1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
9 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
10 related applications. Previously an "H" was used where available info
11 says that "M" should be, so change to match.
13 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
14 as arguments, so an implementation trying to copy these into a local
15 buffer was taking a taint-enforcement trap. Fix by using dynamically
16 created buffers. Similar fix for radius expansion condition.
18 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
19 reasonable, eg. to count headers. Fix by using dynamically created
20 buffers rather than a local. Do similar fixes for ACL actions "dcc",
21 "log_reject_target", "malware" and "spam"; the arguments are expanded
22 so could be handling tainted values.
24 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
25 broken the (no-op) support for this sendmail command. Restore it
26 to doing nothing, silently, and returning good status.
28 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
29 record path was given (or the default used) without a leading directory
30 path, an error occurred on trying to open it. Use the transport's working
33 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
34 transport. Previously it was the name on the DNS A-record; use instead
35 the head of the CNAME chain leading there (if there is one). This seems
36 to align better with RFC 6125.
38 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
39 smtp_accept_max_per_host allocated resources which were not released
40 when the limit was exceeded. This eventually crashed the daemon. Fix
41 by adding a relase action in that path.
43 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
44 expanded; previously using tainted values was rejected. Fix by using
45 dynamically-created buffers.
47 JH/09 Relax restrictions on ACL verify condition needing access to message
48 headers. Previously they were only permitted in data and non-smtp ACLs;
49 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
50 not_blind, header_sender and header_names_ascii verification.
52 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
53 Previously a macro used one argument twice; when called with the
54 argument as an expression having side-effects, incorrect operation
55 resulted. Use an inlineable function.
57 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
58 held open for a verify callout. Previously this wan not accounted for
59 and a corrupt onward SMTP conversation resulted.
61 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
62 Previously they were embedded naively in the constructed address; when
63 needed, strip the quoting and quote the entire local_part.
64 Also make the inbound_srs expansion condition handle quoting.
66 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
67 excluded, not matching the documentation.
69 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
70 was given for the sqlite_dbfile a trap resulted.
72 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
73 "name" argument resulted in a trap. There is no reason to disallow such;
74 this was a coding error.
76 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
77 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
78 spent suspended, ignoring the Posix definition. Previously we assumed
79 it did and a constant offset from real time could be used as a correction.
80 Change to using the same clock source for the start-of-message and the
81 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
82 exists, just to get a clock slightly more aligned to reality.
84 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
85 RFC says it is optional some validators care. The missing char was not
86 intended but triggered by a line-wrap alignement. Discovery and fix by
87 Guillaume Outters, hacked on by JH.
89 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
90 name being quoted was tainted a trap would be taken. Fix by using
91 dynamicaly created buffers. The routine could have been called by a
92 rewrite with the "h" flag, by using the "-F" command-line option, or
93 by using a "name=" option on a control=submission ACL modifier.
95 JH/19 SPF: change the Authentication-Results expansion component to give
96 smtp.helo when the sender domain is empty. Previously it gave
99 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
100 not in 127.0.0.0/8 to help in spotting list domains taken over by a
101 domain-parking registrar.
103 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
104 Previously when a whitespace character was specified it was not inserted
105 after removing the newline.
107 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
108 the domain part of the recipient address. This overrides any tls_sni
109 option set, which was previously used.
111 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
114 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
115 is_tainted() had an off-by-one error in the overenthusiastic direction.
116 Find and fix by Gavan. Although NetBSD is not a supported platform for
117 4.94 this bug could affect other platforms.
119 PP/01 Fix default prime selection to be consistent.
120 One path used ike23 still, instead of exim.dev.20160529.3; now both
121 execution flows will use the same DH primes (currently
122 exim.dev.20160529.3).
124 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
125 option in smtp transport, to match the documentation. Previously
126 verification was not being done in some cases where it should have been.
128 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
129 than one server was defined and depending on the platform memory layout
130 details, an internal consistency trap could be hit while walking the list
133 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
134 files. The value is used by the authresults expansion item. Previously
135 if this was used in a router or transport, a crash could result.
137 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
138 have a sufficient incidence of EINTR returns from select that an
139 interaction with other operations done by the main daemon loop exposed
140 a bug in the error-handling. This was benign apart from the log
143 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
144 for consistency with delivered "=>" and failed "**" lines. While we're
145 there, handle PRX and TFO.
147 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
148 applied. This resulted, if any header-line rewrite rules were configured,
149 in a panic-log trigerrable by sending a message with a long address in
150 a header. Fix by increaing the arbitrary limit to larger than a single
151 (dewrapped) 5322 header line maximum size.
153 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
154 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
155 RFC 6648 which deprecates X- options in protocols as a general practice.
156 Changeover between the implementations is handled by the mechanisms
159 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
160 a remote which uses callout/hold. Previously the local was queued.
162 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
169 JH/01 Avoid costly startup code when not strictly needed. This reduces time
170 for some exim process initialisations. It does mean that the logging
171 of TLS configuration problems is only done for the daemon startup.
173 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
175 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
176 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
179 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
180 DKIM signing is being done. Previously a transport_filter would always
181 disable CHUNKING, falling back to traditional DATA.
183 JH/05 Regard command-line recipients as tainted.
185 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
187 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
188 PAM library frees one of the arguments given to it, despite the
189 documentation. Therefore a plain malloc must be used.
191 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
192 on-stack buffers were used, resulting in a taint trap when DSN information
193 copied from a received message was written into the buffer.
195 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
196 the ordering of its ARC headers. This caused a crash.
198 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
199 a new record was being constructed with information from the peer, a trap
202 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
203 installation would get error messages from DMARC verify, when it hit the
204 nonexistent file indicated by the default. Distros wanting DMARC enabled
205 should both provide the file and set the option.
206 Also enforce no DMARC verification for command-line sourced messages.
208 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
209 could, depending on the platform, hang at the STARTTLS response.
211 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
212 message on a connection. Previously if one message had ARC headers and
213 the following one did not, a crash could result when adding an
214 Authentication-Results: header.
216 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
217 Exim main code and Exim-related utities. The introduction of taint
218 tracking also did many adjustments to string handling. Since then, eximon
219 frequently terminated with an assert failure.
221 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
222 check for 452 responses. This slightly helps the inefficieny of doing
223 a large alias-expansion into a recipient-limited target. The max_rcpt
224 transport option still applies (and at the current default, will override
225 the new feature). The check is done for either cause of synch, and forces
226 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
227 connection. The new facility is not tunable at this time.
229 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
230 library live data was being used, so the results became garbage. Make
231 copies while it is still usable.
233 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
234 on delivery deferred (==) and failed (**) lines (if a delivery was
235 attemtped). Previously it was only on completion (=>) lines.
237 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
238 for the expansion of the server_scram_iter and server_scram_salt options.
240 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
241 are now specifically given a NO_DATA response without hitting the system
242 resolver. The library goes on to do the now-standard TXT lookup.
243 Use of dnsdb lookups is not affected.
245 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
246 only retrieve the errormessage once. Previously two calls to dlerror()
247 were used, and the second one (for mainlog/paniclog) retrieved null
250 JH/20 Taint checking: disallow use of tainted data for
251 - the appendfile transport file and directory options
252 - the pipe transport command
253 - the autoreply transport file, log and once options
254 - file names used by the redirect router (including filter files)
256 - paths used by single-key lookups
257 Previously this was permitted.
259 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
260 adjusted the size of a major service buffer; this failed because the
261 buffer was in use at the time. Change to a compile-time increase in the
262 buffer size, when this authenticator is compiled into exim.
264 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
265 previous fast-mode was untenable in the face of glibs using mmap to
266 support larger malloc requests.
268 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
269 New values supported, if defined on system where compiled:
270 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
271 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
273 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
274 running a limited number of proceses in parallel, a benefit is gained. The
275 amount varies with the platform hardware and load. The use of the option
276 queue_run_in_order means we cannot do this, as ordering becomes
279 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
280 had introduced a string-copy (for ensuring NUL-termination) which was not
281 appropriate for that case, which can include embedded NUL bytes in the
282 block of data. Investigation showed the copy to actually be needless, the
283 data being length-specified.
285 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
286 done during a receiving connection, and both used TLS, global info was
287 used rather than per-connection info for tracking the state of data
288 queued for transmission. This could result in a connection hang.
290 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
291 Previously, when delivering serveral messages down a single connection
292 only the first would provide a SIZE. This was due to the size information
293 not being properly tracked.
295 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
296 TAI (at 37 seconds currently), pretend to be in UTC for time-related
297 expansion and logging. Previously, spurious values such as a future
298 minute could be seen.
300 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
301 it could crash from a null-deref. This could also affect the
302 ${addresses: } operator and ${readsock } item.
304 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
305 message following a mime one, the variable was not reset.
307 JH/30 When an pipelined-connect fails at the first response, assume incorrect
308 cached capability (perhaps the peer reneged?) and immediately retry in
311 JH/31 Fix spurious detection of timeout while writing to transport filter.
313 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
314 an attempt to copy the string was made before checking it.
316 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
317 taint of the lookup key was maintained; we now regard the presence in the
318 filesystem as sufficient validation.
320 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
321 argument is supplied.
323 JH/35 The dsearch lookup now requires that the directory is an absolute path.
324 Previously this was not checked, and nonempty relative paths made an
325 access under Exim's current working directory.
327 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
328 Previously no event was raised.
330 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
331 parameter supplied by the sender MAIL FROM command. Previously it was
332 ignored, and only the check_spool_space option value for the required
335 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
336 the size of the signing public-key. Previously it was instead giving
337 the size of the signature hash.
339 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
340 the default. See the (new) dkim_verify_min_keysizes option.
342 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
343 and an ACL use a lookup for checking either the local_part or domain,
344 stale data could be accessed. Ensure that variable references are
345 dropped between messages.
347 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
348 by the client was not checked as pointing within response data before
349 being used. A malicious client could thus cause an out-of-bounds read and
350 possibly gain authentication. Fix by adding the check.
352 JH/42 Internationalisation: change the default for downconversion in the smtp
353 transport to be "if needed". Previously it was "as previously set" for
354 the message, which usually meant "if needed" for message-submission but
355 "no" for everything else. However, MTAs have been seen using SMTPUTF8
356 even when the envelope addresses did not need it, resulting in forwarding
357 failures to non-supporting MTAs. A downconvert in such cases will be
358 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
359 transport. The change does mean that addresses needing conversion will
360 be converted when previously a delivery failure would occur.
362 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
363 response was received it would be used unchecked in a fail-DSN, violating
364 standards on line-length limits. Truncate if needed.
366 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
367 doesn't work. (Additionally add a new main config option to configure the
374 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
375 side, to decode a TLS 1.3 packet capture.
377 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
378 Previously the default library behaviour applied, sending two, each in
381 JH/03 Debug output for ACL now gives the config file name and line number for
384 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
386 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
388 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
389 buffer overrun for (non-chunking) other transports.
391 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
392 TLS1.3, means that a server rejecting a client certificate is not visible
393 to the client until the first read of encrypted data (typically the
394 response to EHLO). Add detection for that case and treat it as a failed
395 TLS connection attempt, so that the normal retry-in-clear can work (if
396 suitably configured).
398 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
399 and/or domain. Found and fixed by Jason Betts.
401 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
402 configuration). If a CNAME target was not a wellformed name pattern, a
405 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
406 the OS reports them interleaved with other addresses.
408 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
409 used both for input and for a verify callout, both encrypted, SMTP
410 responses being sent by the server could be lost. This resulted in
411 dropped connections and sometimes bounces generated by a peer sending
414 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
415 string. Previously it was possible to cause undefined behaviour in a
416 library routine (usually a crash). Found by "zerons".
418 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
421 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
422 API was removed, so update to use the newer ones.
424 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
425 any timeout set, is taking a long time. Previously we would hang on to a
426 rotated logfile "forever" if the input was arriving with long gaps
427 (a previous attempt to fix addressed lack, for a long time, of initial
430 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
431 shared (NFS) environment. The length of the tempfile name is now
432 4 + 16 ("hdr.$message_exim_id") which might break on file
433 systems which restrict the file name length to lower values.
436 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
437 shared (NFS) environment.
439 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
440 did for all versions <4.90). Notably -M, -m, --invert, -I may be
443 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
444 on some platforms for bit 31.
446 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
447 to changes apparently associated with TLS1.3 handling some of the APIs
448 previously used were either nonfunctional or inappropriate. Strings
449 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
450 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
451 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
452 This affects log line X= elements, the $tls_{in,out}_cipher variables,
453 and the use of specific cipher names in the encrypted= ACL condition.
455 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
457 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
458 verification result was not updated unless hosts_require_ocsp applied.
460 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
461 queue_list_requires_admin set to false, non-admin users were denied the
464 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
465 directory-of-certs mode. Previously they were advertised despite the
468 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
469 A single TCP connection by a client will now hold a TLS connection open
470 for multiple message deliveries, by default. Previously the default was to
473 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
474 default. If built with the facility, DANE will be used. The facility
475 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
477 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
478 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
479 must be defined and you must still, unless you define DISABLE_TLS, manage
480 the the include-dir and library-file requirements that go with that
481 choice. Non-TLS builds are still supported.
483 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
486 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
487 default. If the platform supports and has the facility enabled, it will
488 be requested on all coneections.
490 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
491 controlled by the build-time option SUPPORT_PIPE_CONNECT.
493 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
495 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
496 success-DSN messages. Previously the From: header was always the default
497 one for these; the option was ignored.
499 JH/28 Fix the timeout on smtp response to apply to the whole response.
500 Previously it was reset for every read, so a teergrubing peer sending
501 single bytes within the time limit could extend the connection for a
502 long time. Credit to Qualsys Security Advisory Team for the discovery.
504 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
505 delivery address, which leaked information of the results of local
506 forwarding. Change to the original envelope recipient address, per
509 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
510 requested. Previously not bounce was generated and a log entry of
511 error ignored was made.
513 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
515 JH/32 Introduce a general tainting mechanism for values read from the input
516 channel, and values derived from them. Refuse to expand any tainted
517 values, to catch one form of exploit.
519 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
520 was unused and the unexpanded text used for the test. Found and
521 fixed by Ruben Jenster.
523 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
524 an attempt to use a TLS library read routine dereffed a nul pointer,
527 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
528 them in our smtp response.
530 JH/36 Have the general router option retry_use_local_part default to true when
531 any of the restrictive preconditions are set (to anything). Previously it
532 was only for check_local user. The change removes one item of manual
533 configuration which is required for proper retries when a remote router
534 handles a subset of addresses for a domain.
536 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
537 link count into consideration.
539 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
540 caused the extension of big_buffer, the following lines were ignored.
542 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
543 accordance with RFC 2308. Previously there was no expiry, so a longlived
544 receive process (eg. due to ACL delays) versus a short SOA value could
547 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
549 JH/39 Promote DMARC support to mainline.
551 JH/40 Bug 2452: Add a References: header to DSNs.
553 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
554 parameters. The relevant library call is documented as "Deprecated: This
555 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
556 3.6.0, DH parameters are negotiated following RFC7919."
558 HS/06 Change the default of dnssec_request_domains to "*"
560 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
561 carried on and emitted a BDAT command, even when PIPELINING was not
564 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
565 buffer was used for the filename, resulting in a trap when tainted
566 arguments (eg. $domain) were used.
568 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
569 recommended to avoid a possible server-load attack. The feature can be
570 re-enabled via the openssl_options main cofiguration option.
572 JH/45 local_scan API: documented the current smtp_printf() call. This changed
573 for version 4.90 - adding a "more data" boolean to the arguments.
574 Bumped the ABI version number also, this having been missed previously;
575 release versions 4.90 to 4.92.3 inclusive were effectively broken in
576 respect of usage of smtp_printf() by either local_scan code or libraries
577 accessed via the ${dlfunc } expansion item. Both will need coding
578 adjustment for any calls to smtp_printf() to match the new function
579 signature; a FALSE value for the new argument is always safe.
581 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
582 the file-offset (which the Linux syscall does, and exim expects); this
583 resulted in an indefinite loop.
585 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
586 to do ARC verification. The Authentication-Results: header line added
587 by the configuration then had no ARC item.
593 JH/01 Remove code calling the customisable local_scan function, unless a new
594 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
596 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
597 non-signal-safe functions being used.
599 JH/03 Bug 2269: When presented with a received message having a stupidly large
600 number of DKIM-Signature headers, disable DKIM verification to avoid
601 a resource-consumption attack. The limit is set at twenty.
603 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
604 report of oldest_pass in ${authres } in consequence, and separate out
605 some descriptions of reasons for verification fail.
607 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
608 files in the spool were present and unlocked. A queue-runner could spot
609 them, resulting in a duplicate delivery. Fix that by doing the unlock
610 after the unlink. Investigation by Tim Stewart. Take the opportunity to
611 add more error-checking on spoolfile handling while that code is being
614 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
615 No known attacks, no CVE, this is defensive hardening.
617 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
618 a queue-runner could start a delivery while other operations were ongoing.
619 Cutthrough delivery was a common victim, resulting in duplicate delivery.
620 Found and investigated by Tim Stewart. Fix by using the open message data
621 file handle rather than opening another, and not locally closing it (which
622 releases a lock) for that case, while creating the temporary .eml format
623 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
625 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
626 $sender_verify_failure/$recipient_verify_failure to "random".
628 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
631 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
632 Previously this would segfault.
634 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
637 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
638 like zero, since the resolver should be doing this for us, But we need one
639 as a CNAME but no MX presence gets the CNAME returned; we need to check
640 that doesn't point to an MX to declare it "no MX returned" rather than
641 "error, loop". A new main option is added so the older capability of
642 following some limited number of chain links is maintained.
644 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
646 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
647 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
648 now-deprecated earlier definitions used only specified the range up to TLS
649 1.2 (in the older-version library docs).
651 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
653 JH/15 Rework TLS client-side context management. Stop using a global, and
654 explicitly pass a context around. This enables future use of TLS for
655 connections to service-daemons (eg. malware scanning) while a client smtp
656 connection is using TLS; with cutthrough connections this is quite likely.
658 JH/16 Fix ARC verification to do AS checks in reverse order.
660 JH/17 Support a "tls" option on the ${readsocket } expansion item.
662 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
663 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
664 Previously the "utf8" would be re-prepended for every additional message.
666 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
667 Previously thery were accepted, resulting in issues when attempting to
668 forward messages to a non-supporting MTA.
670 PP/02 Let -n work with printing macros too, not just options.
672 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
673 one parent address was copied, and bogus data was used at delivery-logging
674 time. Either a crash (after delivery) or bogus log data could result.
675 Discovery and analysis by Tim Stewart.
677 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
678 Previously if the string ended mid-character, we did not insert the
679 promised '?' replacement.
681 PP/04 Documentation: current string operators work on bytes, not codepoints.
683 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
684 should pack well giving a smaller memory footprint so better caching and
685 therefore performance. Group the declarations where this can't be done so
686 that the byte-sized flag variables are not interspersed among pointer
687 variables, giving a better chance of good packing by the compiler.
689 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
690 non-null, to avoid issues with sites running BATV. Previously reports were
691 sent with an empty envelope sender so looked like bounces.
693 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
694 The ignore_error flag wasn't being returned from the filter subprocess so
695 was not set for later routers. Investigation and fix by Matthias Kurz.
697 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
698 and a msg:complete for the whole, when a message is manually removed using
699 -Mrm. Developement by Matthias Kurz, hacked on by JH.
701 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
702 a "Gnu special" function, asprintf() in the DB utility binary builds; I
703 hope that is portable enough.
705 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
706 requiring a known-CA anchor certificate; make it now rely entirely on the
707 TLSA as an anchor. Checking the name on the leaf cert against the name
708 on the A-record for the host is still done for TA (but not for EE mode).
710 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
711 included in delivery lines for non-proxied connections, when compiled with
712 SUPPORT_SOCKS and running with proxy logging enabled.
714 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
715 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
716 move the existing event to fire before the normal logging of message
717 failure so that custom logging is bracketed by normal logging.
719 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
720 msg:fail:internal event. Developement by Matthias Kurz.
722 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
723 far too small for todays use of crypto signatures stored there. Go all
724 the way to the max DNS message size of 64kB, even though this might be
725 overmuch for IOT constrained device use.
727 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
728 copy a string over itself. The library routine is documented as not
729 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
731 JH/32 For main options check_spool_space and check_inode_space, where the
732 platform supports 64b integers, support more than the previous 2^31 kB
733 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
734 the previous G, M, k.
736 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
737 $authenticated_fail_id variable on authentication failure. Previously
740 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
741 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
742 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
743 GNUTLS_SEC_PARAM_MEDIUM.
745 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
746 side. Previously we would continue as if no SNI had been received.
748 JH/36 Harden the handling of string-lists. When a list consisted of a sole
749 "<" character, which should be a list-separator specification, we walked
750 off past the nul-terimation.
752 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
753 causes) even when the retry time is not yet met. Previously they were
754 not, meaning that when (say) an account was over-quota and temp-rejecting,
755 and multiple senders' messages were queued, only one sender would get
756 notified on each configured delay_warning cycle.
758 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
760 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
761 cipher-suites, an error can be left on the stack even for a succeeding
762 accept; this results in impossible error messages when a later operation
765 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
766 return error codes indicating retry. Under TLS1.3 this becomes required.
768 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
769 it only wrote the new authenticators, resulting in a lack of tracking of
770 peer changes of ESMTP extensions until the next cache flush.
772 JH/41 Fix the loop reading a message header line to check for integer overflow,
773 and more-often against header_maxsize. Previously a crafted message could
774 induce a crash of the recive process; now the message is cleanly rejected.
776 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
777 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
783 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
784 When redis_servers is set to a list of > 1 element, and the Redis servers
785 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
786 case of MOVED into a DEFER case instead, thus moving the query onto the
787 next server in the list. For a cluster of N elements, all N servers must
788 be defined in redis_servers.
790 GF/02 Catch and remove uninitialized value warning in exiqsumm
791 Check for existence of @ARGV before looking at $ARGV[0]
793 JH/01 Replace the store_release() internal interface with store_newblock(),
794 which internalises the check required to safely use the old one, plus
795 the allocate and data copy operations duplicated in both (!) of the
796 extant use locations.
798 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
799 modifier. This matches the restriction on the commandline.
801 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
802 Previously only the last row was returned.
804 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
805 we assumed that tags in the header were well-formed, and parsed the
806 element content after inspecting only the first char of the tag.
807 Assumptions at that stage could crash the receive process on malformed
810 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
811 While running the DKIM ACL we operate on the Permanent memory pool so that
812 variables created with "set" persist to the DATA ACL. Also (at any time)
813 DNS lookups that fail create cache records using the Permanent pool. But
814 expansions release any allocations made on the current pool - so a dnsdb
815 lookup expansion done in the DKIM ACL releases the memory used for the
816 DNS negative-cache, and bad things result. Solution is to switch to the
817 Main pool for expansions.
818 While we're in that code, add checks on the DNS cache during store_reset,
819 active in the testsuite.
820 Problem spotted, and debugging aided, by Wolfgang Breyha.
822 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
823 When none of the hosts presented to a transport match an already-open
824 connection, close it and proceed with the list. Previously we would
825 queue the message. Spotted by Lena with Yahoo, probably involving
828 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
829 Previously a spurious "250 OK id=" response was appended to the proper
832 JH/08 The "support for" informational output now, which built with Content
833 Scanning support, has a line for the malware scanner interfaces compiled
834 in. Interface can be individually included or not at build time.
836 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
837 by the template makefile "src/EDITME". The "STREAM" support for an older
838 ClamAV interface method is removed.
840 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
841 rows affected is given instead).
843 JH/11 The runtime Berkeley DB library version is now additionally output by
844 "exim -d -bV". Previously only the compile-time version was shown.
846 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
847 SMTP connection. Previously, when one had more recipients than the
848 first, an abortive onward connection was made. Move to full support for
849 multiple onward connections in sequence, handling cutthrough connection
850 for all multi-message initiating connections.
852 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
853 routers. Previously, a multi-recipient message would fail to match the
854 onward-connection opened for the first recipient, and cause its closure.
856 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
857 a timeout on read on a GnuTLS initiating connection, resulting in the
858 initiating connection being dropped. This mattered most when the callout
859 was marked defer_ok. Fix to keep the two timeout-detection methods
862 JH/15 Relax results from ACL control request to enable cutthrough, in
863 unsupported situations, from error to silently (except under debug)
864 ignoring. This covers use with PRDR, frozen messages, queue-only and
867 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
869 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
870 metadata, resulting in a crash in free().
872 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
873 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
874 Broken also in d185889f4, with init system revamp.
876 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
877 Previously we abruptly closed the connection after reading a malware-
878 found indication; now we go on to read the "scan ok" response line,
881 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
882 ACL. Previously, a crash would result.
884 JH/19 Speed up macro lookups during configuration file read, by skipping non-
885 macro text after a replacement (previously it was only once per line) and
886 by skipping builtin macros when searching for an uppercase lead character.
888 JH/20 DANE support moved from Experimental to mainline. The Makefile control
889 for the build is renamed.
891 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
892 was allocated for every new TLS startup, meaning one per message. Fix
893 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
895 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
896 reported the original. Fix to report (as far as possible) the ACL
897 result replacing the original.
899 JH/23 Fix memory leak during multi-message connections using STARTTLS under
900 OpenSSL. Certificate information is loaded for every new TLS startup,
901 and the resources needed to be freed.
903 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
905 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
908 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
909 DATA response info to the (existing) per-recipient response info for
910 the "C=" log element. It can have useful tracking info from the
911 destination system. Patch from Simon Arlott.
913 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
914 length value. Previously this would segfault.
916 HS/02 Support Avast multiline protoocol, this allows passing flags to
917 newer versions of the scanner.
919 JH/28 Ensure that variables possibly set during message acceptance are marked
920 dead before release of memory in the daemon loop. This stops complaints
921 about them when the debug_store option is enabled. Discovered specifically
922 for sender_rate_period, but applies to a whole set of variables.
923 Do the same for the queue-runner and queue-list loops, for variables set
924 from spool message files. Do the same for the SMTP per-message loop, for
925 certain variables indirectly set in ACL operations.
927 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
928 as a multi-recipient message from a mailinglist manager). The coding had
929 an arbitrary cutoff number of characters while checking for more input;
930 enforced by writing a NUL into the buffer. This corrupted long / fast
931 input. The problem was exposed more widely when more pipelineing of SMTP
932 responses was introduced, and one Exim system was feeding another.
933 The symptom is log complaints of SMTP syntax error (NUL chars) on the
934 receiving system, and refused recipients seen by the sending system
935 (propating to people being dropped from mailing lists).
936 Discovered and pinpointed by David Carter.
938 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
939 replaced by the ${authresults } expansion.
941 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
943 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
944 allows proper process termination in container environments.
946 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
947 Previously the "final dot" had a newline after it; ensure it is CR,LF.
949 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
950 and "err_perm", deprecated since 4.83 when the RFC-defined words
951 "temperror" and "permerror" were introduced.
953 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
954 transport-filters or DKIM-signing. The restriction was lost in the
955 consolidation of verify-callout and delivery SMTP handling.
956 Extend the restriction to also cover ARC-signing.
958 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
959 in defer=pass mode supply a 450 to the initiator. Previously the message
962 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
963 tls_require_ciphers is used as before.
965 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
966 "pass_unscanned". Only tmpfails from the scanner are written to
967 the paniclog, as they may require admin intervention (permission
968 denied, license issues). Other scanner errors (like decompression
969 bombs) do not cause a paniclog entry.
971 JH/36 Fix reinitialisation of DKIM logging variable between messages.
972 Previously it was possible to log spurious information in receive log
975 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
976 triggered odd behaviour from Outlook Express clients.
978 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
981 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
982 since the IETF WG has not yet settled on that versus the original
983 "bare" representation.
985 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
986 Previously the millisecond value corrupted the output.
987 Fix also for syslog_pid=no and log_selector +pid, for which the pid
988 corrupted the output.
994 JH/01 Rework error string handling in TLS interface so that the caller in
995 more cases is responsible for logging. This permits library-sourced
996 string to be attached to addresses during delivery, and collapses
997 pairs of long lines into single ones.
999 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1000 during configuration. Wildcards are allowed and expanded.
1002 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1003 This permits better logging.
1005 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1006 do not close it down and have the child transport start it up again on
1007 the passed-on TCP connection. Instead, proxy the child (and any
1008 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1009 affected: the continued delivery log lines do not have any DNSSEC, TLS
1010 Certificate or OCSP information. TLS cipher information is still logged.
1012 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1013 identical IP addresses on different listening ports. Will also affect
1016 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1017 add noisy ifdef guards to special-case this sillyness.
1018 Patch from Bernd Kuhls.
1020 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1021 than 255 are no longer allowed.
1023 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1024 Disable the session-cache too, which might reduce our load. Since we
1025 currrectly use a new context for every connection, both as server and
1026 client, there is no benefit for these.
1027 GnuTLS appears to not support tickets server-side by default (we don't
1028 call gnutls_session_ticket_enable_server()) but client side is enabled
1029 by default on recent versions (3.1.3 +) unless the PFS priority string
1032 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1033 <https://reproducible-builds.org/specs/source-date-epoch/>.
1035 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1036 the check for any unsuccessful recipients did not notice the limit, and
1037 erroneously found still-pending ones.
1039 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1040 MSG_MORE. Only in-clear (not on TLS connections).
1042 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1043 transport-filter is involved we can buffer the headers in memory for
1044 creating the signature, and read the spool data file once for the
1045 signature and again for transmission.
1047 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1048 4.77 as the kernel support then wasn't solid, having issues in 64bit
1049 mode. Now, it's been long enough. Add support for FreeBSD also.
1051 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1052 case where the routing stage had gathered several addresses to send to
1053 a host before calling the transport for the first, we previously failed
1054 to close down TLS in the old transport process before passing the TCP
1055 connection to the new process. The new one sent a STARTTLS command
1056 which naturally failed, giving a failed delivery and bloating the retry
1057 database. Investigation and fix prototype from Wolfgang Breyha.
1059 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1060 false-negatives in the check that the sender had not preempted a response
1061 or prompt from Exim (running as a server), due to that code's lack of
1062 awareness of the SMTP input buffering.
1064 PP/04 Add commandline_checks_require_admin option.
1065 Exim drops privileges sanely, various checks such as -be aren't a
1066 security problem, as long as you trust local users with access to their
1067 own account. When invoked by services which pass untrusted data to
1068 Exim, this might be an issue. Set this option in main configuration
1069 AND make fixes to the calling application, such as using `--` to stop
1072 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1073 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1074 responses to those, into a single TLS record each way (this usually means
1075 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1078 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1079 affects you only if you're dancing at the edge of the param size limits.
1080 If you are, and this message makes sense to you, then: raise the
1081 configured limit or use OpenSSL 1.1. Nothing we can do for older
1084 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1085 cmdline element to get the documented default one. Previously it was
1088 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1089 are unsuable for non-message-specific reasons - eg. connection timeout,
1090 banner-time rejection.
1092 JH/16 Fix logging of delivery remote port, when specified by router, under
1095 PP/06 Repair manualroute's ability to take options in any order, even if one
1096 is the name of a transport.
1099 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1101 JH/17 Change the list-building routines interface to use the expanding-string
1102 triplet model, for better allocation and copying behaviour.
1104 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1105 Previously it was constructed the first time a possibly-matching string
1106 was met in the configuration file input during startup; now it is done
1109 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1110 compatible one, to avoid the (poorly documented) possibility of a config
1111 file in the working directory redirecting the DB files, possibly correpting
1112 some existing file. CVE-2017-10140 assigned for BDB.
1114 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1115 cache-hot. Previously, although the result was properly cached, the
1116 initial verify call returned a defer.
1118 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1119 the main verify for receipient in uncached-mode.
1121 JH/22 Retire historical build files to an "unsupported" subdir. These are
1122 defined as "ones for which we have no current evidence of testing".
1124 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1125 if present. Previously it was ignored.
1127 JH/24 Start using specified-initialisers in C structure init coding. This is
1128 a C99 feature (it's 2017, so now considered safe).
1130 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1131 if was a fixed-sized field and bitmask ops via macros; it is now more
1134 PP/07 GitHub PR 56: Apply MariaDB build fix.
1135 Patch provided by Jaroslav Škarvada.
1137 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1138 during Coverity cleanups [4.87 JH/47]
1139 Diagnosis and fix provided by Michael Fischer v. Mollard.
1141 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1142 the right size to place the terminating semicolon on its own folded
1143 line, the header hash was calculated to an incorrect value thanks to
1144 the (relaxed) space the fold became.
1146 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1147 and confused the parent.
1149 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1150 which could crash as a result. This could lead to undeliverable messages.
1152 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1155 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1156 matters on fast-turnover and PID-randomising systems, which were getting
1157 out-of-order delivery.
1159 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1160 a possibly-overlapping copy. The symptom was that "Remote host closed
1161 connection in response to HELO" was logged instead of the actual 4xx
1164 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1165 Previously only that bufferd was discarded, resulting in SYMTP command
1168 JH/32 DKIM: when a message has multiple signatures matching an identity given
1169 in dkim_verify_signers, run the dkim acl once for each. Previously only
1170 one run was done. Bug 2189.
1172 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1173 "panic the current process" to "deliberately defer". The panic log is
1174 still written with the problem list name; the mail and reject logs now
1175 get a temp-reject line for the message that was being handled, saying
1176 something like "domains check lookup or other defer". The SMTP 451
1177 message is still "Temporary local problem".
1179 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1180 A crafted sequence of BDAT commands could result in in-use memory beeing
1181 freed. CVE-2017-16943.
1183 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1184 from SMTP input. Previously it was always done; now only done for DATA
1185 and not BDAT commands. CVE-2017-16944.
1187 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1188 to the message (such as an overlong header line). Previously this was
1189 not done and we did not exit BDAT mode. Followon from the previous item
1190 though a different problem.
1196 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1197 than -2003 did; needs libidn2 in addition to libidn.
1199 JH/02 The path option on a pipe transport is now expanded before use.
1201 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1202 Patch provided by "Björn", documentation fix added too.
1204 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1205 missing a wire-to-host endian conversion.
1207 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1208 close after a BDAT command line could be taken as a following command,
1209 giving a synch failure. Fix by only checking for synch immediately
1210 before acknowledging the chunk.
1212 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1213 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1214 macro. Patches provided by Josh Soref.
1216 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1217 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1218 by IANA as a service extension. However, John Klensin suggests that we
1221 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1222 the dkim code may be unix-mode line endings rather than smtp wire-format
1223 CRLF, so prepend a CR to any bare LF.
1225 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1226 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1228 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1229 the first were themselves being wrongly included in the feed into dkim
1230 processing; with most chunk sizes in use this resulted in an incorrect
1231 body hash calculated value.
1233 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1234 DKIM signature block, for verification. Although advised against by
1235 standards it is specifically not ruled illegal.
1237 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1239 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1240 missing a body hash (the bh= tag).
1242 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1243 It seems that HAProxy sends the Proxy Protocol information in clear and
1244 only then does a TLS startup, so do the same.
1246 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1247 TCP connections (such as for Spamd) unless the daemon successfully set
1248 Fast Open mode on its listening sockets. This fixes breakage seen on
1249 too-old kernels or those not configured for Fast Open, at the cost of
1250 requiring both directions being enabled for TFO, and TFO never being used
1251 by non-daemon-related Exim processes.
1253 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1254 endings, at least on the first header line. Try to canonify any that get
1255 past that check, despite the cost.
1257 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1258 now limited to an arbitrary five deep, while parsing addresses with the
1259 strip_excess_angle_brackets option enabled.
1261 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1262 instead leave the unprompted TLS handshake in socket buffer for the
1263 TLS library to consume.
1265 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1267 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1269 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1270 region in one operation, for speed, and this leaves assigned pointers
1271 dangling. Add checks run only under the testsuite which checks all
1272 variables at a store-reset and panics on a dangling pointer; add code
1273 explicitly nulling out all the variables discovered. Fixes one known
1274 bug: a transport crash, where a dangling pointer for $sending_ip_address
1275 originally assigned in a verify callout, is re-used.
1277 PP/06 Drop '.' from @INC in various Perl scripts.
1279 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1281 PP/08 Reduce a number of compilation warnings under clang; building with
1282 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1283 should be warning-free.
1285 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1287 HS/01 Fix portability problems introduced by PP/08 for platforms where
1288 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1290 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1291 chunk. This allows us to accept broken chunked messages. We need a more
1292 general solution here.
1294 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1295 already-broken messages in the queue.
1297 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1299 JH/19 Fix reference counting bug in routing-generated-address tracking.
1305 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1306 supports it and a size is available (ie. the sending peer gave us one).
1308 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1309 years of being deprecated). The replacements are the ACLs
1310 acl_smtp_mime and acl_not_smtp_mime.
1312 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1313 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1314 or even an in-clear connection were permitted. Now, if the host lookup
1315 was dnssec and dane was requested then the host is only used if the
1316 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1317 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1318 if one fails this test.
1319 This means that a poorly-configured remote DNS will make it incommunicado;
1320 but it protects against a DNS-interception attack on it.
1322 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1323 non-noisy when a race steals the message being considered.
1325 JH/05 If main configuration option tls_certificate is unset, generate a
1326 self-signed certificate for inbound TLS connections.
1328 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1329 in rewrites and routers.
1331 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1332 and logged a warning sing 4.83; now they are a configuration file error.
1334 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1335 (lacking @domain). Apply the same qualification processing as RCPT.
1337 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1339 JH/10 Support ${sha256:} applied to a string (as well as the previous
1342 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1343 a cutthrough deliver is pending, as we always want to make a connection.
1344 This also avoids re-routing the message when later placing the cutthrough
1345 connection after a verify cache hit.
1346 Do not update it with the verify result either.
1348 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1349 when routing results in more than one destination address.
1351 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1352 signing (which inhibits the cutthrough capability). Previously only
1353 the presence of an option was tested; now an expansion evaluating as
1354 empty is permissible (obviously it should depend only on data available
1355 when the cutthrough connection is made).
1357 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1358 the relevant preceding SMTP command did not note the pipelining mode.
1360 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1361 Previously they were not counted.
1363 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1364 as one having no matching records. Previously we deferred the message
1365 that needed the lookup.
1367 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1368 distinguished as "(=".
1370 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1371 for missing MX records. Previously it only worked for missing A records.
1373 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1375 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1376 after the data-go-ahead and data-ack. Patch from Jason Betts.
1378 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1379 even for a "none" policy. Patch from Tony Meyer.
1381 JH/22 Fix continued use of a connection for further deliveries. If a port was
1382 specified by a router, it must also match for the delivery to be
1385 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1386 When one of the recipients of a message was unsuitable for the connection
1387 (has no matching addresses), we lost track of needing to mark it
1388 deferred. As a result mail would be lost.
1390 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1392 JH/25 Decoding ACL controls is now done using a binary search; the source code
1393 takes up less space and should be simpler to maintain. Merge the ACL
1394 condition decode tables also, with similar effect.
1396 JH/26 Fix problem with one_time used on a redirect router which returned the
1397 parent address unchanged. A retry would see the parent address marked as
1398 delivered, so not attempt the (identical) child. As a result mail would
1401 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1402 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1403 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1406 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1407 Main config options check_{log,spool}_{inodes,space} are now
1408 100 inodes, 10MB unless set otherwise in the configuration.
1410 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1411 Previously it only applied to the main-section connection policy
1414 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1416 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1417 by me. Added RFC7919 DH primes as an alternative.
1419 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1420 are not in the system include path.
1422 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1423 GnuTLS, when a session startup failed (eg because the client disconnected)
1424 Exim did stdio operations after fclose. This was exposed by a recent
1425 change which nulled out the file handle after the fclose.
1427 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1428 signed directly by the cert-signing cert, rather than an intermediate
1429 OCSP-signing cert. This is the model used by LetsEncrypt.
1431 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1433 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1434 an incoming connection.
1436 HS/02 Bug 1802: Do not half-close the connection after sending a request
1439 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1440 fallback to "prime256v1".
1442 JH/34 SECURITY: Use proper copy of DATA command in error message.
1443 Could leak key material. Remotely exploitable. CVE-2016-9963.
1449 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1450 and 3.4.4 - once the server is enabled to respond to an OCSP request
1451 it does even when not requested, resulting in a stapling non-aware
1452 client dropping the TLS connection.
1454 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1455 support variable-length bit vectors. No functional change.
1457 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1458 The I= interface field on outgoing lines is now after the H= remote
1459 host field, same as incoming lines. There is a separate
1460 outgoing_interface log selector which allows you to disable the
1463 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1464 If not running log_selector +smtp_connection the mainlog would be held
1465 open indefinitely after a "too many connections" event, including to a
1466 deleted file after a log rotate. Leave the per net connection logging
1467 leaving it open for efficiency as that will be quickly detected by the
1468 check on the next write.
1470 HS/01 Bug 1671: Fix post transport crash.
1471 Processing the wait-<transport> messages could crash the delivery
1472 process if the message IDs didn't exist for some reason. When
1473 using 'split_spool_directory=yes' the construction of the spool
1474 file name failed already, exposing the same netto behaviour.
1476 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1477 mime_regex ACL conditions.
1479 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1480 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1481 to HELO, local diagnostic string.
1483 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1484 log line to debug. Even when configured with a tls authenticator many
1485 client connections are expected to not authenticate in this way, so
1486 an authenticate fail is not an error.
1488 HS/02 Add the Exim version string to the process info. This way exiwhat
1489 gives some more detail about the running daemon.
1491 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1492 matter for fast-change records such as DNSBLs.
1494 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1495 retry record, even if constant. There may be multiple transports with
1496 different interface settings and the retry behaviour needs to be kept
1499 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1501 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1503 JH/10 Bug 840: fix log_defer_output option of pipe transport
1505 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1506 hosts_randomize. This matters a lot when combined with mua_wrapper.
1508 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1509 ${quote_pgsql:<string>} operator.
1511 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1513 JH/14 Change header file name for freeradius-client. Relevant if compiling
1514 with Radius support; from the Gentoo tree and checked under Fedora.
1516 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1518 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1519 option transport_filter, meaning no filtering.
1521 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1523 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1524 defaults to "*" (all hosts). The variable is now available when not built
1525 with TLS, default unset, mainly to enable keeping the testsuite sane.
1526 If a server certificate is not supplied (via tls_certificate) an error is
1527 logged, and clients will find TLS connections fail on startup. Presumably
1528 they will retry in-clear.
1529 Packagers of Exim are strongly encouraged to create a server certificate
1530 at installation time.
1532 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1533 with the $config_file variable.
1535 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1536 in transport context, after the attempt, and per-recipient. The latter type
1537 is per host attempted. The event data is the error message, and the errno
1538 information encodes the lookup type (A vs. MX) used for the (first) host,
1539 and the trailing two digits of the smtp 4xx response.
1541 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1542 to write to mainlog (or rejectlog, paniclog) in the window between file
1543 creation and permissions/ownership being changed. Particularly affects
1544 installations where exicyclog is run as root, rather than exim user;
1545 result is that the running daemon panics and dies.
1547 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1549 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1550 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1551 "pri" and "weight". Note that the previous implicit priority given by the
1552 list order is no longer honoured.
1554 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1555 for DKIM processing.
1557 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1558 by defining SUPPORT_SOCKS.
1560 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1561 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1562 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1563 variables are renamed to proxy_{local,external}_{address,port}.
1565 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1566 for a build by defining SUPPORT_I18N
1568 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1569 of the query string, and make ${quote_redis:} do that quoting.
1571 JH/29 Move Events support from Experimental to mainline, enabled by default
1572 and removable for a build by defining DISABLE_EVENT.
1574 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1576 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1577 cached by the daemon.
1579 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1580 by defining LOOKUP_REDIS. The libhiredis library is required.
1582 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1583 keys are given for lookup.
1585 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1586 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1587 only supported when built with TLS support. The PolarSSL SHA routines
1588 are still used when the TLS library is too old for convenient support.
1590 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1591 openssl_options), for security. OpenSSL forces this from version 1.1.0
1592 server-side so match that on older versions.
1594 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1595 allocation for $value could be released as the expansion processing
1596 concluded, but leaving the global pointer active for it.
1598 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1599 and to use the domains and local_parts ACL conditions.
1601 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1602 incorrectly not doubled on cutthrough transmission, hence seen as a
1603 body-termination at the receiving system - resulting in truncated mails.
1604 Commonly the sender saw a TCP-level error, and retransmitted the message
1605 via the normal store-and-forward channel. This could result in duplicates
1606 received - but deduplicating mailstores were liable to retain only the
1607 initial truncated version.
1609 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1611 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1613 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1614 we're in there, support oversigning also; bug 1309.
1616 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1618 HS/04 Add support for keep_environment and add_environment options.
1620 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1621 either intentional arithmetic overflow during PRNG, or testing config-
1624 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1625 delivery resulted in actual delivery. Cancel cutthrough before DATA
1628 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1629 rejects a recipient: pass the reject to the originator.
1631 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1632 Many were false-positives and ignorable, but it's worth fixing the
1635 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1636 for the new environment-manipulation done at startup. Move the routines
1637 from being local to tls.c to being global via the os.c file.
1639 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1640 an extract embedded as result-arg for a map, the first arg for extract
1641 is unavailable so we cannot tell if this is a numbered or keyed
1642 extraction. Accept either.
1648 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1651 JH/02 The smtp transport option "multi_domain" is now expanded.
1653 JH/03 The smtp transport now requests PRDR by default, if the server offers
1656 JH/04 Certificate name checking on server certificates, when exim is a client,
1657 is now done by default. The transport option tls_verify_cert_hostnames
1658 can be used to disable this per-host. The build option
1659 EXPERIMENTAL_CERTNAMES is withdrawn.
1661 JH/05 The value of the tls_verify_certificates smtp transport and main options
1662 default to the word "system" to access the system default CA bundle.
1663 For GnuTLS, only version 3.0.20 or later.
1665 JH/06 Verification of the server certificate for a TLS connection is now tried
1666 (but not required) by default. The verification status is now logged by
1667 default, for both outbound TLS and client-certificate supplying inbound
1670 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1673 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1674 Status Notification (bounce) messages are now MIME format per RFC 3464.
1675 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1676 under the control of the dsn_advertise_hosts option, and routers may
1677 have a dsn_lasthop option.
1679 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1680 default, modifiable by a malware= option. The list separator for
1681 the options can now be changed in the usual way. Bug 68.
1683 JH/10 The smtp_receive_timeout main option is now expanded before use.
1685 JH/11 The incoming_interface log option now also enables logging of the
1686 local interface on delivery outgoing connections.
1688 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1689 if the interface and destination host and port all match.
1691 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1694 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1695 Patch from Andrew Lewis.
1697 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1698 now supports optional time-restrictions, weighting, and priority
1699 modifiers per server. Patch originally by <rommer@active.by>.
1701 JH/16 The spamd_address main option now supports a mixed list of local
1702 and remote servers. Remote servers can be IPv6 addresses, and
1703 specify a port-range.
1705 JH/17 Bug 68: The spamd_address main option now supports an optional
1706 timeout value per server.
1708 JH/18 Bug 1581: Router and transport options headers_add/remove can
1709 now have the list separator specified.
1711 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1714 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1717 JH/21 Support for the A6 type of dns record is withdrawn.
1719 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1720 rather than the verbs used.
1722 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1723 from 255 to 1024 chars.
1725 JH/24 Verification callouts now attempt to use TLS by default.
1727 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1728 are generic router options now. The defaults didn't change.
1730 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1731 Original patch from Alexander Shikoff, worked over by JH.
1733 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1734 files if ZCAT_COMMAND is not executable.
1736 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1738 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1740 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1741 Normally benign, it bites when the pair was led to by a CNAME;
1742 modern usage is to not canonicalize the domain to a CNAME target
1743 (and we were inconsistent anyway for A-only vs AAAA+A).
1745 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1747 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1748 when evaluating $sender_host_dnssec.
1750 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1751 $sender_helo_dnssec variable.
1753 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1755 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1757 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1759 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1760 documented as working, but never had. Support all but $spam_report.
1762 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1763 added for tls authenticator.
1765 HS/03 Add perl_taintmode main config option
1771 TL/01 When running the test suite, the README says that variables such as
1772 no_msglog_check are global and can be placed anywhere in a specific
1773 test's script, however it was observed that placement needed to be near
1774 the beginning for it to behave that way. Changed the runtest perl
1775 script to read through the entire script once to detect and set these
1776 variables, reset to the beginning of the script, and then run through
1777 the script parsing/test process like normal.
1779 TL/02 The BSD's have an arc4random API. One of the functions to induce
1780 adding randomness was arc4random_stir(), but it has been removed in
1781 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1782 function when detected.
1784 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1785 cause callback expansion.
1787 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1788 syntax errors in an expansion can be treated as a string instead of
1789 logging or causing an error, due to the internal use of bool_lax
1790 instead of bool when processing it.
1792 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1793 server certificates when making smtp deliveries.
1795 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1797 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1799 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1801 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1802 Merged patch from Sebastian Wiedenroth.
1804 JH/05 Fix results-pipe from transport process. Several recipients, combined
1805 with certificate use, exposed issues where response data items split
1806 over buffer boundaries were not parsed properly. This eventually
1807 resulted in duplicates being sent. This issue only became common enough
1808 to notice due to the introduction of connection certificate information,
1809 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1811 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1812 size buffer was used, resulting in syntax errors when an expansion
1815 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1816 version 3.3.6 or later.
1818 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1819 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1820 both become "event_action", the variables become $event_name, $event_data
1821 and $event_defer_errno. There is a new variable $verify_mode, usable in
1822 routers, transports and related events. The tls:cert event is now also
1823 raised for inbound connections, if the main configuration event_action
1826 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1827 early OCSP support, but no stapling (appears to be less than 1.0.0).
1829 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1830 server certificate names available under the smtp transport option
1831 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1834 JH/10 Time-related extraction expansions from certificates now use the main
1835 option "timezone" setting for output formatting, and are consistent
1836 between OpenSSL and GnuTLS compilations. Bug 1541.
1838 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1839 encoded parameter in the incoming message. Bug 1558.
1841 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1842 include certificate info, eximon was claiming there were spoolfile
1845 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1847 JH/14 Log delivery-related information more consistently, using the sequence
1848 "H=<name> [<ip>]" wherever possible.
1850 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1851 are problematic for Debian distribution, omit them from the release
1854 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1856 JH/16 Fix string representation of time values on 64bit time_t architectures.
1859 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1860 output list separator was used.
1865 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1866 checkers that were complaining about end of non-void function with no
1869 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1870 This was a regression introduced in 4.83 by another bugfix.
1872 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1874 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1875 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1881 TF/01 Correctly close the server side of TLS when forking for delivery.
1883 When a message was received over SMTP with TLS, Exim failed to clear up
1884 the incoming connection properly after forking off the child process to
1885 deliver the message. In some situations the subsequent outgoing
1886 delivery connection happened to have the same fd number as the incoming
1887 connection previously had. Exim would try to use TLS and fail, logging
1888 a "Bad file descriptor" error.
1890 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1891 utilities have not been installed.
1893 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1894 temporary space as the ACL may create new global variables.
1896 TL/01 LDAP support uses per connection or global context settings, depending
1897 upon the detected version of the libraries at build time.
1899 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1900 to extract and use the src ip:port in logging and expansions as if it
1901 were a direct connection from the outside internet. PPv2 support was
1902 updated based on HAProxy spec change in May 2014.
1904 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1906 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1907 Properly escape header and check for NULL return.
1909 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1912 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1914 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1915 characters in header names, implemented as a verify condition.
1916 Contributed by Michael Fischer v. Mollard.
1918 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1919 results permerror and temperror. Previous values are deprecated but
1920 still accepted. In a future release, err_perm and err_temp will be
1921 completely removed, which will be a backward incompatibility if the
1922 ACL tests for either of these two old results. Patch contributed by
1923 user bes-internal on the mailing list.
1925 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1927 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1928 selectors, in both main and reject logs.
1930 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1933 JH/07 Add malware type "sock" for talking to simple daemon.
1935 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1937 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1938 routers/transports under cutthrough routing.
1940 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1941 numbers. Touch up "bool" conditional to keep the same definition.
1943 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1945 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1947 JH/12 Expand items in router/transport headers_add or headers_remove lists
1948 individually rather than the list as a whole. Bug 1452.
1950 Required for reasonable handling of multiple headers_ options when
1951 they may be empty; requires that headers_remove items with embedded
1952 colons must have them doubled (or the list-separator changed).
1954 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1955 view the policy declared in the DMARC record. Currently, $dmarc_status
1956 is a combined value of both the record presence and the result of the
1959 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1961 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1962 dnslookup router and the smtp transport (applying to the forward
1965 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1966 of ldap servers used for a specific lookup. Patch provided by Heiko
1969 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1970 New variable $lookup_dnssec_authenticated for observability.
1972 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1973 Patch submitted by Lars Timman.
1975 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1977 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1978 Requires trusted mode and valid format message id, aborts otherwise.
1979 Patch contributed by Heiko Schlichting.
1981 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1982 certextract with support for various fields. Bug 1358.
1984 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1985 is requested by default, modifiable by smtp transport option
1988 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1989 operate on certificate variables to give certificate fingerprints
1990 Also new ${sha256:cert_variable}.
1992 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1994 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1995 Christian Aistleitner.
1997 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1999 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2000 file. Patch from Wolfgang Breyha.
2002 JH/25 Expand the coverage of the delivery $host and $host_address to
2003 client authenticators run in verify callout. Bug 1476.
2005 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2006 align with daemon_smtp_ports. Bug 72.
2008 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2009 support and error reporting did not work properly.
2011 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2012 and is readable. Patch from Andrew Colin Kissa.
2014 TL/14 Enhance documentation of ${run expansion and how it parses the
2015 commandline after expansion, particularly in the case when an
2016 unquoted variable expansion results in an empty value.
2018 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2020 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2021 by renaming away from it. Addresses GH issue 15, reported by
2024 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2026 TL/15 SECURITY: prevent double expansion in math comparison functions
2027 (can expand unsanitized data). Not remotely exploitable.
2034 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2036 PP/02 Make -n do something, by making it not do something.
2037 When combined with -bP, the name of an option is not output.
2039 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2042 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2043 $sender_host_name and config options to manage this, and basic check
2046 PP/05 DSCP support for outbound connections and control modifier for inbound.
2048 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2049 (Only plugin which currently uses this is kerberos4, which nobody should
2050 be using, but we should make it available and other future plugins might
2051 conceivably use it, even though it would break NAT; stuff *should* be
2052 using channel bindings instead).
2054 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2055 name; added for Sendmail compatibility; requires admin caller.
2056 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2057 just ignore it); requires trusted caller.
2058 Also parse but ignore: -Ac -Am -X<logfile>
2061 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2063 TL/02 Add +smtp_confirmation as a default logging option.
2065 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2066 Patch by Magnus Holmgren from 2007-02-20.
2068 TL/04 Bugzilla 1281 - Spec typo.
2069 Bugzilla 1283 - Spec typo.
2070 Bugzilla 1290 - Spec grammar fixes.
2072 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2074 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2076 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2077 mailing list by by Dmitry Isaikin.
2079 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2081 JH/02 Support "G" suffix to numbers in ${if comparisons.
2083 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2085 NM/01 Bugzilla 1197 - Spec typo
2086 Bugzilla 1196 - Spec examples corrections
2088 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2090 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2091 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2094 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2095 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2097 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2098 "acl {{name}{arg}...}", and optional args on acl condition
2101 JH/05 Permit multiple router/transport headers_add/remove lines.
2103 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2105 JH/07 Avoid using a waiting database for a single-message-only transport.
2106 Performance patch from Paul Fisher. Bugzilla 1262.
2108 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2111 JH/09 Add $headers_added variable, with content from use of ACL modifier
2112 add_header (but not yet added to the message). Bugzilla 199.
2114 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2115 Pulled from Bugzilla 817 by Wolfgang Breyha.
2117 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2119 (nb: this is the same fix as in Exim 4.80.1)
2121 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2124 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2127 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2129 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2130 advertise SMTP AUTH mechanism to us, instead of a generic
2131 protocol violation error. Also, make Exim more robust to bad
2132 data from the Dovecot auth socket.
2134 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2136 When a queue runner is handling a message, Exim first routes the
2137 recipient addresses, during which it prunes them based on the retry
2138 hints database. After that it attempts to deliver the message to
2139 any remaining recipients. It then updates the hints database using
2142 So if a recipient address works intermittently, it can get repeatedly
2143 deferred at routing time. The retry hints record remains fresh so the
2144 address never reaches the final cutoff time.
2146 This is a fairly common occurrence when a user is bumping up against
2147 their storage quota. Exim had some logic in its local delivery code
2148 to deal with this. However it did not apply to per-recipient defers
2149 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2151 This change adds a proper retry rule check during routing so that the
2152 final cutoff time is checked against the message's age. We only do
2153 this check if there is an address retry record and there is not a
2154 domain retry record; this implies that previous attempts to handle
2155 the address had the retry_use_local_parts option turned on. We use
2156 this as an approximation for the destination being like a local
2157 delivery, as in LMTP.
2159 I suspect this new check makes the old local delivery cutoff check
2160 redundant, but I have not verified this so I left the code in place.
2162 TF/02 Correct gecos expansion when From: is a prefix of the username.
2164 Test 0254 submits a message to Exim with the header
2168 When I ran the test suite under the user fanf2, Exim expanded
2169 the header to contain my full name, whereas it should have added
2170 a Resent-Sender: header. It erroneously treats any prefix of the
2171 username as equal to the username.
2173 This change corrects that bug.
2175 GF/01 DCC debug and logging tidyup
2176 Error conditions log to paniclog rather than rejectlog.
2177 Debug lines prefixed by "DCC: " to remove any ambiguity.
2179 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2181 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2182 Bug spotted by Jeremy Harris; was flawed since initial commit.
2183 Would have resulted in OCSP responses post-SNI triggering an Exim
2184 NULL dereference and crash.
2186 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2188 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2189 Bug detection, analysis and fix by Samuel Thibault.
2190 Bugzilla 1331, Debian bug #698092.
2192 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2194 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2195 Server implementation by Todd Lyons, client by JH.
2196 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2197 config variable "prdr_enable" controls whether the server
2198 advertises the facility. If the client requests PRDR a new
2199 acl_data_smtp_prdr ACL is called once for each recipient, after
2200 the body content is received and before the acl_smtp_data ACL.
2201 The client is controlled by both of: a hosts_try_prdr option
2202 on the smtp transport, and the server advertisement.
2203 Default client logging of deliveries and rejections involving
2204 PRDR are flagged with the string "PRDR".
2206 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2207 fclose(). Diagnosis by Todd Lyons.
2209 PP/17 Update configure.default to handle IPv6 localhost better.
2210 Patch by Alain Williams (plus minor tweaks).
2213 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2214 This is now consistent with GnuTLS, and is now documented: the
2215 previous undocumented portable approach to treating the option as
2216 unset was to force an expansion failure. That still works, and
2217 an empty string is now equivalent.
2219 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2220 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2221 not performing validation itself.
2223 PP/20 Added force_command boolean option to pipe transport.
2224 Patch from Nick Koston, of cPanel Inc.
2226 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2229 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2231 PP/21 Fix eximon continuous updating with timestamped log-files.
2232 Broken in a format-string cleanup in 4.80, missed when I repaired the
2233 other false fix of the same issue.
2234 Report and fix from Heiko Schlichting.
2237 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2238 Report from Prashanth Katuri.
2240 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2241 It's SecureTransport, so affects any MacOS clients which use the
2242 system-integrated TLS libraries, including email clients.
2244 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2245 using a MIME ACL for non-SMTP local injection.
2246 Report and assistance in diagnosis by Warren Baker.
2248 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2250 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2252 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2253 last id that failed so it may be referenced in subsequent ACL's.
2255 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2258 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2259 ldap library initialization, allowing self-signed CA's to be
2260 used. Also properly sets require_cert option later in code by
2261 using NULL (global ldap config) instead of ldap handle (per
2262 session). Bug diagnosis and testing by alxgomz.
2264 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2265 the src/util/ subdirectory.
2267 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2268 renamed to Transport Post Delivery Action by Jeremy Harris, as
2271 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2272 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2273 redis_servers = needs to be configured which will be used by the redis
2274 lookup. Patch from Warren Baker, of The Packet Hub.
2276 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2278 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2279 hostname or reverse DNS when processing a host list. Used suggestions
2280 from multiple comments on this bug.
2282 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2284 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2285 Missed a few lines, added it to make the runtest require no keyboard
2288 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2289 contains upper case chars. Make router use caseful_local_part.
2291 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2292 support when GnuTLS has been built with p11-kit.
2298 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2300 This, or similar/improved, will also be change PP/11 of 4.82.
2306 PP/01 Handle short writes when writing local log-files.
2307 In practice, only affects FreeBSD (8 onwards).
2308 Bugzilla 1053, with thanks to Dmitry Isaikin.
2310 NM/01 Bugzilla 949 - Documentation tweak
2312 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2315 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2317 PP/02 Implemented gsasl authenticator.
2319 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2321 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2322 `pkg-config foo` for cflags/libs.
2324 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2325 with rest of GSASL and with heimdal_gssapi.
2327 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2328 `pkg-config foo` for cflags/libs for the TLS implementation.
2330 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2331 properties get this fed in as external SSF. A number of robustness
2332 and debugging improvements to the cyrus_sasl authenticator.
2334 PP/08 cyrus_sasl server now expands the server_realm option.
2336 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2337 Patch by Jeremy Harris.
2339 PP/10 Added dbmjz lookup type.
2341 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2343 PP/12 MAIL args handles TAB as well as SP, for better interop with
2344 non-compliant senders.
2345 Analysis and variant patch by Todd Lyons.
2347 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2348 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2349 Patch from Dirk Mueller <dmueller@suse.com>
2351 PP/13 tls_peerdn now print-escaped for spool files.
2352 Observed some $tls_peerdn in wild which contained \n, which resulted
2353 in spool file corruption.
2355 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2356 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2357 or write after TLS renegotiation, which otherwise led to messages
2360 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2361 as a tracking header (ie: a signed header comes before the signature).
2362 Patch from Wolfgang Breyha.
2364 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2365 comma-sep list; embedded commas doubled.
2367 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2369 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2371 Report and patch from Dmitry Banschikov.
2373 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2374 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2375 failures. We appear to now support TLS1.1+ with Exim.
2377 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2378 lets Exim select keys and certificates based upon TLS SNI from client.
2379 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2380 before an outbound SMTP session. New log_selector, +tls_sni.
2382 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2383 NULL dereference. Report and patch from Alun Jones.
2385 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2386 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2387 Not seeing resolver debug output on NetBSD, but suspect this is a
2388 resolver implementation change.
2390 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2391 Left warnings. Added "eximon gdb" invocation mode.
2393 PP/21 Defaulting "accept_8bitmime" to true, not false.
2395 PP/22 Added -bw for inetd wait mode support.
2397 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2398 locate the relevant includes and libraries. Made this the default.
2400 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2401 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2403 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2404 This may cause build issues on older platforms.
2406 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2407 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2408 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2409 Added SNI support via GnuTLS too.
2410 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2412 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2414 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2415 Applied second patch from Janne, implementing suggestion to default
2416 multiple-strings-in-record handling to match SPF spec.
2418 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2420 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2421 read-only, out of scope).
2422 Patch from Wolfgang Breyha, report from Stuart Northfield.
2424 PP/29 Fix three issues highlighted by clang analyser static analysis.
2425 Only crash-plausible issue would require the Cambridge-specific
2426 iplookup router and a misconfiguration.
2427 Report from Marcin Mirosław.
2429 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2431 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2432 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2433 As part of this, removing so much warning spew let me fix some minor
2434 real issues in debug logging.
2436 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2437 assignment on my part. Fixed.
2439 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2440 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2441 Janne Snabb (who went above and beyond: thank you).
2443 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2444 string otherwise requires a connection and a bunch more work and it's
2445 relatively easy to get wrong. Should also expose TLS library linkage
2448 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2449 64-bit ${eval} (JH/03).
2451 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2452 GNU libc to support some of the 64-bit stuff, should not lead to
2453 conflicts. Defined before os.h is pulled in, so if a given platform
2454 needs to override this, it can.
2456 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2457 protection layer was required, which is not implemented.
2458 Bugzilla 1254, patch from Wolfgang Breyha.
2460 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2461 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2462 tls_dhparam take prime identifiers. Also unbreak combination of
2463 OpenSSL+DH_params+TLSSNI.
2465 PP/39 Disable SSLv2 by default in OpenSSL support.
2471 PP/01 Solaris build fix for Oracle's LDAP libraries.
2472 Bugzilla 1109, patch from Stephen Usher.
2474 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2476 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2479 TF/02 Fix a couple more cases where we did not log the error message
2480 when unlink() failed. See also change 4.74-TF/03.
2482 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2483 lock up or crash if it happened to be inside a call to libc when it
2484 got a SIGUSR1 from exiwhat.
2486 The SIGUSR1 handler appends the current process status to the process
2487 log which is later printed by exiwhat. It used to use the general
2488 purpose logging code to do this, but several functions it calls are
2489 not safe for signals.
2491 The new output code in the SIGUSR1 handler is specific to the process
2492 log, and simple enough that it's easy to inspect for signal safety.
2493 Removing some special cases also simplifies the general logging code.
2494 Removing the spurious timestamps from the process log simplifies
2497 TF/04 Improved ratelimit ACL condition.
2499 The /noupdate option has been deprecated in favour of /readonly which
2500 has clearer semantics. The /leaky, /strict, and /readonly update modes
2501 are mutually exclusive. The update mode is no longer included in the
2502 database key; it just determines when the database is updated. (This
2503 means that when you upgrade Exim will forget old rate measurements.)
2505 Exim now checks that the per_* options are used with an update mode that
2506 makes sense for the current ACL. For example, when Exim is processing a
2507 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2508 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2509 must specify per_mail/readonly. If you omit the update mode it defaults to
2510 /leaky where that makes sense (as before) or /readonly where required.
2512 The /noupdate option is now undocumented but still supported for
2513 backwards compatibility. It is equivalent to /readonly except that in
2514 ACLs where /readonly is required you may specify /leaky/noupdate or
2515 /strict/noupdate which are treated the same as /readonly.
2517 A useful new feature is the /count= option. This is a generalization
2518 of the per_byte option, so that you can measure the throughput of other
2519 aggregate values. For example, the per_byte option is now equivalent
2520 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2522 The per_rcpt option has been generalized using the /count= mechanism
2523 (though it's more complicated than the per_byte equivalence). When it is
2524 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2525 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2526 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2527 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2528 Note that using per_rcpt with a non-readonly update mode in more than
2529 one ACL will cause the recipients to be double-counted. (The per_mail
2530 and per_byte options don't have this problem.)
2532 The handling of very low rates has changed slightly. If the computed rate
2533 is less than the event's count (usually one) then this event is the first
2534 after a long gap. In this case the rate is set to the same as this event's
2535 count, so that the first message of a spam run is counted properly.
2537 The major new feature is a mechanism for counting the rate of unique
2538 events. The new per_addr option counts the number of different
2539 recipients that someone has sent messages to in the last time period. It
2540 behaves like per_rcpt if all the recipient addresses are different, but
2541 duplicate recipient addresses do not increase the measured rate. Like
2542 the /count= option this is a general mechanism, so the per_addr option
2543 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2544 example, measure the rate that a client uses different sender addresses
2545 with the options per_mail/unique=$sender_address. There are further
2546 details in the main documentation.
2548 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2550 TF/06 Removed a few PCRE remnants.
2552 TF/07 Automatically extract Exim's version number from tags in the git
2553 repository when doing development or release builds.
2555 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2556 Bugzilla 879. Patch from Paul Fisher.
2558 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2559 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2562 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2564 PP/05 Variable $av_failed, true if the AV scanner deferred.
2565 Bugzilla 1078. Patch from John Horne.
2567 PP/06 Stop make process more reliably on build failure.
2568 Bugzilla 1087. Patch from Heiko Schlittermann.
2570 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2571 Bugzilla 1089. Patch from Heiko Schlittermann.
2573 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2574 Bugzilla 1131. Patch from Holger Weiß.
2576 PP/09 Handle IPv6 addresses with SPF.
2577 Bugzilla 860. Patch from Wolfgang Breyha.
2579 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2581 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2584 PP/11 match_* no longer expand right-hand-side by default.
2585 New compile-time build option, EXPAND_LISTMATCH_RHS.
2586 New expansion conditions, "inlist", "inlisti".
2588 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2590 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2592 PP/14 fix log_write() format string regression from TF/03.
2593 Bugzilla 1152. Patch from Dmitry Isaikin.
2599 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2601 PP/02 Harmonised TLS library version reporting; only show if debugging.
2602 Layout now matches that introduced for other libraries in 4.74 PP/03.
2604 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2606 PP/04 New "dns_use_edns0" global option.
2608 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2611 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2612 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2614 TK/01 Updated PolarSSL code to 0.14.2.
2615 Bugzilla 1097. Patch from Andreas Metzler.
2617 PP/07 Catch divide-by-zero in ${eval:...}.
2618 Fixes bugzilla 1102.
2620 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2623 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2624 format-string attack -- SECURITY: remote arbitrary code execution.
2626 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2627 time unintentionally subject to list matching rules, letting the header
2628 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2629 arbitrary string expansion). This allowed for information disclosure.
2631 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2632 INT_MIN/-1 -- value coerced to INT_MAX.
2638 NM/01 Workaround for PCRE version dependency in version reporting
2641 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2642 This fixes portability to compilers other than gcc, notably
2643 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2645 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2646 makefiles for portability to HP-UX and POSIX correctness.
2648 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2649 Also via indented variable definition in the Makefile.
2650 (Debugging by Oliver Heesakkers).
2652 PP/02 Restore caching of spamd results with expanded spamd_address.
2653 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2655 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2656 Improves build reliability. Fix from: Frank Elsner
2658 NM/02 Fix wide character breakage in the rfc2047 coding
2659 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2661 NM/03 Allow underscore in dnslist lookups
2662 Fixes bug 1026. Patch from Graeme Fowler
2664 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2665 Code patches from Adam Ciarcinski of NetBSD.
2667 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2670 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2671 is logged, to avoid truncation. Patch from John Horne.
2673 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2674 Patch from Jakob Hirsch.
2676 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2677 SQL string expansion failure details.
2678 Patch from Andrey Oktyabrski.
2680 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2681 Patch from Simon Arlott.
2683 PP/09 New lookups functionality failed to compile on old gcc which rejects
2684 extern declarations in function scope.
2685 Patch from Oliver Fleischmann
2687 PP/10 Use sig_atomic_t for flags set from signal handlers.
2688 Check getgroups() return and improve debugging.
2689 Fixed developed for diagnosis in bug 927 (which turned out to be
2692 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2693 Patch from Mark Zealey.
2695 PP/12 Bugzilla 1056: Improved spamd server selection.
2696 Patch from Mark Zealey.
2698 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2699 Based on patch from Heiko Schlittermann.
2701 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2702 Patch from Uwe Doering, sign-off by Michael Haardt.
2704 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2705 variable declaration deep within a block. Bug and patch from
2708 PP/15 lookups-Makefile IRIX compatibility coercion.
2710 PP/16 Make DISABLE_DKIM build knob functional.
2712 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2713 Patch by Simon Arlott
2715 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2716 variable argument macros. Our copy now differs from upstream.
2722 TF/01 Failure to get a lock on a hints database can have serious
2723 consequences so log it to the panic log.
2725 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2726 controlled using the smtp_confirmation log selector.
2728 TF/03 Include the error message when we fail to unlink a spool file.
2730 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2731 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2732 for maintaining out-of-tree patches for some time.
2734 PP/01 Bugzilla 139: Documentation and portability issues.
2735 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2736 Handle per-OS dynamic-module compilation flags.
2738 PP/02 Let /dev/null have normal permissions.
2739 The 4.73 fixes were a little too stringent and complained about the
2740 permissions on /dev/null. Exempt it from some checks.
2741 Reported by Andreas M. Kirchwitz.
2743 PP/03 Report version information for many libraries, including
2744 Exim version information for dynamically loaded libraries. Created
2745 version.h, now support a version extension string for distributors
2746 who patch heavily. Dynamic module ABI change.
2748 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2749 privilege escalation vulnerability whereby the Exim run-time user
2750 can cause root to append content of the attacker's choosing to
2753 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2756 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2757 If dropping privileges for untrusted macros, we disabled normal logging
2758 on the basis that it would fail; for the Exim run-time user, this is not
2759 the case, and it resulted in successful deliveries going unlogged.
2760 Fixed. Reported by Andreas Metzler.
2766 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2767 only prepend for the Resent-* case. Fixes regression introduced in
2768 Exim 4.70 by NM/22 for Bugzilla 607.
2770 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2771 increasing numbers of administrators be bitten by this.
2773 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2775 PP/03 Bugzilla 994: added openssl_options main configuration option.
2777 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2779 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2781 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2783 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2784 without a peer certificate, leading to a segfault because of an
2785 assumption that peers always have certificates. Be a little more
2786 paranoid. Problem reported by Martin Tscholak.
2788 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2789 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2790 NB: ClamAV planning to remove STREAM in "middle of 2010".
2791 CL also introduces -bmalware, various -d+acl logging additions and
2792 more caution in buffer sizes.
2794 PP/09 Implemented reverse_ip expansion operator.
2796 PP/10 Bugzilla 937: provide a "debug" ACL control.
2798 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2800 PP/12 Bugzilla 973: Implement --version.
2802 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2804 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2806 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2808 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2809 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2810 ignore trailing whitespace.
2812 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2814 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2817 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2818 Notification from Dr Andrew Aitchison.
2820 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2821 ExtendedDetectionInfo response format.
2822 Notification from John Horne.
2824 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2827 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2828 XSL and documented dependency on system catalogs, with examples of how
2831 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2834 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2835 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2836 configuration file which is writeable by the Exim user or group.
2838 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2839 of configuration files to cover files specified with the -C option if
2840 they are going to be used with root privileges, not just the default
2843 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2844 option (effectively making it always true).
2846 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2847 files to be used while preserving root privileges.
2849 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2850 that rogue child processes cannot use them.
2852 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2853 run-time user, instead of root.
2855 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2856 Exim run-time user without dropping privileges.
2858 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2859 result string, instead of calling string_vformat() twice with the same
2862 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2863 for other users. Others should always drop root privileges if they use
2864 -C on the command line, even for a whitelisted configure file.
2866 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2868 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2874 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2875 $data_path, and $header_path variables; fixed documentation bugs and
2878 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2879 exipick to access non-standard spools, including the "frozen" queue
2882 NM/01 Bugzilla 965: Support mysql stored procedures.
2883 Patch from Alain Williams
2885 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2887 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2888 Patch from Andreas Metzler
2890 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2891 Patch from Kirill Miazine
2893 NM/05 Bugzilla 671: Added umask to procmail example.
2895 JJ/03 installed exipick 20100323.0, fixing doc bug
2897 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2898 directory. Notification and patch from Dan Rosenberg.
2900 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2902 TK/02 Improve log output when DKIM signing operation fails.
2904 MH/01 Treat the transport option dkim_domain as a colon separated
2905 list, not as a single string, and sign the message with each element,
2906 omitting multiple occurences of the same signer.
2908 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2909 Bugzilla 985, 986. Patch by Simon Arlott
2911 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2912 Patch by Simon Arlott
2914 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2915 MBX locking. Notification from Dan Rosenberg.
2921 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2923 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2925 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2927 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2929 NM/04 Bugzilla 915: Flag broken perl installation during build.
2935 TK/01 Added patch by Johannes Berg that expands the main option
2936 "spamd_address" if it starts with a dollar sign.
2938 TK/02 Write list of recipients to X-Envelope-Sender header when building
2939 the mbox-format spool file for content scanning (suggested by Jakob
2942 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2943 (http://www.dcc-servers.net/) support via dccifd. Activated by
2944 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2946 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2947 by Mark Daniel Reidel <mr@df.eu>.
2949 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2950 When building exim an external PCRE library is now needed -
2951 PCRE is a system library on the majority of modern systems.
2952 See entry on PCRE_LIBS in EDITME file.
2954 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2955 conversation. Added nologin parameter to request.
2956 Patch contributed by Kirill Miazine.
2958 TF/01 Do not log submission mode rewrites if they do not change the address.
2960 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2962 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2963 log files in place. Contributed by Roberto Lima.
2965 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2967 TF/03 Bugzilla 615: When checking the local_parts router precondition
2968 after a local_part_suffix or local_part_prefix option, Exim now
2969 does not use the address's named list lookup cache, since this
2970 contains cached lookups for the whole local part.
2972 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2973 Robert Millan. Documentation is in experimental-spec.txt.
2975 TF/04 Bugzilla 668: Fix parallel build (make -j).
2977 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2979 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2980 Patch provided by Jan Srzednicki.
2982 TF/05 Leading white space used to be stripped from $spam_report which
2983 wrecked the formatting. Now it is preserved.
2985 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2986 that they are available at delivery time.
2988 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2990 TF/08 TLS error reporting now respects the incoming_interface and
2991 incoming_port log selectors.
2993 TF/09 Produce a more useful error message if an SMTP transport's hosts
2994 setting expands to an empty string.
2996 NM/06 Bugzilla 744: EXPN did not work under TLS.
2997 Patch provided by Phil Pennock.
2999 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3000 Patch provided by Richard Godbee.
3002 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3003 acl_smtp_notquit, added index entry.
3005 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3006 Patch provided by Eugene Bujak.
3008 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3009 accept(). Patch provided by Maxim Dounin.
3011 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3012 Patch provided by Phil Pennock.
3014 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3016 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3017 Patch provided by Brad "anomie" Jorsch.
3019 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3020 Patch provided by Dean Brooks.
3022 TK/05 Add native DKIM support (does not depend on external libraries).
3024 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3025 Patch provided by Graeme Fowler.
3027 NM/16 Bugzilla 851: Documentation example syntax fix.
3029 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3031 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3034 NM/19 Bugzilla 745: TLS version reporting.
3035 Patch provided by Phil Pennock.
3037 NM/20 Bugzilla 167: bool: condition support.
3038 Patch provided by Phil Pennock.
3040 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3041 clients. Patch provided by Phil Pennock.
3043 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3044 Patch provided by Brad "anomie" Jorsch.
3046 NM/23 Bugzilla 687: Fix misparses in eximstats.
3047 Patch provided by Heiko Schlittermann.
3049 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3050 Patch provided by Heiko Schlittermann.
3052 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3053 plus update to original patch.
3055 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3057 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3058 Patch provided by David Brownlee.
3060 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3062 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3064 NM/30 Bugzilla 888: TLS documentation bugfixes.
3066 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3068 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3069 Unlike the original bugzilla I have changed all shell scripts in src tree.
3071 NM/33 Bugzilla 898: Transport filter timeout fix.
3072 Patch by Todd Rinaldo.
3074 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3075 Patch by Serge Demonchaux.
3077 NM/35 Bugzilla 39: Base64 decode bug fixes.
3078 Patch by Jakob Hirsch.
3080 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3082 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3084 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3086 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3092 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3093 ALT-N's libdkim that I have put here:
3094 http://duncanthrax.net/exim-experimental/
3096 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3097 were called 'true' and it seems that C99 defines that as a reserved
3098 keyword to be used with 'bool' variable types. That means you could
3099 not include C99-style headers which use bools without triggering
3100 build errors in sieve.c.
3102 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3103 as mailq or other aliases. Changed the --help handling significantly
3104 to do whats expected. exim_usage() emits usage/help information.
3106 SC/01 Added the -bylocaldomain option to eximstats.
3108 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3110 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3112 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3118 PH/01 Another patch from the Sieve maintainer.
3120 PH/02 When an IPv6 address is converted to a string for single-key lookup
3121 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3122 dots are used instead of colons so that keys in lsearch files need not
3123 contain colons. This was done some time before quoting was made available
3124 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3125 (notated using the quote facility) so as to distinguish them from IPv4
3126 keys. This meant that lookups for IP addresses in host lists did not work
3127 for iplsearch lookups.
3129 This has been fixed by arranging for IPv6 addresses to be expressed with
3130 colons if the lookup type is iplsearch. This is not incompatible, because
3131 previously such lookups could never work.
3133 The situation is now rather anomalous, since one *can* have colons in
3134 ordinary lsearch keys. However, making the change in all cases is
3135 incompatible and would probably break a number of configurations.
3137 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3140 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3141 conversion specification without a maximum field width, thereby enabling
3142 a rogue spamd server to cause a buffer overflow. While nobody in their
3143 right mind would setup Exim to query an untrusted spamd server, an
3144 attacker that gains access to a server running spamd could potentially
3145 exploit this vulnerability to run arbitrary code as the Exim user.
3147 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3148 $primary_hostname instead of what libspf2 thinks the hosts name is.
3150 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3151 a directory entry by the name of the lookup key. Previously, if a
3152 symlink pointed to a non-existing file or a file in a directory that
3153 Exim lacked permissions to read, a lookup for a key matching that
3154 symlink would fail. Now it is enough that a matching directory entry
3155 exists, symlink or not. (Bugzilla 503.)
3157 PH/03 The body_linecount and body_zerocount variables are now exported in the
3160 PH/04 Added the $dnslist_matched variable.
3162 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3163 This means they are set thereafter only if the connection becomes
3166 PH/06 Added the client_condition to authenticators so that some can be skipped
3167 by clients under certain conditions.
3169 PH/07 The error message for a badly-placed control=no_multiline_responses left
3170 "_responses" off the end of the name.
3172 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3174 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3175 (without spaces) instead of just copying the configuration text.
3177 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3179 PH/11 Added $max_received_linelength.
3181 PH/12 Added +ignore_defer and +include_defer to host lists.
3183 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3184 way in which PCRE > 7.0 is built.
3186 PH/14 Implemented queue_only_load_latch.
3188 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3189 MAIL command. The effect was to mangle the value on 64-bit systems.
3191 PH/16 Another patch from the Sieve maintainer.
3193 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3195 PH/18 If a system quota error occurred while trying to create the file for
3196 a maildir delivery, the message "Mailbox is full" was not appended to the
3197 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3198 only to a quota excession during the actual writing of the file.
3200 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3201 characters?) which causes problems in log lines. The DN values are now
3202 passed through string_printing() before being added to log lines.
3204 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3205 and InterBase are left for another time.)
3207 PH/21 Added message_body_newlines option.
3209 PH/22 Guard against possible overflow in moan_check_errorcopy().
3211 PH/23 POSIX allows open() to be a macro; guard against that.
3213 PH/24 If the recipient of an error message contained an @ in the local part
3214 (suitably quoted, of course), incorrect values were put in $domain and
3215 $local_part during the evaluation of errors_copy.
3221 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3222 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3225 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3226 issue a MAIL command.
3228 PH/02 In an ACL statement such as
3230 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3232 if a client was not listed at all, or was listed with a value other than
3233 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3234 the condition was not true (as it should be), so access was not denied.
3235 The bug was that the ! inversion was incorrectly passed on to the second
3236 item. This has been fixed.
3238 PH/03 Added additional dnslists conditions == and =& which are different from
3239 = and & when the dns lookup returns more than one IP address.
3241 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3242 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3244 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3245 FSYNC, which compiles an option called disable_fsync that allows for
3246 bypassing fsync(). The documentation is heavily laced with warnings.
3248 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3250 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3251 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3252 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3253 including adding "make clean"; (3) Added -fPIC when compiling the test
3254 dynamically loaded module, to get rid of a warning.
3256 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3257 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3258 = 0s. The bug is otherwise harmless.
3260 PH/07 There was a bug in the dovecot authenticator such that the value of
3261 $auth1 could be overwritten, and so not correctly preserved, after a
3262 successful authentication. This usually meant that the value preserved by
3263 the server_setid option was incorrect.
3265 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3267 PH/09 Installed PCRE release 7.0.
3269 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3270 run for batched SMTP input. It is now run at the start of every message
3271 in the batch. While fixing this I discovered that the process information
3272 (output by running exiwhat) was not always getting set for -bs and -bS
3273 input. This is fixed, and it now also says "batched" for BSMTP.
3275 PH/11 Added control=no_pipelining.
3277 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3278 patch, slightly modified), and move the expansion of helo_data till after
3279 the connection is made in the smtp transport (so it can use these
3282 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3284 PH/14 Added log_selector = +pid.
3286 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3288 PH/16 Add ${if forany and ${if forall.
3290 PH/17 Added dsn_from option to vary the From: line in DSNs.
3292 PH/18 Flush SMTP output before performing a callout, unless control =
3293 no_callout_flush is set.
3295 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3296 was true (the default) a successful delivery failed to delete the retry
3297 item, thus causing premature timeout of the address. The bug is now
3300 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3302 PH/21 Long custom messages for fakedefer and fakereject are now split up
3303 into multiline responses in the same way that messages for "deny" and
3304 other ACL rejections are.
3306 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3307 with slight modification.
3309 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3310 draft, changing the syntax and factoring some duplicate code".
3312 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3313 for deliveries of the second and subsequent messages over the same SMTP
3316 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3317 ${reduce, with only minor "tidies".
3319 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3321 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3322 expansion side effects.
3324 PH/27 When a message times out after an over-quota error from an Exim-imposed
3325 quota, the bounce message says "mailbox is full". This message was not
3326 being given when it was a system quota that was exceeded. It now should
3329 MH/03 Made $recipients available in local_scan(). local_scan() already has
3330 better access to the recipient list through recipients_list[], but
3331 $recipients can be useful in postmaster-provided expansion strings.
3333 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3334 in the case of a MAIL command with additional options following the
3335 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3336 were accidentally chopped off.
3338 PH/29 SMTP synchronization checks are implemented when a command is read -
3339 there is a check that no more input is waiting when there shouldn't be
3340 any. However, for some commands, a delay in an ACL can mean that it is
3341 some time before the response is written. In this time, more input might
3342 arrive, invalidly. So now there are extra checks after an ACL has run for
3343 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3344 pipelining has not been advertised.
3346 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3348 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3349 returned in the SMTP response when smtp_return_error_details was false.
3350 This has been fixed.
3352 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3353 instead of the C I/O that was originally supplied, because problems were
3354 reported on Solaris.
3356 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3357 Exim which did not show up earlier: it was assuming that a call to
3358 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3359 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3360 was a macro that became an assignment, so it seemed to work. This has
3361 changed to a proper function call with a void return, hence the compile
3362 error. Exim's code has been fixed.
3364 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3367 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3369 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3371 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3372 because it turns out to be too big or there is a timeout) there is no
3373 "Completed" line in the log. When some messages of this type were
3374 selected by exigrep, they were listed as "not completed". Others were
3375 picked up by some special patterns. I have improved the selection
3376 criteria to be more general.
3378 PH/38 The host_find_failed option in the manualroute router can now be set
3379 to "ignore", to completely ignore a host whose IP address cannot be
3380 found. If all hosts are ignored, the behaviour is controlled by the new
3381 host_all_ignored option.
3383 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3384 homing or because of multiple MX records with /mx) generated more than
3385 one IP address, and the following item turned out to be the local host,
3386 all the secondary addresses of the first item were incorrectly removed
3387 from the list, along with the local host and any following hosts (which
3388 is what is supposed to happen).
3390 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3391 whoever called Exim into the -H file. In the case of the daemon it was
3392 behaving confusingly. When first started, it used values for whoever
3393 started the daemon, but after a SIGHUP it used the Exim user (because it
3394 calls itself on a restart). I have changed the code so that it now always
3397 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3398 message are rejected with the same error (e.g. no authentication or bad
3399 sender address), and a DATA command is nevertheless sent (as can happen
3400 with PIPELINING or a stupid MUA), the error message that was given to the
3401 RCPT commands is included in the rejection of the DATA command. This is
3402 intended to be helpful for MUAs that show only the final error to their
3405 PH/42 Another patch from the Sieve maintainer.
3407 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3408 Eximstats - Fixed some broken HTML links and added missing column headers
3410 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3411 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3413 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3415 PH/43 Yet another patch from the Sieve maintainer.
3417 PH/44 I found a way to check for a TCP/IP connection going away before sending
3418 the response to the final '.' that terminates a message, but only in the
3419 case where the client has not sent further data following the '.'
3420 (unfortunately, this is allowed). However, in many cases there won't be
3421 any further data because there won't be any more messages to send. A call
3422 to select() can be used: if it shows that the input is "ready", there is
3423 either input waiting, or the socket has been closed. An attempt to read
3424 the next input character can distinguish the two cases. Previously, Exim
3425 would have sent an OK response which the client would never have see.
3426 This could lead to message repetition. This fix should cure that, at
3427 least in a lot of common cases.
3429 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3430 advertised in response to EHLO.
3436 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3437 fixed by 4.65/MH/01 (is this a record?) are fixed:
3439 (i) An empty string was always treated as zero by the numeric comparison
3440 operators. This behaviour has been restored.
3442 (ii) It is documented that the numeric comparison operators always treat
3443 their arguments as decimal numbers. This was broken in that numbers
3444 starting with 0 were being interpreted as octal.
3446 While fixing these problems I realized that there was another issue that
3447 hadn't been noticed. Values of message_size_limit (both the global option
3448 and the transport option) were treated as octal if they started with 0.
3449 The documentation was vague. These values are now always treated as
3450 decimal, and I will make that clear in the documentation.
3456 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3457 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3460 MH/01 Don't check that the operands of numeric comparison operators are
3461 integers when their expansion is in "skipping" mode (fixes bug
3462 introduced by 4.64-PH/07).
3464 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3465 child addresses, Exim now panics and dies. Previously, because the count
3466 is held in a short int, deliveries were likely to be lost. As such a
3467 large number of recipients for a single message is ridiculous
3468 (performance will be very, very poor), I have chosen to impose a limit
3469 rather than extend the field.
3475 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3476 leftover -K file (the existence of which was triggered by #402).
3477 While we were at it, introduced process PID as part of the -K
3478 filename. This should rule out race conditions when creating
3481 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3482 processing considerably. Previous code took too long for large mails,
3483 triggering a timeout which in turn triggers #401.
3485 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3486 in the DK code in transports.c. sendfile() is not really portable,
3487 hence the _LINUX specificness.
3489 TF/01 In the add_headers option to the mail command in an Exim filter,
3490 there was a bug that Exim would claim a syntax error in any
3491 header after the first one which had an odd number of characters
3494 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3495 callout verification, Exim cached a "reject" for the entire domain. This
3496 is correct for most verifications, but it is not correct for a recipient
3497 verification with use_sender or use_postmaster set, because in that case
3498 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3499 case of MAIL FROM:<> rejection from other early rejections (e.g.
3500 rejection of HELO). When verifying a recipient using a non-null MAIL
3501 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3502 Whatever the result of the callout, the value of the domain cache is
3503 left unchanged (for any other kind of callout, getting as far as trying
3504 RCPT means that the domain itself is ok).
3506 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3509 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3510 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3511 session, the daemon ignores ECONNECT errors and logs others; it now
3512 ignores EPIPE as well.
3514 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3515 (quoted-printable decoding).
3517 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3518 later the small subsequent patch to fix an introduced bug.
3520 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3522 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3524 PH/08 An error is now given if message_size_limit is specified negative.
3526 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3527 to be given (somewhat) arbitrary names.
3529 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3532 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3533 miscellaneous code fixes
3535 PH/10 Added the log_reject_target ACL modifier to specify where to log
3538 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3539 hostname. This is wrong, because it relates to the incoming message (and
3540 probably the interface on which it is arriving) and not to the outgoing
3541 callout (which could be using a different interface). This has been
3542 changed to use the value of the helo_data option from the smtp transport
3543 instead - this is what is used when a message is actually being sent. If
3544 there is no remote transport (possible with a router that sets up host
3545 addresses), $smtp_active_hostname is used.
3547 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3548 tweaks were necessary in order to get it to work (see also 21 below):
3549 (a) The code assumed that strncpy() returns a negative number on buffer
3550 overflow, which isn't the case. Replaced with Exim's string_format()
3552 (b) There were several signed/unsigned issues. I just did the minimum
3553 hacking in of casts. There is scope for a larger refactoring.
3554 (c) The code used strcasecmp() which is not a standard C function.
3555 Replaced with Exim's strcmpic() function.
3556 (d) The code set only $1; it now sets $auth1 as well.
3557 (e) A simple test gave the error "authentication client didn't specify
3558 service in request". It would seem that Dovecot has changed its
3559 interface. Fortunately there's a specification; I followed it and
3560 changed what the client sends and it appears to be working now.
3562 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3565 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3566 address A is aliased to B and C, where B exists and C does not. Without
3567 -v the output is "A verified" because verification stops after a
3568 successful redirection if more than one address is generated. However,
3569 with -v the child addresses are also verified. Exim was outputting "A
3570 failed to verify" and then showing the successful verification for C,
3571 with its parentage. It now outputs "B failed to verify", showing B's
3572 parentage before showing the successful verification of C.
3574 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3575 look up a TXT record in a specific list after matching in a combined
3578 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3579 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3580 they consult the DNS. I had assumed they would set it the way they
3581 wanted; and indeed my experiments on Linux seem to show that in some
3582 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3583 To be on the safe side, however, I have now made the interface to
3584 host_find_byname() similar to host_find_bydns(), with an argument
3585 containing the DNS resolver options. The host_find_byname() function now
3586 sets these options at its start, just as host_find_bydns() does. The smtp
3587 transport options dns_qualify_single and dns_search_parents are passed to
3588 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3589 of host_find_byname() use the default settings of RES_DEFNAMES
3590 (qualify_single) but not RES_DNSRCH (search_parents).
3592 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3593 spool_read_header() do less string testing, by means of a preliminary
3594 switch on the second character of optional "-foo" lines. (This is
3595 overdue, caused by the large number of possibilities that now exist.
3596 Originally there were few.) While I was there, I also converted the
3597 str(n)cmp tests so they don't re-test the leading "-" and the first
3598 character, in the hope this might squeeze out yet more improvement.
3600 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3601 flag allowing group syntax was set by the header_syntax check but not
3602 turned off, possible causing trouble later; (2) The flag was not being
3603 set at all for the header_verify test, causing "group"-style headers to
3604 be rejected. I have now set it in this case, and also caused header_
3605 verify to ignore an empty address taken from a group. While doing this, I
3606 came across some other cases where the code for allowing group syntax
3607 while scanning a header line wasn't quite right (mostly, not resetting
3608 the flag correctly in the right place). These bugs could have caused
3609 trouble for malformed header lines. I hope it is now all correct.
3611 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3612 with the "reply" argument non-NULL. The code, however (which originally
3613 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3614 but it didn't always do it. This confused somebody who was copying the
3615 code for some other use. I have removed all the tests.
3617 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3618 feature that was used to support insecure browsers during the U.S. crypto
3619 embargo. It requires special client support, and Exim is probably the
3620 only MTA that supported it -- and would never use it because real RSA is
3621 always available. This code has been removed, because it had the bad
3622 effect of slowing Exim down by computing (never used) parameters for the
3623 RSA_EXPORT functionality.
3625 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3626 authenticator to fail if there's a tab character in the incoming data
3627 (there should never be unless someone is messing about, as it's supposed
3628 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3629 if the connection is using TLS or if the remote IP is the same as the
3630 local IP, and the "valid-client-cert option" if a client certificate has
3633 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3634 authenticators. This can be used for authorization after authentication
3635 succeeds. (In the case of plaintext, it servers for both authentication
3638 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3639 if any retry times were supplied.
3641 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3642 connection, where there is no client IP address to check. In this
3643 situation, the verify now always succeeds.
3645 PH/25 Applied John Jetmore's -Mset patch.
3647 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3649 PH/27 In a string expansion for a processed (not raw) header when multiple
3650 headers of the same name were present, leading whitespace was being
3651 removed from all of them, but trailing whitespace was being removed only
3652 from the last one. Now trailing whitespace is removed from each header
3653 before concatenation. Completely empty headers in a concatenation (as
3654 before) are ignored.
3656 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3657 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3659 PH/29 [Removed. This was a change that I later backed out, and forgot to
3660 correct the ChangeLog entry (that I had efficiently created) before
3661 committing the later change.]
3663 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3664 address errors (4xx response to RCPT) over the same connection as other
3665 messages routed to the same hosts. Such deliveries are always "forced",
3666 so retry times are not inspected. This resulted in far too many retries
3667 for the affected addresses. The effect occurred only when there were more
3668 hosts than the hosts_max_try setting in the smtp transport when it had
3669 the 4xx errors. Those hosts that it had tried were not added to the list
3670 of hosts for which the message was waiting, so if all were tried, there
3671 was no problem. Two fixes have been applied:
3673 (i) If there are any address or message errors in an SMTP delivery, none
3674 of the hosts (tried or untried) are now added to the list of hosts
3675 for which the message is waiting, so the message should not be a
3676 candidate for sending over the same connection that was used for a
3677 successful delivery of some other message. This seems entirely
3678 reasonable: after all the message is NOT "waiting for some host".
3679 This is so "obvious" that I'm not sure why it wasn't done
3680 previously. Hope I haven't missed anything, but it can't do any
3681 harm, as the worst effect is to miss an optimization.
3683 (ii) If, despite (i), such a delivery is accidentally attempted, the
3684 routing retry time is respected, so at least it doesn't keep
3685 hammering the server.
3687 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3688 in ${readsocket because some servers need this prod.
3690 PH/32 Added some extra debug output when updating a wait-xxx database.
3692 PH/33 The hint "could be header name not terminated by colon", which has been
3693 given for certain expansion errors for a long time, was not being given
3694 for the ${if def:h_colon_omitted{... case.
3696 PH/34 The spec says: "With one important exception, whenever a domain list is
3697 being scanned, $domain contains the subject domain." There was at least
3698 one case where this was not true.
3700 PH/35 The error "getsockname() failed: connection reset by peer" was being
3701 written to the panic log as well as the main log, but it isn't really
3702 panic-worthy as it just means the connection died rather early on. I have
3703 removed the panic log writing for the ECONNRESET error when getsockname()
3706 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3707 runs only) independently of the message's sender address. This meant
3708 that, if the 4xx error was in fact related to the sender, a different
3709 message to the same recipient with a different sender could confuse
3710 things. In particular, this can happen when sending to a greylisting
3711 server, but other circumstances could also provoke similar problems.
3712 I have changed the default so that the retry time for these errors is now
3713 based a combination of the sender and recipient addresses. This change
3714 can be overridden by setting address_retry_include_sender=false in the
3717 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3718 remote server are returned as part of bounce messages. This was not
3719 happening for LMTP over a pipe (the lmtp transport), but now it is the
3720 same for both kinds of LMTP.
3722 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3723 in header lines that were in fact CNAMEs. This is no longer the case.
3725 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3726 and queue runs started by the daemon processed all messages. This has
3727 been fixed so that -R and -S can now usefully be given with -q<time>.
3729 PH/40 Import PCRE release 6.7 (fixes some bugs).
3731 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3733 PH/42 Give an error if -q is specified more than once.
3735 PH/43 Renamed the variables $interface_address and $interface_port as
3736 $received_ip_address and $received_port, to make it clear that these
3737 values apply to message reception, and not to the outgoing interface when
3738 a message is delivered. (The old names remain recognized, of course.)
3740 PH/44 There was no timeout on the connect() call when using a Unix domain
3741 socket in the ${readsocket expansion. There now is.
3743 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3744 be meaningful with "accept".
3746 SC/01 Eximstats V1.43
3747 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3749 SC/02 Eximstats V1.44
3750 Use a glob alias rather than an array ref in the generated
3751 parser. This improves both readability and performance.
3753 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3754 Collect SpamAssassin and rejection statistics.
3755 Don't display local sender or destination tables unless
3756 there is data to show.
3757 Added average volumes into the top table text output.
3759 SC/04 Eximstats V1.46
3760 Collect data on the number of addresses (recipients)
3761 as well as the number of messages.
3763 SC/05 Eximstats V1.47
3764 Added 'Message too big' to the list of mail rejection
3765 reasons (thanks to Marco Gaiarin).
3767 SC/06 Eximstats V1.48
3768 Mainlog lines which have GMT offsets and are too short to
3769 have a flag are now skipped.
3771 SC/07 Eximstats V1.49 (Alain Williams)
3772 Added the -emptyok flag.
3774 SC/08 Eximstats V1.50
3775 Fixes for obtaining the IP address from reject messages.
3777 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3778 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3779 whitespace changes from 4.64-PH/27
3781 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3784 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3785 are found, allow negative numbers in numeric criteria)
3787 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3789 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3790 to match changes made in 4.64-PH/43
3792 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3794 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3795 "domain" is actually an IP address, and force a failure. This locks out
3796 those revolvers/nameservers that support "A-for-A" lookups, in
3797 contravention of the specifications.
3799 PH/48 When a host name was looked up from an IP address, and the subsequent
3800 forward lookup of the name timed out, the host name was left in
3801 $sender_host_name, contrary to the specification.
3803 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3804 restricted to single-key lookups, Exim was not diagnosing an error if
3805 * or *@ was used with a query-style lookup.
3807 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3809 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3810 long ago, but noone interested enough thought of it. Let's just say that
3811 the "1.1" means that there are some new functions that weren't there at
3812 some point in the past.
3814 PH/51 Error processing for expansion failure of helo_data from an smtp
3815 transport during callout processing was broken.
3817 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3818 tested/used via the -bh/-bhc/-bs options.
3820 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3821 bug, fixed in subsequent PCRE releases).
3823 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3824 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3826 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3832 SC/01 Use a glob alias rather than an array ref in eximstats generated
3833 parser. This improves both readability and performance.
3835 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3836 Don't display local sender or destination tables in eximstats unless
3837 there is data to show.
3838 Added average volumes into the eximstats top table text output.
3840 SC/03 Collect data on the number of addresses (recipients) as well
3841 as the number of messages in eximstats.
3843 TF/01 Correct an error in the documentation for the redirect router. Exim
3844 does (usually) call initgroups() when daemonizing.
3846 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3847 with consistent privilege compared to when running as a daemon.
3849 TF/03 Note in the spec that $authenticated_id is not set for local
3850 submissions from trusted users.
3852 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3853 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3855 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3856 by adding some example configuration directives to the default
3857 configuration file. A little bit of work is required to uncomment the
3858 directives and define how usernames and passwords are checked, but
3859 there is now a framework to start from.
3861 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3862 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3863 without this. I don't know how relevant this is to other LDAP libraries.
3865 PH/02 Add the verb name to the "unknown ACL verb" error.
3867 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3869 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3871 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3872 directory not expanded when it should be if an expanded home directory
3873 was set for the address (which is overridden by the transport).
3875 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3878 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3879 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3880 because it is too late at that time, and has no effect.
3882 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3883 security issue with \' (bugzilla #107). I could not use the
3884 PQescapeStringConn() function, because it needs a PGconn value as one of
3887 PH/08 When testing addresses using -bt, indicate those final addresses that
3888 are duplicates that would not cause an additional delivery. At least one
3889 person was confused, thinking that -bt output corresponded to deliveries.
3890 (Suppressing duplicates isn't a good idea as you lose the information
3891 about possibly different redirections that led to the duplicates.)
3893 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3894 systems where poll() doesn't work, in particular OS X.
3896 PH/10 Added more information to debugging output for retry time not reached.
3898 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3899 operations in malware.c.
3901 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3904 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3905 syslog_duplication set false, log lines that would normally be written
3906 both the the main log and to the reject log were not written to syslog at
3909 PH/14 In the default configuration, change the use of "message" in ACL warn
3910 statements to "add_header".
3912 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3913 not followed by a command (e.g. "seen endif").
3915 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3916 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3919 PH/17 Added extra conditions to the default value of delay_warning_condition
3923 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3924 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3925 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3928 The Auto-Submitted: and various List- headers are standardised, whereas I
3929 don't think Precedence: ever was.
3931 PH/18 Refactored debugging code in route_finduser() to show more information,
3932 in particular, the error code if getpwnam() issues one.
3934 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3935 This is apparently needed in addition to the PH/07 change above to avoid
3936 any possible encoding problems.
3938 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3939 but not after initializing Perl.
3941 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3942 output them only if debugging. By default they are written stderr,
3943 apparently, which is not desirable.
3945 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3948 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3951 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3953 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3954 authenticated or an ident call has been made. Suppress the default
3955 values for $authenticated_id and $authenticated_sender (but permit -oMai
3956 and -oMas) when testing with -bh.
3958 PH/24 Re-jigged the order of the tests in the default configuration so that the
3959 tests for valid domains and recipients precede the DNS black list and CSA
3960 tests, on the grounds that those ones are more expensive.
3962 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3963 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3964 This bug exists in every version of Exim that I still have, right back to
3967 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3968 However, an attempt to turn on case-sensitivity in a regex key by
3969 including (?-i) didn't work because the subject string was already
3970 lowercased, and the effects were non-intuitive. It turns out that a
3971 one-line patch can be used to allow (?-i) to work as expected.
3977 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3978 other effects) broke the use of negated acl sub-conditions.
3980 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3983 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3984 "Deny" causes Exim to reject the incoming connection with a 554 error.
3985 Unfortunately, if there is a major crisis, such as a disk failure,
3986 tcp-wrappers gives "deny", whereas what one would like would be some
3987 kind of temporary error. A kludge has been added to help with this.
3988 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3989 554 error is used if errno is still zero or contains ENOENT (which occurs
3990 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3993 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3995 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3996 errors. Otherwise a message that provokes a temporary error (when other
3997 messages do not) can cause a whole host to time out.
3999 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4000 addresses were routed directly to files or pipes from a redirect router.
4001 File deliveries just didn't batch; pipe deliveries might have suffered
4004 PH/06 A failure to get a lock for a hints database would erroneously always say
4005 "Failed to get write lock", even when it was really a read lock.
4007 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4008 of 0600. This has been changed to use the value of the lockfile_mode
4009 option (which defaults to 0600).
4011 PH/08 Applied small patch from the Sieve maintainer.
4013 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4014 folder from quota calculations, a direct delivery into this folder messed
4015 up the contents of the maildirsize file. This was because the regex was
4016 used only to exclude .Trash (or whatever) when the size of the mailbox
4017 was calculated. There was no check that a delivery was happening into an
4018 excluded directory. This bug has been fixed by ignoring all quota
4019 processing for deliveries into excluded directories.
4021 PH/10 Added the maildirfolder_create_regex option to appendfile.
4027 PH/01 The code for finding all the local interface addresses on a FreeBSD
4028 system running IPv6 was broken. This may well have applied to all BSD
4029 systems, as well as to others that have similar system calls. The broken
4030 code found IPv4 interfaces correctly, but gave incorrect values for the
4031 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4032 that it would not match correctly against @[] and not recognize the IPv6
4035 PH/02 The ipliteral router was not recognizing addresses of the form user@
4036 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4038 PH/03 Added disable_ipv6.
4040 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4041 decoded form, because it is most often used to construct To: headers
4042 lines in autoreplies, and the decoded form may well be syntactically
4043 invalid. However, $reply_address has leading white space removed, and all
4044 newlines turned into spaces so that the autoreply transport does not
4047 PH/05 If group was specified without a user on a router, and no group or user
4048 was specified on a transport, the group from the router was ignored.
4050 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4051 for visible compile-time settings that can be used to change these
4052 numbers, for those that want even more. Backwards compatibility with old
4053 spool files has been maintained. However, going back to a previous Exim
4054 release will lost any variables that are in spool files.
4056 PH/07 Two small changes when running in the test harness: increase delay when
4057 passing a TCP/IP connection to a new process, in case the original
4058 process has to generate a bounce, and remove special handling of
4059 127.0.0.2 (sic), which is no longer necessary.
4061 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4062 be the same on different OS.
4064 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4067 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4068 whether --show-vars was specified or not
4070 JJ/02 exipick: Added support for new ACL variable spool format introduced
4073 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4074 syntactically invalid From: or Reply-to: line, and a filter used this to
4075 generate an autoreply, and therefore failed to obtain an address for the
4076 autoreply, Exim could try to deliver to a non-existent relative file
4077 name, causing unrelated and misleading errors. What now happens is that
4078 it logs this as a hard delivery error, but does not attempt to create a
4081 PH/11 The exinext utility has a -C option for testing purposes, but although
4082 the given file was scanned by exinext itself; it wasn't being passed on
4083 when Exim was called.
4085 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4086 an end-of-file indication when reading a command response.
4088 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4089 compiled. In many other places in Exim, IPv6 addresses are always
4090 recognized, so I have changed this. It also means that IPv4 domain
4091 literals of the form [IPV4:n.n.n.n] are now always recognized.
4093 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4094 used if the router is not running as root, for example, when verifying at
4095 ACL time, or when using -bh. The debugging output from this situation was
4096 non-existent - all you got was a failure to exec. I have made two
4099 (a) Failures to set uid/gid, the current directory, or a process leader
4100 in a subprocess such as that created by queryprogram now generate
4101 suitable debugging output when -d is set.
4103 (b) The queryprogram router detects when it is not running as root,
4104 outputs suitable debugging information if -d is set, and then runs
4105 the subprocess without attempting to change uid/gid.
4107 PH/15 Minor change to Makefile for building test_host (undocumented testing
4110 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4111 additional section of a DNS packet that returns MX or SRV records.
4112 Instead, it always explicitly searches for A/AAAA records. This avoids
4113 major problems that occur when a DNS server includes only records of one
4114 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4115 fixed another bug: if SRV records were looked up and the corresponding
4116 address records were *not* found in the additional section, the port
4117 values from the SRV records were lost.
4119 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4120 using the correct key (the original address) when searching the retry
4121 rules in order to find which one to use for generating the retry hint.
4123 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4124 adding the default one. Similarly, if it contains a Reply-To: header, the
4125 errors_reply_to option, if set, is not used.
4127 PH/19 When calculating a retry time, Exim used to measure the "time since
4128 failure" by looking at the "first failed" field in the retry record. Now
4129 it does not use this if it is later than than the arrival time of the
4130 message. Instead it uses the arrival time. This makes for better
4131 behaviour in cases where some deliveries succeed, thus re-setting the
4132 "first failed" field. An example is a quota failure for a huge message
4133 when small messages continue to be delivered. Without this change, the
4134 "time since failure" will always be short, possible causing more frequent
4135 delivery attempts for the huge message than are intended.
4136 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4138 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4139 $1, $2, $3) because the numerical variables can be reset during some
4140 expansion items (e.g. "match"), thereby losing the authentication data.
4142 PH/21 Make -bV show the size of off_t variables so that the test suite can
4143 decide whether to run tests for quotas > 2G.
4145 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4146 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4147 filecount value is greater than 2G or if a quota value is greater than 2G
4148 on a system where the size of off_t is not greater than 4, a panic error
4151 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4152 never match. The debug and -bh output now contains an explicit error
4153 message indicating a malformed IPv4 address or mask.
4155 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4156 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4157 PH/23 above applies.
4159 PH/25 Do not write to syslog when running in the test harness. The only
4160 occasion when this arises is a failure to open the main or panic logs
4161 (for which there is an explicit test).
4163 PH/26 Added the /no_tell option to "control=freeze".
4165 PH/27 If a host name lookup failed very early in a connection, for example, if
4166 the IP address matched host_lookup and the reverse lookup yielded a name
4167 that did not have a forward lookup, an error message of the form "no IP
4168 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4169 could be logged. Now it outputs the IP address instead of "NULL".
4171 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4172 allows the sender and the authenticated sender to be set when
4173 submitting a message from within Exim. Since child_open_exim() is
4174 documented for local_scan(), the new function should be too.
4176 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4177 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4178 results in an empty string is now treated as unset.
4180 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4182 PH/31 Added sender_verify_fail logging option.
4184 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4185 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4186 tidied the source and removed it altogether.
4188 PH/33 When a queue run was abandoned because the load average was too high, a
4189 log line was always written; now it is written only if the queue_run log
4190 selector is set. In addition, the log line for abandonment now contains
4191 information about the queue run such as the pid. This is always present
4192 in "start" and "stop" lines but was omitted from the "abandon" line.
4194 PH/34 Omit spaces between a header name and the colon in the error message that
4195 is given when verify = headers_syntax fails (if there are lots of them,
4196 the message gets confusing).
4198 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4199 names, as there are now some PTR records that contain slashes. This check
4200 is only to protect against broken name servers that fall over on strange
4201 characters, so the fact that it applies to all lookups doesn't matter.
4203 PH/36 Now that the new test suite is complete, we can remove some of the
4204 special code in Exim that was needed for the old test suite. For example,
4205 sorting DNS records because real resolvers return them in an arbitrary
4206 order. The new test suite's fake resolver always returns records in the
4209 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4210 bounces) except when queue_only is set, to avoid logging races between
4211 the different processes.
4213 PH/38 Panic-die if .include specifies a non-absolute path.
4215 PH/39 A tweak to the "H" retry rule from its user.
4217 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4218 a label. They prevented compilation on older perls.
4220 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4221 a warning to be raised on newish perls.
4223 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4224 on queue. Changes to match documented behaviour of showing count of
4225 messages matching specified criteria.
4227 PH/40 Changed the default ident timeout from 30s to 5s.
4229 PH/41 Added support for the use of login_cap features, on those BSD systems
4230 that have them, for controlling the resources used by pipe deliveries.
4232 PH/42 The content-scanning code uses fopen() to create files in which to put
4233 message data. Previously it was not paying any attention to the mode of
4234 the files. Exim runs with umask(0) because the rest of the code creates
4235 files with open(), and sets the required mode explicitly. Thus, these
4236 files were ending up world-writeable. This was not a big issue, because,
4237 being within the spool directory, they were not world-accessible. I have
4238 created a function called modefopen, which takes an additional mode
4239 argument. It sets umask(777), creates the file, chmods it to the required
4240 mode, then resets the umask. All the relevant calls to fopen() in the
4241 content scanning code have been changed to use this function.
4243 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4244 to 24 hours. This avoids potential overflow problems when processing G
4245 and H retry rules. I suspect nobody ever tinkers with this value.
4247 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4249 PH/45 When the plaintext authenticator is running as a client, the server's
4250 challenges are checked to ensure they are valid base64 strings. By
4251 default, the authentication attempt is cancelled if an invalid string is
4252 received. Setting client_ignore_invalid_base64 true ignores these errors.
4253 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4254 they are received. Thus, the responses can be made to depend on the
4255 challenges. If an invalid string is ignored, an empty string is placed in
4258 PH/46 Messages that are created by the autoreply transport now contains a
4259 References: header, in accordance with RFCs 2822 and 3834.
4261 PH/47 Added authenticated_sender_force to the smtp transport.
4263 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4265 PH/49 Installed latest patch from the Sieve maintainer.
4267 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4268 was also set, the appendfile transport was unnecessarily scanning a
4269 directory of message files (e.g. for maildir delivery) to find the count
4270 of files (along with the size), even though it did not need this
4271 information. It now does the scan only if it needs to find either the
4272 size of the count of files.
4274 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4276 PH/52 Two bugs concerned with error handling when the smtp transport is
4279 (i) Exim was not creating retry information for temporary errors given
4280 for individual recipients after the DATA command when the smtp transport
4281 was used in LMTP mode. This meant that they could be retried too
4282 frequently, and not timed out correctly.
4284 (ii) Exim was setting the flag that allows error details to be returned
4285 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4286 recipients that were returned after the DATA command.
4288 PH/53 This is related to PH/52, but is more general: for any failing address,
4289 when detailed error information was permitted to be returned to the
4290 sender, but the error was temporary, then after the final timeout, only
4291 "retry timeout exceeded" was returned. Now it returns the full error as
4292 well as "retry timeout exceeded".
4294 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4295 do this, and (what is worse) MTAs that accept it.
4297 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4298 will now be deprecated.
4300 PH/56 New os.c-cygwin from the Cygwin maintainer.
4302 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4303 formats (previously only available in exim formats via -bpr, -bpru,
4304 and -bpra. Now also available in native and exiqgrep formats)
4306 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4307 with very large, slow to parse queues
4309 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4311 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4313 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4314 responses to authentication challenges, though it was showing the
4315 challenges; (ii) I've removed the CR characters from the debug output for
4318 PH/58 Allow for the insertion of a newline as well as a space when a string
4319 is turned into more than one encoded-word during RFC 2047 encoding. The
4320 Sieve code now uses this.
4322 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4323 data_4xx, lost_connection, tls_required.
4325 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4326 message was being sent as an SMTP response.
4328 PH/61 Add -l and -k options to exicyclog.
4330 PH/62 When verifying, if an address was redirected to one new address, so that
4331 verification continued, and the new address failed or deferred after
4332 having set something in $address_data, the value of $address_data was not
4333 passed back to the ACL. This was different to the case when no
4334 redirection occurred. The value is now passed back in both cases.
4336 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4337 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4338 use login_cap.h, so on its own it isn't the distinguishing feature. The
4339 new name refers directly to the setclassresources() function.
4341 PH/65 Added configuration files for NetBSD3.
4343 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4345 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4346 is preferred over IPv4.
4348 PH/68 The bounce_return_message and bounce_return_body options were not being
4349 honoured for bounces generated during the reception of non-SMTP messages.
4350 In particular, this applied to messages rejected by the ACL. This bug has
4351 been fixed. However, if bounce_return_message is true and bounce_return_
4352 body is false, the headers that are returned for a non-SMTP message
4353 include only those that have been read before the error was detected.
4354 (In the case of an ACL rejection, they have all been read.)
4356 PH/69 The HTML version of the specification is now built in a directory called
4357 spec_html instead of spec.html, because the latter looks like a path with
4358 a MIME-type, and this confuses some software.
4360 PH/70 Catch two compiler warnings in sieve.c.
4362 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4363 function verify_get_ident() calls ip_connect() to connect a socket, but
4364 if the "connect()" function timed out, ip_connect() used to close the
4365 socket. However, verify_get_ident() also closes the socket later, and in
4366 between Exim writes to the log, which may get opened at this point. When
4367 the socket was closed in ip_connect(), the log could get the same file
4368 descriptor number as the socket. This naturally causes chaos. The fix is
4369 not to close the socket in ip_connect(); the socket should be closed by
4370 the function that creates it. There was only one place in the code where
4371 this was missing, in the iplookup router, which I don't think anybody now
4372 uses, but I've fixed it anyway.
4374 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4375 well as to direct DNS lookups. Otherwise the handling of names in host
4376 lists is inconsistent and therefore confusing.
4382 PH/01 Two changes to the default runtime configuration:
4384 (1) Move the checks for relay_from_hosts and authenticated clients from
4385 after to before the (commented out) DNS black list checks.
4387 (2) Add control=submission to the relay_from_hosts and authenticated
4388 clients checks, on the grounds that messages accepted by these
4389 statements are most likely to be submissions.
4391 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4393 (1) Generate an error if the third argument for the ${prvs expansion is
4396 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4399 (3) Reset the variables that are obtained from the first argument of
4400 ${prvscheck and used in the second argument before leaving the code,
4401 because their memory is reclaimed, so using them afterwards may do
4404 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4405 one (it's much easier than Tom thought :-).
4407 (5) Because of (4), we can now allow for the use of $prvscheck_result
4408 inside the third argument.
4410 PH/03 For some reason, the default setting of PATH when running a command from
4411 a pipe transport was just "/usr/bin". I have changed it to
4414 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4415 anything to be listed in the output from -bV.
4417 PH/05 When a filter generated an autoreply, the entire To: header line was
4418 quoted in the delivery log line, like this:
4420 => >A.N.Other <ano@some.domain> <original@ddress> ...
4422 This has been changed so that it extracts the operative address. There
4423 may be more than one such address. If so, they are comma-separated, like
4426 => >ano@some.domain,ona@other.domain <original@ddress> ...
4428 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4429 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4430 not being looked up in the rDNS to get a host name, Exim was showing the
4431 IP address twice in Received: lines, even though the IP addresses were
4432 identical. For example:
4434 Received: from [1.2.3.4] (helo=[1.2.3.4])
4436 However, if the real host name was known, it was omitting the HELO data
4437 if it matched the actual IP address. This has been tidied up so that it
4438 doesn't show the same IP address twice.
4440 PH/07 When both +timestamp and +memory debugging was on, the value given by
4441 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4442 called by the debug printing, thereby overwriting the timestamp buffer.
4443 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4445 PH/08 When the original message was included in an autoreply transport, it
4446 always said "this is a copy of the message, including all the headers",
4447 even if body_only or headers_only was set. It now gives an appropriate
4450 PH/09 Applied a patch from the Sieve maintainer which:
4452 o fixes some comments
4453 o adds the (disabled) notify extension core
4454 o adds some debug output for the result of if/elsif tests
4455 o points to the current vacation draft in the documentation
4456 and documents the missing references header update
4460 o fixes a bug in processing the envelope test (when testing
4461 multiple envelope elements, the last element determined the
4464 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4465 Electronic Mail") by including:
4467 Auto-submitted: auto-generated
4469 in the messages that it generates (bounce messages and others, such as
4470 warnings). In the case of bounce messages for non-SMTP messages, there was
4471 also a typo: it was using "Auto_submitted" (underscore instead of
4472 hyphen). Since every message generated by Exim is necessarily in response
4473 to another message, thes have all been changed to:
4475 Auto-Submitted: auto-replied
4477 in accordance with these statements in the RFC:
4479 The auto-replied keyword:
4481 - SHOULD be used on messages sent in direct response to another
4482 message by an automatic process,
4484 - MUST NOT be used on manually-generated messages,
4486 - MAY be used on Delivery Status Notifications (DSNs) and Message
4487 Disposition Notifications (MDNs),
4489 - MUST NOT be used on messages generated by automatic or periodic
4490 processes, except for messages which are automatic responses to
4493 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4494 to the default Received: header definition.
4496 PH/12 Added log selector acl_warn_skipped (default on).
4498 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4499 variables because (a) they are in the wrong storage pool and (b) even if
4500 they were copied, it wouldn't work properly because of the caching.
4502 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4503 checking when decoding. Apparently there are clients that generate
4504 overlong encoded strings. Why am I not surprised?
4506 PH/15 If the first argument of "${if match_address" was not empty, but did not
4507 contain an "@" character, Exim crashed. Now it writes a panic log message
4508 and treats the condition as false.
4510 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4512 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4513 extension "envelope-auth". The code is finished and in agreement with
4514 other implementations, but there is no documentation so far and in fact,
4515 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4516 not changing the active code.
4518 Print executed "if" and "elsif" statements when debugging is used. This
4519 helps a great deal to understand what a filter does.
4521 Document more things not specified clearly in RFC3028. I had all this
4522 sorted out, when out of a sudden new issues came to my mind. Oops."
4524 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4527 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4528 canonical form (as documented). However, after a host name lookup from
4529 the IP address, check_host() was doing a simple string comparison with
4530 addresses acquired from the DNS when checking that the found name did
4531 have the original IP as one of its addresses. Since any found IPv6
4532 addresses are likely to be in abbreviated form, the comparison could
4533 fail. Luckily, there already exists a function for doing the comparison
4534 by converting both addresses to binary, so now that is used instead of
4535 the text comparison.
4537 PH/20 There was another similar case to PH/19, when a complete host name was
4538 given in a host list; looking up its IP address could give an abbreviated
4539 form, whereas the current host's name might or might not be abbreviated.
4540 The same fix has been applied.
4546 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4547 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4550 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4551 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4553 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4555 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4556 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4557 Also, for "Auto-Submitted", treat anything other than "no" as
4558 non-personal, in accordance with RFC 3834. (Previously it treated
4559 anything starting "auto-" as non-personal.)
4561 TF/01 The control=submission/name=... option had a problem with syntax
4562 errors if the name included a slash character. The /name= option
4563 now slurps the rest of the string, so it can include any characters
4564 but it must come last in the list of options (after /sender_retain
4567 PH/05 Some modifications to the interface to the fake nameserver for the new
4575 TK/01 Added the "success_on_redirect" address verification option. See
4576 NewStuff for rationale and an example.
4578 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4580 PH/02 Patch to exigrep to allow it to work on syslog lines.
4582 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4583 fread() to skip over the body file's header line, because in Cygwin the
4584 header line is locked and is inaccessible.
4586 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4587 co-exist for some time) to make it clear that it is the Exim ID that is
4588 referenced, not the Message-ID: header line.
4590 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4591 string_format() function, because snprintf() does not exist on all
4594 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4595 ${stat: expansion item.
4597 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4598 protocol synchronization error", to keep the pedants happy.
4600 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4601 well as for IRIX systems, when gcc is being used. See the host.c source
4604 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4606 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4609 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4610 transport and to the smtp transport in LMTP mode.
4612 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4614 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4615 run a filter in a subprocess. This could lead to confusion in subsequent
4616 lookups in the parent process. There should also be a search_tidyup() at
4617 the end of the subprocess.
4619 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4620 only if the host matched helo_try_verify_hosts, which caused the
4621 verification to occur when the EHLO/HELO command was issued. The ACL just
4622 tested the remembered result. Now, if a previous verification attempt has
4623 not happened, "verify = helo" does it there and then.
4625 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4627 TK/03 Fix log output including CR from clamd.
4629 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4630 exist provoked a memory error which could cause a segfault.
4632 PH/15 Installed PCRE 6.2
4634 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4636 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4637 of the problem. Specifically, suggested +O2 rather than +O1 for the
4640 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4642 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4643 record, the port was not correctly logged when the outgoing_port log
4644 selector was set (it logged the transort's default port).
4646 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4647 fallback_hosts, and "hosts" in the smtp transport.
4649 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4650 host errors such as "Connection refused".
4652 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4653 authentication with radiusclient 0.4.9:
4655 - Error returned from rc_read_config was caught wrongly
4656 - Username/password not passed on to radius server due to wrong length.
4658 The presumption is that some radiusclient API changes for 4.51/PH/17
4659 were not taken care of correctly. The code is still untested by me (my
4660 Linux distribution still has 0.3.2 of radiusclient), but it was
4661 contributed by a Radius user.
4663 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4664 expanding the "port" option of the smtp transport.
4666 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4667 while reading a MIME header. Thanks to Tom Hughes for a patch.
4669 PH/24 Include config.h inside local_scan.h so that configuration settings are
4672 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4673 that in an ACL for RCPT (for example), you can examine exactly what was
4676 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4677 commands, but it was not correctly comparing the address with the actual
4678 client host address. Thus, it would show the EHLO address in Received:
4679 header lines when this was not necessary.
4681 PH/27 Added the % operator to ${eval:}.
4683 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4684 it should be ignoring failures (because with -C, for example, it has lost
4685 privilege). It wasn't ignoring creation failures other than "already
4688 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4691 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4692 because an input line was too long, either on its own, or by virtue of
4693 too many continuations, the temporary file was not being removed, and the
4694 return code was incorrect.
4696 PH/31 Missing "BOOL" in function definition in filtertest.c.
4698 PH/32 Applied Sieve patches from the maintainer.
4700 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4702 PH/33 Added "verify = not_blind".
4704 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4705 Local/Makefile (with some defaults set). These are used in built scripts
4706 such as exicyclog, but they have never been used in the exim_install
4707 script (though there are many overriding facilities there). I have
4708 arranged that the exim_install script now takes note of these two
4711 PH/35 Installed configuration files for Dragonfly.
4713 PH/36 When a locally submitted message by a trusted user did not contain a
4714 From: header, and the sender address was obtained from -f or from an SMTP
4715 MAIL command, and the trusted user did not use -F to supply a sender
4716 name, $originator_name was incorrectly used when constructing a From:
4717 header. Furthermore, $originator_name was used for submission mode
4718 messages from external hosts without From: headers in a similar way,
4719 which is clearly wrong.
4721 PH/37 Added control=suppress_local_fixups.
4723 PH/38 When log_selector = +received_sender was set, and the addition of the
4724 sender made the log line's construction buffer exactly full, or one byte
4725 less than full, an overflow happened when the terminating "\n" was
4728 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4729 when the result of a list match is failure because a DNS lookup failed.
4731 PH/40 RM_COMMAND is now used in the building process.
4733 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4734 the "build-* directories that it finds.
4736 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4737 address in a domain literal was a prefix of an interface address.
4739 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4740 when verifying a sender address, unless rewrite_headers is false.
4742 PH/44 Wrote a long comment about why errors_to addresses are verified as
4743 recipients, not senders.
4745 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4746 the ratelimit ACL was added.
4748 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4750 PH/46 Added extra information about PostgreSQL errors to the error string.
4752 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4753 avoiding the need to install special zones in a real server. This is
4754 backwards compatible; if it can't find the fake resolver, it drops back.
4755 Thus, both old and new test suites can be run.
4757 TF/02 Added util/ratelimit.pl
4759 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4760 clock is set back in time.
4762 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4763 Candler <B.Candler@pobox.com>.
4765 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4766 OK for addresses that are the result of redirections.
4768 PH/48 A number of further additions for the benefit of the new test suite,
4769 including a fake gethostbyname() that interfaces to the fake DNS resolver
4772 TF/06 The fix for widen_domains has also been applied to qualify_single and
4773 search_parents which are the other dnslookup options that can cause
4776 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4779 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4781 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4782 service" if it encountered an MX record with an empty target hostname.
4783 The message is now "an MX or SRV record indicated no SMTP service".
4785 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4786 if the DNS of the sending site is misconfigured. This is quite a
4787 common situation. This change restores the behaviour of treating a
4788 helo verification defer as a failure.
4790 PH/51 If self=fail was set on a router, the bounce message did not include the
4791 actual error message.
4797 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4799 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4800 command itself ended in error, the underlying message about the transport
4801 filter timeout was being overwritten with the pipe command error. Now the
4802 underlying error message should be appended to the second error message.
4804 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4806 PH/02 Reduce the amount of output that "make" produces by default. Full output
4807 can still be requested.
4809 PH/03 The warning log line about a condition test deferring for a "warn" verb
4810 was being output only once per connection, rather than after each
4811 occurrence (because it was using the same function as for successful
4812 "warn" verbs). This seems wrong, so I have changed it.
4814 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4815 it should not have, which might have caused a crash in the right
4816 circumstances, but probably never did.
4818 PH/04 Installed a modified version of Tony Finch's patch to make submission
4819 mode fix the return path as well as the Sender: header line, and to
4820 add a /name= option so that you can make the user's friendly name appear
4823 TF/03 Added the control = fakedefer ACL modifier.
4825 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4826 Mark Lowes for thorough testing.
4828 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4830 TK/03 Merged latest SRS patch from Miles Wilton.
4832 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4833 with the definition in sysexits.h (which is #included earlier).
4834 Fortunately, Exim does not actually use EX_OK. The code used to try to
4835 preserve the sysexits.h value, by assuming that macro definitions were
4836 scanned for macro replacements. I have been disabused of this notion,
4837 so now the code just undefines EX_OK before #including unistd.h.
4839 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4840 in the smtp transport. When a block could not be written in a single
4841 write() function, the timeout was being re-applied to each part-write.
4842 This seems wrong - if the receiver was accepting one byte at a time it
4843 would take for ever. The timeout is now adjusted when this happens. It
4844 doesn't have to be particularly precise.
4846 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4847 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4849 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4850 without a domain if the check to <postmaster@domain> fails.
4852 SC/01 Eximstats: added -xls and the ability to specify output files
4853 (patch written by Frank Heydlauf).
4855 SC/02 Eximstats: use FileHandles for outputting results.
4857 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4859 SC/04 Eximstats: fixed display of large numbers with -nvr option
4861 SC/05 Eximstats: fixed merging of reports with empty tables.
4863 SC/06 Eximstats: added the -include_original_destination flag
4865 SC/07 Eximstats: removed tabs and trailing whitespace.
4867 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4869 TK/06 MBOX spool code: Add real "From " MBOX separator line
4870 so the .eml file is really in mbox format (even though
4871 most programs do not really care). Patch from Alex Miller.
4873 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4874 The latter is generated from $received_to and is only set if the
4875 message has one envelope recipient. SA can use these headers,
4876 obviously out-of-the-box. Patch from Alex Miller.
4878 PH/08 The ${def test on a variable was returning false if the variable's
4879 value was "0", contrary to what the specification has always said!
4880 The result should be true unless the variable is empty.
4882 PH/09 The syntax error of a character other than { following "${if
4883 def:variable_name" (after optional whitespace) was not being diagnosed.
4884 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4885 accidental colon was present, for example, could give incorrect results.
4887 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4888 result is used (not including appendfile, where other changes are about
4891 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4892 This involved changing a lot of size variables from int to off_t. It
4893 should work with maildirs and everything.
4895 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4896 spamd dying while we are connected to it.
4898 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4901 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4902 function for BDB 4.3.
4904 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4906 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4907 the sources for PCRE have been re-arranged and more files are now
4910 PH/15 The code I had for printing potentially long long variables in PH/11
4911 above was not the best (it lost precision). The length of off_t variables
4912 is now inspected at build time, and an appropriate printing format (%ld
4913 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4914 to be "long long int" or "long int". This is needed for the internal
4915 formatting function string_vformat().
4917 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4918 the configuration file to be ":syslog", then the script "guesses" where
4919 the logs files are, rather than using the compiled in default. In our
4920 case the guess is not the same as the compiled default, so the script
4921 suddenly stopped working when I started to use syslog. The patch checks
4922 to see if log_file_path is "". If so, it attempts to read it from exim
4923 with no configuration file to get the compiled in version, before it
4924 falls back to the previous guessing code."
4926 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4927 implementing BATV in an Exim configuration. See NewStuff for the gory
4930 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4931 Makefile that are specific to HP-UX.
4933 PH/18 If the "use_postmaster" option was set for a recipient callout together
4934 with the "random" option, the postmaster address was used as the MAIL
4935 FROM address for the random test, but not for the subsequent recipient
4936 test. It is now used for both.
4938 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4939 patch removes a few documentation additions to RFC 3028, because the
4940 latest draft now contains them. It adds the new en;ascii-case comparator
4941 and a new error check for 8bit text in MIME parts. Comparator and
4942 require names are now matched exactly. I enabled the subaddress
4943 extension, but it is not well tested yet (read: it works for me)."
4945 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4946 rework some of the code of TK/09 above to avoid the hardwired use of
4947 "%lld" and "long long". Replaced the call to snprintf() with a call to
4950 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4951 records point to non-existent hosts", "retry timeout exceeded", and
4952 "retry time not reached for any host after a long failure period".
4954 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4955 experimental DomainKeys support:
4957 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4958 (2) On an error such as an illegally used "control", the wrong name for
4959 the control was given.
4961 These problems did NOT occur unless DomainKeys support was compiled.
4963 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4965 PH/24 Added ${if match_ip condition.
4967 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4968 databases so that it will be absolutely obvious if a crash occurs in the
4969 DB library. This is a regular occurrence (often caused by mis-matched
4972 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4973 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4974 picked up on a user's system that detects such things. There doesn't seem
4975 to be a gcc warning option for this - only an attribute that has to be
4976 put on the function's prototype. It seems that in Fedora Core 4 they have
4977 set this on a number of new functions. No doubt there will be more in due
4980 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4981 specify a transport. However, if an address that was verified by such a
4982 router was the subject of a callout, Exim crashed because it tried to
4983 read the rcpt_include_affixes from the non-existent transport. Now it
4984 just assumes that the setting of that option is false. This bug was
4985 introduced by 4.51/PH/31.
4987 PH/28 Changed -d+all to exclude +memory, because that information is very
4988 rarely of interest, but it makes the output a lot bigger. People tend to
4989 do -d+all out of habit.
4991 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4992 code in os-type was giving problems when libc.so lives in lib64, like on
4995 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4996 aren't the modern standard, and it seems that some systems' include files
4997 don't always have them. Exim was already checking for some of the newer
4998 ones like T_AAAA, and defining it itself. I've added checks for all the
4999 record types that Exim uses.
5001 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5002 not automatically generating a new one, as it is supposed to. This
5003 prevented TLS from working. If the file did exist, but contained invalid
5004 data, a new version was generated, as expected. It was only the case of a
5005 non-existent file that was broken.
5007 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5008 with a change in libdomainkeys > 0.64.
5010 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5011 from DNS. If the selector record carries the flag, it now has
5012 precedence over the domain-wide flag.
5014 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5016 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5017 the use of a port name that isn't defined in /etc/services) occurred, the
5018 message was deferred as in a normal delivery, and thus remained on the
5019 spool, instead of being failed because of the mua_wrapper setting. This
5020 is now fixed, and I tidied up some of the mua_wrapper messages at the
5023 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5024 the messages in a hash of arrays rather than using individual hashes.
5025 This is a bit cleaner and results in dramatic memory savings, albeit
5026 at a slight CPU cost.
5028 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5029 as requested by Marc Sherman.
5031 SC/10 Eximstats: added histograms for user specified patterns as requested
5034 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5036 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5037 fopen() in the content-scanning modules that did not already have it.
5043 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5044 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5046 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5048 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5050 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5051 to test Sieve filters that use "vacation".
5053 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5054 that changes the way the GnuTLS parameters are stored in the cache file.
5055 The new format can be generated externally. For backward compatibility,
5056 if the data in the cache doesn't make sense, Exim assumes it has read an
5057 old-format file, and it generates new data and writes a new file. This
5058 means that you can't go back to an older release without removing the
5061 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5062 work if there are any delivery delays because "one_time" forces the
5063 parent to be marked "delivered", so its unseen clone is never tried
5064 again. For this reason, Exim now forbids the simultaneous setting of
5067 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5068 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5069 entry if you want to know the details. The fix, however, neglected to
5070 consider the case when local delivery batching is involved. The test for
5071 "previously delivered" was not happening when checking to see if an
5072 address could be batched with a previous (undelivered) one; under
5073 certain circumstances this could lead to multiple deliveries to the same
5076 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5077 in its include files, and this causes problems building Exim.
5079 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5080 header_syntax) but Exim was just ignoring anything given after a slash.
5081 In particular, this caused confusion with an attempt to use "verify =
5082 reverse_host_lookup/defer_ok". An error is now given when options are
5083 supplied for verify items that do not have them. (Maybe reverse_host_
5084 lookup should have a defer_ok option, but that's a different point.)
5086 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5087 defined by RFC 821) to 2048, because there were problems with some AUTH
5088 commands, and RFC 1869 says the size should be increased for extended
5089 SMTP commands that take arguments.
5091 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5094 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5095 "unknown" error; now it says that the functionality isn't in the binary.
5097 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5098 an address' error message when a string expansion fails (syntax or
5099 whatever). Otherwise the password may appear in the log. Following change
5100 PH/42 below, there is no longer a chance of it appearing in a bounce
5103 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5105 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5106 of its addresses was ever tried. (Bugzilla bug #2.)
5108 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5109 the result incorrectly in the debug output. (It correctly added a newline
5110 to what was transported.)
5112 TF/01 Added $received_time.
5114 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5115 commented out examples of how to interface to a virus scanner and to
5116 SpamAssassin. Also added commented examples of av_scanner and
5117 spamd_address settings.
5119 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5120 and controls are allowed in which ACLs. There were a couple of minor
5121 errors. Some of the entries in the conditions table (which is a table of
5122 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5123 negation of where the condition IS allowed.
5125 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5127 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5128 header file does not have a version number, so I've had to invent a new
5129 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5130 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5131 radiusclient), but it was contributed by a Radius user.
5133 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5134 files or directories, for OpenSSL.
5136 PH/19 When an Exim process that is running as root has to create an Exim log
5137 file, it does so in a subprocess that runs as exim:exim so as to get the
5138 ownership right at creation (otherwise, other Exim processes might see
5139 the file with the wrong ownership). There was no test for failure of this
5140 fork() call, which would lead to the process getting stuck as it waited
5141 for a non-existent subprocess. Forks do occasionally fail when resources
5142 run out. I reviewed all the other calls to fork(); they all seem to check
5145 PH/20 When checking for unexpected SMTP input at connect time (before writing
5146 the banner), Exim was not dealing correctly with a non-positive return
5147 from the read() function. If the client had disconnected by this time,
5148 the result was a log entry for a synchronization error with an empty
5149 string after "input=" when read() returned zero. If read() returned -1
5150 (an event I could not check), uninitialized data bytes were printed.
5151 There were reports of junk text (parts of files, etc) appearing after
5154 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5156 PH/22 Added support for macro redefinition, and (re)definition in between
5157 driver and ACL definitions.
5159 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5160 forgetting to use the resulting value; it was using the unexpanded value.
5162 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5163 hadn't been configured. The fix is from Juergen Kreileder, who
5164 understands it better than I do:
5166 "Here's what I see happening with three configured cyrus_sasl
5167 authenticators configured (plain, login, cram-md5):
5169 On startup auth_cyrus_sasl_init() gets called for each of these.
5170 This means three calls to sasl_listmech() without a specified mech_list.
5171 => SASL tests which mechs of all available mechs actually work
5172 => three warnings about OTP not working
5173 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5175 With the patch, sasl_listmech() also gets called three times. But now
5176 SASL's mech_list option is set to the server_mech specified in the the
5177 authenticator. Or in other words, the answer from sasl_listmech()
5178 gets limited to just the mech you're testing for (which is different
5180 => the return list contains just 'plain' or 'login', 'cram-md5' or
5181 nothing depending on the value of ob->server_mech.
5183 I've just tested the patch: Authentication still works fine,
5184 unavailable mechs specified in the exim configuration are still
5185 caught, and the auth.log warnings about OTP are gone."
5187 PH/25 When debugging is enabled, the contents of the command line are added
5188 to the debugging output, even when log_selector=+arguments is not
5191 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5192 answer is "GNU", and only if the return is "GNU/something" is the answer
5195 PH/27 $acl_verify_message is now set immediately after the failure of a
5196 verification in an ACL, and so is available in subsequent modifiers. In
5197 particular, the message can be preserved by coding like this:
5199 warn !verify = sender
5200 set acl_m0 = $acl_verify_message
5202 Previously, $acl_verify_message was set only while expanding "message"
5203 and "log_message" when a very denied access.
5205 PH/28 Modified OS/os.c-Linux with
5207 -#ifndef OS_LOAD_AVERAGE
5208 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5210 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5211 nomenclature these days.)
5213 PH/29 Installed patch from the Sieve maintainer that adds the options
5214 sieve_useraddress and sieve_subaddress to the redirect router.
5216 PH/30 In these circumstances:
5217 . Two addresses routed to the same list of hosts;
5218 . First host does not offer TLS;
5219 . First host accepts first address;
5220 . First host gives temporary error to second address;
5221 . Second host offers TLS and a TLS session is established;
5222 . Second host accepts second address.
5223 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5224 and peerdn, if requested) that were in fact used only for the second
5227 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5228 attention to any local part prefix or suffix that was matched by the
5229 router that accepted the address. It now behaves in the same way as it
5230 does for delivery: the affixes are removed from the local part unless
5231 rcpt_include_affixes is set on the transport.
5233 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5234 timeout during the DATA phase of an incoming message.
5236 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5237 applied a patch sanctioned by the Sieve maintainer.
5239 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5240 the uid or gid is negative. A case of a negative gid caused this to be
5241 noticed. The fix allows for either to be negative.
5243 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5244 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5247 PH/36 The change PH/12 above was broken. Fixed it.
5249 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5250 the grounds that routing the same address twice would always produce the
5251 same answer. This might have been true once, but it is certainly no
5252 longer true now. Routing a child address may depend on the previous
5253 routing that produced that child. Some complicated redirection strategies
5254 went wrong when messages had multiple recipients, and made Exim's
5255 behaviour dependent on the order in which the addresses were given.
5257 I have moved the duplicate checking until after the routing is complete.
5258 Exim scans the addresses that are assigned to local and remote
5259 transports, and removes any duplicates. This means that more work will be
5260 done, as duplicates will always all be routed, but duplicates are
5261 presumably rare, so I don't expect this is of any significance.
5263 For deliveries to pipes, files, and autoreplies, the duplicate checking
5264 still happens during the routing process, since they are not going to be
5267 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5268 It corrects a timeout issue with spamd. This is Ian's comment: "The
5269 background is that sometimes spamd either never reads data from a
5270 connection it has accepted, or it never writes response data. The exiscan
5271 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5272 blindly assumes that writes won't block so it may never time out."
5274 PH/39 Allow G after quota size as well as K and M.
5276 PH/40 The value set for $authenticated_id in an authenticator may not contain
5277 binary zeroes or newlines because the value is written to log lines and
5278 to spool files. There was no check on this. Now the value is run through
5279 the string_printing() function so that such characters are converted to
5280 printable escape sequences.
5282 PH/41 $message_linecount is a new variable that contains the total number of
5283 lines in the message. Compare $body_linecount, which is the count for the
5286 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5287 bounce and delay warning messages, except in certain special cases, which
5290 (a) An SMTP error message from a remote host;
5291 (b) A message specified in a :fail: redirection;
5292 (c) A message specified in a "fail" command in a system filter;
5293 (d) A message specified in a FAIL return from the queryprogram router;
5294 (e) A message specified by the cannot_route_message router option.
5296 In these cases only, Exim does include the error details in bounce and
5297 warning messages. There are also a few cases where bland messages such
5298 as "unrouteable address" or "local delivery error" are given.
5300 PH/43 $value is now also set for the "else" part of a ${run expansion.
5302 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5303 being worked on, but at least Exim now implements the latest version to
5306 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5307 process to complete was treated as a delivery failure, a timeout while
5308 writing the message to the pipe was logged, but erroneously treated as a
5309 successful delivery. Such timeouts include transport filter timeouts. For
5310 consistency with the overall process timeout, these timeouts are now
5311 treated as errors, giving rise to delivery failures by default. However,
5312 there is now a new Boolean option for the pipe transport called
5313 timeout_defer, which, if set TRUE, converts the failures into defers for
5314 both kinds of timeout. A transport filter timeout is now identified in
5317 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5318 systems where "make" and "gmake" are different, calling "gmake" at top
5319 level broke things. I've arranged for the value of $(MAKE) to be passed
5320 from the Makefile to this script so that it can call the same version of
5324 A note about Exim versions 4.44 and 4.50
5325 ----------------------------------------
5327 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5328 changes of various kinds. As a consequence, a big documentation update was
5329 needed. This delayed the release for rather longer than seemed good, especially
5330 in the light of a couple of (minor) security issues. Therefore, the changes
5331 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5332 release. So 4.44 and 4.50 are in effect two different branches that both start
5335 I have left the 4.50 change log unchanged; it contains all the changes since
5336 4.43. The change log for 4.44 is below; many of its items are identical to
5337 those for 4.50. This seems to be the most sensible way to preserve the
5338 historical information.
5344 1. Minor wording change to the doc/README.SIEVE file.
5346 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5347 computation of the current number of files was incorrect.
5349 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5350 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5353 4. Give more explanation in the error message when the command for a transport
5354 filter fails to execute.
5356 5. There are several places where Exim runs a non-Exim command in a
5357 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5358 was being done only for the command run by the queryprogram router. It is
5359 now done for all such subprocesses. The other cases are: ${run, transport
5360 filters, and the commands run by the lmtp and pipe transports.
5362 6. Added CONFIGURE_GROUP build-time option.
5364 7. Some older OS have a limit of 256 on the maximum number of file
5365 descriptors. Exim was using setrlimit() to set 1000 as a large value
5366 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5367 systems. I've change it so that if it can't get 1000, it tries for 256.
5369 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5370 was an oversight, and furthermore, ever since the addition of extra
5371 controls (e.g. 4.43/32), the checks on when to allow different forms of
5372 "control" were broken. There should now be diagnostics for all cases when a
5373 control that does not make sense is encountered.
5375 9. Added the /retain_sender option to "control=submission".
5377 10. $recipients is now available in the predata ACL (oversight).
5379 11. Tidy the search cache before the fork to do a delivery from a message
5380 received from the command line. Otherwise the child will trigger a lookup
5381 failure and thereby defer the delivery if it tries to use (for example) a
5382 cached ldap connection that the parent has called unbind on.
5384 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5385 of $address_data from the recipient verification was clobbered by the
5386 sender verification.
5388 13. The value of address_data from a sender verification is now available in
5389 $sender_address_data in subsequent conditions in the ACL statement.
5391 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5393 15. Added a new option "connect=<time>" to callout options, to set a different
5396 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5397 was its contents. (It was OK if the option was not defined at all.)
5399 17. A "Completed" log line is now written for messages that are removed from
5400 the spool by the -Mrm option.
5402 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5403 information about exactly what failed.
5405 19. Added -dd to debug only the daemon process.
5407 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5408 handles timeouts, both on the server side and network timeouts. Renamed the
5409 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5411 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5412 It is now set to "smtps".
5414 22. $host_address is now set to the target address during the checking of
5415 ignore_target_hosts.
5417 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5418 being passed; this would have caused $sender_host_name to have been used if
5419 matching the list had actually called for a host name (not very likely,
5420 since this list is usually IP addresses). A host name is now passed as
5423 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5424 code that specifies a non-restarting handler (typically sigaction() in
5425 modern systems) in an attempt to fix a rare and obscure crash bug.
5427 25. Narrowed the window for a race in the daemon that could cause it to ignore
5428 SIGCHLD signals. This is not a major problem, because they are used only to
5429 wake it up if nothing else does.
5431 26. A malformed maildirsize file could cause Exim to calculate negative values
5432 for the mailbox size or file count. Odd effects could occur as a result.
5433 The maildirsize information is now recalculated if the size or filecount
5436 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5437 support for a long time. Removed HAVE_SYS_VFS_H.
5439 28. Installed the latest version of exipick from John Jetmore.
5441 29. In an address list, if the pattern was not a regular expression, an empty
5442 subject address (from a bounce message) matched only if the pattern was an
5443 empty string. Non-empty patterns were not even tested. This was the wrong
5444 because it is perfectly reasonable to use an empty address as part of a
5445 database query. An empty address is now tested by patterns that are
5446 lookups. However, all the other forms of pattern expect the subject to
5447 contain a local part and a domain, and therefore, for them, an empty
5448 address still always fails if the pattern is not itself empty.
5450 30. Exim went into a mad DNS loop when attempting to do a callout where the
5451 host was specified on an smtp transport, and looking it up yielded more
5452 than one IP address.
5454 31. Re-factored the code for checking spool and log partition space into a
5455 function that finds that data and another that does the check. The former
5456 is then used to implement four new variables: $spool_space, $log_space,
5457 $spool_inodes, and $log_inodes.
5459 32. The RFC2047 encoding function was originally intended for short strings
5460 such as real names; it was not keeping to the 75-character limit for
5461 encoded words that the RFC imposes. It now respects the limit, and
5462 generates multiple encoded words if necessary. To be on the safe side, I
5463 have increased the buffer size for the ${rfc2047: expansion operator from
5466 33. It is now permitted to omit both strings after an "if" condition; if the
5467 condition is true, the result is "true". As before, when the second string
5468 is omitted, a false condition yields an empty string. This makes it less
5469 cumbersome to write custom ACL and router conditions.
5471 34. Failure to deliver a bounce message always caused it to be frozen, even if
5472 there was an errors_to setting on the router. The errors_to setting is now
5475 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5476 canonical form (fully expanded) before being placed in
5477 $sender_host_address.
5479 36. The table in the code that translates DNS record types into text (T_A to
5480 "A" for instance) was missing entries for NS and CNAME. It is just possible
5481 that this could have caused confusion if both these types were looked up
5482 for the same domain, because the text type is used as part of Exim's
5483 per-process caching. But the chance of anyone hitting this buglet seems
5486 37. The dnsdb lookup has been extended in a number of ways.
5488 (1) There is a new type, "zns", which walks up the domain tree until it
5489 finds some nameserver records. It should be used with care.
5491 (2) There is a new type, "mxh", which is like "mx" except that it returns
5492 just the host names, not the priorities.
5494 (3) It is now possible to give a list of domains (or IP addresses) to be
5495 looked up. The behaviour when one of the lookups defers can be
5496 controlled by a keyword.
5498 (4) It is now possible to specify the separator character for use when
5499 multiple records are returned.
5501 38. The dnslists ACL condition has been extended: it is now possible to supply
5502 a list of IP addresses and/or domains to be looked up in a particular DNS
5505 39. Added log_selector=+queue_time_overall.
5507 40. When running the queue in the test harness, wait just a tad after forking a
5508 delivery process, to get repeatability of debugging output.
5510 41. Include certificate and key file names in error message when GnuTLS fails
5511 to set them up, because the GnuTLS error message doesn't include the name
5512 of the failing file when there is a problem reading it.
5514 42. Allow both -bf and -bF in the same test run.
5516 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5518 44. The "Exiscan patch" is now merged into the mainline Exim source.
5520 45. Sometimes the final signoff response after QUIT could fail to get
5521 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5522 < 0 before doing a fflush(). This bug looks as though it goes back to the
5523 introduction of TLS in release 3.20, but "sometimes" must have been rare
5524 because the tests only now provoked it.
5526 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5527 (this can affect the format of dates).
5529 47. exim_tidydb, when checking for the continued existence of a message for
5530 which it has found a message-specific retry record, was not finding
5531 messages that were in split spool directories. Consequently, it was
5532 deleting retry records that should have stayed in existence.
5534 48. Steve fixed some bugs in eximstats.
5536 49. The SPA authentication driver was not abandoning authentication and moving
5537 on to the next authenticator when an expansion was forced to fail,
5538 contradicting the general specification for all authenticators. Instead it
5539 was generating a temporary error. It now behaves as specified.
5541 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5542 (the order specifies the preference for clients). The order is now AES256,
5543 AES128, 3DES, ARCFOUR128.
5545 51. Small patch to Sieve code - explicitly set From: when generating an
5548 52. Exim crashed if a remote delivery caused a very long error message to be
5549 recorded - for instance if somebody sent an entire SpamAssassin report back
5550 as a large number of 550 error lines. This bug was coincidentally fixed by
5551 increasing the size of one of Exim's internal buffers (big_buffer) that
5552 happened as part of the Exiscan merge. However, to be on the safe side, I
5553 have made the code more robust (and fixed the comments that describe what
5556 53. Now that there can be additional text after "Completed" in log lines (if
5557 the queue_time_overall log selector is set), a one-byte patch to exigrep
5558 was needed to allow it to recognize "Completed" as not the last thing in
5561 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5562 patch that reportedly fixes this has been added. I am not expert enough to
5563 create a test for it. This is what the patch creator wrote:
5565 "I found a little strange behaviour of ldap code when working with
5566 Windows 2003 AD Domain, where users was placed in more than one
5567 Organization Units. When I tried to give exim partial DN, the exit code
5568 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5569 But simultaneously result of request was absolutely normal ldap result,
5570 so I produce this patch..."
5572 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5573 REFERENCE, so I have modified the code to exclude the patch when that macro
5576 55. Some experimental protocols are using DNS PTR records for new purposes. The
5577 keys for these records are domain names, not reversed IP addresses. The
5578 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5579 leaves it alone. Component reversal etc. now happens only for IP addresses.
5582 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5584 57. Double the size of the debug message buffer (to 2048) so that more of very
5585 long debug lines gets shown.
5587 58. The exicyclog utility now does better if the number of log files to keep
5588 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5590 59. Two changes related to the smtp_active_hostname option:
5592 (1) $smtp_active_hostname is now available as a variable.
5593 (2) The default for smtp_banner uses $smtp_active_hostname instead
5594 of $primary_hostname.
5596 60. The host_aton() function is supposed to be passed a string that is known
5597 to be a valid IP address. However, in the case of IPv6 addresses, it was
5598 not checking this. This is a hostage to fortune. Exim now panics and dies
5599 if the condition is not met. A case was found where this could be provoked
5600 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5601 components; fortuitously, this particular loophole had already been fixed
5602 by change 4.50/55 above.
5604 If there are any other similar loopholes, the new check in host_aton()
5605 itself should stop them being exploited. The report I received stated that
5606 data on the command line could provoke the exploit when Exim was running as
5607 exim, but did not say which command line option was involved. All I could
5608 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5609 running as the user.
5612 61. There was a buffer overflow vulnerability in the SPA authentication code
5613 (which came originally from the Samba project). I have added a test to the
5614 spa_base64_to_bits() function which I hope fixes it.
5617 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5618 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5620 63. The daemon start-up calls getloadavg() while still root for those OS that
5621 need the first call to be done as root, but it missed one case: when
5622 deliver_queue_load_max is set with deliver_drop_privilege. This is
5623 necessary for the benefit of the queue runner, because there is no re-exec
5624 when deliver_drop_privilege is set.
5626 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5627 This has been fixed.
5629 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5630 was in use, was not putting the data itself into the right store pool;
5631 consequently, it could be overwritten for a subsequent message in the same
5632 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5635 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5637 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5638 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5639 the function were treating the return as a boolean value, which happened to
5640 work because 0=false and not-0=true, but is not correct code.
5642 68. The host_aton() function was not handling scoped IPv6 addresses (those
5643 with, for example, "%eth0" on the end) correctly.
5645 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5646 negated items (that is, ~something) in unsigned ints. Some compilers
5647 apparently mutter when there is no cast.
5649 70. If an address verification called from an ACL failed, and did not produce a
5650 user-specific message (i.e. there was only a "system" message), nothing was
5651 put in $acl_verify_message. In this situation, it now puts the system
5654 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5655 session; change 4.31/43 added the unwanted input to the log line - except
5656 that it did not do this in the start of session case. It now does.
5658 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5659 This is wrong and can cause the other end to generate a synchronization
5660 error if it is another Exim or anything else that does the synchronization
5661 check. A QUIT command is no longer sent after a timeout.
5663 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5664 during host lookups.
5666 74. The defer_ok option of callout verification was not working if it was used
5667 when verifying addresses in header lines, that is, for this case:
5669 verify = header_sender/callout=defer_ok
5671 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5672 those file descriptors could be used for SMTP connections. If anything
5673 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5674 could be sent to the SMTP client, causing chaos. The daemon now opens
5675 stdin, stdout, and stderr to /dev/null when it puts itself into the
5678 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5679 log by default. The user can override this with suitable Perl magic.
5681 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5682 the log message when discard triggers, was not working for the DATA ACL or
5683 for the non-SMTP ACL.
5685 78. Error message wording change in sieve.c.
5687 79. If smtp_accept_max_per_host was set, the number of connections could be
5688 restricted to fewer than expected, because the daemon was trying to set up
5689 a new connection before checking whether the processes handling previous
5690 connections had finished. The check for completed processes is now done
5691 earlier. On busy systems, this bug wouldn't be noticed because something
5692 else would have woken the daemon, and it would have reaped the completed
5695 80. If a message was submitted locally by a user whose login name contained one
5696 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5697 It caused a spool format error. I have fixed the spool reading code. A
5698 related problem was that the "from" clause in the Received: line became
5699 illegal because of the space(s). It is now covered by ${quote_local_part.
5701 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5704 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5706 83. Patch from Sieve maintainer to fix unterminated string problem in
5707 "vacation" handling.
5709 84. Some minor changes to the Linux configuration files to help with other
5710 OS variants using glibc.
5712 85. One more patch for Sieve to update vacation handling to latest spec.
5715 ----------------------------------------------------
5716 See the note above about the 4.44 and 4.50 releases.
5717 ----------------------------------------------------
5723 1. Change 4.43/35 introduced a bug that caused file counts to be
5724 incorrectly computed when quota_filecount was set in an appendfile
5727 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5728 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5731 3. Give more explanation in the error message when the command for a transport
5732 filter fails to execute.
5734 4. There are several places where Exim runs a non-Exim command in a
5735 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5736 was being done only for the command run by the queryprogram router. It is
5737 now done for all such subprocesses. The other cases are: ${run, transport
5738 filters, and the commands run by the lmtp and pipe transports.
5740 5. Some older OS have a limit of 256 on the maximum number of file
5741 descriptors. Exim was using setrlimit() to set 1000 as a large value
5742 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5743 systems. I've change it so that if it can't get 1000, it tries for 256.
5745 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5746 was an oversight, and furthermore, ever since the addition of extra
5747 controls (e.g. 4.43/32), the checks on when to allow different forms of
5748 "control" were broken. There should now be diagnostics for all cases when a
5749 control that does not make sense is encountered.
5751 7. $recipients is now available in the predata ACL (oversight).
5753 8. Tidy the search cache before the fork to do a delivery from a message
5754 received from the command line. Otherwise the child will trigger a lookup
5755 failure and thereby defer the delivery if it tries to use (for example) a
5756 cached ldap connection that the parent has called unbind on.
5758 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5759 of $address_data from the recipient verification was clobbered by the
5760 sender verification.
5762 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5763 was its contents. (It was OK if the option was not defined at all.)
5765 11. A "Completed" log line is now written for messages that are removed from
5766 the spool by the -Mrm option.
5768 12. $host_address is now set to the target address during the checking of
5769 ignore_target_hosts.
5771 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5772 being passed; this would have caused $sender_host_name to have been used if
5773 matching the list had actually called for a host name (not very likely,
5774 since this list is usually IP addresses). A host name is now passed as
5777 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5778 code that specifies a non-restarting handler (typically sigaction() in
5779 modern systems) in an attempt to fix a rare and obscure crash bug.
5781 15. Narrowed the window for a race in the daemon that could cause it to ignore
5782 SIGCHLD signals. This is not a major problem, because they are used only to
5783 wake it up if nothing else does.
5785 16. A malformed maildirsize file could cause Exim to calculate negative values
5786 for the mailbox size or file count. Odd effects could occur as a result.
5787 The maildirsize information is now recalculated if the size or filecount
5790 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5791 support for a long time. Removed HAVE_SYS_VFS_H.
5793 18. Updated exipick to current release from John Jetmore.
5795 19. Allow an empty sender to be matched against a lookup in an address list.
5796 Previously the only cases considered were a regular expression, or an
5799 20. Exim went into a mad DNS lookup loop when doing a callout where the
5800 host was specified on the transport, if the DNS lookup yielded more than
5803 21. The RFC2047 encoding function was originally intended for short strings
5804 such as real names; it was not keeping to the 75-character limit for
5805 encoded words that the RFC imposes. It now respects the limit, and
5806 generates multiple encoded words if necessary. To be on the safe side, I
5807 have increased the buffer size for the ${rfc2047: expansion operator from
5810 22. Failure to deliver a bounce message always caused it to be frozen, even if
5811 there was an errors_to setting on the router. The errors_to setting is now
5814 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5815 canonical form (fully expanded) before being placed in
5816 $sender_host_address.
5818 24. Updated eximstats to version 1.33
5820 25. Include certificate and key file names in error message when GnuTLS fails
5821 to set them up, because the GnuTLS error message doesn't include the name
5822 of the failing file when there is a problem reading it.
5824 26. Expand error message when OpenSSL has problems setting up cert/key files.
5827 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5828 (this can affect the format of dates).
5830 28. exim_tidydb, when checking for the continued existence of a message for
5831 which it has found a message-specific retry record, was not finding
5832 messages that were in split spool directories. Consequently, it was
5833 deleting retry records that should have stayed in existence.
5835 29. eximstats updated to version 1.35
5836 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5837 1.35 - bugfix such that pie charts by volume are generated correctly
5839 30. The SPA authentication driver was not abandoning authentication and moving
5840 on to the next authenticator when an expansion was forced to fail,
5841 contradicting the general specification for all authenticators. Instead it
5842 was generating a temporary error. It now behaves as specified.
5844 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5845 (the order specifies the preference for clients). The order is now AES256,
5846 AES128, 3DES, ARCFOUR128.
5848 31. Small patch to Sieve code - explicitly set From: when generating an
5851 32. Exim crashed if a remote delivery caused a very long error message to be
5852 recorded - for instance if somebody sent an entire SpamAssassin report back
5853 as a large number of 550 error lines. This bug was coincidentally fixed by
5854 increasing the size of one of Exim's internal buffers (big_buffer) that
5855 happened as part of the Exiscan merge. However, to be on the safe side, I
5856 have made the code more robust (and fixed the comments that describe what
5859 33. Some experimental protocols are using DNS PTR records for new purposes. The
5860 keys for these records are domain names, not reversed IP addresses. The
5861 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5862 leaves it alone. Component reversal etc. now happens only for IP addresses.
5865 34. The host_aton() function is supposed to be passed a string that is known
5866 to be a valid IP address. However, in the case of IPv6 addresses, it was
5867 not checking this. This is a hostage to fortune. Exim now panics and dies
5868 if the condition is not met. A case was found where this could be provoked
5869 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5870 components; fortuitously, this particular loophole had already been fixed
5871 by change 4.50/55 or 4.44/33 above.
5873 If there are any other similar loopholes, the new check in host_aton()
5874 itself should stop them being exploited. The report I received stated that
5875 data on the command line could provoke the exploit when Exim was running as
5876 exim, but did not say which command line option was involved. All I could
5877 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5878 running as the user.
5881 35. There was a buffer overflow vulnerability in the SPA authentication code
5882 (which came originally from the Samba project). I have added a test to the
5883 spa_base64_to_bits() function which I hope fixes it.
5886 36. The daemon start-up calls getloadavg() while still root for those OS that
5887 need the first call to be done as root, but it missed one case: when
5888 deliver_queue_load_max is set with deliver_drop_privilege. This is
5889 necessary for the benefit of the queue runner, because there is no re-exec
5890 when deliver_drop_privilege is set.
5892 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5893 was in use, was not putting the data itself into the right store pool;
5894 consequently, it could be overwritten for a subsequent message in the same
5895 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5898 38. Sometimes the final signoff response after QUIT could fail to get
5899 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5900 < 0 before doing a fflush(). This bug looks as though it goes back to the
5901 introduction of TLS in release 3.20, but "sometimes" must have been rare
5902 because the tests only now provoked it.
5908 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5909 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5910 Now it returns an integer. A number of calls were still expecting a T/F
5911 return. Fortuitously, in all cases, the tests worked in OK situations,
5912 which is the norm. However, things would have gone wrong on any write
5913 failures on the smtp file descriptor. This function is used when sending
5914 messages over SMTP and also when doing verify callouts.
5916 2. When Exim is called to do synchronous delivery of a locally submitted
5917 message (the -odf or -odi options), it no longer closes stderr before doing
5920 3. Implemented the mua_wrapper option.
5922 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5924 5. Implemented the functions header_remove(), header_testname(),
5925 header_add_at_position(), and receive_remove_recipient(), and exported them
5928 6. If an ACL "warn" statement specified the addition of headers, Exim already
5929 inserted X-ACL-Warn: at the start if there was no header name. However, it
5930 was not making this test for the second and subsequent header lines if
5931 there were newlines in the string. This meant that an invalid header could
5932 be inserted if Exim was badly configured.
5934 7. Allow an ACL "warn" statement to add header lines at the start or after all
5935 the Received: headers, as well as at the end.
5937 8. Added the rcpt_4xx retry error code.
5939 9. Added postmaster_mailfrom=xxx to callout verification option.
5941 10. Added mailfrom=xxxx to the callout verification option, for verify=
5944 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5945 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5947 12. Inserted some casts to stop certain compilers warning when using pointer
5948 differences as field lengths or precisions in printf-type calls (mostly
5949 affecting debugging statements).
5951 13. Added optional readline() support for -be (dynamically loaded).
5953 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5954 same clock tick as a message's arrival, so that its received time was the
5955 same as the "first fail" time on the retry record, and that message
5956 remained on the queue past the ultimate address timeout, every queue runner
5957 would try a delivery (because it was past the ultimate address timeout) but
5958 after another failure, the ultimate address timeout, which should have then
5959 bounced the address, did not kick in. This was a "< instead of <=" error;
5960 in most cases the first failure would have been in the next clock tick
5961 after the received time, and all would be well.
5963 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5964 being recognized when the domain list was tested by the match_domain
5965 condition in an expansion string.
5967 16. Added the ${str2b64: operator.
5969 17. Exim was always calling setrlimit() to set a large limit for the number of
5970 processes, without checking whether the existing limit was already
5971 adequate. (It did check for the limit on file descriptors.) Furthermore,
5972 errors from getrlimit() and setrlimit() were being ignored. Now they are
5973 logged to the main and panic logs, but Exim does carry on, to try to do its
5974 job under whatever limits there are.
5976 18. Imported PCRE 5.0.
5978 19. Trivial typo in log message " temporarily refused connection" (the leading
5981 20. If the log selector return_path_on_delivery was set and an address was
5982 redirected to /dev/null, the delivery process crashed because it assumed
5983 that a return path would always be set for a "successful" delivery. In this
5984 case, the whole delivery is bypassed as an optimization, and therefore no
5987 21. Internal re-arrangement: the function for sending a challenge and reading
5988 a response while authentication was assuming a zero-terminated challenge
5989 string. It's now changed to take a pointer and a length, to allow for
5990 binary data in such strings.
5992 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5994 23. Exim was not respecting finduser_retries when seeking the login of the
5995 uid under which it was called; it was always trying 10 times. (The default
5996 setting of finduser_retries is zero.) Also, it was sleeping after the final
5997 failure, which is pointless.
5999 24. Implemented tls_on_connect_ports.
6001 25. Implemented acl_smtp_predata.
6003 26. If the domain in control=submission is set empty, Exim assumes that the
6004 authenticated id is a complete email address when it generates From: or
6005 Sender: header lines.
6007 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6008 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6009 chown and chgrp in /bin and hostname in /usr/bin.
6011 28. Exim was keeping the "process log" file open after each use, just as it
6012 does for the main log. This opens the possibility of it remaining open for
6013 long periods when the USR1 signal hits a daemon. Occasional processlog
6014 errors were reported, that could have been caused by this. Anyway, it seems
6015 much more sensible not to leave this file open at all, so that is what now
6018 29. The long-running daemon process does not normally write to the log once it
6019 has entered its main loop, and it closes the log before doing so. This is
6020 so that log files can straightforwardly be renamed and moved. However,
6021 there are a couple of unusual error situations where the daemon does write
6022 log entries, and I had neglected to close the log afterwards.
6024 30. The text of an SMTP error response that was received during a remote
6025 delivery was being truncated at 512 bytes. This is too short for some of
6026 the long messages that one sometimes sees. I've increased the limit to
6029 31. It is now possible to make retry rules that apply only when a message has a
6030 specific sender, in particular, an empty sender.
6032 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6033 it possible to be selective about when SMTP synchronization is enforced.
6035 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6037 32. Implemented hosts_connection_nolog.
6039 33. Added an ACL for QUIT.
6041 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6044 35. Added mailbox_size and mailbox_filecount to appendfile.
6046 36. Added control = no_multiline_responses to ACLs.
6048 37. There was a bug in the logic of the code that waits for the clock to tick
6049 in the case where the clock went backwards by a substantial amount such
6050 that the microsecond fraction of "now" was more than the microsecond
6051 fraction of "then" (but the whole seconds number was less).
6053 38. Added support for the libradius Radius client library this is found on
6054 FreeBSD (previously only the radiusclient library was supported).
6060 1. When certain lookups returned multiple values in the form name=value, the
6061 quoting of the values was not always being done properly. Specifically:
6062 (a) If the value started with a double quote, but contained no whitespace,
6064 (b) If the value contained whitespace other than a space character (i.e.
6065 tabs or newlines or carriage returns) it was not quoted.
6066 This fix has been applied to the mysql and pgsql lookups by writing a
6067 separate quoting function and calling it from the lookup code. The fix
6068 should probably also be applied to nisplus, ibase and oracle lookups, but
6069 since I cannot test any of those, I have not disturbed their existing code.
6071 2. A hit in the callout cache for a specific address caused a log line with no
6072 reason for rejecting RCPT. Now it says "Previous (cached) callout
6073 verification failure".
6075 3. There was an off-by-one bug in the queryprogram router. An over-long
6076 return line was truncated at 256 instead of 255 characters, thereby
6077 overflowing its buffer with the terminating zero. As well as fixing this, I
6078 have increased the buffer size to 1024 (and made a note to document this).
6080 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6081 when Exim is waiting for an SMTP response from a remote server, Exim
6082 restarts its select() call on the socket, thereby resetting its timeout.
6083 This is not a problem when such interrupts are rare. Somebody set up a cron
6084 job to run exiwhat every 2 minutes, which is less than the normal select()
6085 timeout (5 or 10 minutes). This meant that the select() timeout never
6086 kicked in because it was always reset. I have fixed this by comparing the
6087 time when an interrupt arrives with the time at the start of the first call
6088 to select(). If more time than the timeout has elapsed, the interrupt is
6089 treated as a timeout.
6091 5. Some internal re-factoring in preparation for the addition of Sieve
6092 extensions (by MH). In particular, the "personal" test is moved to a
6093 separate function, and given an option for scanning Cc: and Bcc: (which is
6094 not set for Exim filters).
6096 6. When Exim created an email address using the login of the caller as the
6097 local part (e.g. when creating a From: or Sender: header line), it was not
6098 quoting the local part when it contained special characters such as @.
6100 7. Installed new OpenBSD configuration files.
6102 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6103 try to make them clearer.
6105 9. Callout options, other than the timeout value, were being ignored when
6106 verifying sender addresses in header lines. For example, when using
6108 verify = header_sender/callout=no_cache
6110 the cache was (incorrectly) being used.
6112 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6113 only the Cygwin environment.
6115 11. When return_path_on_delivery was set as a log selector, if different remote
6116 addresses in the same message used different return paths and parallel
6117 remote delivery occurred, the wrong values would sometimes be logged.
6118 (Whenever a remote delivery process finished, the return path value from
6119 the most recently started remote delivery process was logged.)
6121 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6122 lines when AUTH and/or TLS are in use. This is the "received protocol"
6123 field. Exim used to use "asmtp" for authenticated SMTP, without any
6124 indication (in the protocol name) for TLS use. Now it follows the RFC and
6125 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6126 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6127 names appear in log lines as well as in Received: header lines.
6129 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6130 extensions, and comparison tests, and to fix some bugs.
6132 14. Changes to the "personal" filter test:
6134 (1) The test was buggy in that it was just doing the equivalent of
6135 "contains" tests on header lines. For example, if a user's address was
6136 anne@some.where, the "personal" test would incorrectly be true for
6138 To: susanne@some.where
6140 This test is now done by extracting each address from the header in turn,
6141 and checking the entire address. Other tests that are part of "personal"
6142 are now done using regular expressions (for example, to check local parts
6143 of addresses in From: header lines).
6145 (2) The list of non-personal local parts in From: addresses has been
6146 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6147 taken from the Sieve specification recommendations.
6149 (3) If the message contains any header line starting with "List-" it is
6150 treated as non-personal.
6152 (4) The test for "circular" in the Subject: header line has been removed
6153 because it now seems ill-conceived.
6155 15. Minor typos in src/EDITME comments corrected.
6157 16. Installed latest exipick from John Jetmore.
6159 17. If headers_add on a router specified a text string that was too long for
6160 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6161 of string_sprintf() is now avoided.
6163 18. $message_body_size was not set (it was always zero) when running the DATA
6164 ACL and the local_scan() function.
6166 19. For the "mail" command in an Exim filter, no default was being set for
6167 the once_repeat time, causing a random time value to be used if "once" was
6168 specified. (If the value happened to be <= 0, no repeat happened.) The
6169 default is now 0s, meaning "never repeat". The "vacation" command was OK
6170 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6171 (I found it when inspecting the code).
6173 20. There is now an overall timeout for performing a callout verification. It
6174 defaults to 4 times the callout timeout, which applies to individual SMTP
6175 commands during the callout. The overall timeout applies when there is more
6176 than one host that can be tried. The timeout is checked before trying the
6177 next host. This prevents very long delays if there are a large number of
6178 hosts and all are timing out (e.g. when the network connections are timing
6179 out). The value of the overall timeout can be changed by specifying an
6180 additional sub-option for "callout", called "maxwait". For example:
6182 verify = sender/callout=5s,maxwait=20s
6184 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6185 to the end before writing, but this should make it even safer).
6187 22. Exim was forgetting that it had advertised PIPELINING for the second and
6188 subsequent messages on an SMTP connection. It was also not resetting its
6189 memory on STARTTLS and an internal HELO.
6191 23. When Exim logs an SMTP synchronization error within a session, it now
6192 records whether PIPELINING has been advertised or not.
6194 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6195 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6196 rather than long int.
6198 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6200 26. Added the never_mail option to autoreply.
6206 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6207 crash if the getsockname() call failed; this can happen if a connection is
6208 closed very soon after it is established. The problem was simply in the
6209 order in which certain operations were done, causing Exim to try to write
6210 to the SMTP stream before it had set up the file descriptor. The bug has
6211 been fixed by making things happen in the correct order.
6217 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6218 before the connection was closed, thus losing the rejection response.
6220 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6221 some early Solaris releases, but causes trouble in current releases where
6222 socklen_t is defined.
6224 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6227 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6230 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6231 that releasing the top of it at the end releases what was used for sub-
6232 expansions (unless the block got too big). However, discard this block if
6233 the first thing is a variable or header, so that we can use its block when
6234 it is dynamic (useful for very large $message_headers, for example).
6236 6. Lookups now cache *every* query, not just the most recent. A new, separate
6237 store pool is used for this. It can be recovered when all lookup caches are
6238 flushed. Lookups now release memory at the end of their result strings.
6239 This has involved some general refactoring of the lookup sources.
6241 7. Some code has been added to the store_xxx() functions to reduce the amount
6242 of flapping under certain conditions.
6244 8. log_incoming_interface used to affect only the <= reception log lines. Now
6245 it causes the local interface and port to be added to several more SMTP log
6246 lines, for example "SMTP connection from", and rejection lines.
6248 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6250 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6252 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6253 could be overwritten at the end of the current message (or the start of a
6254 new message if it was set in a HELO ACL). The value is now preserved for
6255 the duration of the SMTP connection.
6257 12. If a transport had a headers_rewrite setting, and a matching header line
6258 contained an unqualified address, that address was qualified, even if it
6259 did not match any rewriting rules. The underlying bug was that the values
6260 of the flags that permit the existence of unqualified sender and recipient
6261 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6262 non-local messages, and by -bnq for local messages) were not being
6263 preserved with the message after it was received.
6265 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6266 "next input=" as part of the text comprising the host identity instead of
6267 the correct text. The code was using the same buffer for two different
6268 strings. However, depending on which order the printing function evaluated
6269 its arguments, the bug did not always show up. Under Linux, for example, my
6270 test suite worked just fine.
6272 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6273 doesn't work with some older versions of Perl. It has been changed to "my",
6274 which in any case is probably the better facility to use.
6276 15. A really picky compiler found some instances of statements for creating
6277 error messages that either had too many or two few arguments for the format
6280 16. The size of the buffer for calls to the DNS resolver has been increased
6281 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6282 for addresses that have a lot of PTR records. This alleviates a problem; it
6283 does not fully solve it.
6285 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6286 buffer now truncates the list and logs the incident, which is the same
6287 action as happens when Exim is looking up a host name and its aliases.
6288 Previously in this situation something unpredictable would happen;
6289 sometimes it was "internal error: store_reset failed".
6291 18. If a server dropped the connection unexpectedly when an Exim client was
6292 using GnuTLS and trying to read a response, the client delivery process
6293 crashed while trying to generate an error log message.
6295 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6296 string, for example:
6298 warn message = H1: something\nH2: something
6300 the text was added as a single header line from Exim's point of view
6301 though it ended up OK in the delivered message. However, searching for the
6302 second and subsequent header lines using $h_h2: did not work. This has been
6303 fixed. Similarly, if a system filter added multiple headers in this way,
6304 the routers could not see them.
6306 20. Expanded the error message when iplsearch is called with an invalid key to
6307 suggest using net-iplsearch in a host list.
6309 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6310 ACLs are no longer actually imposed (and a message to that effect is
6313 22. If a "gecos" field in a passwd entry contained escaped characters, in
6314 particular, if it contained a \" sequence, Exim got it wrong when building
6315 a From: or a Sender: header from that name. A second bug also caused
6316 incorrect handling when an unquoted " was present following a character
6317 that needed quoting.
6319 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6320 was not being matched caselessly.
6322 24. Arranged for all hyphens in the exim.8 source to be escaped with
6325 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6326 properly. Recipient callout cache records were still being keyed to include
6327 the sender, even when use_sender was set false. This led to far more
6328 callouts that were necessary. The sender is no longer included in the key
6329 when use_sender is false.
6331 26. Added "control = submission" modifier to ACLs.
6333 27. Added the ${base62d: operator to decode base 62 numbers.
6335 28. dnsdb lookups can now access SRV records.
6337 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6338 the configuration file.
6340 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6341 (-v) mode. This makes the output for a verbose queue run more intelligible.
6343 31. Added a use_postmaster feature to recipient callouts.
6345 32. Added the $body_zerocount variable, containing the number of binary zero
6346 bytes in the message body.
6348 33. The time of last modification of the "new" subdirectory is now used as the
6349 "mailbox time last read" when there is a quota error for a maildir
6352 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6354 35. Added +ignore_unknown as a special item in host lists.
6356 36. Code for decoding IPv6 addresses in host lists is now included, even if
6357 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6358 address was recognized as an IP address, but was then not correctly decoded
6359 into binary, causing unexpected and incorrect effects when compared with
6366 1. Very minor rewording of debugging text in manualroute to say "list of
6367 hosts" instead of "hostlist".
6369 2. If verify=header_syntax was set, and a header line with an unqualified
6370 address (no domain) and a large number of spaces between the end of the
6371 name and the colon was received, the reception process suffered a buffer
6372 overflow, and (when I tested it) crashed. This was caused by some obsolete
6373 code that should have been removed. The fix is to remove it!
6375 3. When running in the test harness, delay a bit after writing a bounce
6376 message to get a bit more predictability in the log output.
6378 4. Added a call to search_tidyup() just before forking a reception process. In
6379 theory, someone could use a lookup in the expansion of smtp_accept_max_
6380 per_host which, without the tidyup, could leave open a database connection.
6382 5. Added the variables $recipient_data and $sender_data which get set from a
6383 lookup success in an ACL "recipients" or "senders" condition, or a router
6384 "senders" option, similar to $domain_data and $local_part_data.
6386 6. Moved the writing of debug_print from before to after the "senders" test
6389 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6390 problems for message scanning, either using a data ACL, or using
6391 local_scan() because the Received: header was not generated till after they
6392 were called (in order to set the time as the time of reception completion).
6393 I have revised the way this works. The header is now generated after the
6394 body is received, but before the ACL or local_scan() are called. After they
6395 are run, the timestamp in the header is updated.
6401 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6402 before starting a queue runner without re-exec. This happened only when
6403 deliver_drop_privilege was set or when the Exim user was set to root. The
6404 effect of the bug was that timeouts during subsequent deliveries caused
6405 crashes instead of being properly handled. The handler is now left at its
6406 default (and expected) setting.
6408 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6409 message, again when deliver_drop_privilege is set or Exim is run as root.
6410 The bug described in (1) was not present in this case, but the tidying up
6411 of the other signals was missing. I have made the two cases consistent.
6413 3. The ignore_target_hosts setting on a manualroute router was being ignored
6414 for hosts that were looked up using the /MX notation.
6416 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6419 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6420 operated on the sender address. After changing the $sender_address to <>
6421 for the sender address verify, Exim was re-instated it as the original
6422 (before rewriting) address, but remembering that it had rewritten it, so it
6423 wasn't rewriting it again. This bug also had the effect of breaking the
6424 sender address verification caching when the sender address was rewritten.
6426 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6427 This has been changed so that if the ip literal address matches
6428 ignore_target_hosts, the router declines.
6430 7. Added expansion conditions match_domain, match_address, and match_local_
6431 part (NOT match_host).
6433 8. The placeholder for the Received: header didn't have a length field set.
6435 9. Added code to Exim itself and to exim_lock to test for a specific race
6436 condition that could lead to file corruption when using MBX delivery. The
6437 issue is with the lockfile that is created in /tmp. If this file is removed
6438 after a process has opened it but before that process has acquired a lock,
6439 there is the potential for a second process to recreate the file and also
6440 acquire a lock. This could lead to two Exim processes writing to the file
6441 at the same time. The added code performs the same test as UW imapd; it
6442 checks after acquiring the lock that its file descriptor still refers to
6443 the same named file.
6445 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6446 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6449 11. Added the smtp_active_hostname option. If used, this will typically be made
6450 to depend on the incoming interface address. Because $interface_address is
6451 not set up until the daemon has forked a reception process, error responses
6452 that can happen earlier (such as "too many connections") no longer contain
6455 12. If an expansion in a condition on a "warn" statement fails because a lookup
6456 defers, the "warn" statement is abandoned, and the next ACL statement is
6457 processed. Previously this caused the whole ACL to be aborted.
6459 13. Added the iplsearch lookup type.
6461 14. Added ident_timeout as a log selector.
6463 15. Added tls_certificate_verified as a log selector.
6465 16. Added a global option tls_require_ciphers (compare the smtp transport
6466 option of the same name). This controls incoming TLS connections.
6468 17. I finally figured out how to make tls_require_ciphers do a similar thing
6469 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6470 before starting the TLS session.
6472 18. Tabs are now shown as \t in -bP output.
6474 19. If the log selector return_path_on_delivery was set, Exim crashed when
6475 bouncing a message because it had too many Received: header lines.
6477 20. If two routers both had headers_remove settings, and the first one included
6478 a superfluous trailing colon, the final name in the first list and the
6479 first name in the second list were incorrectly joined into one item (with a
6480 colon in the middle).
6486 1. Added -C and -D options to the exinext utility, mainly to make it easier
6487 to include in the automated testing, but these could be helpful when
6488 multiple configurations are in use.
6490 2. The exinext utility was not formatting the output nicely when there was
6491 an alternate port involved in the retry record key, nor when there was a
6492 message id as well (for retries that were specific to a specific message
6493 and a specific host). It was also confused by IPv6 addresses, because of
6494 the additional colons they contain. I have fixed the IPv4 problem, and
6495 patched it up to do a reasonable job for IPv6.
6497 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6498 delivery, the log line now contains "pipelined" if PIPELINING was used.
6500 4. An SMTP transport process used to panic and die if the bind() call to set
6501 an explicit outgoing interface failed. This has been changed; it is now
6502 treated in the same way as a connect() failure.
6504 5. A reference to $sender_host_name in the part of a conditional expansion
6505 that was being skipped was still causing a DNS lookup. This no longer
6508 6. The def: expansion condition was not recognizing references to header lines
6509 that used bh_ and bheader_.
6511 7. Added the _cache feature to named lists.
6513 8. The code for checking quota_filecount in the appendfile transport was
6514 allowing one more file than it should have been.
6516 9. For compatibility with Sendmail, the command line option
6524 and sets the incoming protocol and host name (for trusted callers). The
6525 host name and its colon can be omitted when only the protocol is to be set.
6526 Note the Exim already has two private options, -pd and -ps, that refer to
6527 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6528 "s", but I don't think that's a major issue.
6530 10. A number of refactoring changes to the code, none of which should affect
6533 (a) The number of logging options was getting close to filling up the
6534 32-bit word that was used as a bit map. I have split them into two classes:
6535 those that are passed in the argument to log_write(), and those that are
6536 only ever tested independently outside of that function. These are now in
6537 separate 32-bit words, so there is plenty of room for expansion again.
6538 There is no change in the user interface or the logging behaviour.
6540 (b) When building, for example, log lines, the code previously used a
6541 macro that called string_cat() twice, in order to add two strings. This is
6542 not really sufficiently general. Furthermore, there was one instance where
6543 it was actually wrong because one of the argument was used twice, and in
6544 one call a function was used. (As it happened, calling the function twice
6545 did not affect the overall behaviour.) The macro has been replaced by a
6546 function that can join an arbitrary number of extra strings onto a growing
6549 (c) The code for expansion conditions now uses a table and a binary chop
6550 instead of a serial search (which was left over from when there were very
6551 few conditions). Also, it now recognizes conditions like "pam" even when
6552 the relevant support is not compiled in: a suitably worded error message is
6553 given if an attempt is made to use such a condition.
6555 11. Added ${time_interval:xxxxx}.
6557 12. A bug was causing one of the ddress fields not to be passed back correctly
6558 from remote delivery subprocesses. The field in question was not being
6559 subsequently used, so this caused to problems in practice.
6561 13. Added new log selectors queue_time and deliver_time.
6563 14. Might have fixed a bug in maildirsizefile handling that threw up
6564 "unexpected character" debug warnings, and recalculated the data
6565 unnecessarily. In any case, I expanded the warning message to give more
6568 15. Added the message "Restricted characters in address" to the statements in
6569 the default ACL that block characters like @ and % in local parts.
6571 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6572 Three changes have been made:
6574 (a) There was a serious bug; a negative response to MAIL caused the whole
6575 recipient domain to be cached as invalid, thereby blocking all messages
6576 to all local parts at the same domain, from all senders. This bug has
6577 been fixed. The domain is no longer cached after a negative response to
6578 MAIL if the sender used is not empty.
6580 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6583 (c) A new callout option, "use_sender" has been added for people who want
6584 the modified behaviour.
6590 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6593 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6594 indeed breaks things for older releases.
6596 3. Added additional logging to the case where there is a problem reading data
6597 from a filter that is running in a subprocess using a pipe, in order to
6598 try to track down a specific problem.
6600 4. Testing facility fudge: when running in the test harness and attempting
6601 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6602 getting "No route to host". Convert this to a timeout.
6604 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6607 6. Some OS don't have socklen_t but use size_t instead. This affects the
6608 fifth argument of getsockopt() amongst other things. This is now
6609 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6610 can be set for individual OS. I have set it for SunOS5, OSF1, and
6611 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6612 some earlier ones do not.
6614 7. Change 4.30/15 was not doing the test caselessly.
6616 8. The standard form for an IPv6 address literal was being rejected by address
6617 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6618 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6619 this, as well as the form without the "IPv6" on the front (but only when
6620 address literals are enabled, of course).
6622 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6624 10. Exim crashed if a message with an empty sender address specified by -f
6625 encountered a router with an errors_to setting. This could be provoked only
6626 by a command such as
6630 where an empty string was supplied; "<>" did not hit this bug.
6632 11. Installed PCRE release 4.5.
6634 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6635 remained set. It is now erased.
6637 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6638 times from message ids (which are base 36 rather than the normal 62).
6640 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6641 were being counted as actual protocol errors, and logged if the log
6642 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6643 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6644 and DATA following a set of rejected RCPTs do not count as protocol errors.
6645 In other words, Exim assumes they were pipelined, though this may not
6646 actually be the case. Of course, in all cases the client gets an
6647 appropriate error code.
6649 15. If a lookup fails in an ACL condition, a message about the failure may
6650 be available; it is used if testing the ACL cannot continue, because most
6651 such messages specify what the cause of the deferral is. However, some
6652 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6653 that caused an old message to be retained and used if a later statement
6654 caused a defer, replacing the real cause of the deferral.
6656 16. If an IP address had so many PTR records that the DNS lookup buffer
6657 was not large enough to hold them, Exim could crash while trying to process
6658 the truncated data. It now detects and logs this case.
6660 17. Further to 4.21/58, another change has been made: if (and only if) the
6661 first line of a message (the first header line) ends with CRLF, a bare LF
6662 in a subsequent header line has a space inserted after it, so as not to
6663 terminate the header.
6665 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6666 unnecessarily, used atoi() instead of strtol(), and didn't check the
6667 termination when getting file sizes from file names by regex.
6669 19. Completely re-implemented the support for maildirsize files, in the light
6670 of a number of problems with the previous contributed implementation
6671 (4.30/29). In particular:
6673 . If the quota is zero, the maildirsize file is maintained, but no quota is
6676 . If the maildir directory does not exist, it is created before any attempt
6677 to write a maildirsize file.
6679 . The quota value in the file is just a cache; if the quota is changed in
6680 the transport, the new value overrides.
6682 . A regular expression is available for excluding directories from the
6685 20. The autoreply transport checks the characters in options that define the
6686 message's headers; it allows continued headers, but it was checking with
6687 isspace() after an embedded newline instead of explicitly looking for a
6690 21. If all the "regular" hosts to which an address was routed had passed their
6691 expiry times, and had not reached their retry times, the address was
6692 bounced, even if fallback hosts were defined. Now Exim should go on to try
6695 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6696 equivalent code in the SMTP transport. Some hosts send humungous responses
6697 to HELO/EHLO, more than 1024 it seems.
6699 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6700 gives compiler warnings in some environments. I've now done it "properly",
6703 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6704 (because of problems with the built-in one) was declared to return uschar *
6705 instead of char *, causing compiler failure.
6707 25. Fixed a file descriptor leak when processing alias/forward files.
6709 26. Fixed a minor format string issue in dbfn.c.
6711 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6713 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6714 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6717 29. When a sender address is verified, it is cached, to save repeating the test
6718 when there is more than one recipient in a message. However, when the
6719 verification involves a callout, it is possible for different callout
6720 options to be set for different recipients. It is too complicated to keep
6721 track of this in the cache, so now Exim always runs a verification when a
6722 callout is required, relying on the callout cache for the optimization.
6723 The overhead is duplication of the address routing, but this should not be
6726 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6727 to be verified with callout=postmaster, and the main callout worked but the
6728 postmaster check failed, the verification correctly failed. However, if a
6729 subsequent RCPT command asked for sender verification *without* the
6730 postmaster check, incorrect caching caused this verification also to fail,
6733 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6734 it was not caching the DNS options (qualify_single, search_parents) that
6735 were used when the lookup failed. A subsequent lookup with different
6736 options therefore always gave the same answer, though there were cases
6737 where it should not have. (Example: a "domains = !$mx_any" option on a
6738 dnslookup router: the "domains" option is always processed without any
6739 widening, but the router might have qualify_single set.) Now Exim uses the
6740 cached value only when the same options are set.
6742 32. Added John Jetmore's "exipick" utility to the distribution.
6744 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6745 than a timeout (e.g. a certificate is required, and is not provided), an
6746 Exim server now closes the connection immediately. Previously it waited for
6747 the client to close - but if the client is SSL, it seems that they each
6748 wait for each other, leading to a delay before one of them times out.
6750 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6751 maintained 0.8.x compatibility because I don't think many are using it, and
6752 it is clearly obsolete.
6754 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6757 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6758 client certificate was expired. A simple patch fixes this, though I don't
6759 understand the full logic of why the verify callback is called multiple
6762 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6763 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6764 which causes problems with some clients (such as the Certicom SSL Plus
6765 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6766 disables the coutermeasure allowing Eudora to connect."
6768 38. Exim was not checking that a write() to a log file succeeded. This could
6769 lead to Bad Things if a log got too big, in particular if it hit a file
6770 size limit. Exim now panics and dies if it cannot write to a log file, just
6771 as it does if it cannot open a log file.
6773 39. Modified OS/Makefile-Linux so that it now contains
6775 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6777 The two -D definitions ensure that Exim is compiled with large file
6778 support, which makes it possible to handle log files that are bigger than
6781 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6782 instance) a domain was checked against a named list that involved a lookup,
6783 causing $domain_data to be set, then another domain was checked against the
6784 same list, then the first domain was re-checked, the value of $domain_data
6785 after the final check could be wrong. In particular, if the second check
6786 failed, it could be set empty. This bug probably also applied to
6789 41. The strip_trailing_dot option was not being applied to the address given
6790 with the -f command-line option.
6792 42. The code for reading a message's header from the spool was incrementing
6793 $received_count, but never initializing it. This meant that the value was
6794 incorrect (doubled) while delivering a message in the same process in which
6795 it was received. In the most common configuration of Exim, this never
6796 happens - a fresh exec is done - but it can happen when
6797 deliver_drop_privilege is set.
6799 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6800 it now includes up to 150 characters of the unexpected data in the log
6803 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6804 and building data strings. The size of both of these buffers was 10 000
6805 bytes - far larger than anybody would *ever* want, thought I. Needless to
6806 say, somebody hit the limit. I have increased the maximum line length to
6807 20 000 and the maximum data length of concatenated lines to 100 000. I have
6808 also fixed two bugs, because there was no checking on these buffers. Tsk,
6809 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6810 buffer is too small.
6812 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6813 lsearch lookups. Now it does.
6815 46. When parsing a route_list item in a manualroute router, a fixed-length
6816 buffer was used for the list of hosts. I made this 1024 bytes long,
6817 thinking that nobody would ever have a list of hosts that long. Wrong.
6818 Somebody had a whole pile of complicated expansion conditions, and the
6819 string was silently truncated, leading to an expansion error. It turns out
6820 that it is easier to change to an unlimited length (owing to other changes
6821 that have happened since this code was originally written) than to build
6822 structure for giving a limitation error. The length of the item that
6823 expands into the list of hosts is now unlimited.
6825 47. The lsearch lookup could not handle data where the length of text line was
6826 more than 4095 characters. Such lines were truncated, leading to shortened
6827 data being returned. It should now handle lines of any length.
6829 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6830 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6833 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6834 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6835 no longer happens in comment lines. A list of replacements is now placed
6836 at the head of all of the source files, except those whose only change is
6837 to replace PERL_COMMAND in the very first #! line.
6839 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6840 messages on the queue, with a bottom-up merge sort, using code contributed
6841 by Michael Haardt. This should make operations like -bp somewhat faster on
6842 large queues. It won't affect queue runners, except when queue_run_in_order
6845 51. Installed eximstats 1.31 in the distribution.
6847 52. Added support for SRV lookups to the dnslookup router.
6849 53. If an ACL referred to $message_body or $message_body_end, the value was not
6850 reset for any messages that followed in the same SMTP session.
6852 54. The store-handling optimization for building very long strings was not
6853 differentiating between the different store pools. I don't think this
6854 actually made any difference in practice, but I've tidied it.
6856 55. While running the routers to verify a sender address, $sender_address
6857 was still set to the sender address. This is wrong, because when routing to
6858 send a bounce to the sender, it would be empty. Therefore, I have changed
6859 it so that, while verifying a sender address, $sender_address is set to <>.
6860 (There is no change to what happens when verifying a recipient address.)
6862 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6863 target A or AAAA records (if not already returned) without resetting the
6864 qualify_single or search_parents options of the DNS resolver. These are
6865 inappropriate in this case because the targets of MX and SRV records must
6866 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6867 target that, when qualified, matched something in the local domain. These
6868 two options are now turned off when doing these lookups.
6870 57. It seems that at least some releases of Reiserfs (which does not have the
6871 concept of a fixed number of inodes) returns zero and not -1 for the
6872 number of available inodes. This interacted badly with check_spool_inodes,
6873 which assumed that -1 was the "no such thing" setting. What I have done is
6874 to check that the total number of inodes is greater than zero before doing
6875 the test of how many are available.
6877 58. When a "warn" ACL statement has a log_message modifier, the message is
6878 remembered, and not repeated. This is to avoid a lot of repetition when a
6879 message has many recipients that cause the same warning to be written.
6880 However, Exim was preserving the list of already written lines for an
6881 entire SMTP session, which doesn't seem right. The memory is now reset if a
6882 new message is started.
6884 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6885 debugging output unless log_rewrite was also set.
6887 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6888 of (int)(handle) when we know that handle contains (void *)(-1).
6890 61. The Exim daemon panic-logs an error return when it closes the incoming
6891 connection. However "connection reset by peer" seems to be common, and
6892 isn't really an error worthy of noting specially, so that particular error
6895 62. When Exim is trying to find all the local interfaces, it used to panic and
6896 die if the ioctl to get the interface flags failed. However, it seems that
6897 on at least one OS (Solaris 9) it is possible to have an interface that is
6898 included in the list of interfaces, but for which you get a failure error
6899 for this call. This happens when the interface is not "plumbed" into a
6900 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6901 failure of the "get flags" call assumes that the interface is down.
6903 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6904 makes life easier for people who are doing arithmetic on fields extracted
6905 from dates, where you often get leading zeros that should not be
6906 interpreted as octal.
6908 64. Added qualify_domain to the redirect router, to override the global
6911 65. If a pathologically long header line contained very many addresses (the
6912 report of this problem mentioned 10 000) and each of them was rewritten,
6913 Exim could use up a very large amount of memory. (It kept on making new
6914 copies of the header line as it rewrote, and never released the old ones.)
6915 At the expense of a bit more processing, the header rewriting function has
6916 been changed so that it no longer eats memory in this way.
6918 66. The generation of the Received: header has been moved from the time that a
6919 message starts to be received, to the time that it finishes. The timestamp
6920 in the Received: header should now be very close to that of the <= log
6921 line. There are two side-effects of this change:
6923 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6924 logged header lines no longer include the local Received: line, because
6925 it has not yet been created. The same applies to a copy of the message
6926 that is returned to a non-SMTP sender when a message is rejected.
6928 (b) When a filter file is tested using -bf, no additional Received: header
6929 is added to the test message. After some thought, I decided that this
6932 This change does not affect the value of $received_for. It is still set
6933 after address rewriting, but before local_scan() is called.
6935 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6937 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6938 gave an unhelpful panic error message, and a defer error. I have managed to
6939 change this behaviour so that it now rejects any supplied certificate,
6940 which seems right, as the list of acceptable certificates is empty.
6942 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6943 gave an unhelpful defer error. I have not managed to make this reject any
6944 supplied certificates, but the error message it gives is "no certificate
6945 supplied", which is not helpful.
6947 70. exigrep's output now also includes lines that are not associated with any
6948 message, but which match the given pattern. Implemented by a patch from
6949 Martin Sluka, which also tidied up the Perl a bit.
6951 71. Recipient callout verification, like sender verification, was using <> in
6952 the MAIL FROM command. This isn't really the right thing, since the actual
6953 sender may affect whether the remote host accepts the recipient or not. I
6954 have changed it to use the actual sender in the callout; this means that
6955 the cache record is now keyed on a recipient/sender pair, not just the
6956 recipient address. There doesn't seem to be a real danger of callout loops,
6957 since a callout by the remote host to check the sender would use <>.
6958 [SEE ABOVE: changed after hitting problems.]
6960 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6961 temporary errors. However, in the case of such a code being given after
6962 the end of a data transmission (i.e. after ".") Exim was failing to write
6963 a retry record for the message. (Yes, there was some broken host that was
6964 actually sending 8xx at this point.)
6966 73. An unknown lookup type in a host list could cause Exim to panic-die when
6967 the list was checked. (An example that provoked this was putting <; in the
6968 middle of a list instead of at the start.) If this happened during a DATA
6969 ACL check, a -D file could be left lying around. This kind of configuration
6970 error no longer causes Exim to die; instead it causes a defer error. The
6971 incident is still logged to the main and panic logs.
6973 74. Buglet left over from Exim 3 conversion. The message "too many messages
6974 in one connection" was written to the rejectlog but not the mainlog, except
6975 when address rewriting (yes!) was being logged.
6977 75. Added write_rejectlog option.
6979 76. When a system filter was run not as root (that is, when system_filter_user
6980 was set), the values of the $n variables were not being returned to the
6981 main process; thus, they were not subsequently available in the $sn
6984 77. Added +return_path_on_delivery log selector.
6986 78. A connection timeout was being treated differently from recipients deferred
6987 when testing hosts_max_try with a message that was older than the host's
6988 retry timeout. (The host should not be counted, thus allowing all hosts to
6989 be tried at least once before bouncing.) This may have been the cause of an
6990 occasionally reported bug whereby a message would remain on the queue
6991 longer than the retry timeout, but would be bounced if a delivery was
6992 forced. I say "may" because I never totally pinned down the problem;
6993 setting up timeout/retry tests is difficult. See also the next item.
6995 79. The ultimate address timeout was not being applied to errors that involved
6996 a combination of host plus message (for example, a timeout on a MAIL
6997 command). When an address resolved to a number of possible hosts, and they
6998 were not all tried for each delivery (e.g. because of hosts_max_try), a
6999 message could remain on the queue longer than the retry timeout.
7001 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7004 81. Fixed an obscure SMTP outgoing bug which required at least the following
7005 conditions: (a) there was another message waiting for the same server;
7006 (b) the server returned 5xx to all RCPT commands in the first message so
7007 that the message was not completed; (c) the server dropped the connection
7008 or gave a negative response to the RSET that Exim sends to abort the
7009 transaction. The observed case was a dropped connection after DATA that had
7010 been sent in pipelining mode. That is, the server had advertised PIPELINING
7011 but was not implementing it correctly. The effect of the bug was incorrect
7012 behaviour, such as trying another host, and this could lead to a crash.
7018 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7019 and daemon.c were passed as pointers to ints; they should have been
7020 pointers to socklen_t variables (which are typically unsigned ints).
7022 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7025 3. Fixed a really odd bug that affected only the testing scheme; patching a
7026 certain fixed string in the binary changed the value of another string that
7027 happened to be identical to the end of the original first string.
7029 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7030 name", it returns that address as the IP address. On some operating
7031 systems (e.g. Solaris), it also passes back the IP address string as the
7032 "host name". However, on others (e.g. Linux), it passes back an empty
7033 string. Exim wasn't checking for this, and was changing the host name to an
7034 empty string, assuming it had been canonicalized.
7036 5. Although rare, it is permitted to have more than one PTR record for a given
7037 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7038 all the names associated with an address, because they do in Solaris.
7039 However, it seems that they do not in Linux for data that comes from the
7040 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7041 I found this out when I moved to a new Linux workstation and tried to run
7042 the Exim test suite.
7044 To get round this problem I have changed the code so that it now does its
7045 own call to the DNS to look up PTR records when searching for a host name.
7046 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7047 addresses that are only in /etc/hosts are still found.
7049 This behaviour is, however, controlled by an option called host_lookup_
7050 order, which defaults to "bydns:byaddr". If people want to use the other
7051 order, or indeed, just use one or the other means of lookup, they can
7052 specify it in this variable.
7054 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7055 some operating systems, this comes back from gethostbyaddr() as an empty
7056 string, and this is what Exim used to test for. However, it seems that in
7057 other systems, "." is yielded. Exim now tests for this case too.
7059 7. The values of check_spool_space and check_log_space are now held internally
7060 as a number of kilobytes instead of an absolute number of bytes. If a
7061 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7062 kilobyte. This means that much larger values can be stored.
7064 8. Exim monitor: an attempt to get the action menu when not actually pointing
7065 at a message produces an empty menu entitled "No message selected". This
7066 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7067 no entries in it ("Shell widget menu has zero width and/or height"). So I
7068 have added a single, blank menu entry in this case.
7070 9. Added ${quote_local_part.
7072 10. MIME decoding is now applied to the contents of Subject: header lines when
7075 11. Now that a reference to $sender_host_address automatically causes a reverse
7076 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7077 host lookup before query-style lookups in lists that might use this
7078 variable. This has therefore been abolished, and the "net-" prefix is no
7079 longer necessary for query-style lookups.
7081 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7082 have been a typo for LFLAGS, so it has been changed.
7084 13. The install script calls Exim with "-C /dev/null" in order to find the
7085 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7086 to be output. However, since Exim outputs its version number before the
7087 error, it didn't break the script. It just looked ugly. I fixed this by
7088 always allowing "-C /dev/null" if the caller is root.
7090 14. Ignore overlarge ACL variable number when reading spool file - insurance
7091 against a later release with more variables having written the file.
7093 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7094 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7095 this, as well as the form without the "IPv6" on the front.
7097 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7098 OS/Makefile-Darwin file.
7100 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7101 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7103 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7104 variable. However, some OS use ints for inodes. I've added cast to long int
7105 to get rid of the compiler warning.
7107 19. I had forgotten to lock out "/../" in configuration file names when
7108 ALT_CONFIG_PREFIX was set.
7110 20. Routers used for verification do not need to specify transports. However,
7111 if such a router generated a host list, and callout was configured, Exim
7112 crashed, because it could not find a port number from the (non-existent)
7113 transport. It now assumes port 25 in this circumstance.
7115 21. Added the -t option to exigrep.
7117 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7118 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7119 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7120 them is set, LOOKUP_LSEARCH is forced.
7122 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7124 24. Added sender and host information to the "rejected by local_scan()" log
7125 line; previously there was no indication of these.
7127 25. Added .include_if_exists.
7129 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7130 Linux. It turns out that not all file systems support this. Apparently some
7131 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7132 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7133 not supported on the file descriptor, is now ignored when Exim is trying to
7134 sync a directory. This applies only to Linux.
7136 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7138 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7139 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7142 29. Added support for maildirsize files from supplied patch (modified a bit).
7144 30. The use of :fail: followed by an empty string could lead Exim to respond to
7145 sender verification failures with (e.g.):
7147 550 Verification failed for <xxx>
7148 550 Sender verify failed
7150 where the first response line was missing the '-' that indicates it is not
7151 the final line of the response.
7153 31. The loop for finding the name of the user that called Exim had a hardwired
7154 limit of 10; it now uses the value of finduser_retries, which is used for
7155 all other user lookups.
7157 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7160 33. Exim was neglecting to zero errno before one call of strtol() when
7161 expanding a string and expecting an integer value. On some systems this
7162 resulted in spurious "integer overflow" errors. Also, it was casting the
7163 result into an int without checking.
7165 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7166 did not work. The code looks as if it has *never* worked, though it appears
7167 to have been documented since at least release 1.62. I have made it work.
7169 35. The "timeout_DNS" error in retry rules, also documented since at least
7170 1.62, also never worked. As it isn't clear exactly what this means, and
7171 clearly it isn't a major issue, I have abolished the feature by treating it
7172 as "timeout", and writing a warning to the main and panic logs.
7174 36. The display of retry rules for -brt wasn't always showing the error code
7177 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7178 timeout_connect_A, timeout_connect_MX.
7180 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7181 to the empty sender.
7183 39. The daemon was not analysing the content of -oX till after it had closed
7184 stderr and disconnected from the controlling terminal. This meant that any
7185 syntax errors were only noted on the panic log, and the return code from
7186 the command was 0. By re-arranging the code a little, I've made the
7187 decoding happen first, so such errors now appear on stderr, and the return
7188 code is 1. However, the actual setting up of the sockets still happens in
7189 the disconnected process, so errors there are still only recorded on the
7192 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7193 connections (as happens on some IP stacks) was logged at start up time as
7194 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7195 it from "IPv6 and IPv4", which means that two separate sockets are being
7198 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7199 says whether AF_INET or AF_INET6 was passed as an argument.
7201 42. Exiwhat output was messed up when time zones were included in log
7204 43. Exiwhat now gives more information about the daemon's listening ports,
7205 and whether -tls-on-connect was used.
7207 44. The "port" option of the smtp transport is now expanded.
7209 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7210 silently ignored. Now an error message is written to the main and panic
7213 46. There's a new ACL modifier called "logwrite" which writes to a log file
7214 as soon as it is encountered.
7216 47. Added $local_user_uid and $local_user_gid at routing time.
7218 48. Exim crashed when trying to verify a sender address that was being
7221 49. Exim was recognizing only a space character after ".include". It now also
7222 recognizes a tab character.
7224 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7225 extracting the relevant information from the specification. The man page no
7226 longer contains scrambled data for the -d option, and I've added a section
7227 at the front about calling Exim under different names.
7229 51. Added "extra_headers" argument to the "mail" command in filter files.
7231 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7234 53. Installed eximstats 1.29.
7236 54. Added transport_filter_timeout as a generic transport option.
7238 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7239 Cc: header lines. This was required by RFC 822, but it not required by RFC
7242 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7243 incoming messages that did not have them. Now it does so only if the
7244 message originates locally, that is, if there is no associated remote host
7245 address. When Resent- header lines are present, this applies to the Resent-
7246 lines rather than the non-Resent- lines.
7248 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7249 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7251 58. Messages for configuration errors now include the name of the main
7252 configuration file - useful now that there may be more than one file in a
7253 list (.included file names were always shown).
7255 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7256 for those rare installations that do not start the daemon as root or run it
7257 setuid root. I've cut out the call to initgroups() if the daemon is not
7260 60. The Exim user and group can now be bound into the binary as text strings
7261 that are looked up at the start of Exim's processing.
7263 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7265 62. Added $mailstore_basename variable.
7267 63. Installed patch to sieve.c from Michael Haardt.
7269 64. When Exim failed to open the panic log after failing to open the main log,
7270 the original message it was trying to log was written to stderr and debug
7271 output, but if they were not available (the usual case in production), it
7272 was lost. Now it is written to syslog before the two lines that record the
7273 failures to open the logs.
7275 65. Users' Exim filters run in subprocesses under the user's uid. It is
7276 possible for a "deliver" command or an alias in a "personal" command to
7277 provoke an address rewrite. If logging of address rewriting is configured,
7278 this fails because the process is not running as root or exim. There may be
7279 a better way of dealing with this, but for the moment (because 4.30 needs
7280 to be released), I have disabled address rewrite logging when running a
7281 filter in a non-root, non-exim process.
7287 1. The buildconfig auxiliary program wasn't quoting the value set for
7288 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7289 not defined. This bug was masked in 4.22 by the effect that was fixed in
7292 2. Some messages that were rejected after a message id was allocated were
7293 shown as "incomplete" by exigrep. It no longer does this for messages that
7294 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7296 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7297 have allow_domain_literals set, the ID did not get logged in the <= line.
7298 Domain literals are now always recognized in Message-ID: header lines.
7300 4. The first argument for a ${extract expansion item is the key name or field
7301 number. Leading and trailing spaces in this item were not being ignored,
7302 causing some misleading effects.
7304 5. When deliver_drop_privilege was set, single queue runner processes started
7305 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7306 same command in the process it spins off) were not dropping privilege.
7308 6. When the daemon running as "exim" started a queue runner, it always
7309 re-executed Exim in the spun-off process. This is a waste of effort when
7310 deliver_drop_privilege is set. The new process now just calls the
7311 queue-runner function directly.
7317 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7320 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7321 ignored. Though the use of -oP was forcing the writing of a pid file, it
7322 was always written to the default place.
7324 3. If the message "no IP address found for host xxxx" is generated during
7325 incoming verification, it is now followed by identification of the incoming
7326 connection (so you can more easily find what provoked it).
7328 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7330 5. Added some features to "harden" Exim a bit more against certain attacks:
7332 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7333 be put in Local/Makefile. This is like the never_users runtime option,
7334 but it cannot be overridden. The default setting is "root".
7336 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7337 prefix string with which any file named in a -C command line option
7340 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7341 is retained for -C and -D only if the caller of Exim is root. Without
7342 it, the exim user may also use -C and -D and retain privilege.
7344 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7345 command line option is disabled.
7347 6. Macro names set by the -D option must start with an upper case letter, just
7348 like macro names defined in the configuration file.
7350 7. Added "dereference=" facility to LDAP.
7352 8. Two instances of the typo "uknown" in the source files are fixed.
7354 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7355 the Configure-Makefile script screwed up while processing it.
7357 10. Incorporated PCRE 4.4.
7359 11. The SMTP synchronization check was not operating right at the start of an
7360 SMTP session. For example, it could not catch a HELO sent before the client
7361 waited for the greeting. There is now a check for outstanding input at the
7362 point when the greeting is written. Because of the duplex, asynchronous
7363 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7364 way, but not yet received, when the check is performed.
7366 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7367 on TCP/IP sockets, because this apparently causes some broken clients to
7370 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7371 unchanged) from the Cygwin maintainer.
7373 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7374 was supported instead of testing for mbx. Effectively a typo.
7376 15. The spa authenticator server code was not checking that the input it
7377 received was valid base64.
7379 16. The debug output line for the "set" modifier in ACLs was not showing the
7380 name of the variable that was being set.
7382 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7384 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7385 lookup on its own. Something else was needed to trigger the lookup. For
7386 example, a match in host_lookup or the need for a host name in a host list.
7387 Now, if $sender_host_name is referenced and the host name has not yet been
7388 looked up, a lookup is performed. If the lookup fails, the variable remains
7389 empty, and $host_lookup_failed is set to "1".
7391 19. Added "eqi" as a case-independent comparison operator.
7393 20. The saslauthd authentication condition could segfault if neither service
7394 nor realm was specified.
7396 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7397 error message that was logged was misleading, and incoming SMTP
7398 connections were dropped. The message is now more accurate, and temporary
7399 errors are given to SMTP connections.
7401 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7402 (except RSET) with a 421 error, until QUIT is received. However, it was
7403 failing to send a response to QUIT.
7405 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7406 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7407 ACL is now run after the helo_try_verify_hosts code.
7409 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7410 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7411 case-independent, so other case variants are also recognized). Apparently
7412 some systems use these upper case variants.
7414 25. If more than two messages were waiting for the same host, and a transport
7415 filter was specified for the transport, Exim sent two messages over the
7416 same TCP/IP connection, and then failed with "socket operation on non-
7417 socket" when it tried to send the third.
7419 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7421 27. The extern definition of crypt16() in expand.c was not being excluded when
7422 the OS had its own crypt16() function.
7424 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7425 as a preferred synonym for return_size_limit, both as an option and as an
7428 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7430 30. Changed the default configuration ACL to relax the local part checking rule
7431 for addresses that are not in any local domains. For these addresses,
7432 slashes and pipe symbols are allowed within local parts, but the sequence
7433 /../ is explicitly forbidden.
7435 31. SPA server authentication was not clearing the challenge buffer before
7438 32. log_message in a "warn" ACL statement was writing to the reject log as
7439 well as to the main log, which contradicts the documentation and doesn't
7440 seem right (because no rejection is happening). So I have stopped it.
7442 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7443 However, I am unable to do any testing of this.
7445 34. Fixed an infelicity in the appendfile transport. When checking directories
7446 for a mailbox, to see if any needed to be created, it was accidentally
7447 using path names with one or more superfluous leading slashes; tracing
7448 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7450 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7451 contents are added to the log line that is written for every discarded
7452 recipient. (Previously a log_message setting was ignored.)
7454 36. The ${quote: operator now quotes the string if it is empty.
7456 37. The install script runs exim in order to find its version number. If for
7457 some reason other than non-existence or emptiness, which it checks, it
7458 could not run './exim', it was installing it with an empty version number,
7459 i.e. as "exim-". This error state is now caught, and the installation is
7462 38. An argument was missing from the function that creates an error message
7463 when Exim fails to connect to the socket for saslauthd authentication.
7464 This could cause Exim to crash, or give a corrupted message.
7466 39. Added isip, isip4, and isip6 to ${if conditions.
7468 40. The ACL variables $acl_xx are now saved with the message, and can be
7469 accessed later in routers, transports, and filters.
7471 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7472 strings in the file are not string-expanded.
7474 42. If a MAIL command specified a SIZE value that was too large to fit into an
7475 int variable, the check against message_size_limit failed. Such values are
7476 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7477 day this will have to be increased, but I don't think I want to be around
7478 when emails are that large.
7485 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7486 iconv() is not standard in FreeBSD.
7488 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7489 IPv6 enabled. The observed symptom was a segmentation fault on return from
7490 the function os_common_find_running_interfaces() in src/os.c.
7492 3. In the check_special_case() function in daemon.c I had used "errno" as an
7493 argument name, which causes warnings on some systems. This was basically a
7494 typo, since it was named "eno" in the comments!
7496 4. The code that waits for the clock to tick (at a resolution of some fraction
7497 of a second) so as to ensure message-id uniqueness was always waiting for
7498 at least one whole tick, when it could have waited for less. [This is
7499 almost certainly not relevant at current processor speeds, where it is
7500 unlikely to ever wait at all. But we try to future-proof.]
7502 5. The function that sleeps for a time interval that includes fractions of a
7503 second contained a race. It did not block SIGALRM between setting the
7504 timer, and suspending (a couple of lines later). If the interval was short
7505 and the sigsuspend() was delayed until after it had expired, the suspension
7506 never ended. On busy systems this could lead to processes getting stuck for
7509 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7510 process, before it forks any delivery processes. The open lookup caching
7511 mechanism meant that the open file or database connection was passed into
7512 the delivery process. The problem was that delivery processes always tidy
7513 up cached lookup data. This could cause a problem for the next delivery
7514 process started by the queue runner, because the external queue runner
7515 process does not know about the closure. So the next delivery process
7516 still has data in the lookup cache. In the case of a file lookup, there was
7517 no problem because closing a file descriptor in a subprocess doesn't affect
7518 the parent. However, if the lookup was caching a connection to a database,
7519 the connection was closed, and the second delivery process was likely to
7520 see errors such as "PGSQL: query failed: server closed the connection
7521 unexpectedly". The problem has been fixed by closing all cached lookups
7522 in a queue runner before running a delivery process.
7524 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7525 seem to have the "const" qualifier which it has on other OS. I've
7528 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7529 *of the same type* (client or server) with the same public name that an
7530 error should be diagnosed.
7532 9. When Exim looked up a host name for an IP address, but failed to find the
7533 original IP address when looking up the host name (a safety check), it
7534 output the message "<ip address> does not match any IP for NULL", which was
7535 confusing, to say the least. The bug was that the host name should have
7536 appeared instead of "NULL".
7538 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7539 user that is built into the binary, and the -C or -D options is used, root
7540 privilege is dropped before the configuration file is read. In addition,
7541 logging is switched to stderr instead of the normal log files. If the
7542 configuration then re-defines the Exim user, the unprivileged environment
7543 is probably not what is expected, so Exim logs a panic warning message (but
7546 However, if deliver_drop_privilege is set, the unprivileged state may well
7547 be exactly what is intended, so the warning has been cut out in that case,
7548 and Exim is allowed to try to write to its normal log files.
7554 1. smtp_return_error_details was not giving details for temporary sender
7555 or receiver verification errors.
7557 2. Diagnose a configuration error if two authenticators have the same public
7560 3. Exim used not to create the message log file for a message until the first
7561 delivery attempt. This could be confusing when incoming messages were held
7562 for policy or load reasons. The message log file is now created at the time
7563 the message is received, and an initial "Received" line is written to it.
7565 4. The automatically generated man page for command line options had a minor
7566 bug that caused no ill effects; however, a more serious problem was that
7567 the procedure for building the man page automatically didn't always
7568 operate. Consequently, release 4.20 contains an out-of-date version. This
7569 shouldn't happen again.
7571 5. When building Exim with embedded Perl support, the script that builds the
7572 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7573 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7575 6. The freeze_tell option was not being used for messages that were frozen on
7576 arrival, either by an ACL or by local_scan().
7578 7. Added the smtp_incomplete_transaction log selector.
7580 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7581 was accepting AUTH without a new EHLO.
7583 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7584 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7587 10. Logging of TCP/IP connections (when configured) now happens in the main
7588 daemon process instead of the child process, so that the TCP/IP connection
7589 count is more accurate (but it can never be perfect).
7591 11. The use of "drop" in a nested ACL was not being handled correctly in the
7592 outer ACL. Now, if condition failure induced by the nested "drop" causes
7593 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7594 or "require"), the connection is dropped.
7596 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7597 that yield "discard" can now be used with an "accept" or a "discard" verb,
7598 but an error is generated for any others (because I can't see a useful way
7599 to define what should happen).
7601 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7602 are now processed in the same way as lines in the Exim configuration file.
7603 In particular, continuation lines are supported.
7605 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7607 15. Added -ti meaning -t -i.
7609 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7610 domains, and warn by logging if others are found.
7612 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7613 returned from GIFCONF when Exim is trying to find the list of interfaces on
7614 a host. The code in os.c has been modified to copy each ifreq to an aligned
7615 structure in all cases.
7617 Also, in some cases, the returned ifreq's were being copied to a 'struct
7618 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7619 means the last couple of bytes of an IPv6 address could be chopped if the
7620 ifreq contained only a normal sockaddr (14 bytes storage).
7622 18. Named domain lists were not supported in the hosts_treat_as_local option.
7623 An entry such as +xxxx was not recognized, and was treated as a literal
7626 19. Ensure that header lines added by a DATA ACL are included in the reject log
7627 if the ACL subsequently rejects the message.
7629 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7630 MD5 (which is deprecated).
7632 21. When testing a filter file using -bf, Exim was writing a message when it
7633 took the sender from a "From " line in the message, but it was not doing so
7634 when it took $return_path from a Return-Path: header line. It now does.
7636 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7637 with a valid header line field name (a series of printing characters
7638 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7640 23. Changed "disc" in the source to "disk" to conform to the documentation and
7641 the book and for uniformity.
7643 24. Ignore Sendmail's -Ooption=value command line item.
7645 25. When execve() failed while trying to run a command in a pipe transport,
7646 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7647 could be confused with a return value of 69 from the command itself. This
7648 has been changed to 127, the value the shell returns if it is asked to run
7649 a non-existent command. The wording for the related log line suggests a
7650 non-existent command as the problem.
7652 26. If received_header_text expands to an empty string, do not add a Received:
7653 header line to the message. (Well, it adds a token one on the spool, but
7654 marks it "old" so that it doesn't get used or transmitted.)
7656 27. Installed eximstats 1.28 (addition of -nt option).
7658 28. There was no check for failure on the call to getsockname() in the daemon
7659 code. This can fail if there is a shortage of resources on the system, with
7660 ENOMEM, for example. A temporary error is now given on failure.
7662 29. Contrary to the C standard, it seems that in some environments, the
7663 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7664 program. Exim now does this explicitly; it affects the formatting of
7665 timestamps using strftime().
7667 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7668 complaints. I've now initialized all the variables, to avoid this.
7670 32. Header lines added by a system filter were not being "seen" during
7671 transport-time rewrites.
7673 33. The info_callback() function passed to OpenSSL is set up with type void
7674 (*)(SSL *, int, int), as described somewhere. However, when calling the
7675 function (actually a macro) that sets it up, the type void(*)() is
7676 expected. I've put in a cast to prevent warnings from picky compilers.
7678 34. If a DNS black list lookup found a CNAME record, but there were no A
7679 records associated with the domain it pointed at, Exim crashed.
7681 35. If a DNS black list lookup returned more than one A record, Exim ignored
7682 all but the first. It now scans all returned addresses if a particular IP
7683 value is being sought. In this situation, the contents of the
7684 $dnslist_value variable are a list of all the addresses, separated by a
7687 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7688 to accept a host name and all its aliases if the forward lookup for any of
7689 them yielded the IP address of the incoming connection. Now it accepts only
7690 those names whose forward lookup yields the correct IP address. Any other
7691 names are discarded. This closes a loophole whereby a rogue DNS
7692 administrator could create reverse DNS records to break through a
7693 wildcarded host restriction in an ACL.
7695 37. If a user filter or a system filter that ran in a subprocess used any of
7696 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7697 the wrong values were passed to the pipe command ($thisaddress had the
7698 value of $0, $0 had the value of $1, etc). This bug was introduced by
7699 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7701 38. Improved the line breaking for long SMTP error messages from ACLs.
7702 Previously, if there was no break point between 40 and 75 characters, Exim
7703 left the rest of the message alone. Two changes have been made: (a) I've
7704 reduced the minimum length to 35 characters; (b) if it can't find a break
7705 point between 35 and 75 characters, it looks ahead and uses the first one
7706 that it finds. This may give the occasional overlong line, but at least the
7707 remaining text gets split now.
7709 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7710 file descriptors might be low, and that setting 1000 would always raise it.
7711 It turns out that in some environments, the limit is already over 1000 and
7712 that lowering it causes trouble. So now Exim takes care not to decrease it.
7714 40. When delivering a message, the value of $return_path is set to $sender_
7715 address at the start of routing (routers may change the value). By an
7716 oversight, this default was not being set up when an address was tested by
7717 -bt or -bv, which affected the outcome if any router or filter referred to
7720 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7721 log only when "message" or "log_message" are set. However, if one of the
7722 conditions was an address verification, or a call to a nested ACL, the
7723 messages generated by the underlying test were being passed through. This
7724 no longer happens. The underlying message is available in $acl_verify_
7725 message for both "message" and "log_message" expansions, so it can be
7726 passed through if needed.
7728 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7729 new expansion $bh_ to give the encoded byte string without charset
7730 translation. Translation happens only if iconv() is available; HAVE_ICONV
7731 indicates this at build time. HEADERS_CHARSET gives the charset to
7732 translate to; headers_charset can change it in the configuration, and
7733 "headers charset" can change it in an individual filter file.
7735 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7736 that creates RFC 2047 encoded "words" labels them as that charset instead
7737 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7738 expansion operator; (ii) when Exim creates a From: line for a local
7739 message; (iii) when a header line is rewritten to include a "phrase" part.
7741 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7742 buggy, causing it to skip the first lines of messages whose message ID
7743 ended in 'D'. This would not have bitten before Exim release 4.14, because
7744 message IDs were unlikely to end in 'D' before then. The effect was to have
7745 incorrect size information for certain domains.
7747 45. #include "config.h" was missing at the start of the crypt16.c module. This
7748 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7751 46. If there was a timeout during a "random" callout check, Exim treated it as
7752 a failure of the random address, and carried on sending RSET and the real
7753 address. If the delay was just some slowness somewhere, the response to the
7754 original RCPT would be taken as a response to RSET and so on, causing
7755 mayhem of various kinds.
7757 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7758 when I implemented it. It didn't allow for the fact that some option values
7759 may legitimately be negative (e.g. size_addition), and it didn't even do
7760 the right test for positive values.
7762 48. Domain names in DNS records are case-independent. Exim always looks them up
7763 in lower case. Some resolvers return domain names in exactly the case they
7764 appear in the zone file, that is, they may contain uppercase letters. Not
7765 all resolvers do this - some return always lower case. Exim was treating a
7766 change of case by a resolver as a change of domain, similar to a widening
7767 of a domain abbreviation. This triggered its re-routing code and so it was
7768 trying to route what was effectively the same domain again. This normally
7769 caused routing to fail (because the router wouldn't handle the domain
7770 twice). Now Exim checks for this case specially, and just changes the
7771 casing of the domain that it ultimately uses when it transmits the message
7774 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7777 50. If a filter generated a file delivery with a non-absolute name (possible if
7778 no home directory exists for the router), the forbid_file option was not
7781 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7782 the existing equality matching.
7784 52. Exim was using ints instead of ino_t variables in some places where it was
7785 dealing with inode numbers.
7787 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7788 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7789 called TMPDIR, and if it finds it is different, it changes its value.
7791 54. The smtp_printf() function is now made available to local_scan() so
7792 additional output lines can be written before returning. There is also an
7793 smtp_fflush() function to enable the detection of a dropped connection.
7794 The variables smtp_input and smtp_batched_input are exported to
7797 55. Changed the default runtime configuration: the message "Unknown user"
7798 has been removed from the ACL, and instead placed on the localuser router,
7799 using the cannot_route_message feature. This means that any verification
7800 failures that generate their own messages won't get overridden. Similarly,
7801 the "Unrouteable address" message that was in the ACL for unverifiable
7802 relay addresses has also been removed.
7804 56. Added hosts_avoid_esmtp to the smtp transport.
7806 57. The exicyclog script was not checking for the esoteric option
7807 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7808 will work only if exicyclog is run under the appropriate euid.
7810 58. Following a discussion on the list, the rules by which Exim recognises line
7811 endings on incoming messages have been changed. The -dropcr and drop_cr
7812 options are now no-ops, retained only for backwards compatibility. The
7813 following line terminators are recognized: LF CRLF CR. However, special
7814 processing applies to CR:
7816 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7817 nor a local message in the state where . is a terminator.
7819 (ii) If a bare CR is encountered in a header line, an extra space is added
7820 after the line terminator so as not to end the header. The reasoning
7821 behind this is that bare CRs in header lines are most likely either
7822 to be mistakes, or people trying to play silly games.
7824 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7825 was being incorrectly given as 18 bytes larger than it should have been.
7826 This is a VOB (very old bug).
7828 60. This may never have affected anything current, but just in case it has:
7829 When the local host is found other than at the start of a list of hosts,
7830 the local host, those with the same MX, and any that follow, are discarded.
7831 When the list in question was part of a longer list of hosts, the following
7832 hosts (not currently being processed) were also being discarded. This no
7833 longer happens. I'm not sure if this situation could ever has previously
7836 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7839 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7840 header. This is something that is recommended in a new Internet Draft, and
7841 is something that is documented as being done by Sendmail. There are two
7842 possible values. For messages generated by the autoreply transport, Exim
7845 Auto-Submitted: auto-replied
7847 whereas for all other generated messages (e.g. bounces) it adds
7849 Auto-Submitted: auto-generated
7851 63. The "personal" condition in filters now includes a test for the
7852 Auto-Submitted: header. If it contains the string "auto-" the message it
7853 not considered personal.
7855 64. Added rcpt_include_affixes as a generic transport option.
7857 65. Added queue_only_override (default true).
7859 66. Added the syslog_duplication option.
7861 67. If what should have been the first header line of a message consisted of
7862 a space followed by a colon, Exim was mis-interpreting it as a header line.
7863 It isn't of course - it is syntactically invalid and should therefore be
7864 treated as the start of the message body. The misbehaviour could have
7865 caused a number of strange effects, including loss of data in subsequent
7866 header lines, and spool format errors.
7868 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7869 client host had authenticated. This control can now be exercised by an ACL
7870 for more flexibility.
7872 69. By default, callouts do not happen when testing with -bh. There is now a
7873 variant, -bhc, which does actually run the callout code, including
7874 consulting and updating the callout cache.
7876 70. Added support for saslauthd authentication, courtesy of Alexander
7879 71. If statvfs() failed on the spool or log directories while checking their
7880 size for availability, Exim confusingly gave the error "space shortage".
7881 Furthermore, in debugging mode it crashed with a floating point exception.
7882 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7883 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7884 serious problem, Exim now writes to the main and panic logs when this
7885 happens, with details of the failure. It then refuses to accept the
7886 incoming message, giving the message "spool directory problem" or "log
7887 directory problem" with a 421 code for SMTP messages.
7889 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7890 0, 1, and 2 exist, because some OS complain for execs without them (see
7891 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7892 descriptors. However, the code omitted to check that the open succeeded,
7893 causing mysterious errors if for some reason the permissions on /dev/null
7894 got screwed. Now Exim writes a message to the main and panic logs, and
7895 bombs out if it can't open /dev/null.
7897 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7898 interact so that it is all more flexible. It is supposed to remain
7899 backwards compatible. Also added extra_local_interfaces.
7901 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7902 to bomb out with an assertion failure - to the client this appears as a
7903 connection drop. This problem occurs in the part of the code that was taken
7904 from the Samba project. Fortunately, the assertion is in a very simple
7905 function, so I have fixed this by reproducing the function inline in the
7906 one place where it is called, and arranging for authentication to fail
7907 instead of killing the process with assert().
7909 75. The SPA client code was not working when the server requested OEM rather
7910 than Unicode encoding.
7912 76. Added code to make require_files with a specific uid setting more usable in
7913 the case where statting the file as root fails - usually a non-root-mounted
7914 NFS file system. When this happens and the failure is EACCES, Exim now
7915 forks a subprocess and does the per-uid checking as the relevant uid.
7917 77. Added process_log_path.
7919 78. If log_file_path was not explicitly set, a setting of check_log_space or
7920 check_log_inodes was ignored.
7922 79. If a space check for the spool or log partitions fails, the incident is now
7923 logged. Of course, in the latter case the data may get lost...
7925 80. Added the %p formatting code to string_format() so that it can be used to
7926 print addresses in debug_print(). Adjusted all the address printing in the
7927 debugging in store.c to use %p rather than %d.
7929 81. There was a concern that a line of code in smtp_in.c could overflow a
7930 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7931 initially expressed, the concern was not well-founded, because trailing
7932 spaces are removed early. However, if the trailing spaces were followed by
7933 a NULL, they did not get removed, so the overflow was possible. Two fixes
7936 (a) I re-wrote the offending code in a cleaner fashion.
7937 (b) If an incoming SMTP command contains a NULL character, it is rejected
7940 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7941 runs initgroups(), so that if the Exim user is in any additional groups,
7942 they will be used during message reception.
7948 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob