can be used for authorization after authentication
3628 succeeds. (In the case of plaintext, it servers for both authentication
3631 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3632 if any retry times were supplied.
3634 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3635 connection, where there is no client IP address to check. In this
3636 situation, the verify now always succeeds.
3638 PH/25 Applied John Jetmore's -Mset patch.
3640 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3642 PH/27 In a string expansion for a processed (not raw) header when multiple
3643 headers of the same name were present, leading whitespace was being
3644 removed from all of them, but trailing whitespace was being removed only
3645 from the last one. Now trailing whitespace is removed from each header
3646 before concatenation. Completely empty headers in a concatenation (as
3647 before) are ignored.
3649 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3650 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3652 PH/29 [Removed. This was a change that I later backed out, and forgot to
3653 correct the ChangeLog entry (that I had efficiently created) before
3654 committing the later change.]
3656 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3657 address errors (4xx response to RCPT) over the same connection as other
3658 messages routed to the same hosts. Such deliveries are always "forced",
3659 so retry times are not inspected. This resulted in far too many retries
3660 for the affected addresses. The effect occurred only when there were more
3661 hosts than the hosts_max_try setting in the smtp transport when it had
3662 the 4xx errors. Those hosts that it had tried were not added to the list
3663 of hosts for which the message was waiting, so if all were tried, there
3664 was no problem. Two fixes have been applied:
3666 (i) If there are any address or message errors in an SMTP delivery, none
3667 of the hosts (tried or untried) are now added to the list of hosts
3668 for which the message is waiting, so the message should not be a
3669 candidate for sending over the same connection that was used for a
3670 successful delivery of some other message. This seems entirely
3671 reasonable: after all the message is NOT "waiting for some host".
3672 This is so "obvious" that I'm not sure why it wasn't done
3673 previously. Hope I haven't missed anything, but it can't do any
3674 harm, as the worst effect is to miss an optimization.
3676 (ii) If, despite (i), such a delivery is accidentally attempted, the
3677 routing retry time is respected, so at least it doesn't keep
3678 hammering the server.
3680 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3681 in ${readsocket because some servers need this prod.
3683 PH/32 Added some extra debug output when updating a wait-xxx database.
3685 PH/33 The hint "could be header name not terminated by colon", which has been
3686 given for certain expansion errors for a long time, was not being given
3687 for the ${if def:h_colon_omitted{... case.
3689 PH/34 The spec says: "With one important exception, whenever a domain list is
3690 being scanned, $domain contains the subject domain." There was at least
3691 one case where this was not true.
3693 PH/35 The error "getsockname() failed: connection reset by peer" was being
3694 written to the panic log as well as the main log, but it isn't really
3695 panic-worthy as it just means the connection died rather early on. I have
3696 removed the panic log writing for the ECONNRESET error when getsockname()
3699 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3700 runs only) independently of the message's sender address. This meant
3701 that, if the 4xx error was in fact related to the sender, a different
3702 message to the same recipient with a different sender could confuse
3703 things. In particular, this can happen when sending to a greylisting
3704 server, but other circumstances could also provoke similar problems.
3705 I have changed the default so that the retry time for these errors is now
3706 based a combination of the sender and recipient addresses. This change
3707 can be overridden by setting address_retry_include_sender=false in the
3710 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3711 remote server are returned as part of bounce messages. This was not
3712 happening for LMTP over a pipe (the lmtp transport), but now it is the
3713 same for both kinds of LMTP.
3715 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3716 in header lines that were in fact CNAMEs. This is no longer the case.
3718 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3719 and queue runs started by the daemon processed all messages. This has
3720 been fixed so that -R and -S can now usefully be given with -q<time>.
3722 PH/40 Import PCRE release 6.7 (fixes some bugs).
3724 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3726 PH/42 Give an error if -q is specified more than once.
3728 PH/43 Renamed the variables $interface_address and $interface_port as
3729 $received_ip_address and $received_port, to make it clear that these
3730 values apply to message reception, and not to the outgoing interface when
3731 a message is delivered. (The old names remain recognized, of course.)
3733 PH/44 There was no timeout on the connect() call when using a Unix domain
3734 socket in the ${readsocket expansion. There now is.
3736 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3737 be meaningful with "accept".
3739 SC/01 Eximstats V1.43
3740 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3742 SC/02 Eximstats V1.44
3743 Use a glob alias rather than an array ref in the generated
3744 parser. This improves both readability and performance.
3746 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3747 Collect SpamAssassin and rejection statistics.
3748 Don't display local sender or destination tables unless
3749 there is data to show.
3750 Added average volumes into the top table text output.
3752 SC/04 Eximstats V1.46
3753 Collect data on the number of addresses (recipients)
3754 as well as the number of messages.
3756 SC/05 Eximstats V1.47
3757 Added 'Message too big' to the list of mail rejection
3758 reasons (thanks to Marco Gaiarin).
3760 SC/06 Eximstats V1.48
3761 Mainlog lines which have GMT offsets and are too short to
3762 have a flag are now skipped.
3764 SC/07 Eximstats V1.49 (Alain Williams)
3765 Added the -emptyok flag.
3767 SC/08 Eximstats V1.50
3768 Fixes for obtaining the IP address from reject messages.
3770 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3771 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3772 whitespace changes from 4.64-PH/27
3774 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3777 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3778 are found, allow negative numbers in numeric criteria)
3780 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3782 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3783 to match changes made in 4.64-PH/43
3785 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3787 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3788 "domain" is actually an IP address, and force a failure. This locks out
3789 those revolvers/nameservers that support "A-for-A" lookups, in
3790 contravention of the specifications.
3792 PH/48 When a host name was looked up from an IP address, and the subsequent
3793 forward lookup of the name timed out, the host name was left in
3794 $sender_host_name, contrary to the specification.
3796 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3797 restricted to single-key lookups, Exim was not diagnosing an error if
3798 * or *@ was used with a query-style lookup.
3800 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3802 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3803 long ago, but noone interested enough thought of it. Let's just say that
3804 the "1.1" means that there are some new functions that weren't there at
3805 some point in the past.
3807 PH/51 Error processing for expansion failure of helo_data from an smtp
3808 transport during callout processing was broken.
3810 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3811 tested/used via the -bh/-bhc/-bs options.
3813 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3814 bug, fixed in subsequent PCRE releases).
3816 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3817 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3819 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3825 SC/01 Use a glob alias rather than an array ref in eximstats generated
3826 parser. This improves both readability and performance.
3828 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3829 Don't display local sender or destination tables in eximstats unless
3830 there is data to show.
3831 Added average volumes into the eximstats top table text output.
3833 SC/03 Collect data on the number of addresses (recipients) as well
3834 as the number of messages in eximstats.
3836 TF/01 Correct an error in the documentation for the redirect router. Exim
3837 does (usually) call initgroups() when daemonizing.
3839 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3840 with consistent privilege compared to when running as a daemon.
3842 TF/03 Note in the spec that $authenticated_id is not set for local
3843 submissions from trusted users.
3845 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3846 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3848 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3849 by adding some example configuration directives to the default
3850 configuration file. A little bit of work is required to uncomment the
3851 directives and define how usernames and passwords are checked, but
3852 there is now a framework to start from.
3854 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3855 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3856 without this. I don't know how relevant this is to other LDAP libraries.
3858 PH/02 Add the verb name to the "unknown ACL verb" error.
3860 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3862 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3864 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3865 directory not expanded when it should be if an expanded home directory
3866 was set for the address (which is overridden by the transport).
3868 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3871 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3872 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3873 because it is too late at that time, and has no effect.
3875 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3876 security issue with \' (bugzilla #107). I could not use the
3877 PQescapeStringConn() function, because it needs a PGconn value as one of
3880 PH/08 When testing addresses using -bt, indicate those final addresses that
3881 are duplicates that would not cause an additional delivery. At least one
3882 person was confused, thinking that -bt output corresponded to deliveries.
3883 (Suppressing duplicates isn't a good idea as you lose the information
3884 about possibly different redirections that led to the duplicates.)
3886 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3887 systems where poll() doesn't work, in particular OS X.
3889 PH/10 Added more information to debugging output for retry time not reached.
3891 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3892 operations in malware.c.
3894 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3897 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3898 syslog_duplication set false, log lines that would normally be written
3899 both the the main log and to the reject log were not written to syslog at
3902 PH/14 In the default configuration, change the use of "message" in ACL warn
3903 statements to "add_header".
3905 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3906 not followed by a command (e.g. "seen endif").
3908 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3909 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3912 PH/17 Added extra conditions to the default value of delay_warning_condition
3916 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3917 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3918 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3921 The Auto-Submitted: and various List- headers are standardised, whereas I
3922 don't think Precedence: ever was.
3924 PH/18 Refactored debugging code in route_finduser() to show more information,
3925 in particular, the error code if getpwnam() issues one.
3927 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3928 This is apparently needed in addition to the PH/07 change above to avoid
3929 any possible encoding problems.
3931 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3932 but not after initializing Perl.
3934 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3935 output them only if debugging. By default they are written stderr,
3936 apparently, which is not desirable.
3938 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3941 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3944 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3946 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3947 authenticated or an ident call has been made. Suppress the default
3948 values for $authenticated_id and $authenticated_sender (but permit -oMai
3949 and -oMas) when testing with -bh.
3951 PH/24 Re-jigged the order of the tests in the default configuration so that the
3952 tests for valid domains and recipients precede the DNS black list and CSA
3953 tests, on the grounds that those ones are more expensive.
3955 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3956 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3957 This bug exists in every version of Exim that I still have, right back to
3960 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3961 However, an attempt to turn on case-sensitivity in a regex key by
3962 including (?-i) didn't work because the subject string was already
3963 lowercased, and the effects were non-intuitive. It turns out that a
3964 one-line patch can be used to allow (?-i) to work as expected.
3970 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3971 other effects) broke the use of negated acl sub-conditions.
3973 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3976 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3977 "Deny" causes Exim to reject the incoming connection with a 554 error.
3978 Unfortunately, if there is a major crisis, such as a disk failure,
3979 tcp-wrappers gives "deny", whereas what one would like would be some
3980 kind of temporary error. A kludge has been added to help with this.
3981 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3982 554 error is used if errno is still zero or contains ENOENT (which occurs
3983 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3986 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3988 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3989 errors. Otherwise a message that provokes a temporary error (when other
3990 messages do not) can cause a whole host to time out.
3992 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3993 addresses were routed directly to files or pipes from a redirect router.
3994 File deliveries just didn't batch; pipe deliveries might have suffered
3997 PH/06 A failure to get a lock for a hints database would erroneously always say
3998 "Failed to get write lock", even when it was really a read lock.
4000 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4001 of 0600. This has been changed to use the value of the lockfile_mode
4002 option (which defaults to 0600).
4004 PH/08 Applied small patch from the Sieve maintainer.
4006 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4007 folder from quota calculations, a direct delivery into this folder messed
4008 up the contents of the maildirsize file. This was because the regex was
4009 used only to exclude .Trash (or whatever) when the size of the mailbox
4010 was calculated. There was no check that a delivery was happening into an
4011 excluded directory. This bug has been fixed by ignoring all quota
4012 processing for deliveries into excluded directories.
4014 PH/10 Added the maildirfolder_create_regex option to appendfile.
4020 PH/01 The code for finding all the local interface addresses on a FreeBSD
4021 system running IPv6 was broken. This may well have applied to all BSD
4022 systems, as well as to others that have similar system calls. The broken
4023 code found IPv4 interfaces correctly, but gave incorrect values for the
4024 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4025 that it would not match correctly against @[] and not recognize the IPv6
4028 PH/02 The ipliteral router was not recognizing addresses of the form user@
4029 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4031 PH/03 Added disable_ipv6.
4033 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4034 decoded form, because it is most often used to construct To: headers
4035 lines in autoreplies, and the decoded form may well be syntactically
4036 invalid. However, $reply_address has leading white space removed, and all
4037 newlines turned into spaces so that the autoreply transport does not
4040 PH/05 If group was specified without a user on a router, and no group or user
4041 was specified on a transport, the group from the router was ignored.
4043 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4044 for visible compile-time settings that can be used to change these
4045 numbers, for those that want even more. Backwards compatibility with old
4046 spool files has been maintained. However, going back to a previous Exim
4047 release will lost any variables that are in spool files.
4049 PH/07 Two small changes when running in the test harness: increase delay when
4050 passing a TCP/IP connection to a new process, in case the original
4051 process has to generate a bounce, and remove special handling of
4052 127.0.0.2 (sic), which is no longer necessary.
4054 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4055 be the same on different OS.
4057 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4060 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4061 whether --show-vars was specified or not
4063 JJ/02 exipick: Added support for new ACL variable spool format introduced
4066 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4067 syntactically invalid From: or Reply-to: line, and a filter used this to
4068 generate an autoreply, and therefore failed to obtain an address for the
4069 autoreply, Exim could try to deliver to a non-existent relative file
4070 name, causing unrelated and misleading errors. What now happens is that
4071 it logs this as a hard delivery error, but does not attempt to create a
4074 PH/11 The exinext utility has a -C option for testing purposes, but although
4075 the given file was scanned by exinext itself; it wasn't being passed on
4076 when Exim was called.
4078 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4079 an end-of-file indication when reading a command response.
4081 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4082 compiled. In many other places in Exim, IPv6 addresses are always
4083 recognized, so I have changed this. It also means that IPv4 domain
4084 literals of the form [IPV4:n.n.n.n] are now always recognized.
4086 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4087 used if the router is not running as root, for example, when verifying at
4088 ACL time, or when using -bh. The debugging output from this situation was
4089 non-existent - all you got was a failure to exec. I have made two
4092 (a) Failures to set uid/gid, the current directory, or a process leader
4093 in a subprocess such as that created by queryprogram now generate
4094 suitable debugging output when -d is set.
4096 (b) The queryprogram router detects when it is not running as root,
4097 outputs suitable debugging information if -d is set, and then runs
4098 the subprocess without attempting to change uid/gid.
4100 PH/15 Minor change to Makefile for building test_host (undocumented testing
4103 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4104 additional section of a DNS packet that returns MX or SRV records.
4105 Instead, it always explicitly searches for A/AAAA records. This avoids
4106 major problems that occur when a DNS server includes only records of one
4107 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4108 fixed another bug: if SRV records were looked up and the corresponding
4109 address records were *not* found in the additional section, the port
4110 values from the SRV records were lost.
4112 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4113 using the correct key (the original address) when searching the retry
4114 rules in order to find which one to use for generating the retry hint.
4116 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4117 adding the default one. Similarly, if it contains a Reply-To: header, the
4118 errors_reply_to option, if set, is not used.
4120 PH/19 When calculating a retry time, Exim used to measure the "time since
4121 failure" by looking at the "first failed" field in the retry record. Now
4122 it does not use this if it is later than than the arrival time of the
4123 message. Instead it uses the arrival time. This makes for better
4124 behaviour in cases where some deliveries succeed, thus re-setting the
4125 "first failed" field. An example is a quota failure for a huge message
4126 when small messages continue to be delivered. Without this change, the
4127 "time since failure" will always be short, possible causing more frequent
4128 delivery attempts for the huge message than are intended.
4129 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4131 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4132 $1, $2, $3) because the numerical variables can be reset during some
4133 expansion items (e.g. "match"), thereby losing the authentication data.
4135 PH/21 Make -bV show the size of off_t variables so that the test suite can
4136 decide whether to run tests for quotas > 2G.
4138 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4139 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4140 filecount value is greater than 2G or if a quota value is greater than 2G
4141 on a system where the size of off_t is not greater than 4, a panic error
4144 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4145 never match. The debug and -bh output now contains an explicit error
4146 message indicating a malformed IPv4 address or mask.
4148 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4149 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4150 PH/23 above applies.
4152 PH/25 Do not write to syslog when running in the test harness. The only
4153 occasion when this arises is a failure to open the main or panic logs
4154 (for which there is an explicit test).
4156 PH/26 Added the /no_tell option to "control=freeze".
4158 PH/27 If a host name lookup failed very early in a connection, for example, if
4159 the IP address matched host_lookup and the reverse lookup yielded a name
4160 that did not have a forward lookup, an error message of the form "no IP
4161 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4162 could be logged. Now it outputs the IP address instead of "NULL".
4164 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4165 allows the sender and the authenticated sender to be set when
4166 submitting a message from within Exim. Since child_open_exim() is
4167 documented for local_scan(), the new function should be too.
4169 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4170 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4171 results in an empty string is now treated as unset.
4173 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4175 PH/31 Added sender_verify_fail logging option.
4177 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4178 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4179 tidied the source and removed it altogether.
4181 PH/33 When a queue run was abandoned because the load average was too high, a
4182 log line was always written; now it is written only if the queue_run log
4183 selector is set. In addition, the log line for abandonment now contains
4184 information about the queue run such as the pid. This is always present
4185 in "start" and "stop" lines but was omitted from the "abandon" line.
4187 PH/34 Omit spaces between a header name and the colon in the error message that
4188 is given when verify = headers_syntax fails (if there are lots of them,
4189 the message gets confusing).
4191 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4192 names, as there are now some PTR records that contain slashes. This check
4193 is only to protect against broken name servers that fall over on strange
4194 characters, so the fact that it applies to all lookups doesn't matter.
4196 PH/36 Now that the new test suite is complete, we can remove some of the
4197 special code in Exim that was needed for the old test suite. For example,
4198 sorting DNS records because real resolvers return them in an arbitrary
4199 order. The new test suite's fake resolver always returns records in the
4202 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4203 bounces) except when queue_only is set, to avoid logging races between
4204 the different processes.
4206 PH/38 Panic-die if .include specifies a non-absolute path.
4208 PH/39 A tweak to the "H" retry rule from its user.
4210 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4211 a label. They prevented compilation on older perls.
4213 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4214 a warning to be raised on newish perls.
4216 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4217 on queue. Changes to match documented behaviour of showing count of
4218 messages matching specified criteria.
4220 PH/40 Changed the default ident timeout from 30s to 5s.
4222 PH/41 Added support for the use of login_cap features, on those BSD systems
4223 that have them, for controlling the resources used by pipe deliveries.
4225 PH/42 The content-scanning code uses fopen() to create files in which to put
4226 message data. Previously it was not paying any attention to the mode of
4227 the files. Exim runs with umask(0) because the rest of the code creates
4228 files with open(), and sets the required mode explicitly. Thus, these
4229 files were ending up world-writeable. This was not a big issue, because,
4230 being within the spool directory, they were not world-accessible. I have
4231 created a function called modefopen, which takes an additional mode
4232 argument. It sets umask(777), creates the file, chmods it to the required
4233 mode, then resets the umask. All the relevant calls to fopen() in the
4234 content scanning code have been changed to use this function.
4236 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4237 to 24 hours. This avoids potential overflow problems when processing G
4238 and H retry rules. I suspect nobody ever tinkers with this value.
4240 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4242 PH/45 When the plaintext authenticator is running as a client, the server's
4243 challenges are checked to ensure they are valid base64 strings. By
4244 default, the authentication attempt is cancelled if an invalid string is
4245 received. Setting client_ignore_invalid_base64 true ignores these errors.
4246 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4247 they are received. Thus, the responses can be made to depend on the
4248 challenges. If an invalid string is ignored, an empty string is placed in
4251 PH/46 Messages that are created by the autoreply transport now contains a
4252 References: header, in accordance with RFCs 2822 and 3834.
4254 PH/47 Added authenticated_sender_force to the smtp transport.
4256 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4258 PH/49 Installed latest patch from the Sieve maintainer.
4260 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4261 was also set, the appendfile transport was unnecessarily scanning a
4262 directory of message files (e.g. for maildir delivery) to find the count
4263 of files (along with the size), even though it did not need this
4264 information. It now does the scan only if it needs to find either the
4265 size of the count of files.
4267 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4269 PH/52 Two bugs concerned with error handling when the smtp transport is
4272 (i) Exim was not creating retry information for temporary errors given
4273 for individual recipients after the DATA command when the smtp transport
4274 was used in LMTP mode. This meant that they could be retried too
4275 frequently, and not timed out correctly.
4277 (ii) Exim was setting the flag that allows error details to be returned
4278 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4279 recipients that were returned after the DATA command.
4281 PH/53 This is related to PH/52, but is more general: for any failing address,
4282 when detailed error information was permitted to be returned to the
4283 sender, but the error was temporary, then after the final timeout, only
4284 "retry timeout exceeded" was returned. Now it returns the full error as
4285 well as "retry timeout exceeded".
4287 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4288 do this, and (what is worse) MTAs that accept it.
4290 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4291 will now be deprecated.
4293 PH/56 New os.c-cygwin from the Cygwin maintainer.
4295 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4296 formats (previously only available in exim formats via -bpr, -bpru,
4297 and -bpra. Now also available in native and exiqgrep formats)
4299 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4300 with very large, slow to parse queues
4302 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4304 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4306 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4307 responses to authentication challenges, though it was showing the
4308 challenges; (ii) I've removed the CR characters from the debug output for
4311 PH/58 Allow for the insertion of a newline as well as a space when a string
4312 is turned into more than one encoded-word during RFC 2047 encoding. The
4313 Sieve code now uses this.
4315 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4316 data_4xx, lost_connection, tls_required.
4318 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4319 message was being sent as an SMTP response.
4321 PH/61 Add -l and -k options to exicyclog.
4323 PH/62 When verifying, if an address was redirected to one new address, so that
4324 verification continued, and the new address failed or deferred after
4325 having set something in $address_data, the value of $address_data was not
4326 passed back to the ACL. This was different to the case when no
4327 redirection occurred. The value is now passed back in both cases.
4329 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4330 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4331 use login_cap.h, so on its own it isn't the distinguishing feature. The
4332 new name refers directly to the setclassresources() function.
4334 PH/65 Added configuration files for NetBSD3.
4336 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4338 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4339 is preferred over IPv4.
4341 PH/68 The bounce_return_message and bounce_return_body options were not being
4342 honoured for bounces generated during the reception of non-SMTP messages.
4343 In particular, this applied to messages rejected by the ACL. This bug has
4344 been fixed. However, if bounce_return_message is true and bounce_return_
4345 body is false, the headers that are returned for a non-SMTP message
4346 include only those that have been read before the error was detected.
4347 (In the case of an ACL rejection, they have all been read.)
4349 PH/69 The HTML version of the specification is now built in a directory called
4350 spec_html instead of spec.html, because the latter looks like a path with
4351 a MIME-type, and this confuses some software.
4353 PH/70 Catch two compiler warnings in sieve.c.
4355 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4356 function verify_get_ident() calls ip_connect() to connect a socket, but
4357 if the "connect()" function timed out, ip_connect() used to close the
4358 socket. However, verify_get_ident() also closes the socket later, and in
4359 between Exim writes to the log, which may get opened at this point. When
4360 the socket was closed in ip_connect(), the log could get the same file
4361 descriptor number as the socket. This naturally causes chaos. The fix is
4362 not to close the socket in ip_connect(); the socket should be closed by
4363 the function that creates it. There was only one place in the code where
4364 this was missing, in the iplookup router, which I don't think anybody now
4365 uses, but I've fixed it anyway.
4367 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4368 well as to direct DNS lookups. Otherwise the handling of names in host
4369 lists is inconsistent and therefore confusing.
4375 PH/01 Two changes to the default runtime configuration:
4377 (1) Move the checks for relay_from_hosts and authenticated clients from
4378 after to before the (commented out) DNS black list checks.
4380 (2) Add control=submission to the relay_from_hosts and authenticated
4381 clients checks, on the grounds that messages accepted by these
4382 statements are most likely to be submissions.
4384 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4386 (1) Generate an error if the third argument for the ${prvs expansion is
4389 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4392 (3) Reset the variables that are obtained from the first argument of
4393 ${prvscheck and used in the second argument before leaving the code,
4394 because their memory is reclaimed, so using them afterwards may do
4397 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4398 one (it's much easier than Tom thought :-).
4400 (5) Because of (4), we can now allow for the use of $prvscheck_result
4401 inside the third argument.
4403 PH/03 For some reason, the default setting of PATH when running a command from
4404 a pipe transport was just "/usr/bin". I have changed it to
4407 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4408 anything to be listed in the output from -bV.
4410 PH/05 When a filter generated an autoreply, the entire To: header line was
4411 quoted in the delivery log line, like this:
4413 => >A.N.Other <ano@some.domain> <original@ddress> ...
4415 This has been changed so that it extracts the operative address. There
4416 may be more than one such address. If so, they are comma-separated, like
4419 => >ano@some.domain,ona@other.domain <original@ddress> ...
4421 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4422 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4423 not being looked up in the rDNS to get a host name, Exim was showing the
4424 IP address twice in Received: lines, even though the IP addresses were
4425 identical. For example:
4427 Received: from [1.2.3.4] (helo=[1.2.3.4])
4429 However, if the real host name was known, it was omitting the HELO data
4430 if it matched the actual IP address. This has been tidied up so that it
4431 doesn't show the same IP address twice.
4433 PH/07 When both +timestamp and +memory debugging was on, the value given by
4434 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4435 called by the debug printing, thereby overwriting the timestamp buffer.
4436 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4438 PH/08 When the original message was included in an autoreply transport, it
4439 always said "this is a copy of the message, including all the headers",
4440 even if body_only or headers_only was set. It now gives an appropriate
4443 PH/09 Applied a patch from the Sieve maintainer which:
4445 o fixes some comments
4446 o adds the (disabled) notify extension core
4447 o adds some debug output for the result of if/elsif tests
4448 o points to the current vacation draft in the documentation
4449 and documents the missing references header update
4453 o fixes a bug in processing the envelope test (when testing
4454 multiple envelope elements, the last element determined the
4457 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4458 Electronic Mail") by including:
4460 Auto-submitted: auto-generated
4462 in the messages that it generates (bounce messages and others, such as
4463 warnings). In the case of bounce messages for non-SMTP messages, there was
4464 also a typo: it was using "Auto_submitted" (underscore instead of
4465 hyphen). Since every message generated by Exim is necessarily in response
4466 to another message, thes have all been changed to:
4468 Auto-Submitted: auto-replied
4470 in accordance with these statements in the RFC:
4472 The auto-replied keyword:
4474 - SHOULD be used on messages sent in direct response to another
4475 message by an automatic process,
4477 - MUST NOT be used on manually-generated messages,
4479 - MAY be used on Delivery Status Notifications (DSNs) and Message
4480 Disposition Notifications (MDNs),
4482 - MUST NOT be used on messages generated by automatic or periodic
4483 processes, except for messages which are automatic responses to
4486 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4487 to the default Received: header definition.
4489 PH/12 Added log selector acl_warn_skipped (default on).
4491 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4492 variables because (a) they are in the wrong storage pool and (b) even if
4493 they were copied, it wouldn't work properly because of the caching.
4495 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4496 checking when decoding. Apparently there are clients that generate
4497 overlong encoded strings. Why am I not surprised?
4499 PH/15 If the first argument of "${if match_address" was not empty, but did not
4500 contain an "@" character, Exim crashed. Now it writes a panic log message
4501 and treats the condition as false.
4503 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4505 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4506 extension "envelope-auth". The code is finished and in agreement with
4507 other implementations, but there is no documentation so far and in fact,
4508 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4509 not changing the active code.
4511 Print executed "if" and "elsif" statements when debugging is used. This
4512 helps a great deal to understand what a filter does.
4514 Document more things not specified clearly in RFC3028. I had all this
4515 sorted out, when out of a sudden new issues came to my mind. Oops."
4517 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4520 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4521 canonical form (as documented). However, after a host name lookup from
4522 the IP address, check_host() was doing a simple string comparison with
4523 addresses acquired from the DNS when checking that the found name did
4524 have the original IP as one of its addresses. Since any found IPv6
4525 addresses are likely to be in abbreviated form, the comparison could
4526 fail. Luckily, there already exists a function for doing the comparison
4527 by converting both addresses to binary, so now that is used instead of
4528 the text comparison.
4530 PH/20 There was another similar case to PH/19, when a complete host name was
4531 given in a host list; looking up its IP address could give an abbreviated
4532 form, whereas the current host's name might or might not be abbreviated.
4533 The same fix has been applied.
4539 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4540 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4543 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4544 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4546 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4548 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4549 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4550 Also, for "Auto-Submitted", treat anything other than "no" as
4551 non-personal, in accordance with RFC 3834. (Previously it treated
4552 anything starting "auto-" as non-personal.)
4554 TF/01 The control=submission/name=... option had a problem with syntax
4555 errors if the name included a slash character. The /name= option
4556 now slurps the rest of the string, so it can include any characters
4557 but it must come last in the list of options (after /sender_retain
4560 PH/05 Some modifications to the interface to the fake nameserver for the new
4568 TK/01 Added the "success_on_redirect" address verification option. See
4569 NewStuff for rationale and an example.
4571 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4573 PH/02 Patch to exigrep to allow it to work on syslog lines.
4575 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4576 fread() to skip over the body file's header line, because in Cygwin the
4577 header line is locked and is inaccessible.
4579 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4580 co-exist for some time) to make it clear that it is the Exim ID that is
4581 referenced, not the Message-ID: header line.
4583 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4584 string_format() function, because snprintf() does not exist on all
4587 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4588 ${stat: expansion item.
4590 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4591 protocol synchronization error", to keep the pedants happy.
4593 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4594 well as for IRIX systems, when gcc is being used. See the host.c source
4597 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4599 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4602 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4603 transport and to the smtp transport in LMTP mode.
4605 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4607 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4608 run a filter in a subprocess. This could lead to confusion in subsequent
4609 lookups in the parent process. There should also be a search_tidyup() at
4610 the end of the subprocess.
4612 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4613 only if the host matched helo_try_verify_hosts, which caused the
4614 verification to occur when the EHLO/HELO command was issued. The ACL just
4615 tested the remembered result. Now, if a previous verification attempt has
4616 not happened, "verify = helo" does it there and then.
4618 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4620 TK/03 Fix log output including CR from clamd.
4622 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4623 exist provoked a memory error which could cause a segfault.
4625 PH/15 Installed PCRE 6.2
4627 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4629 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4630 of the problem. Specifically, suggested +O2 rather than +O1 for the
4633 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4635 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4636 record, the port was not correctly logged when the outgoing_port log
4637 selector was set (it logged the transort's default port).
4639 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4640 fallback_hosts, and "hosts" in the smtp transport.
4642 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4643 host errors such as "Connection refused".
4645 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4646 authentication with radiusclient 0.4.9:
4648 - Error returned from rc_read_config was caught wrongly
4649 - Username/password not passed on to radius server due to wrong length.
4651 The presumption is that some radiusclient API changes for 4.51/PH/17
4652 were not taken care of correctly. The code is still untested by me (my
4653 Linux distribution still has 0.3.2 of radiusclient), but it was
4654 contributed by a Radius user.
4656 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4657 expanding the "port" option of the smtp transport.
4659 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4660 while reading a MIME header. Thanks to Tom Hughes for a patch.
4662 PH/24 Include config.h inside local_scan.h so that configuration settings are
4665 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4666 that in an ACL for RCPT (for example), you can examine exactly what was
4669 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4670 commands, but it was not correctly comparing the address with the actual
4671 client host address. Thus, it would show the EHLO address in Received:
4672 header lines when this was not necessary.
4674 PH/27 Added the % operator to ${eval:}.
4676 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4677 it should be ignoring failures (because with -C, for example, it has lost
4678 privilege). It wasn't ignoring creation failures other than "already
4681 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4684 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4685 because an input line was too long, either on its own, or by virtue of
4686 too many continuations, the temporary file was not being removed, and the
4687 return code was incorrect.
4689 PH/31 Missing "BOOL" in function definition in filtertest.c.
4691 PH/32 Applied Sieve patches from the maintainer.
4693 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4695 PH/33 Added "verify = not_blind".
4697 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4698 Local/Makefile (with some defaults set). These are used in built scripts
4699 such as exicyclog, but they have never been used in the exim_install
4700 script (though there are many overriding facilities there). I have
4701 arranged that the exim_install script now takes note of these two
4704 PH/35 Installed configuration files for Dragonfly.
4706 PH/36 When a locally submitted message by a trusted user did not contain a
4707 From: header, and the sender address was obtained from -f or from an SMTP
4708 MAIL command, and the trusted user did not use -F to supply a sender
4709 name, $originator_name was incorrectly used when constructing a From:
4710 header. Furthermore, $originator_name was used for submission mode
4711 messages from external hosts without From: headers in a similar way,
4712 which is clearly wrong.
4714 PH/37 Added control=suppress_local_fixups.
4716 PH/38 When log_selector = +received_sender was set, and the addition of the
4717 sender made the log line's construction buffer exactly full, or one byte
4718 less than full, an overflow happened when the terminating "\n" was
4721 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4722 when the result of a list match is failure because a DNS lookup failed.
4724 PH/40 RM_COMMAND is now used in the building process.
4726 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4727 the "build-* directories that it finds.
4729 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4730 address in a domain literal was a prefix of an interface address.
4732 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4733 when verifying a sender address, unless rewrite_headers is false.
4735 PH/44 Wrote a long comment about why errors_to addresses are verified as
4736 recipients, not senders.
4738 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4739 the ratelimit ACL was added.
4741 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4743 PH/46 Added extra information about PostgreSQL errors to the error string.
4745 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4746 avoiding the need to install special zones in a real server. This is
4747 backwards compatible; if it can't find the fake resolver, it drops back.
4748 Thus, both old and new test suites can be run.
4750 TF/02 Added util/ratelimit.pl
4752 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4753 clock is set back in time.
4755 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4756 Candler <B.Candler@pobox.com>.
4758 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4759 OK for addresses that are the result of redirections.
4761 PH/48 A number of further additions for the benefit of the new test suite,
4762 including a fake gethostbyname() that interfaces to the fake DNS resolver
4765 TF/06 The fix for widen_domains has also been applied to qualify_single and
4766 search_parents which are the other dnslookup options that can cause
4769 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4772 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4774 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4775 service" if it encountered an MX record with an empty target hostname.
4776 The message is now "an MX or SRV record indicated no SMTP service".
4778 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4779 if the DNS of the sending site is misconfigured. This is quite a
4780 common situation. This change restores the behaviour of treating a
4781 helo verification defer as a failure.
4783 PH/51 If self=fail was set on a router, the bounce message did not include the
4784 actual error message.
4790 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4792 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4793 command itself ended in error, the underlying message about the transport
4794 filter timeout was being overwritten with the pipe command error. Now the
4795 underlying error message should be appended to the second error message.
4797 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4799 PH/02 Reduce the amount of output that "make" produces by default. Full output
4800 can still be requested.
4802 PH/03 The warning log line about a condition test deferring for a "warn" verb
4803 was being output only once per connection, rather than after each
4804 occurrence (because it was using the same function as for successful
4805 "warn" verbs). This seems wrong, so I have changed it.
4807 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4808 it should not have, which might have caused a crash in the right
4809 circumstances, but probably never did.
4811 PH/04 Installed a modified version of Tony Finch's patch to make submission
4812 mode fix the return path as well as the Sender: header line, and to
4813 add a /name= option so that you can make the user's friendly name appear
4816 TF/03 Added the control = fakedefer ACL modifier.
4818 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4819 Mark Lowes for thorough testing.
4821 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4823 TK/03 Merged latest SRS patch from Miles Wilton.
4825 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4826 with the definition in sysexits.h (which is #included earlier).
4827 Fortunately, Exim does not actually use EX_OK. The code used to try to
4828 preserve the sysexits.h value, by assuming that macro definitions were
4829 scanned for macro replacements. I have been disabused of this notion,
4830 so now the code just undefines EX_OK before #including unistd.h.
4832 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4833 in the smtp transport. When a block could not be written in a single
4834 write() function, the timeout was being re-applied to each part-write.
4835 This seems wrong - if the receiver was accepting one byte at a time it
4836 would take for ever. The timeout is now adjusted when this happens. It
4837 doesn't have to be particularly precise.
4839 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4840 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4842 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4843 without a domain if the check to <postmaster@domain> fails.
4845 SC/01 Eximstats: added -xls and the ability to specify output files
4846 (patch written by Frank Heydlauf).
4848 SC/02 Eximstats: use FileHandles for outputting results.
4850 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4852 SC/04 Eximstats: fixed display of large numbers with -nvr option
4854 SC/05 Eximstats: fixed merging of reports with empty tables.
4856 SC/06 Eximstats: added the -include_original_destination flag
4858 SC/07 Eximstats: removed tabs and trailing whitespace.
4860 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4862 TK/06 MBOX spool code: Add real "From " MBOX separator line
4863 so the .eml file is really in mbox format (even though
4864 most programs do not really care). Patch from Alex Miller.
4866 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4867 The latter is generated from $received_to and is only set if the
4868 message has one envelope recipient. SA can use these headers,
4869 obviously out-of-the-box. Patch from Alex Miller.
4871 PH/08 The ${def test on a variable was returning false if the variable's
4872 value was "0", contrary to what the specification has always said!
4873 The result should be true unless the variable is empty.
4875 PH/09 The syntax error of a character other than { following "${if
4876 def:variable_name" (after optional whitespace) was not being diagnosed.
4877 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4878 accidental colon was present, for example, could give incorrect results.
4880 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4881 result is used (not including appendfile, where other changes are about
4884 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4885 This involved changing a lot of size variables from int to off_t. It
4886 should work with maildirs and everything.
4888 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4889 spamd dying while we are connected to it.
4891 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4894 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4895 function for BDB 4.3.
4897 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4899 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4900 the sources for PCRE have been re-arranged and more files are now
4903 PH/15 The code I had for printing potentially long long variables in PH/11
4904 above was not the best (it lost precision). The length of off_t variables
4905 is now inspected at build time, and an appropriate printing format (%ld
4906 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4907 to be "long long int" or "long int". This is needed for the internal
4908 formatting function string_vformat().
4910 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4911 the configuration file to be ":syslog", then the script "guesses" where
4912 the logs files are, rather than using the compiled in default. In our
4913 case the guess is not the same as the compiled default, so the script
4914 suddenly stopped working when I started to use syslog. The patch checks
4915 to see if log_file_path is "". If so, it attempts to read it from exim
4916 with no configuration file to get the compiled in version, before it
4917 falls back to the previous guessing code."
4919 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4920 implementing BATV in an Exim configuration. See NewStuff for the gory
4923 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4924 Makefile that are specific to HP-UX.
4926 PH/18 If the "use_postmaster" option was set for a recipient callout together
4927 with the "random" option, the postmaster address was used as the MAIL
4928 FROM address for the random test, but not for the subsequent recipient
4929 test. It is now used for both.
4931 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4932 patch removes a few documentation additions to RFC 3028, because the
4933 latest draft now contains them. It adds the new en;ascii-case comparator
4934 and a new error check for 8bit text in MIME parts. Comparator and
4935 require names are now matched exactly. I enabled the subaddress
4936 extension, but it is not well tested yet (read: it works for me)."
4938 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4939 rework some of the code of TK/09 above to avoid the hardwired use of
4940 "%lld" and "long long". Replaced the call to snprintf() with a call to
4943 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4944 records point to non-existent hosts", "retry timeout exceeded", and
4945 "retry time not reached for any host after a long failure period".
4947 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4948 experimental DomainKeys support:
4950 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4951 (2) On an error such as an illegally used "control", the wrong name for
4952 the control was given.
4954 These problems did NOT occur unless DomainKeys support was compiled.
4956 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4958 PH/24 Added ${if match_ip condition.
4960 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4961 databases so that it will be absolutely obvious if a crash occurs in the
4962 DB library. This is a regular occurrence (often caused by mis-matched
4965 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4966 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4967 picked up on a user's system that detects such things. There doesn't seem
4968 to be a gcc warning option for this - only an attribute that has to be
4969 put on the function's prototype. It seems that in Fedora Core 4 they have
4970 set this on a number of new functions. No doubt there will be more in due
4973 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4974 specify a transport. However, if an address that was verified by such a
4975 router was the subject of a callout, Exim crashed because it tried to
4976 read the rcpt_include_affixes from the non-existent transport. Now it
4977 just assumes that the setting of that option is false. This bug was
4978 introduced by 4.51/PH/31.
4980 PH/28 Changed -d+all to exclude +memory, because that information is very
4981 rarely of interest, but it makes the output a lot bigger. People tend to
4982 do -d+all out of habit.
4984 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4985 code in os-type was giving problems when libc.so lives in lib64, like on
4988 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4989 aren't the modern standard, and it seems that some systems' include files
4990 don't always have them. Exim was already checking for some of the newer
4991 ones like T_AAAA, and defining it itself. I've added checks for all the
4992 record types that Exim uses.
4994 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4995 not automatically generating a new one, as it is supposed to. This
4996 prevented TLS from working. If the file did exist, but contained invalid
4997 data, a new version was generated, as expected. It was only the case of a
4998 non-existent file that was broken.
5000 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5001 with a change in libdomainkeys > 0.64.
5003 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5004 from DNS. If the selector record carries the flag, it now has
5005 precedence over the domain-wide flag.
5007 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5009 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5010 the use of a port name that isn't defined in /etc/services) occurred, the
5011 message was deferred as in a normal delivery, and thus remained on the
5012 spool, instead of being failed because of the mua_wrapper setting. This
5013 is now fixed, and I tidied up some of the mua_wrapper messages at the
5016 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5017 the messages in a hash of arrays rather than using individual hashes.
5018 This is a bit cleaner and results in dramatic memory savings, albeit
5019 at a slight CPU cost.
5021 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5022 as requested by Marc Sherman.
5024 SC/10 Eximstats: added histograms for user specified patterns as requested
5027 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5029 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5030 fopen() in the content-scanning modules that did not already have it.
5036 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5037 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5039 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5041 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5043 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5044 to test Sieve filters that use "vacation".
5046 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5047 that changes the way the GnuTLS parameters are stored in the cache file.
5048 The new format can be generated externally. For backward compatibility,
5049 if the data in the cache doesn't make sense, Exim assumes it has read an
5050 old-format file, and it generates new data and writes a new file. This
5051 means that you can't go back to an older release without removing the
5054 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5055 work if there are any delivery delays because "one_time" forces the
5056 parent to be marked "delivered", so its unseen clone is never tried
5057 again. For this reason, Exim now forbids the simultaneous setting of
5060 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5061 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5062 entry if you want to know the details. The fix, however, neglected to
5063 consider the case when local delivery batching is involved. The test for
5064 "previously delivered" was not happening when checking to see if an
5065 address could be batched with a previous (undelivered) one; under
5066 certain circumstances this could lead to multiple deliveries to the same
5069 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5070 in its include files, and this causes problems building Exim.
5072 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5073 header_syntax) but Exim was just ignoring anything given after a slash.
5074 In particular, this caused confusion with an attempt to use "verify =
5075 reverse_host_lookup/defer_ok". An error is now given when options are
5076 supplied for verify items that do not have them. (Maybe reverse_host_
5077 lookup should have a defer_ok option, but that's a different point.)
5079 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5080 defined by RFC 821) to 2048, because there were problems with some AUTH
5081 commands, and RFC 1869 says the size should be increased for extended
5082 SMTP commands that take arguments.
5084 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5087 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5088 "unknown" error; now it says that the functionality isn't in the binary.
5090 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5091 an address' error message when a string expansion fails (syntax or
5092 whatever). Otherwise the password may appear in the log. Following change
5093 PH/42 below, there is no longer a chance of it appearing in a bounce
5096 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5098 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5099 of its addresses was ever tried. (Bugzilla bug #2.)
5101 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5102 the result incorrectly in the debug output. (It correctly added a newline
5103 to what was transported.)
5105 TF/01 Added $received_time.
5107 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5108 commented out examples of how to interface to a virus scanner and to
5109 SpamAssassin. Also added commented examples of av_scanner and
5110 spamd_address settings.
5112 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5113 and controls are allowed in which ACLs. There were a couple of minor
5114 errors. Some of the entries in the conditions table (which is a table of
5115 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5116 negation of where the condition IS allowed.
5118 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5120 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5121 header file does not have a version number, so I've had to invent a new
5122 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5123 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5124 radiusclient), but it was contributed by a Radius user.
5126 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5127 files or directories, for OpenSSL.
5129 PH/19 When an Exim process that is running as root has to create an Exim log
5130 file, it does so in a subprocess that runs as exim:exim so as to get the
5131 ownership right at creation (otherwise, other Exim processes might see
5132 the file with the wrong ownership). There was no test for failure of this
5133 fork() call, which would lead to the process getting stuck as it waited
5134 for a non-existent subprocess. Forks do occasionally fail when resources
5135 run out. I reviewed all the other calls to fork(); they all seem to check
5138 PH/20 When checking for unexpected SMTP input at connect time (before writing
5139 the banner), Exim was not dealing correctly with a non-positive return
5140 from the read() function. If the client had disconnected by this time,
5141 the result was a log entry for a synchronization error with an empty
5142 string after "input=" when read() returned zero. If read() returned -1
5143 (an event I could not check), uninitialized data bytes were printed.
5144 There were reports of junk text (parts of files, etc) appearing after
5147 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5149 PH/22 Added support for macro redefinition, and (re)definition in between
5150 driver and ACL definitions.
5152 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5153 forgetting to use the resulting value; it was using the unexpanded value.
5155 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5156 hadn't been configured. The fix is from Juergen Kreileder, who
5157 understands it better than I do:
5159 "Here's what I see happening with three configured cyrus_sasl
5160 authenticators configured (plain, login, cram-md5):
5162 On startup auth_cyrus_sasl_init() gets called for each of these.
5163 This means three calls to sasl_listmech() without a specified mech_list.
5164 => SASL tests which mechs of all available mechs actually work
5165 => three warnings about OTP not working
5166 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5168 With the patch, sasl_listmech() also gets called three times. But now
5169 SASL's mech_list option is set to the server_mech specified in the the
5170 authenticator. Or in other words, the answer from sasl_listmech()
5171 gets limited to just the mech you're testing for (which is different
5173 => the return list contains just 'plain' or 'login', 'cram-md5' or
5174 nothing depending on the value of ob->server_mech.
5176 I've just tested the patch: Authentication still works fine,
5177 unavailable mechs specified in the exim configuration are still
5178 caught, and the auth.log warnings about OTP are gone."
5180 PH/25 When debugging is enabled, the contents of the command line are added
5181 to the debugging output, even when log_selector=+arguments is not
5184 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5185 answer is "GNU", and only if the return is "GNU/something" is the answer
5188 PH/27 $acl_verify_message is now set immediately after the failure of a
5189 verification in an ACL, and so is available in subsequent modifiers. In
5190 particular, the message can be preserved by coding like this:
5192 warn !verify = sender
5193 set acl_m0 = $acl_verify_message
5195 Previously, $acl_verify_message was set only while expanding "message"
5196 and "log_message" when a very denied access.
5198 PH/28 Modified OS/os.c-Linux with
5200 -#ifndef OS_LOAD_AVERAGE
5201 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5203 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5204 nomenclature these days.)
5206 PH/29 Installed patch from the Sieve maintainer that adds the options
5207 sieve_useraddress and sieve_subaddress to the redirect router.
5209 PH/30 In these circumstances:
5210 . Two addresses routed to the same list of hosts;
5211 . First host does not offer TLS;
5212 . First host accepts first address;
5213 . First host gives temporary error to second address;
5214 . Second host offers TLS and a TLS session is established;
5215 . Second host accepts second address.
5216 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5217 and peerdn, if requested) that were in fact used only for the second
5220 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5221 attention to any local part prefix or suffix that was matched by the
5222 router that accepted the address. It now behaves in the same way as it
5223 does for delivery: the affixes are removed from the local part unless
5224 rcpt_include_affixes is set on the transport.
5226 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5227 timeout during the DATA phase of an incoming message.
5229 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5230 applied a patch sanctioned by the Sieve maintainer.
5232 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5233 the uid or gid is negative. A case of a negative gid caused this to be
5234 noticed. The fix allows for either to be negative.
5236 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5237 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5240 PH/36 The change PH/12 above was broken. Fixed it.
5242 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5243 the grounds that routing the same address twice would always produce the
5244 same answer. This might have been true once, but it is certainly no
5245 longer true now. Routing a child address may depend on the previous
5246 routing that produced that child. Some complicated redirection strategies
5247 went wrong when messages had multiple recipients, and made Exim's
5248 behaviour dependent on the order in which the addresses were given.
5250 I have moved the duplicate checking until after the routing is complete.
5251 Exim scans the addresses that are assigned to local and remote
5252 transports, and removes any duplicates. This means that more work will be
5253 done, as duplicates will always all be routed, but duplicates are
5254 presumably rare, so I don't expect this is of any significance.
5256 For deliveries to pipes, files, and autoreplies, the duplicate checking
5257 still happens during the routing process, since they are not going to be
5260 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5261 It corrects a timeout issue with spamd. This is Ian's comment: "The
5262 background is that sometimes spamd either never reads data from a
5263 connection it has accepted, or it never writes response data. The exiscan
5264 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5265 blindly assumes that writes won't block so it may never time out."
5267 PH/39 Allow G after quota size as well as K and M.
5269 PH/40 The value set for $authenticated_id in an authenticator may not contain
5270 binary zeroes or newlines because the value is written to log lines and
5271 to spool files. There was no check on this. Now the value is run through
5272 the string_printing() function so that such characters are converted to
5273 printable escape sequences.
5275 PH/41 $message_linecount is a new variable that contains the total number of
5276 lines in the message. Compare $body_linecount, which is the count for the
5279 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5280 bounce and delay warning messages, except in certain special cases, which
5283 (a) An SMTP error message from a remote host;
5284 (b) A message specified in a :fail: redirection;
5285 (c) A message specified in a "fail" command in a system filter;
5286 (d) A message specified in a FAIL return from the queryprogram router;
5287 (e) A message specified by the cannot_route_message router option.
5289 In these cases only, Exim does include the error details in bounce and
5290 warning messages. There are also a few cases where bland messages such
5291 as "unrouteable address" or "local delivery error" are given.
5293 PH/43 $value is now also set for the "else" part of a ${run expansion.
5295 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5296 being worked on, but at least Exim now implements the latest version to
5299 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5300 process to complete was treated as a delivery failure, a timeout while
5301 writing the message to the pipe was logged, but erroneously treated as a
5302 successful delivery. Such timeouts include transport filter timeouts. For
5303 consistency with the overall process timeout, these timeouts are now
5304 treated as errors, giving rise to delivery failures by default. However,
5305 there is now a new Boolean option for the pipe transport called
5306 timeout_defer, which, if set TRUE, converts the failures into defers for
5307 both kinds of timeout. A transport filter timeout is now identified in
5310 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5311 systems where "make" and "gmake" are different, calling "gmake" at top
5312 level broke things. I've arranged for the value of $(MAKE) to be passed
5313 from the Makefile to this script so that it can call the same version of
5317 A note about Exim versions 4.44 and 4.50
5318 ----------------------------------------
5320 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5321 changes of various kinds. As a consequence, a big documentation update was
5322 needed. This delayed the release for rather longer than seemed good, especially
5323 in the light of a couple of (minor) security issues. Therefore, the changes
5324 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5325 release. So 4.44 and 4.50 are in effect two different branches that both start
5328 I have left the 4.50 change log unchanged; it contains all the changes since
5329 4.43. The change log for 4.44 is below; many of its items are identical to
5330 those for 4.50. This seems to be the most sensible way to preserve the
5331 historical information.
5337 1. Minor wording change to the doc/README.SIEVE file.
5339 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5340 computation of the current number of files was incorrect.
5342 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5343 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5346 4. Give more explanation in the error message when the command for a transport
5347 filter fails to execute.
5349 5. There are several places where Exim runs a non-Exim command in a
5350 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5351 was being done only for the command run by the queryprogram router. It is
5352 now done for all such subprocesses. The other cases are: ${run, transport
5353 filters, and the commands run by the lmtp and pipe transports.
5355 6. Added CONFIGURE_GROUP build-time option.
5357 7. Some older OS have a limit of 256 on the maximum number of file
5358 descriptors. Exim was using setrlimit() to set 1000 as a large value
5359 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5360 systems. I've change it so that if it can't get 1000, it tries for 256.
5362 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5363 was an oversight, and furthermore, ever since the addition of extra
5364 controls (e.g. 4.43/32), the checks on when to allow different forms of
5365 "control" were broken. There should now be diagnostics for all cases when a
5366 control that does not make sense is encountered.
5368 9. Added the /retain_sender option to "control=submission".
5370 10. $recipients is now available in the predata ACL (oversight).
5372 11. Tidy the search cache before the fork to do a delivery from a message
5373 received from the command line. Otherwise the child will trigger a lookup
5374 failure and thereby defer the delivery if it tries to use (for example) a
5375 cached ldap connection that the parent has called unbind on.
5377 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5378 of $address_data from the recipient verification was clobbered by the
5379 sender verification.
5381 13. The value of address_data from a sender verification is now available in
5382 $sender_address_data in subsequent conditions in the ACL statement.
5384 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5386 15. Added a new option "connect=<time>" to callout options, to set a different
5389 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5390 was its contents. (It was OK if the option was not defined at all.)
5392 17. A "Completed" log line is now written for messages that are removed from
5393 the spool by the -Mrm option.
5395 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5396 information about exactly what failed.
5398 19. Added -dd to debug only the daemon process.
5400 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5401 handles timeouts, both on the server side and network timeouts. Renamed the
5402 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5404 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5405 It is now set to "smtps".
5407 22. $host_address is now set to the target address during the checking of
5408 ignore_target_hosts.
5410 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5411 being passed; this would have caused $sender_host_name to have been used if
5412 matching the list had actually called for a host name (not very likely,
5413 since this list is usually IP addresses). A host name is now passed as
5416 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5417 code that specifies a non-restarting handler (typically sigaction() in
5418 modern systems) in an attempt to fix a rare and obscure crash bug.
5420 25. Narrowed the window for a race in the daemon that could cause it to ignore
5421 SIGCHLD signals. This is not a major problem, because they are used only to
5422 wake it up if nothing else does.
5424 26. A malformed maildirsize file could cause Exim to calculate negative values
5425 for the mailbox size or file count. Odd effects could occur as a result.
5426 The maildirsize information is now recalculated if the size or filecount
5429 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5430 support for a long time. Removed HAVE_SYS_VFS_H.
5432 28. Installed the latest version of exipick from John Jetmore.
5434 29. In an address list, if the pattern was not a regular expression, an empty
5435 subject address (from a bounce message) matched only if the pattern was an
5436 empty string. Non-empty patterns were not even tested. This was the wrong
5437 because it is perfectly reasonable to use an empty address as part of a
5438 database query. An empty address is now tested by patterns that are
5439 lookups. However, all the other forms of pattern expect the subject to
5440 contain a local part and a domain, and therefore, for them, an empty
5441 address still always fails if the pattern is not itself empty.
5443 30. Exim went into a mad DNS loop when attempting to do a callout where the
5444 host was specified on an smtp transport, and looking it up yielded more
5445 than one IP address.
5447 31. Re-factored the code for checking spool and log partition space into a
5448 function that finds that data and another that does the check. The former
5449 is then used to implement four new variables: $spool_space, $log_space,
5450 $spool_inodes, and $log_inodes.
5452 32. The RFC2047 encoding function was originally intended for short strings
5453 such as real names; it was not keeping to the 75-character limit for
5454 encoded words that the RFC imposes. It now respects the limit, and
5455 generates multiple encoded words if necessary. To be on the safe side, I
5456 have increased the buffer size for the ${rfc2047: expansion operator from
5459 33. It is now permitted to omit both strings after an "if" condition; if the
5460 condition is true, the result is "true". As before, when the second string
5461 is omitted, a false condition yields an empty string. This makes it less
5462 cumbersome to write custom ACL and router conditions.
5464 34. Failure to deliver a bounce message always caused it to be frozen, even if
5465 there was an errors_to setting on the router. The errors_to setting is now
5468 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5469 canonical form (fully expanded) before being placed in
5470 $sender_host_address.
5472 36. The table in the code that translates DNS record types into text (T_A to
5473 "A" for instance) was missing entries for NS and CNAME. It is just possible
5474 that this could have caused confusion if both these types were looked up
5475 for the same domain, because the text type is used as part of Exim's
5476 per-process caching. But the chance of anyone hitting this buglet seems
5479 37. The dnsdb lookup has been extended in a number of ways.
5481 (1) There is a new type, "zns", which walks up the domain tree until it
5482 finds some nameserver records. It should be used with care.
5484 (2) There is a new type, "mxh", which is like "mx" except that it returns
5485 just the host names, not the priorities.
5487 (3) It is now possible to give a list of domains (or IP addresses) to be
5488 looked up. The behaviour when one of the lookups defers can be
5489 controlled by a keyword.
5491 (4) It is now possible to specify the separator character for use when
5492 multiple records are returned.
5494 38. The dnslists ACL condition has been extended: it is now possible to supply
5495 a list of IP addresses and/or domains to be looked up in a particular DNS
5498 39. Added log_selector=+queue_time_overall.
5500 40. When running the queue in the test harness, wait just a tad after forking a
5501 delivery process, to get repeatability of debugging output.
5503 41. Include certificate and key file names in error message when GnuTLS fails
5504 to set them up, because the GnuTLS error message doesn't include the name
5505 of the failing file when there is a problem reading it.
5507 42. Allow both -bf and -bF in the same test run.
5509 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5511 44. The "Exiscan patch" is now merged into the mainline Exim source.
5513 45. Sometimes the final signoff response after QUIT could fail to get
5514 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5515 < 0 before doing a fflush(). This bug looks as though it goes back to the
5516 introduction of TLS in release 3.20, but "sometimes" must have been rare
5517 because the tests only now provoked it.
5519 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5520 (this can affect the format of dates).
5522 47. exim_tidydb, when checking for the continued existence of a message for
5523 which it has found a message-specific retry record, was not finding
5524 messages that were in split spool directories. Consequently, it was
5525 deleting retry records that should have stayed in existence.
5527 48. Steve fixed some bugs in eximstats.
5529 49. The SPA authentication driver was not abandoning authentication and moving
5530 on to the next authenticator when an expansion was forced to fail,
5531 contradicting the general specification for all authenticators. Instead it
5532 was generating a temporary error. It now behaves as specified.
5534 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5535 (the order specifies the preference for clients). The order is now AES256,
5536 AES128, 3DES, ARCFOUR128.
5538 51. Small patch to Sieve code - explicitly set From: when generating an
5541 52. Exim crashed if a remote delivery caused a very long error message to be
5542 recorded - for instance if somebody sent an entire SpamAssassin report back
5543 as a large number of 550 error lines. This bug was coincidentally fixed by
5544 increasing the size of one of Exim's internal buffers (big_buffer) that
5545 happened as part of the Exiscan merge. However, to be on the safe side, I
5546 have made the code more robust (and fixed the comments that describe what
5549 53. Now that there can be additional text after "Completed" in log lines (if
5550 the queue_time_overall log selector is set), a one-byte patch to exigrep
5551 was needed to allow it to recognize "Completed" as not the last thing in
5554 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5555 patch that reportedly fixes this has been added. I am not expert enough to
5556 create a test for it. This is what the patch creator wrote:
5558 "I found a little strange behaviour of ldap code when working with
5559 Windows 2003 AD Domain, where users was placed in more than one
5560 Organization Units. When I tried to give exim partial DN, the exit code
5561 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5562 But simultaneously result of request was absolutely normal ldap result,
5563 so I produce this patch..."
5565 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5566 REFERENCE, so I have modified the code to exclude the patch when that macro
5569 55. Some experimental protocols are using DNS PTR records for new purposes. The
5570 keys for these records are domain names, not reversed IP addresses. The
5571 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5572 leaves it alone. Component reversal etc. now happens only for IP addresses.
5575 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5577 57. Double the size of the debug message buffer (to 2048) so that more of very
5578 long debug lines gets shown.
5580 58. The exicyclog utility now does better if the number of log files to keep
5581 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5583 59. Two changes related to the smtp_active_hostname option:
5585 (1) $smtp_active_hostname is now available as a variable.
5586 (2) The default for smtp_banner uses $smtp_active_hostname instead
5587 of $primary_hostname.
5589 60. The host_aton() function is supposed to be passed a string that is known
5590 to be a valid IP address. However, in the case of IPv6 addresses, it was
5591 not checking this. This is a hostage to fortune. Exim now panics and dies
5592 if the condition is not met. A case was found where this could be provoked
5593 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5594 components; fortuitously, this particular loophole had already been fixed
5595 by change 4.50/55 above.
5597 If there are any other similar loopholes, the new check in host_aton()
5598 itself should stop them being exploited. The report I received stated that
5599 data on the command line could provoke the exploit when Exim was running as
5600 exim, but did not say which command line option was involved. All I could
5601 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5602 running as the user.
5605 61. There was a buffer overflow vulnerability in the SPA authentication code
5606 (which came originally from the Samba project). I have added a test to the
5607 spa_base64_to_bits() function which I hope fixes it.
5610 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5611 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5613 63. The daemon start-up calls getloadavg() while still root for those OS that
5614 need the first call to be done as root, but it missed one case: when
5615 deliver_queue_load_max is set with deliver_drop_privilege. This is
5616 necessary for the benefit of the queue runner, because there is no re-exec
5617 when deliver_drop_privilege is set.
5619 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5620 This has been fixed.
5622 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5623 was in use, was not putting the data itself into the right store pool;
5624 consequently, it could be overwritten for a subsequent message in the same
5625 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5628 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5630 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5631 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5632 the function were treating the return as a boolean value, which happened to
5633 work because 0=false and not-0=true, but is not correct code.
5635 68. The host_aton() function was not handling scoped IPv6 addresses (those
5636 with, for example, "%eth0" on the end) correctly.
5638 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5639 negated items (that is, ~something) in unsigned ints. Some compilers
5640 apparently mutter when there is no cast.
5642 70. If an address verification called from an ACL failed, and did not produce a
5643 user-specific message (i.e. there was only a "system" message), nothing was
5644 put in $acl_verify_message. In this situation, it now puts the system
5647 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5648 session; change 4.31/43 added the unwanted input to the log line - except
5649 that it did not do this in the start of session case. It now does.
5651 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5652 This is wrong and can cause the other end to generate a synchronization
5653 error if it is another Exim or anything else that does the synchronization
5654 check. A QUIT command is no longer sent after a timeout.
5656 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5657 during host lookups.
5659 74. The defer_ok option of callout verification was not working if it was used
5660 when verifying addresses in header lines, that is, for this case:
5662 verify = header_sender/callout=defer_ok
5664 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5665 those file descriptors could be used for SMTP connections. If anything
5666 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5667 could be sent to the SMTP client, causing chaos. The daemon now opens
5668 stdin, stdout, and stderr to /dev/null when it puts itself into the
5671 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5672 log by default. The user can override this with suitable Perl magic.
5674 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5675 the log message when discard triggers, was not working for the DATA ACL or
5676 for the non-SMTP ACL.
5678 78. Error message wording change in sieve.c.
5680 79. If smtp_accept_max_per_host was set, the number of connections could be
5681 restricted to fewer than expected, because the daemon was trying to set up
5682 a new connection before checking whether the processes handling previous
5683 connections had finished. The check for completed processes is now done
5684 earlier. On busy systems, this bug wouldn't be noticed because something
5685 else would have woken the daemon, and it would have reaped the completed
5688 80. If a message was submitted locally by a user whose login name contained one
5689 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5690 It caused a spool format error. I have fixed the spool reading code. A
5691 related problem was that the "from" clause in the Received: line became
5692 illegal because of the space(s). It is now covered by ${quote_local_part.
5694 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5697 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5699 83. Patch from Sieve maintainer to fix unterminated string problem in
5700 "vacation" handling.
5702 84. Some minor changes to the Linux configuration files to help with other
5703 OS variants using glibc.
5705 85. One more patch for Sieve to update vacation handling to latest spec.
5708 ----------------------------------------------------
5709 See the note above about the 4.44 and 4.50 releases.
5710 ----------------------------------------------------
5716 1. Change 4.43/35 introduced a bug that caused file counts to be
5717 incorrectly computed when quota_filecount was set in an appendfile
5720 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5721 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5724 3. Give more explanation in the error message when the command for a transport
5725 filter fails to execute.
5727 4. There are several places where Exim runs a non-Exim command in a
5728 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5729 was being done only for the command run by the queryprogram router. It is
5730 now done for all such subprocesses. The other cases are: ${run, transport
5731 filters, and the commands run by the lmtp and pipe transports.
5733 5. Some older OS have a limit of 256 on the maximum number of file
5734 descriptors. Exim was using setrlimit() to set 1000 as a large value
5735 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5736 systems. I've change it so that if it can't get 1000, it tries for 256.
5738 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5739 was an oversight, and furthermore, ever since the addition of extra
5740 controls (e.g. 4.43/32), the checks on when to allow different forms of
5741 "control" were broken. There should now be diagnostics for all cases when a
5742 control that does not make sense is encountered.
5744 7. $recipients is now available in the predata ACL (oversight).
5746 8. Tidy the search cache before the fork to do a delivery from a message
5747 received from the command line. Otherwise the child will trigger a lookup
5748 failure and thereby defer the delivery if it tries to use (for example) a
5749 cached ldap connection that the parent has called unbind on.
5751 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5752 of $address_data from the recipient verification was clobbered by the
5753 sender verification.
5755 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5756 was its contents. (It was OK if the option was not defined at all.)
5758 11. A "Completed" log line is now written for messages that are removed from
5759 the spool by the -Mrm option.
5761 12. $host_address is now set to the target address during the checking of
5762 ignore_target_hosts.
5764 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5765 being passed; this would have caused $sender_host_name to have been used if
5766 matching the list had actually called for a host name (not very likely,
5767 since this list is usually IP addresses). A host name is now passed as
5770 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5771 code that specifies a non-restarting handler (typically sigaction() in
5772 modern systems) in an attempt to fix a rare and obscure crash bug.
5774 15. Narrowed the window for a race in the daemon that could cause it to ignore
5775 SIGCHLD signals. This is not a major problem, because they are used only to
5776 wake it up if nothing else does.
5778 16. A malformed maildirsize file could cause Exim to calculate negative values
5779 for the mailbox size or file count. Odd effects could occur as a result.
5780 The maildirsize information is now recalculated if the size or filecount
5783 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5784 support for a long time. Removed HAVE_SYS_VFS_H.
5786 18. Updated exipick to current release from John Jetmore.
5788 19. Allow an empty sender to be matched against a lookup in an address list.
5789 Previously the only cases considered were a regular expression, or an
5792 20. Exim went into a mad DNS lookup loop when doing a callout where the
5793 host was specified on the transport, if the DNS lookup yielded more than
5796 21. The RFC2047 encoding function was originally intended for short strings
5797 such as real names; it was not keeping to the 75-character limit for
5798 encoded words that the RFC imposes. It now respects the limit, and
5799 generates multiple encoded words if necessary. To be on the safe side, I
5800 have increased the buffer size for the ${rfc2047: expansion operator from
5803 22. Failure to deliver a bounce message always caused it to be frozen, even if
5804 there was an errors_to setting on the router. The errors_to setting is now
5807 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5808 canonical form (fully expanded) before being placed in
5809 $sender_host_address.
5811 24. Updated eximstats to version 1.33
5813 25. Include certificate and key file names in error message when GnuTLS fails
5814 to set them up, because the GnuTLS error message doesn't include the name
5815 of the failing file when there is a problem reading it.
5817 26. Expand error message when OpenSSL has problems setting up cert/key files.
5820 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5821 (this can affect the format of dates).
5823 28. exim_tidydb, when checking for the continued existence of a message for
5824 which it has found a message-specific retry record, was not finding
5825 messages that were in split spool directories. Consequently, it was
5826 deleting retry records that should have stayed in existence.
5828 29. eximstats updated to version 1.35
5829 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5830 1.35 - bugfix such that pie charts by volume are generated correctly
5832 30. The SPA authentication driver was not abandoning authentication and moving
5833 on to the next authenticator when an expansion was forced to fail,
5834 contradicting the general specification for all authenticators. Instead it
5835 was generating a temporary error. It now behaves as specified.
5837 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5838 (the order specifies the preference for clients). The order is now AES256,
5839 AES128, 3DES, ARCFOUR128.
5841 31. Small patch to Sieve code - explicitly set From: when generating an
5844 32. Exim crashed if a remote delivery caused a very long error message to be
5845 recorded - for instance if somebody sent an entire SpamAssassin report back
5846 as a large number of 550 error lines. This bug was coincidentally fixed by
5847 increasing the size of one of Exim's internal buffers (big_buffer) that
5848 happened as part of the Exiscan merge. However, to be on the safe side, I
5849 have made the code more robust (and fixed the comments that describe what
5852 33. Some experimental protocols are using DNS PTR records for new purposes. The
5853 keys for these records are domain names, not reversed IP addresses. The
5854 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5855 leaves it alone. Component reversal etc. now happens only for IP addresses.
5858 34. The host_aton() function is supposed to be passed a string that is known
5859 to be a valid IP address. However, in the case of IPv6 addresses, it was
5860 not checking this. This is a hostage to fortune. Exim now panics and dies
5861 if the condition is not met. A case was found where this could be provoked
5862 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5863 components; fortuitously, this particular loophole had already been fixed
5864 by change 4.50/55 or 4.44/33 above.
5866 If there are any other similar loopholes, the new check in host_aton()
5867 itself should stop them being exploited. The report I received stated that
5868 data on the command line could provoke the exploit when Exim was running as
5869 exim, but did not say which command line option was involved. All I could
5870 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5871 running as the user.
5874 35. There was a buffer overflow vulnerability in the SPA authentication code
5875 (which came originally from the Samba project). I have added a test to the
5876 spa_base64_to_bits() function which I hope fixes it.
5879 36. The daemon start-up calls getloadavg() while still root for those OS that
5880 need the first call to be done as root, but it missed one case: when
5881 deliver_queue_load_max is set with deliver_drop_privilege. This is
5882 necessary for the benefit of the queue runner, because there is no re-exec
5883 when deliver_drop_privilege is set.
5885 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5886 was in use, was not putting the data itself into the right store pool;
5887 consequently, it could be overwritten for a subsequent message in the same
5888 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5891 38. Sometimes the final signoff response after QUIT could fail to get
5892 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5893 < 0 before doing a fflush(). This bug looks as though it goes back to the
5894 introduction of TLS in release 3.20, but "sometimes" must have been rare
5895 because the tests only now provoked it.
5901 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5902 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5903 Now it returns an integer. A number of calls were still expecting a T/F
5904 return. Fortuitously, in all cases, the tests worked in OK situations,
5905 which is the norm. However, things would have gone wrong on any write
5906 failures on the smtp file descriptor. This function is used when sending
5907 messages over SMTP and also when doing verify callouts.
5909 2. When Exim is called to do synchronous delivery of a locally submitted
5910 message (the -odf or -odi options), it no longer closes stderr before doing
5913 3. Implemented the mua_wrapper option.
5915 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5917 5. Implemented the functions header_remove(), header_testname(),
5918 header_add_at_position(), and receive_remove_recipient(), and exported them
5921 6. If an ACL "warn" statement specified the addition of headers, Exim already
5922 inserted X-ACL-Warn: at the start if there was no header name. However, it
5923 was not making this test for the second and subsequent header lines if
5924 there were newlines in the string. This meant that an invalid header could
5925 be inserted if Exim was badly configured.
5927 7. Allow an ACL "warn" statement to add header lines at the start or after all
5928 the Received: headers, as well as at the end.
5930 8. Added the rcpt_4xx retry error code.
5932 9. Added postmaster_mailfrom=xxx to callout verification option.
5934 10. Added mailfrom=xxxx to the callout verification option, for verify=
5937 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5938 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5940 12. Inserted some casts to stop certain compilers warning when using pointer
5941 differences as field lengths or precisions in printf-type calls (mostly
5942 affecting debugging statements).
5944 13. Added optional readline() support for -be (dynamically loaded).
5946 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5947 same clock tick as a message's arrival, so that its received time was the
5948 same as the "first fail" time on the retry record, and that message
5949 remained on the queue past the ultimate address timeout, every queue runner
5950 would try a delivery (because it was past the ultimate address timeout) but
5951 after another failure, the ultimate address timeout, which should have then
5952 bounced the address, did not kick in. This was a "< instead of <=" error;
5953 in most cases the first failure would have been in the next clock tick
5954 after the received time, and all would be well.
5956 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5957 being recognized when the domain list was tested by the match_domain
5958 condition in an expansion string.
5960 16. Added the ${str2b64: operator.
5962 17. Exim was always calling setrlimit() to set a large limit for the number of
5963 processes, without checking whether the existing limit was already
5964 adequate. (It did check for the limit on file descriptors.) Furthermore,
5965 errors from getrlimit() and setrlimit() were being ignored. Now they are
5966 logged to the main and panic logs, but Exim does carry on, to try to do its
5967 job under whatever limits there are.
5969 18. Imported PCRE 5.0.
5971 19. Trivial typo in log message " temporarily refused connection" (the leading
5974 20. If the log selector return_path_on_delivery was set and an address was
5975 redirected to /dev/null, the delivery process crashed because it assumed
5976 that a return path would always be set for a "successful" delivery. In this
5977 case, the whole delivery is bypassed as an optimization, and therefore no
5980 21. Internal re-arrangement: the function for sending a challenge and reading
5981 a response while authentication was assuming a zero-terminated challenge
5982 string. It's now changed to take a pointer and a length, to allow for
5983 binary data in such strings.
5985 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5987 23. Exim was not respecting finduser_retries when seeking the login of the
5988 uid under which it was called; it was always trying 10 times. (The default
5989 setting of finduser_retries is zero.) Also, it was sleeping after the final
5990 failure, which is pointless.
5992 24. Implemented tls_on_connect_ports.
5994 25. Implemented acl_smtp_predata.
5996 26. If the domain in control=submission is set empty, Exim assumes that the
5997 authenticated id is a complete email address when it generates From: or
5998 Sender: header lines.
6000 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6001 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6002 chown and chgrp in /bin and hostname in /usr/bin.
6004 28. Exim was keeping the "process log" file open after each use, just as it
6005 does for the main log. This opens the possibility of it remaining open for
6006 long periods when the USR1 signal hits a daemon. Occasional processlog
6007 errors were reported, that could have been caused by this. Anyway, it seems
6008 much more sensible not to leave this file open at all, so that is what now
6011 29. The long-running daemon process does not normally write to the log once it
6012 has entered its main loop, and it closes the log before doing so. This is
6013 so that log files can straightforwardly be renamed and moved. However,
6014 there are a couple of unusual error situations where the daemon does write
6015 log entries, and I had neglected to close the log afterwards.
6017 30. The text of an SMTP error response that was received during a remote
6018 delivery was being truncated at 512 bytes. This is too short for some of
6019 the long messages that one sometimes sees. I've increased the limit to
6022 31. It is now possible to make retry rules that apply only when a message has a
6023 specific sender, in particular, an empty sender.
6025 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6026 it possible to be selective about when SMTP synchronization is enforced.
6028 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6030 32. Implemented hosts_connection_nolog.
6032 33. Added an ACL for QUIT.
6034 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6037 35. Added mailbox_size and mailbox_filecount to appendfile.
6039 36. Added control = no_multiline_responses to ACLs.
6041 37. There was a bug in the logic of the code that waits for the clock to tick
6042 in the case where the clock went backwards by a substantial amount such
6043 that the microsecond fraction of "now" was more than the microsecond
6044 fraction of "then" (but the whole seconds number was less).
6046 38. Added support for the libradius Radius client library this is found on
6047 FreeBSD (previously only the radiusclient library was supported).
6053 1. When certain lookups returned multiple values in the form name=value, the
6054 quoting of the values was not always being done properly. Specifically:
6055 (a) If the value started with a double quote, but contained no whitespace,
6057 (b) If the value contained whitespace other than a space character (i.e.
6058 tabs or newlines or carriage returns) it was not quoted.
6059 This fix has been applied to the mysql and pgsql lookups by writing a
6060 separate quoting function and calling it from the lookup code. The fix
6061 should probably also be applied to nisplus, ibase and oracle lookups, but
6062 since I cannot test any of those, I have not disturbed their existing code.
6064 2. A hit in the callout cache for a specific address caused a log line with no
6065 reason for rejecting RCPT. Now it says "Previous (cached) callout
6066 verification failure".
6068 3. There was an off-by-one bug in the queryprogram router. An over-long
6069 return line was truncated at 256 instead of 255 characters, thereby
6070 overflowing its buffer with the terminating zero. As well as fixing this, I
6071 have increased the buffer size to 1024 (and made a note to document this).
6073 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6074 when Exim is waiting for an SMTP response from a remote server, Exim
6075 restarts its select() call on the socket, thereby resetting its timeout.
6076 This is not a problem when such interrupts are rare. Somebody set up a cron
6077 job to run exiwhat every 2 minutes, which is less than the normal select()
6078 timeout (5 or 10 minutes). This meant that the select() timeout never
6079 kicked in because it was always reset. I have fixed this by comparing the
6080 time when an interrupt arrives with the time at the start of the first call
6081 to select(). If more time than the timeout has elapsed, the interrupt is
6082 treated as a timeout.
6084 5. Some internal re-factoring in preparation for the addition of Sieve
6085 extensions (by MH). In particular, the "personal" test is moved to a
6086 separate function, and given an option for scanning Cc: and Bcc: (which is
6087 not set for Exim filters).
6089 6. When Exim created an email address using the login of the caller as the
6090 local part (e.g. when creating a From: or Sender: header line), it was not
6091 quoting the local part when it contained special characters such as @.
6093 7. Installed new OpenBSD configuration files.
6095 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6096 try to make them clearer.
6098 9. Callout options, other than the timeout value, were being ignored when
6099 verifying sender addresses in header lines. For example, when using
6101 verify = header_sender/callout=no_cache
6103 the cache was (incorrectly) being used.
6105 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6106 only the Cygwin environment.
6108 11. When return_path_on_delivery was set as a log selector, if different remote
6109 addresses in the same message used different return paths and parallel
6110 remote delivery occurred, the wrong values would sometimes be logged.
6111 (Whenever a remote delivery process finished, the return path value from
6112 the most recently started remote delivery process was logged.)
6114 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6115 lines when AUTH and/or TLS are in use. This is the "received protocol"
6116 field. Exim used to use "asmtp" for authenticated SMTP, without any
6117 indication (in the protocol name) for TLS use. Now it follows the RFC and
6118 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6119 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6120 names appear in log lines as well as in Received: header lines.
6122 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6123 extensions, and comparison tests, and to fix some bugs.
6125 14. Changes to the "personal" filter test:
6127 (1) The test was buggy in that it was just doing the equivalent of
6128 "contains" tests on header lines. For example, if a user's address was
6129 anne@some.where, the "personal" test would incorrectly be true for
6131 To: susanne@some.where
6133 This test is now done by extracting each address from the header in turn,
6134 and checking the entire address. Other tests that are part of "personal"
6135 are now done using regular expressions (for example, to check local parts
6136 of addresses in From: header lines).
6138 (2) The list of non-personal local parts in From: addresses has been
6139 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6140 taken from the Sieve specification recommendations.
6142 (3) If the message contains any header line starting with "List-" it is
6143 treated as non-personal.
6145 (4) The test for "circular" in the Subject: header line has been removed
6146 because it now seems ill-conceived.
6148 15. Minor typos in src/EDITME comments corrected.
6150 16. Installed latest exipick from John Jetmore.
6152 17. If headers_add on a router specified a text string that was too long for
6153 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6154 of string_sprintf() is now avoided.
6156 18. $message_body_size was not set (it was always zero) when running the DATA
6157 ACL and the local_scan() function.
6159 19. For the "mail" command in an Exim filter, no default was being set for
6160 the once_repeat time, causing a random time value to be used if "once" was
6161 specified. (If the value happened to be <= 0, no repeat happened.) The
6162 default is now 0s, meaning "never repeat". The "vacation" command was OK
6163 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6164 (I found it when inspecting the code).
6166 20. There is now an overall timeout for performing a callout verification. It
6167 defaults to 4 times the callout timeout, which applies to individual SMTP
6168 commands during the callout. The overall timeout applies when there is more
6169 than one host that can be tried. The timeout is checked before trying the
6170 next host. This prevents very long delays if there are a large number of
6171 hosts and all are timing out (e.g. when the network connections are timing
6172 out). The value of the overall timeout can be changed by specifying an
6173 additional sub-option for "callout", called "maxwait". For example:
6175 verify = sender/callout=5s,maxwait=20s
6177 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6178 to the end before writing, but this should make it even safer).
6180 22. Exim was forgetting that it had advertised PIPELINING for the second and
6181 subsequent messages on an SMTP connection. It was also not resetting its
6182 memory on STARTTLS and an internal HELO.
6184 23. When Exim logs an SMTP synchronization error within a session, it now
6185 records whether PIPELINING has been advertised or not.
6187 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6188 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6189 rather than long int.
6191 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6193 26. Added the never_mail option to autoreply.
6199 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6200 crash if the getsockname() call failed; this can happen if a connection is
6201 closed very soon after it is established. The problem was simply in the
6202 order in which certain operations were done, causing Exim to try to write
6203 to the SMTP stream before it had set up the file descriptor. The bug has
6204 been fixed by making things happen in the correct order.
6210 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6211 before the connection was closed, thus losing the rejection response.
6213 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6214 some early Solaris releases, but causes trouble in current releases where
6215 socklen_t is defined.
6217 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6220 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6223 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6224 that releasing the top of it at the end releases what was used for sub-
6225 expansions (unless the block got too big). However, discard this block if
6226 the first thing is a variable or header, so that we can use its block when
6227 it is dynamic (useful for very large $message_headers, for example).
6229 6. Lookups now cache *every* query, not just the most recent. A new, separate
6230 store pool is used for this. It can be recovered when all lookup caches are
6231 flushed. Lookups now release memory at the end of their result strings.
6232 This has involved some general refactoring of the lookup sources.
6234 7. Some code has been added to the store_xxx() functions to reduce the amount
6235 of flapping under certain conditions.
6237 8. log_incoming_interface used to affect only the <= reception log lines. Now
6238 it causes the local interface and port to be added to several more SMTP log
6239 lines, for example "SMTP connection from", and rejection lines.
6241 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6243 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6245 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6246 could be overwritten at the end of the current message (or the start of a
6247 new message if it was set in a HELO ACL). The value is now preserved for
6248 the duration of the SMTP connection.
6250 12. If a transport had a headers_rewrite setting, and a matching header line
6251 contained an unqualified address, that address was qualified, even if it
6252 did not match any rewriting rules. The underlying bug was that the values
6253 of the flags that permit the existence of unqualified sender and recipient
6254 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6255 non-local messages, and by -bnq for local messages) were not being
6256 preserved with the message after it was received.
6258 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6259 "next input=" as part of the text comprising the host identity instead of
6260 the correct text. The code was using the same buffer for two different
6261 strings. However, depending on which order the printing function evaluated
6262 its arguments, the bug did not always show up. Under Linux, for example, my
6263 test suite worked just fine.
6265 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6266 doesn't work with some older versions of Perl. It has been changed to "my",
6267 which in any case is probably the better facility to use.
6269 15. A really picky compiler found some instances of statements for creating
6270 error messages that either had too many or two few arguments for the format
6273 16. The size of the buffer for calls to the DNS resolver has been increased
6274 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6275 for addresses that have a lot of PTR records. This alleviates a problem; it
6276 does not fully solve it.
6278 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6279 buffer now truncates the list and logs the incident, which is the same
6280 action as happens when Exim is looking up a host name and its aliases.
6281 Previously in this situation something unpredictable would happen;
6282 sometimes it was "internal error: store_reset failed".
6284 18. If a server dropped the connection unexpectedly when an Exim client was
6285 using GnuTLS and trying to read a response, the client delivery process
6286 crashed while trying to generate an error log message.
6288 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6289 string, for example:
6291 warn message = H1: something\nH2: something
6293 the text was added as a single header line from Exim's point of view
6294 though it ended up OK in the delivered message. However, searching for the
6295 second and subsequent header lines using $h_h2: did not work. This has been
6296 fixed. Similarly, if a system filter added multiple headers in this way,
6297 the routers could not see them.
6299 20. Expanded the error message when iplsearch is called with an invalid key to
6300 suggest using net-iplsearch in a host list.
6302 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6303 ACLs are no longer actually imposed (and a message to that effect is
6306 22. If a "gecos" field in a passwd entry contained escaped characters, in
6307 particular, if it contained a \" sequence, Exim got it wrong when building
6308 a From: or a Sender: header from that name. A second bug also caused
6309 incorrect handling when an unquoted " was present following a character
6310 that needed quoting.
6312 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6313 was not being matched caselessly.
6315 24. Arranged for all hyphens in the exim.8 source to be escaped with
6318 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6319 properly. Recipient callout cache records were still being keyed to include
6320 the sender, even when use_sender was set false. This led to far more
6321 callouts that were necessary. The sender is no longer included in the key
6322 when use_sender is false.
6324 26. Added "control = submission" modifier to ACLs.
6326 27. Added the ${base62d: operator to decode base 62 numbers.
6328 28. dnsdb lookups can now access SRV records.
6330 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6331 the configuration file.
6333 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6334 (-v) mode. This makes the output for a verbose queue run more intelligible.
6336 31. Added a use_postmaster feature to recipient callouts.
6338 32. Added the $body_zerocount variable, containing the number of binary zero
6339 bytes in the message body.
6341 33. The time of last modification of the "new" subdirectory is now used as the
6342 "mailbox time last read" when there is a quota error for a maildir
6345 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6347 35. Added +ignore_unknown as a special item in host lists.
6349 36. Code for decoding IPv6 addresses in host lists is now included, even if
6350 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6351 address was recognized as an IP address, but was then not correctly decoded
6352 into binary, causing unexpected and incorrect effects when compared with
6359 1. Very minor rewording of debugging text in manualroute to say "list of
6360 hosts" instead of "hostlist".
6362 2. If verify=header_syntax was set, and a header line with an unqualified
6363 address (no domain) and a large number of spaces between the end of the
6364 name and the colon was received, the reception process suffered a buffer
6365 overflow, and (when I tested it) crashed. This was caused by some obsolete
6366 code that should have been removed. The fix is to remove it!
6368 3. When running in the test harness, delay a bit after writing a bounce
6369 message to get a bit more predictability in the log output.
6371 4. Added a call to search_tidyup() just before forking a reception process. In
6372 theory, someone could use a lookup in the expansion of smtp_accept_max_
6373 per_host which, without the tidyup, could leave open a database connection.
6375 5. Added the variables $recipient_data and $sender_data which get set from a
6376 lookup success in an ACL "recipients" or "senders" condition, or a router
6377 "senders" option, similar to $domain_data and $local_part_data.
6379 6. Moved the writing of debug_print from before to after the "senders" test
6382 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6383 problems for message scanning, either using a data ACL, or using
6384 local_scan() because the Received: header was not generated till after they
6385 were called (in order to set the time as the time of reception completion).
6386 I have revised the way this works. The header is now generated after the
6387 body is received, but before the ACL or local_scan() are called. After they
6388 are run, the timestamp in the header is updated.
6394 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6395 before starting a queue runner without re-exec. This happened only when
6396 deliver_drop_privilege was set or when the Exim user was set to root. The
6397 effect of the bug was that timeouts during subsequent deliveries caused
6398 crashes instead of being properly handled. The handler is now left at its
6399 default (and expected) setting.
6401 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6402 message, again when deliver_drop_privilege is set or Exim is run as root.
6403 The bug described in (1) was not present in this case, but the tidying up
6404 of the other signals was missing. I have made the two cases consistent.
6406 3. The ignore_target_hosts setting on a manualroute router was being ignored
6407 for hosts that were looked up using the /MX notation.
6409 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6412 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6413 operated on the sender address. After changing the $sender_address to <>
6414 for the sender address verify, Exim was re-instated it as the original
6415 (before rewriting) address, but remembering that it had rewritten it, so it
6416 wasn't rewriting it again. This bug also had the effect of breaking the
6417 sender address verification caching when the sender address was rewritten.
6419 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6420 This has been changed so that if the ip literal address matches
6421 ignore_target_hosts, the router declines.
6423 7. Added expansion conditions match_domain, match_address, and match_local_
6424 part (NOT match_host).
6426 8. The placeholder for the Received: header didn't have a length field set.
6428 9. Added code to Exim itself and to exim_lock to test for a specific race
6429 condition that could lead to file corruption when using MBX delivery. The
6430 issue is with the lockfile that is created in /tmp. If this file is removed
6431 after a process has opened it but before that process has acquired a lock,
6432 there is the potential for a second process to recreate the file and also
6433 acquire a lock. This could lead to two Exim processes writing to the file
6434 at the same time. The added code performs the same test as UW imapd; it
6435 checks after acquiring the lock that its file descriptor still refers to
6436 the same named file.
6438 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6439 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6442 11. Added the smtp_active_hostname option. If used, this will typically be made
6443 to depend on the incoming interface address. Because $interface_address is
6444 not set up until the daemon has forked a reception process, error responses
6445 that can happen earlier (such as "too many connections") no longer contain
6448 12. If an expansion in a condition on a "warn" statement fails because a lookup
6449 defers, the "warn" statement is abandoned, and the next ACL statement is
6450 processed. Previously this caused the whole ACL to be aborted.
6452 13. Added the iplsearch lookup type.
6454 14. Added ident_timeout as a log selector.
6456 15. Added tls_certificate_verified as a log selector.
6458 16. Added a global option tls_require_ciphers (compare the smtp transport
6459 option of the same name). This controls incoming TLS connections.
6461 17. I finally figured out how to make tls_require_ciphers do a similar thing
6462 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6463 before starting the TLS session.
6465 18. Tabs are now shown as \t in -bP output.
6467 19. If the log selector return_path_on_delivery was set, Exim crashed when
6468 bouncing a message because it had too many Received: header lines.
6470 20. If two routers both had headers_remove settings, and the first one included
6471 a superfluous trailing colon, the final name in the first list and the
6472 first name in the second list were incorrectly joined into one item (with a
6473 colon in the middle).
6479 1. Added -C and -D options to the exinext utility, mainly to make it easier
6480 to include in the automated testing, but these could be helpful when
6481 multiple configurations are in use.
6483 2. The exinext utility was not formatting the output nicely when there was
6484 an alternate port involved in the retry record key, nor when there was a
6485 message id as well (for retries that were specific to a specific message
6486 and a specific host). It was also confused by IPv6 addresses, because of
6487 the additional colons they contain. I have fixed the IPv4 problem, and
6488 patched it up to do a reasonable job for IPv6.
6490 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6491 delivery, the log line now contains "pipelined" if PIPELINING was used.
6493 4. An SMTP transport process used to panic and die if the bind() call to set
6494 an explicit outgoing interface failed. This has been changed; it is now
6495 treated in the same way as a connect() failure.
6497 5. A reference to $sender_host_name in the part of a conditional expansion
6498 that was being skipped was still causing a DNS lookup. This no longer
6501 6. The def: expansion condition was not recognizing references to header lines
6502 that used bh_ and bheader_.
6504 7. Added the _cache feature to named lists.
6506 8. The code for checking quota_filecount in the appendfile transport was
6507 allowing one more file than it should have been.
6509 9. For compatibility with Sendmail, the command line option
6517 and sets the incoming protocol and host name (for trusted callers). The
6518 host name and its colon can be omitted when only the protocol is to be set.
6519 Note the Exim already has two private options, -pd and -ps, that refer to
6520 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6521 "s", but I don't think that's a major issue.
6523 10. A number of refactoring changes to the code, none of which should affect
6526 (a) The number of logging options was getting close to filling up the
6527 32-bit word that was used as a bit map. I have split them into two classes:
6528 those that are passed in the argument to log_write(), and those that are
6529 only ever tested independently outside of that function. These are now in
6530 separate 32-bit words, so there is plenty of room for expansion again.
6531 There is no change in the user interface or the logging behaviour.
6533 (b) When building, for example, log lines, the code previously used a
6534 macro that called string_cat() twice, in order to add two strings. This is
6535 not really sufficiently general. Furthermore, there was one instance where
6536 it was actually wrong because one of the argument was used twice, and in
6537 one call a function was used. (As it happened, calling the function twice
6538 did not affect the overall behaviour.) The macro has been replaced by a
6539 function that can join an arbitrary number of extra strings onto a growing
6542 (c) The code for expansion conditions now uses a table and a binary chop
6543 instead of a serial search (which was left over from when there were very
6544 few conditions). Also, it now recognizes conditions like "pam" even when
6545 the relevant support is not compiled in: a suitably worded error message is
6546 given if an attempt is made to use such a condition.
6548 11. Added ${time_interval:xxxxx}.
6550 12. A bug was causing one of the ddress fields not to be passed back correctly
6551 from remote delivery subprocesses. The field in question was not being
6552 subsequently used, so this caused to problems in practice.
6554 13. Added new log selectors queue_time and deliver_time.
6556 14. Might have fixed a bug in maildirsizefile handling that threw up
6557 "unexpected character" debug warnings, and recalculated the data
6558 unnecessarily. In any case, I expanded the warning message to give more
6561 15. Added the message "Restricted characters in address" to the statements in
6562 the default ACL that block characters like @ and % in local parts.
6564 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6565 Three changes have been made:
6567 (a) There was a serious bug; a negative response to MAIL caused the whole
6568 recipient domain to be cached as invalid, thereby blocking all messages
6569 to all local parts at the same domain, from all senders. This bug has
6570 been fixed. The domain is no longer cached after a negative response to
6571 MAIL if the sender used is not empty.
6573 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6576 (c) A new callout option, "use_sender" has been added for people who want
6577 the modified behaviour.
6583 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6586 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6587 indeed breaks things for older releases.
6589 3. Added additional logging to the case where there is a problem reading data
6590 from a filter that is running in a subprocess using a pipe, in order to
6591 try to track down a specific problem.
6593 4. Testing facility fudge: when running in the test harness and attempting
6594 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6595 getting "No route to host". Convert this to a timeout.
6597 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6600 6. Some OS don't have socklen_t but use size_t instead. This affects the
6601 fifth argument of getsockopt() amongst other things. This is now
6602 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6603 can be set for individual OS. I have set it for SunOS5, OSF1, and
6604 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6605 some earlier ones do not.
6607 7. Change 4.30/15 was not doing the test caselessly.
6609 8. The standard form for an IPv6 address literal was being rejected by address
6610 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6611 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6612 this, as well as the form without the "IPv6" on the front (but only when
6613 address literals are enabled, of course).
6615 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6617 10. Exim crashed if a message with an empty sender address specified by -f
6618 encountered a router with an errors_to setting. This could be provoked only
6619 by a command such as
6623 where an empty string was supplied; "<>" did not hit this bug.
6625 11. Installed PCRE release 4.5.
6627 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6628 remained set. It is now erased.
6630 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6631 times from message ids (which are base 36 rather than the normal 62).
6633 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6634 were being counted as actual protocol errors, and logged if the log
6635 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6636 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6637 and DATA following a set of rejected RCPTs do not count as protocol errors.
6638 In other words, Exim assumes they were pipelined, though this may not
6639 actually be the case. Of course, in all cases the client gets an
6640 appropriate error code.
6642 15. If a lookup fails in an ACL condition, a message about the failure may
6643 be available; it is used if testing the ACL cannot continue, because most
6644 such messages specify what the cause of the deferral is. However, some
6645 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6646 that caused an old message to be retained and used if a later statement
6647 caused a defer, replacing the real cause of the deferral.
6649 16. If an IP address had so many PTR records that the DNS lookup buffer
6650 was not large enough to hold them, Exim could crash while trying to process
6651 the truncated data. It now detects and logs this case.
6653 17. Further to 4.21/58, another change has been made: if (and only if) the
6654 first line of a message (the first header line) ends with CRLF, a bare LF
6655 in a subsequent header line has a space inserted after it, so as not to
6656 terminate the header.
6658 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6659 unnecessarily, used atoi() instead of strtol(), and didn't check the
6660 termination when getting file sizes from file names by regex.
6662 19. Completely re-implemented the support for maildirsize files, in the light
6663 of a number of problems with the previous contributed implementation
6664 (4.30/29). In particular:
6666 . If the quota is zero, the maildirsize file is maintained, but no quota is
6669 . If the maildir directory does not exist, it is created before any attempt
6670 to write a maildirsize file.
6672 . The quota value in the file is just a cache; if the quota is changed in
6673 the transport, the new value overrides.
6675 . A regular expression is available for excluding directories from the
6678 20. The autoreply transport checks the characters in options that define the
6679 message's headers; it allows continued headers, but it was checking with
6680 isspace() after an embedded newline instead of explicitly looking for a
6683 21. If all the "regular" hosts to which an address was routed had passed their
6684 expiry times, and had not reached their retry times, the address was
6685 bounced, even if fallback hosts were defined. Now Exim should go on to try
6688 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6689 equivalent code in the SMTP transport. Some hosts send humungous responses
6690 to HELO/EHLO, more than 1024 it seems.
6692 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6693 gives compiler warnings in some environments. I've now done it "properly",
6696 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6697 (because of problems with the built-in one) was declared to return uschar *
6698 instead of char *, causing compiler failure.
6700 25. Fixed a file descriptor leak when processing alias/forward files.
6702 26. Fixed a minor format string issue in dbfn.c.
6704 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6706 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6707 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6710 29. When a sender address is verified, it is cached, to save repeating the test
6711 when there is more than one recipient in a message. However, when the
6712 verification involves a callout, it is possible for different callout
6713 options to be set for different recipients. It is too complicated to keep
6714 track of this in the cache, so now Exim always runs a verification when a
6715 callout is required, relying on the callout cache for the optimization.
6716 The overhead is duplication of the address routing, but this should not be
6719 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6720 to be verified with callout=postmaster, and the main callout worked but the
6721 postmaster check failed, the verification correctly failed. However, if a
6722 subsequent RCPT command asked for sender verification *without* the
6723 postmaster check, incorrect caching caused this verification also to fail,
6726 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6727 it was not caching the DNS options (qualify_single, search_parents) that
6728 were used when the lookup failed. A subsequent lookup with different
6729 options therefore always gave the same answer, though there were cases
6730 where it should not have. (Example: a "domains = !$mx_any" option on a
6731 dnslookup router: the "domains" option is always processed without any
6732 widening, but the router might have qualify_single set.) Now Exim uses the
6733 cached value only when the same options are set.
6735 32. Added John Jetmore's "exipick" utility to the distribution.
6737 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6738 than a timeout (e.g. a certificate is required, and is not provided), an
6739 Exim server now closes the connection immediately. Previously it waited for
6740 the client to close - but if the client is SSL, it seems that they each
6741 wait for each other, leading to a delay before one of them times out.
6743 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6744 maintained 0.8.x compatibility because I don't think many are using it, and
6745 it is clearly obsolete.
6747 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6750 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6751 client certificate was expired. A simple patch fixes this, though I don't
6752 understand the full logic of why the verify callback is called multiple
6755 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6756 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6757 which causes problems with some clients (such as the Certicom SSL Plus
6758 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6759 disables the coutermeasure allowing Eudora to connect."
6761 38. Exim was not checking that a write() to a log file succeeded. This could
6762 lead to Bad Things if a log got too big, in particular if it hit a file
6763 size limit. Exim now panics and dies if it cannot write to a log file, just
6764 as it does if it cannot open a log file.
6766 39. Modified OS/Makefile-Linux so that it now contains
6768 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6770 The two -D definitions ensure that Exim is compiled with large file
6771 support, which makes it possible to handle log files that are bigger than
6774 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6775 instance) a domain was checked against a named list that involved a lookup,
6776 causing $domain_data to be set, then another domain was checked against the
6777 same list, then the first domain was re-checked, the value of $domain_data
6778 after the final check could be wrong. In particular, if the second check
6779 failed, it could be set empty. This bug probably also applied to
6782 41. The strip_trailing_dot option was not being applied to the address given
6783 with the -f command-line option.
6785 42. The code for reading a message's header from the spool was incrementing
6786 $received_count, but never initializing it. This meant that the value was
6787 incorrect (doubled) while delivering a message in the same process in which
6788 it was received. In the most common configuration of Exim, this never
6789 happens - a fresh exec is done - but it can happen when
6790 deliver_drop_privilege is set.
6792 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6793 it now includes up to 150 characters of the unexpected data in the log
6796 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6797 and building data strings. The size of both of these buffers was 10 000
6798 bytes - far larger than anybody would *ever* want, thought I. Needless to
6799 say, somebody hit the limit. I have increased the maximum line length to
6800 20 000 and the maximum data length of concatenated lines to 100 000. I have
6801 also fixed two bugs, because there was no checking on these buffers. Tsk,
6802 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6803 buffer is too small.
6805 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6806 lsearch lookups. Now it does.
6808 46. When parsing a route_list item in a manualroute router, a fixed-length
6809 buffer was used for the list of hosts. I made this 1024 bytes long,
6810 thinking that nobody would ever have a list of hosts that long. Wrong.
6811 Somebody had a whole pile of complicated expansion conditions, and the
6812 string was silently truncated, leading to an expansion error. It turns out
6813 that it is easier to change to an unlimited length (owing to other changes
6814 that have happened since this code was originally written) than to build
6815 structure for giving a limitation error. The length of the item that
6816 expands into the list of hosts is now unlimited.
6818 47. The lsearch lookup could not handle data where the length of text line was
6819 more than 4095 characters. Such lines were truncated, leading to shortened
6820 data being returned. It should now handle lines of any length.
6822 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6823 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6826 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6827 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6828 no longer happens in comment lines. A list of replacements is now placed
6829 at the head of all of the source files, except those whose only change is
6830 to replace PERL_COMMAND in the very first #! line.
6832 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6833 messages on the queue, with a bottom-up merge sort, using code contributed
6834 by Michael Haardt. This should make operations like -bp somewhat faster on
6835 large queues. It won't affect queue runners, except when queue_run_in_order
6838 51. Installed eximstats 1.31 in the distribution.
6840 52. Added support for SRV lookups to the dnslookup router.
6842 53. If an ACL referred to $message_body or $message_body_end, the value was not
6843 reset for any messages that followed in the same SMTP session.
6845 54. The store-handling optimization for building very long strings was not
6846 differentiating between the different store pools. I don't think this
6847 actually made any difference in practice, but I've tidied it.
6849 55. While running the routers to verify a sender address, $sender_address
6850 was still set to the sender address. This is wrong, because when routing to
6851 send a bounce to the sender, it would be empty. Therefore, I have changed
6852 it so that, while verifying a sender address, $sender_address is set to <>.
6853 (There is no change to what happens when verifying a recipient address.)
6855 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6856 target A or AAAA records (if not already returned) without resetting the
6857 qualify_single or search_parents options of the DNS resolver. These are
6858 inappropriate in this case because the targets of MX and SRV records must
6859 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6860 target that, when qualified, matched something in the local domain. These
6861 two options are now turned off when doing these lookups.
6863 57. It seems that at least some releases of Reiserfs (which does not have the
6864 concept of a fixed number of inodes) returns zero and not -1 for the
6865 number of available inodes. This interacted badly with check_spool_inodes,
6866 which assumed that -1 was the "no such thing" setting. What I have done is
6867 to check that the total number of inodes is greater than zero before doing
6868 the test of how many are available.
6870 58. When a "warn" ACL statement has a log_message modifier, the message is
6871 remembered, and not repeated. This is to avoid a lot of repetition when a
6872 message has many recipients that cause the same warning to be written.
6873 However, Exim was preserving the list of already written lines for an
6874 entire SMTP session, which doesn't seem right. The memory is now reset if a
6875 new message is started.
6877 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6878 debugging output unless log_rewrite was also set.
6880 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6881 of (int)(handle) when we know that handle contains (void *)(-1).
6883 61. The Exim daemon panic-logs an error return when it closes the incoming
6884 connection. However "connection reset by peer" seems to be common, and
6885 isn't really an error worthy of noting specially, so that particular error
6888 62. When Exim is trying to find all the local interfaces, it used to panic and
6889 die if the ioctl to get the interface flags failed. However, it seems that
6890 on at least one OS (Solaris 9) it is possible to have an interface that is
6891 included in the list of interfaces, but for which you get a failure error
6892 for this call. This happens when the interface is not "plumbed" into a
6893 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6894 failure of the "get flags" call assumes that the interface is down.
6896 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6897 makes life easier for people who are doing arithmetic on fields extracted
6898 from dates, where you often get leading zeros that should not be
6899 interpreted as octal.
6901 64. Added qualify_domain to the redirect router, to override the global
6904 65. If a pathologically long header line contained very many addresses (the
6905 report of this problem mentioned 10 000) and each of them was rewritten,
6906 Exim could use up a very large amount of memory. (It kept on making new
6907 copies of the header line as it rewrote, and never released the old ones.)
6908 At the expense of a bit more processing, the header rewriting function has
6909 been changed so that it no longer eats memory in this way.
6911 66. The generation of the Received: header has been moved from the time that a
6912 message starts to be received, to the time that it finishes. The timestamp
6913 in the Received: header should now be very close to that of the <= log
6914 line. There are two side-effects of this change:
6916 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6917 logged header lines no longer include the local Received: line, because
6918 it has not yet been created. The same applies to a copy of the message
6919 that is returned to a non-SMTP sender when a message is rejected.
6921 (b) When a filter file is tested using -bf, no additional Received: header
6922 is added to the test message. After some thought, I decided that this
6925 This change does not affect the value of $received_for. It is still set
6926 after address rewriting, but before local_scan() is called.
6928 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6930 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6931 gave an unhelpful panic error message, and a defer error. I have managed to
6932 change this behaviour so that it now rejects any supplied certificate,
6933 which seems right, as the list of acceptable certificates is empty.
6935 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6936 gave an unhelpful defer error. I have not managed to make this reject any
6937 supplied certificates, but the error message it gives is "no certificate
6938 supplied", which is not helpful.
6940 70. exigrep's output now also includes lines that are not associated with any
6941 message, but which match the given pattern. Implemented by a patch from
6942 Martin Sluka, which also tidied up the Perl a bit.
6944 71. Recipient callout verification, like sender verification, was using <> in
6945 the MAIL FROM command. This isn't really the right thing, since the actual
6946 sender may affect whether the remote host accepts the recipient or not. I
6947 have changed it to use the actual sender in the callout; this means that
6948 the cache record is now keyed on a recipient/sender pair, not just the
6949 recipient address. There doesn't seem to be a real danger of callout loops,
6950 since a callout by the remote host to check the sender would use <>.
6951 [SEE ABOVE: changed after hitting problems.]
6953 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6954 temporary errors. However, in the case of such a code being given after
6955 the end of a data transmission (i.e. after ".") Exim was failing to write
6956 a retry record for the message. (Yes, there was some broken host that was
6957 actually sending 8xx at this point.)
6959 73. An unknown lookup type in a host list could cause Exim to panic-die when
6960 the list was checked. (An example that provoked this was putting <; in the
6961 middle of a list instead of at the start.) If this happened during a DATA
6962 ACL check, a -D file could be left lying around. This kind of configuration
6963 error no longer causes Exim to die; instead it causes a defer error. The
6964 incident is still logged to the main and panic logs.
6966 74. Buglet left over from Exim 3 conversion. The message "too many messages
6967 in one connection" was written to the rejectlog but not the mainlog, except
6968 when address rewriting (yes!) was being logged.
6970 75. Added write_rejectlog option.
6972 76. When a system filter was run not as root (that is, when system_filter_user
6973 was set), the values of the $n variables were not being returned to the
6974 main process; thus, they were not subsequently available in the $sn
6977 77. Added +return_path_on_delivery log selector.
6979 78. A connection timeout was being treated differently from recipients deferred
6980 when testing hosts_max_try with a message that was older than the host's
6981 retry timeout. (The host should not be counted, thus allowing all hosts to
6982 be tried at least once before bouncing.) This may have been the cause of an
6983 occasionally reported bug whereby a message would remain on the queue
6984 longer than the retry timeout, but would be bounced if a delivery was
6985 forced. I say "may" because I never totally pinned down the problem;
6986 setting up timeout/retry tests is difficult. See also the next item.
6988 79. The ultimate address timeout was not being applied to errors that involved
6989 a combination of host plus message (for example, a timeout on a MAIL
6990 command). When an address resolved to a number of possible hosts, and they
6991 were not all tried for each delivery (e.g. because of hosts_max_try), a
6992 message could remain on the queue longer than the retry timeout.
6994 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6997 81. Fixed an obscure SMTP outgoing bug which required at least the following
6998 conditions: (a) there was another message waiting for the same server;
6999 (b) the server returned 5xx to all RCPT commands in the first message so
7000 that the message was not completed; (c) the server dropped the connection
7001 or gave a negative response to the RSET that Exim sends to abort the
7002 transaction. The observed case was a dropped connection after DATA that had
7003 been sent in pipelining mode. That is, the server had advertised PIPELINING
7004 but was not implementing it correctly. The effect of the bug was incorrect
7005 behaviour, such as trying another host, and this could lead to a crash.
7011 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7012 and daemon.c were passed as pointers to ints; they should have been
7013 pointers to socklen_t variables (which are typically unsigned ints).
7015 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7018 3. Fixed a really odd bug that affected only the testing scheme; patching a
7019 certain fixed string in the binary changed the value of another string that
7020 happened to be identical to the end of the original first string.
7022 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7023 name", it returns that address as the IP address. On some operating
7024 systems (e.g. Solaris), it also passes back the IP address string as the
7025 "host name". However, on others (e.g. Linux), it passes back an empty
7026 string. Exim wasn't checking for this, and was changing the host name to an
7027 empty string, assuming it had been canonicalized.
7029 5. Although rare, it is permitted to have more than one PTR record for a given
7030 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7031 all the names associated with an address, because they do in Solaris.
7032 However, it seems that they do not in Linux for data that comes from the
7033 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7034 I found this out when I moved to a new Linux workstation and tried to run
7035 the Exim test suite.
7037 To get round this problem I have changed the code so that it now does its
7038 own call to the DNS to look up PTR records when searching for a host name.
7039 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7040 addresses that are only in /etc/hosts are still found.
7042 This behaviour is, however, controlled by an option called host_lookup_
7043 order, which defaults to "bydns:byaddr". If people want to use the other
7044 order, or indeed, just use one or the other means of lookup, they can
7045 specify it in this variable.
7047 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7048 some operating systems, this comes back from gethostbyaddr() as an empty
7049 string, and this is what Exim used to test for. However, it seems that in
7050 other systems, "." is yielded. Exim now tests for this case too.
7052 7. The values of check_spool_space and check_log_space are now held internally
7053 as a number of kilobytes instead of an absolute number of bytes. If a
7054 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7055 kilobyte. This means that much larger values can be stored.
7057 8. Exim monitor: an attempt to get the action menu when not actually pointing
7058 at a message produces an empty menu entitled "No message selected". This
7059 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7060 no entries in it ("Shell widget menu has zero width and/or height"). So I
7061 have added a single, blank menu entry in this case.
7063 9. Added ${quote_local_part.
7065 10. MIME decoding is now applied to the contents of Subject: header lines when
7068 11. Now that a reference to $sender_host_address automatically causes a reverse
7069 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7070 host lookup before query-style lookups in lists that might use this
7071 variable. This has therefore been abolished, and the "net-" prefix is no
7072 longer necessary for query-style lookups.
7074 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7075 have been a typo for LFLAGS, so it has been changed.
7077 13. The install script calls Exim with "-C /dev/null" in order to find the
7078 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7079 to be output. However, since Exim outputs its version number before the
7080 error, it didn't break the script. It just looked ugly. I fixed this by
7081 always allowing "-C /dev/null" if the caller is root.
7083 14. Ignore overlarge ACL variable number when reading spool file - insurance
7084 against a later release with more variables having written the file.
7086 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7087 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7088 this, as well as the form without the "IPv6" on the front.
7090 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7091 OS/Makefile-Darwin file.
7093 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7094 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7096 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7097 variable. However, some OS use ints for inodes. I've added cast to long int
7098 to get rid of the compiler warning.
7100 19. I had forgotten to lock out "/../" in configuration file names when
7101 ALT_CONFIG_PREFIX was set.
7103 20. Routers used for verification do not need to specify transports. However,
7104 if such a router generated a host list, and callout was configured, Exim
7105 crashed, because it could not find a port number from the (non-existent)
7106 transport. It now assumes port 25 in this circumstance.
7108 21. Added the -t option to exigrep.
7110 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7111 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7112 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7113 them is set, LOOKUP_LSEARCH is forced.
7115 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7117 24. Added sender and host information to the "rejected by local_scan()" log
7118 line; previously there was no indication of these.
7120 25. Added .include_if_exists.
7122 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7123 Linux. It turns out that not all file systems support this. Apparently some
7124 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7125 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7126 not supported on the file descriptor, is now ignored when Exim is trying to
7127 sync a directory. This applies only to Linux.
7129 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7131 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7132 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7135 29. Added support for maildirsize files from supplied patch (modified a bit).
7137 30. The use of :fail: followed by an empty string could lead Exim to respond to
7138 sender verification failures with (e.g.):
7140 550 Verification failed for <xxx>
7141 550 Sender verify failed
7143 where the first response line was missing the '-' that indicates it is not
7144 the final line of the response.
7146 31. The loop for finding the name of the user that called Exim had a hardwired
7147 limit of 10; it now uses the value of finduser_retries, which is used for
7148 all other user lookups.
7150 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7153 33. Exim was neglecting to zero errno before one call of strtol() when
7154 expanding a string and expecting an integer value. On some systems this
7155 resulted in spurious "integer overflow" errors. Also, it was casting the
7156 result into an int without checking.
7158 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7159 did not work. The code looks as if it has *never* worked, though it appears
7160 to have been documented since at least release 1.62. I have made it work.
7162 35. The "timeout_DNS" error in retry rules, also documented since at least
7163 1.62, also never worked. As it isn't clear exactly what this means, and
7164 clearly it isn't a major issue, I have abolished the feature by treating it
7165 as "timeout", and writing a warning to the main and panic logs.
7167 36. The display of retry rules for -brt wasn't always showing the error code
7170 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7171 timeout_connect_A, timeout_connect_MX.
7173 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7174 to the empty sender.
7176 39. The daemon was not analysing the content of -oX till after it had closed
7177 stderr and disconnected from the controlling terminal. This meant that any
7178 syntax errors were only noted on the panic log, and the return code from
7179 the command was 0. By re-arranging the code a little, I've made the
7180 decoding happen first, so such errors now appear on stderr, and the return
7181 code is 1. However, the actual setting up of the sockets still happens in
7182 the disconnected process, so errors there are still only recorded on the
7185 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7186 connections (as happens on some IP stacks) was logged at start up time as
7187 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7188 it from "IPv6 and IPv4", which means that two separate sockets are being
7191 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7192 says whether AF_INET or AF_INET6 was passed as an argument.
7194 42. Exiwhat output was messed up when time zones were included in log
7197 43. Exiwhat now gives more information about the daemon's listening ports,
7198 and whether -tls-on-connect was used.
7200 44. The "port" option of the smtp transport is now expanded.
7202 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7203 silently ignored. Now an error message is written to the main and panic
7206 46. There's a new ACL modifier called "logwrite" which writes to a log file
7207 as soon as it is encountered.
7209 47. Added $local_user_uid and $local_user_gid at routing time.
7211 48. Exim crashed when trying to verify a sender address that was being
7214 49. Exim was recognizing only a space character after ".include". It now also
7215 recognizes a tab character.
7217 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7218 extracting the relevant information from the specification. The man page no
7219 longer contains scrambled data for the -d option, and I've added a section
7220 at the front about calling Exim under different names.
7222 51. Added "extra_headers" argument to the "mail" command in filter files.
7224 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7227 53. Installed eximstats 1.29.
7229 54. Added transport_filter_timeout as a generic transport option.
7231 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7232 Cc: header lines. This was required by RFC 822, but it not required by RFC
7235 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7236 incoming messages that did not have them. Now it does so only if the
7237 message originates locally, that is, if there is no associated remote host
7238 address. When Resent- header lines are present, this applies to the Resent-
7239 lines rather than the non-Resent- lines.
7241 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7242 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7244 58. Messages for configuration errors now include the name of the main
7245 configuration file - useful now that there may be more than one file in a
7246 list (.included file names were always shown).
7248 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7249 for those rare installations that do not start the daemon as root or run it
7250 setuid root. I've cut out the call to initgroups() if the daemon is not
7253 60. The Exim user and group can now be bound into the binary as text strings
7254 that are looked up at the start of Exim's processing.
7256 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7258 62. Added $mailstore_basename variable.
7260 63. Installed patch to sieve.c from Michael Haardt.
7262 64. When Exim failed to open the panic log after failing to open the main log,
7263 the original message it was trying to log was written to stderr and debug
7264 output, but if they were not available (the usual case in production), it
7265 was lost. Now it is written to syslog before the two lines that record the
7266 failures to open the logs.
7268 65. Users' Exim filters run in subprocesses under the user's uid. It is
7269 possible for a "deliver" command or an alias in a "personal" command to
7270 provoke an address rewrite. If logging of address rewriting is configured,
7271 this fails because the process is not running as root or exim. There may be
7272 a better way of dealing with this, but for the moment (because 4.30 needs
7273 to be released), I have disabled address rewrite logging when running a
7274 filter in a non-root, non-exim process.
7280 1. The buildconfig auxiliary program wasn't quoting the value set for
7281 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7282 not defined. This bug was masked in 4.22 by the effect that was fixed in
7285 2. Some messages that were rejected after a message id was allocated were
7286 shown as "incomplete" by exigrep. It no longer does this for messages that
7287 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7289 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7290 have allow_domain_literals set, the ID did not get logged in the <= line.
7291 Domain literals are now always recognized in Message-ID: header lines.
7293 4. The first argument for a ${extract expansion item is the key name or field
7294 number. Leading and trailing spaces in this item were not being ignored,
7295 causing some misleading effects.
7297 5. When deliver_drop_privilege was set, single queue runner processes started
7298 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7299 same command in the process it spins off) were not dropping privilege.
7301 6. When the daemon running as "exim" started a queue runner, it always
7302 re-executed Exim in the spun-off process. This is a waste of effort when
7303 deliver_drop_privilege is set. The new process now just calls the
7304 queue-runner function directly.
7310 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7313 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7314 ignored. Though the use of -oP was forcing the writing of a pid file, it
7315 was always written to the default place.
7317 3. If the message "no IP address found for host xxxx" is generated during
7318 incoming verification, it is now followed by identification of the incoming
7319 connection (so you can more easily find what provoked it).
7321 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7323 5. Added some features to "harden" Exim a bit more against certain attacks:
7325 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7326 be put in Local/Makefile. This is like the never_users runtime option,
7327 but it cannot be overridden. The default setting is "root".
7329 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7330 prefix string with which any file named in a -C command line option
7333 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7334 is retained for -C and -D only if the caller of Exim is root. Without
7335 it, the exim user may also use -C and -D and retain privilege.
7337 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7338 command line option is disabled.
7340 6. Macro names set by the -D option must start with an upper case letter, just
7341 like macro names defined in the configuration file.
7343 7. Added "dereference=" facility to LDAP.
7345 8. Two instances of the typo "uknown" in the source files are fixed.
7347 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7348 the Configure-Makefile script screwed up while processing it.
7350 10. Incorporated PCRE 4.4.
7352 11. The SMTP synchronization check was not operating right at the start of an
7353 SMTP session. For example, it could not catch a HELO sent before the client
7354 waited for the greeting. There is now a check for outstanding input at the
7355 point when the greeting is written. Because of the duplex, asynchronous
7356 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7357 way, but not yet received, when the check is performed.
7359 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7360 on TCP/IP sockets, because this apparently causes some broken clients to
7363 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7364 unchanged) from the Cygwin maintainer.
7366 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7367 was supported instead of testing for mbx. Effectively a typo.
7369 15. The spa authenticator server code was not checking that the input it
7370 received was valid base64.
7372 16. The debug output line for the "set" modifier in ACLs was not showing the
7373 name of the variable that was being set.
7375 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7377 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7378 lookup on its own. Something else was needed to trigger the lookup. For
7379 example, a match in host_lookup or the need for a host name in a host list.
7380 Now, if $sender_host_name is referenced and the host name has not yet been
7381 looked up, a lookup is performed. If the lookup fails, the variable remains
7382 empty, and $host_lookup_failed is set to "1".
7384 19. Added "eqi" as a case-independent comparison operator.
7386 20. The saslauthd authentication condition could segfault if neither service
7387 nor realm was specified.
7389 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7390 error message that was logged was misleading, and incoming SMTP
7391 connections were dropped. The message is now more accurate, and temporary
7392 errors are given to SMTP connections.
7394 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7395 (except RSET) with a 421 error, until QUIT is received. However, it was
7396 failing to send a response to QUIT.
7398 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7399 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7400 ACL is now run after the helo_try_verify_hosts code.
7402 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7403 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7404 case-independent, so other case variants are also recognized). Apparently
7405 some systems use these upper case variants.
7407 25. If more than two messages were waiting for the same host, and a transport
7408 filter was specified for the transport, Exim sent two messages over the
7409 same TCP/IP connection, and then failed with "socket operation on non-
7410 socket" when it tried to send the third.
7412 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7414 27. The extern definition of crypt16() in expand.c was not being excluded when
7415 the OS had its own crypt16() function.
7417 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7418 as a preferred synonym for return_size_limit, both as an option and as an
7421 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7423 30. Changed the default configuration ACL to relax the local part checking rule
7424 for addresses that are not in any local domains. For these addresses,
7425 slashes and pipe symbols are allowed within local parts, but the sequence
7426 /../ is explicitly forbidden.
7428 31. SPA server authentication was not clearing the challenge buffer before
7431 32. log_message in a "warn" ACL statement was writing to the reject log as
7432 well as to the main log, which contradicts the documentation and doesn't
7433 seem right (because no rejection is happening). So I have stopped it.
7435 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7436 However, I am unable to do any testing of this.
7438 34. Fixed an infelicity in the appendfile transport. When checking directories
7439 for a mailbox, to see if any needed to be created, it was accidentally
7440 using path names with one or more superfluous leading slashes; tracing
7441 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7443 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7444 contents are added to the log line that is written for every discarded
7445 recipient. (Previously a log_message setting was ignored.)
7447 36. The ${quote: operator now quotes the string if it is empty.
7449 37. The install script runs exim in order to find its version number. If for
7450 some reason other than non-existence or emptiness, which it checks, it
7451 could not run './exim', it was installing it with an empty version number,
7452 i.e. as "exim-". This error state is now caught, and the installation is
7455 38. An argument was missing from the function that creates an error message
7456 when Exim fails to connect to the socket for saslauthd authentication.
7457 This could cause Exim to crash, or give a corrupted message.
7459 39. Added isip, isip4, and isip6 to ${if conditions.
7461 40. The ACL variables $acl_xx are now saved with the message, and can be
7462 accessed later in routers, transports, and filters.
7464 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7465 strings in the file are not string-expanded.
7467 42. If a MAIL command specified a SIZE value that was too large to fit into an
7468 int variable, the check against message_size_limit failed. Such values are
7469 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7470 day this will have to be increased, but I don't think I want to be around
7471 when emails are that large.
7478 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7479 iconv() is not standard in FreeBSD.
7481 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7482 IPv6 enabled. The observed symptom was a segmentation fault on return from
7483 the function os_common_find_running_interfaces() in src/os.c.
7485 3. In the check_special_case() function in daemon.c I had used "errno" as an
7486 argument name, which causes warnings on some systems. This was basically a
7487 typo, since it was named "eno" in the comments!
7489 4. The code that waits for the clock to tick (at a resolution of some fraction
7490 of a second) so as to ensure message-id uniqueness was always waiting for
7491 at least one whole tick, when it could have waited for less. [This is
7492 almost certainly not relevant at current processor speeds, where it is
7493 unlikely to ever wait at all. But we try to future-proof.]
7495 5. The function that sleeps for a time interval that includes fractions of a
7496 second contained a race. It did not block SIGALRM between setting the
7497 timer, and suspending (a couple of lines later). If the interval was short
7498 and the sigsuspend() was delayed until after it had expired, the suspension
7499 never ended. On busy systems this could lead to processes getting stuck for
7502 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7503 process, before it forks any delivery processes. The open lookup caching
7504 mechanism meant that the open file or database connection was passed into
7505 the delivery process. The problem was that delivery processes always tidy
7506 up cached lookup data. This could cause a problem for the next delivery
7507 process started by the queue runner, because the external queue runner
7508 process does not know about the closure. So the next delivery process
7509 still has data in the lookup cache. In the case of a file lookup, there was
7510 no problem because closing a file descriptor in a subprocess doesn't affect
7511 the parent. However, if the lookup was caching a connection to a database,
7512 the connection was closed, and the second delivery process was likely to
7513 see errors such as "PGSQL: query failed: server closed the connection
7514 unexpectedly". The problem has been fixed by closing all cached lookups
7515 in a queue runner before running a delivery process.
7517 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7518 seem to have the "const" qualifier which it has on other OS. I've
7521 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7522 *of the same type* (client or server) with the same public name that an
7523 error should be diagnosed.
7525 9. When Exim looked up a host name for an IP address, but failed to find the
7526 original IP address when looking up the host name (a safety check), it
7527 output the message "<ip address> does not match any IP for NULL", which was
7528 confusing, to say the least. The bug was that the host name should have
7529 appeared instead of "NULL".
7531 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7532 user that is built into the binary, and the -C or -D options is used, root
7533 privilege is dropped before the configuration file is read. In addition,
7534 logging is switched to stderr instead of the normal log files. If the
7535 configuration then re-defines the Exim user, the unprivileged environment
7536 is probably not what is expected, so Exim logs a panic warning message (but
7539 However, if deliver_drop_privilege is set, the unprivileged state may well
7540 be exactly what is intended, so the warning has been cut out in that case,
7541 and Exim is allowed to try to write to its normal log files.
7547 1. smtp_return_error_details was not giving details for temporary sender
7548 or receiver verification errors.
7550 2. Diagnose a configuration error if two authenticators have the same public
7553 3. Exim used not to create the message log file for a message until the first
7554 delivery attempt. This could be confusing when incoming messages were held
7555 for policy or load reasons. The message log file is now created at the time
7556 the message is received, and an initial "Received" line is written to it.
7558 4. The automatically generated man page for command line options had a minor
7559 bug that caused no ill effects; however, a more serious problem was that
7560 the procedure for building the man page automatically didn't always
7561 operate. Consequently, release 4.20 contains an out-of-date version. This
7562 shouldn't happen again.
7564 5. When building Exim with embedded Perl support, the script that builds the
7565 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7566 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7568 6. The freeze_tell option was not being used for messages that were frozen on
7569 arrival, either by an ACL or by local_scan().
7571 7. Added the smtp_incomplete_transaction log selector.
7573 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7574 was accepting AUTH without a new EHLO.
7576 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7577 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7580 10. Logging of TCP/IP connections (when configured) now happens in the main
7581 daemon process instead of the child process, so that the TCP/IP connection
7582 count is more accurate (but it can never be perfect).
7584 11. The use of "drop" in a nested ACL was not being handled correctly in the
7585 outer ACL. Now, if condition failure induced by the nested "drop" causes
7586 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7587 or "require"), the connection is dropped.
7589 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7590 that yield "discard" can now be used with an "accept" or a "discard" verb,
7591 but an error is generated for any others (because I can't see a useful way
7592 to define what should happen).
7594 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7595 are now processed in the same way as lines in the Exim configuration file.
7596 In particular, continuation lines are supported.
7598 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7600 15. Added -ti meaning -t -i.
7602 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7603 domains, and warn by logging if others are found.
7605 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7606 returned from GIFCONF when Exim is trying to find the list of interfaces on
7607 a host. The code in os.c has been modified to copy each ifreq to an aligned
7608 structure in all cases.
7610 Also, in some cases, the returned ifreq's were being copied to a 'struct
7611 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7612 means the last couple of bytes of an IPv6 address could be chopped if the
7613 ifreq contained only a normal sockaddr (14 bytes storage).
7615 18. Named domain lists were not supported in the hosts_treat_as_local option.
7616 An entry such as +xxxx was not recognized, and was treated as a literal
7619 19. Ensure that header lines added by a DATA ACL are included in the reject log
7620 if the ACL subsequently rejects the message.
7622 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7623 MD5 (which is deprecated).
7625 21. When testing a filter file using -bf, Exim was writing a message when it
7626 took the sender from a "From " line in the message, but it was not doing so
7627 when it took $return_path from a Return-Path: header line. It now does.
7629 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7630 with a valid header line field name (a series of printing characters
7631 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7633 23. Changed "disc" in the source to "disk" to conform to the documentation and
7634 the book and for uniformity.
7636 24. Ignore Sendmail's -Ooption=value command line item.
7638 25. When execve() failed while trying to run a command in a pipe transport,
7639 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7640 could be confused with a return value of 69 from the command itself. This
7641 has been changed to 127, the value the shell returns if it is asked to run
7642 a non-existent command. The wording for the related log line suggests a
7643 non-existent command as the problem.
7645 26. If received_header_text expands to an empty string, do not add a Received:
7646 header line to the message. (Well, it adds a token one on the spool, but
7647 marks it "old" so that it doesn't get used or transmitted.)
7649 27. Installed eximstats 1.28 (addition of -nt option).
7651 28. There was no check for failure on the call to getsockname() in the daemon
7652 code. This can fail if there is a shortage of resources on the system, with
7653 ENOMEM, for example. A temporary error is now given on failure.
7655 29. Contrary to the C standard, it seems that in some environments, the
7656 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7657 program. Exim now does this explicitly; it affects the formatting of
7658 timestamps using strftime().
7660 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7661 complaints. I've now initialized all the variables, to avoid this.
7663 32. Header lines added by a system filter were not being "seen" during
7664 transport-time rewrites.
7666 33. The info_callback() function passed to OpenSSL is set up with type void
7667 (*)(SSL *, int, int), as described somewhere. However, when calling the
7668 function (actually a macro) that sets it up, the type void(*)() is
7669 expected. I've put in a cast to prevent warnings from picky compilers.
7671 34. If a DNS black list lookup found a CNAME record, but there were no A
7672 records associated with the domain it pointed at, Exim crashed.
7674 35. If a DNS black list lookup returned more than one A record, Exim ignored
7675 all but the first. It now scans all returned addresses if a particular IP
7676 value is being sought. In this situation, the contents of the
7677 $dnslist_value variable are a list of all the addresses, separated by a
7680 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7681 to accept a host name and all its aliases if the forward lookup for any of
7682 them yielded the IP address of the incoming connection. Now it accepts only
7683 those names whose forward lookup yields the correct IP address. Any other
7684 names are discarded. This closes a loophole whereby a rogue DNS
7685 administrator could create reverse DNS records to break through a
7686 wildcarded host restriction in an ACL.
7688 37. If a user filter or a system filter that ran in a subprocess used any of
7689 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7690 the wrong values were passed to the pipe command ($thisaddress had the
7691 value of $0, $0 had the value of $1, etc). This bug was introduced by
7692 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7694 38. Improved the line breaking for long SMTP error messages from ACLs.
7695 Previously, if there was no break point between 40 and 75 characters, Exim
7696 left the rest of the message alone. Two changes have been made: (a) I've
7697 reduced the minimum length to 35 characters; (b) if it can't find a break
7698 point between 35 and 75 characters, it looks ahead and uses the first one
7699 that it finds. This may give the occasional overlong line, but at least the
7700 remaining text gets split now.
7702 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7703 file descriptors might be low, and that setting 1000 would always raise it.
7704 It turns out that in some environments, the limit is already over 1000 and
7705 that lowering it causes trouble. So now Exim takes care not to decrease it.
7707 40. When delivering a message, the value of $return_path is set to $sender_
7708 address at the start of routing (routers may change the value). By an
7709 oversight, this default was not being set up when an address was tested by
7710 -bt or -bv, which affected the outcome if any router or filter referred to
7713 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7714 log only when "message" or "log_message" are set. However, if one of the
7715 conditions was an address verification, or a call to a nested ACL, the
7716 messages generated by the underlying test were being passed through. This
7717 no longer happens. The underlying message is available in $acl_verify_
7718 message for both "message" and "log_message" expansions, so it can be
7719 passed through if needed.
7721 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7722 new expansion $bh_ to give the encoded byte string without charset
7723 translation. Translation happens only if iconv() is available; HAVE_ICONV
7724 indicates this at build time. HEADERS_CHARSET gives the charset to
7725 translate to; headers_charset can change it in the configuration, and
7726 "headers charset" can change it in an individual filter file.
7728 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7729 that creates RFC 2047 encoded "words" labels them as that charset instead
7730 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7731 expansion operator; (ii) when Exim creates a From: line for a local
7732 message; (iii) when a header line is rewritten to include a "phrase" part.
7734 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7735 buggy, causing it to skip the first lines of messages whose message ID
7736 ended in 'D'. This would not have bitten before Exim release 4.14, because
7737 message IDs were unlikely to end in 'D' before then. The effect was to have
7738 incorrect size information for certain domains.
7740 45. #include "config.h" was missing at the start of the crypt16.c module. This
7741 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7744 46. If there was a timeout during a "random" callout check, Exim treated it as
7745 a failure of the random address, and carried on sending RSET and the real
7746 address. If the delay was just some slowness somewhere, the response to the
7747 original RCPT would be taken as a response to RSET and so on, causing
7748 mayhem of various kinds.
7750 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7751 when I implemented it. It didn't allow for the fact that some option values
7752 may legitimately be negative (e.g. size_addition), and it didn't even do
7753 the right test for positive values.
7755 48. Domain names in DNS records are case-independent. Exim always looks them up
7756 in lower case. Some resolvers return domain names in exactly the case they
7757 appear in the zone file, that is, they may contain uppercase letters. Not
7758 all resolvers do this - some return always lower case. Exim was treating a
7759 change of case by a resolver as a change of domain, similar to a widening
7760 of a domain abbreviation. This triggered its re-routing code and so it was
7761 trying to route what was effectively the same domain again. This normally
7762 caused routing to fail (because the router wouldn't handle the domain
7763 twice). Now Exim checks for this case specially, and just changes the
7764 casing of the domain that it ultimately uses when it transmits the message
7767 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7770 50. If a filter generated a file delivery with a non-absolute name (possible if
7771 no home directory exists for the router), the forbid_file option was not
7774 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7775 the existing equality matching.
7777 52. Exim was using ints instead of ino_t variables in some places where it was
7778 dealing with inode numbers.
7780 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7781 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7782 called TMPDIR, and if it finds it is different, it changes its value.
7784 54. The smtp_printf() function is now made available to local_scan() so
7785 additional output lines can be written before returning. There is also an
7786 smtp_fflush() function to enable the detection of a dropped connection.
7787 The variables smtp_input and smtp_batched_input are exported to
7790 55. Changed the default runtime configuration: the message "Unknown user"
7791 has been removed from the ACL, and instead placed on the localuser router,
7792 using the cannot_route_message feature. This means that any verification
7793 failures that generate their own messages won't get overridden. Similarly,
7794 the "Unrouteable address" message that was in the ACL for unverifiable
7795 relay addresses has also been removed.
7797 56. Added hosts_avoid_esmtp to the smtp transport.
7799 57. The exicyclog script was not checking for the esoteric option
7800 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7801 will work only if exicyclog is run under the appropriate euid.
7803 58. Following a discussion on the list, the rules by which Exim recognises line
7804 endings on incoming messages have been changed. The -dropcr and drop_cr
7805 options are now no-ops, retained only for backwards compatibility. The
7806 following line terminators are recognized: LF CRLF CR. However, special
7807 processing applies to CR:
7809 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7810 nor a local message in the state where . is a terminator.
7812 (ii) If a bare CR is encountered in a header line, an extra space is added
7813 after the line terminator so as not to end the header. The reasoning
7814 behind this is that bare CRs in header lines are most likely either
7815 to be mistakes, or people trying to play silly games.
7817 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7818 was being incorrectly given as 18 bytes larger than it should have been.
7819 This is a VOB (very old bug).
7821 60. This may never have affected anything current, but just in case it has:
7822 When the local host is found other than at the start of a list of hosts,
7823 the local host, those with the same MX, and any that follow, are discarded.
7824 When the list in question was part of a longer list of hosts, the following
7825 hosts (not currently being processed) were also being discarded. This no
7826 longer happens. I'm not sure if this situation could ever has previously
7829 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7832 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7833 header. This is something that is recommended in a new Internet Draft, and
7834 is something that is documented as being done by Sendmail. There are two
7835 possible values. For messages generated by the autoreply transport, Exim
7838 Auto-Submitted: auto-replied
7840 whereas for all other generated messages (e.g. bounces) it adds
7842 Auto-Submitted: auto-generated
7844 63. The "personal" condition in filters now includes a test for the
7845 Auto-Submitted: header. If it contains the string "auto-" the message it
7846 not considered personal.
7848 64. Added rcpt_include_affixes as a generic transport option.
7850 65. Added queue_only_override (default true).
7852 66. Added the syslog_duplication option.
7854 67. If what should have been the first header line of a message consisted of
7855 a space followed by a colon, Exim was mis-interpreting it as a header line.
7856 It isn't of course - it is syntactically invalid and should therefore be
7857 treated as the start of the message body. The misbehaviour could have
7858 caused a number of strange effects, including loss of data in subsequent
7859 header lines, and spool format errors.
7861 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7862 client host had authenticated. This control can now be exercised by an ACL
7863 for more flexibility.
7865 69. By default, callouts do not happen when testing with -bh. There is now a
7866 variant, -bhc, which does actually run the callout code, including
7867 consulting and updating the callout cache.
7869 70. Added support for saslauthd authentication, courtesy of Alexander
7872 71. If statvfs() failed on the spool or log directories while checking their
7873 size for availability, Exim confusingly gave the error "space shortage".
7874 Furthermore, in debugging mode it crashed with a floating point exception.
7875 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7876 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7877 serious problem, Exim now writes to the main and panic logs when this
7878 happens, with details of the failure. It then refuses to accept the
7879 incoming message, giving the message "spool directory problem" or "log
7880 directory problem" with a 421 code for SMTP messages.
7882 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7883 0, 1, and 2 exist, because some OS complain for execs without them (see
7884 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7885 descriptors. However, the code omitted to check that the open succeeded,
7886 causing mysterious errors if for some reason the permissions on /dev/null
7887 got screwed. Now Exim writes a message to the main and panic logs, and
7888 bombs out if it can't open /dev/null.
7890 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7891 interact so that it is all more flexible. It is supposed to remain
7892 backwards compatible. Also added extra_local_interfaces.
7894 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7895 to bomb out with an assertion failure - to the client this appears as a
7896 connection drop. This problem occurs in the part of the code that was taken
7897 from the Samba project. Fortunately, the assertion is in a very simple
7898 function, so I have fixed this by reproducing the function inline in the
7899 one place where it is called, and arranging for authentication to fail
7900 instead of killing the process with assert().
7902 75. The SPA client code was not working when the server requested OEM rather
7903 than Unicode encoding.
7905 76. Added code to make require_files with a specific uid setting more usable in
7906 the case where statting the file as root fails - usually a non-root-mounted
7907 NFS file system. When this happens and the failure is EACCES, Exim now
7908 forks a subprocess and does the per-uid checking as the relevant uid.
7910 77. Added process_log_path.
7912 78. If log_file_path was not explicitly set, a setting of check_log_space or
7913 check_log_inodes was ignored.
7915 79. If a space check for the spool or log partitions fails, the incident is now
7916 logged. Of course, in the latter case the data may get lost...
7918 80. Added the %p formatting code to string_format() so that it can be used to
7919 print addresses in debug_print(). Adjusted all the address printing in the
7920 debugging in store.c to use %p rather than %d.
7922 81. There was a concern that a line of code in smtp_in.c could overflow a
7923 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7924 initially expressed, the concern was not well-founded, because trailing
7925 spaces are removed early. However, if the trailing spaces were followed by
7926 a NULL, they did not get removed, so the overflow was possible. Two fixes
7929 (a) I re-wrote the offending code in a cleaner fashion.
7930 (b) If an incoming SMTP command contains a NULL character, it is rejected
7933 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7934 runs initgroups(), so that if the Exim user is in any additional groups,
7935 they will be used during message reception.
7941 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob