1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
9 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
10 related applications. Previously an "H" was used where available info
11 says that "M" should be, so change to match.
13 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
14 as arguments, so an implementation trying to copy these into a local
15 buffer was taking a taint-enforcement trap. Fix by using dynamically
16 created buffers. Similar fix for radius expansion condition.
18 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
19 reasonable, eg. to count headers. Fix by using dynamically created
20 buffers rather than a local. Do similar fixes for ACL actions "dcc",
21 "log_reject_target", "malware" and "spam"; the arguments are expanded
22 so could be handling tainted values.
24 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
25 broken the (no-op) support for this sendmail command. Restore it
26 to doing nothing, silently, and returning good status.
28 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
29 record path was given (or the default used) without a leading directory
30 path, an error occurred on trying to open it. Use the transport's working
33 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
34 transport. Previously it was the name on the DNS A-record; use instead
35 the head of the CNAME chain leading there (if there is one). This seems
36 to align better with RFC 6125.
38 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
39 smtp_accept_max_per_host allocated resources which were not released
40 when the limit was exceeded. This eventually crashed the daemon. Fix
41 by adding a relase action in that path.
43 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
44 expanded; previously using tainted values was rejected. Fix by using
45 dynamically-created buffers.
47 JH/09 Relax restrictions on ACL verify condition needing access to message
48 headers. Previously they were only permitted in data and non-smtp ACLs;
49 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
50 not_blind, header_sender and header_names_ascii verification.
52 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
53 Previously a macro used one argument twice; when called with the
54 argument as an expression having side-effects, incorrect operation
55 resulted. Use an inlineable function.
57 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
58 held open for a verify callout. Previously this wan not accounted for
59 and a corrupt onward SMTP conversation resulted.
61 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
62 Previously they were embedded naively in the constructed address; when
63 needed, strip the quoting and quote the entire local_part.
64 Also make the inbound_srs expansion condition handle quoting.
66 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
67 excluded, not matching the documentation.
69 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
70 was given for the sqlite_dbfile a trap resulted.
72 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
73 "name" argument resulted in a trap. There is no reason to disallow such;
74 this was a coding error.
76 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
77 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
78 spent suspended, ignoring the Posix definition. Previously we assumed
79 it did and a constant offset from real time could be used as a correction.
80 Change to using the same clock source for the start-of-message and the
81 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
82 exists, just to get a clock slightly more aligned to reality.
84 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
85 RFC says it is optional some validators care. The missing char was not
86 intended but triggered by a line-wrap alignement. Discovery and fix by
87 Guillaume Outters, hacked on by JH.
89 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
90 name being quoted was tainted a trap would be taken. Fix by using
91 dynamicaly created buffers. The routine could have been called by a
92 rewrite with the "h" flag, by using the "-F" command-line option, or
93 by using a "name=" option on a control=submission ACL modifier.
95 JH/19 SPF: change the Authentication-Results expansion component to give
96 smtp.helo when the sender domain is empty. Previously it gave
99 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
100 not in 127.0.0.0/8 to help in spotting list domains taken over by a
101 domain-parking registrar.
103 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
104 Previously when a whitespace character was specified it was not inserted
105 after removing the newline.
107 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
108 the domain part of the recipient address. This overrides any tls_sni
109 option set, which was previously used.
111 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
114 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
115 is_tainted() had an off-by-one error in the overenthusiastic direction.
116 Find and fix by Gavan. Although NetBSD is not a supported platform for
117 4.94 this bug could affect other platforms.
119 PP/01 Fix default prime selection to be consistent.
120 One path used ike23 still, instead of exim.dev.20160529.3; now both
121 execution flows will use the same DH primes (currently
122 exim.dev.20160529.3).
124 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
125 option in smtp transport, to match the documentation. Previously
126 verification was not being done in some cases where it should have been.
128 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
129 than one server was defined and depending on the platform memory layout
130 details, an internal consistency trap could be hit while walking the list
133 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
134 files. The value is used by the authresults expansion item. Previously
135 if this was used in a router or transport, a crash could result.
137 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
138 have a sufficient incidence of EINTR returns from select that an
139 interaction with other operations done by the main daemon loop exposed
140 a bug in the error-handling. This was benign apart from the log
143 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
144 for consistency with delivered "=>" and failed "**" lines. While we're
145 there, handle PRX and TFO.
147 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
148 applied. This resulted, if any header-line rewrite rules were configured,
149 in a panic-log trigerrable by sending a message with a long address in
150 a header. Fix by increaing the arbitrary limit to larger than a single
151 (dewrapped) 5322 header line maximum size.
153 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
154 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
155 RFC 6648 which deprecates X- options in protocols as a general practice.
156 Changeover between the implementations is handled by the mechanisms
159 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
160 a remote which uses callout/hold. Previously the local was queued.
162 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
165 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
166 log line. The asterisk is now consistently appended to the remote IP
167 (and port, if given), and will also be provided on defer and fail log
168 lines. Previously it could be placed on the local IP if that was being
169 logged, and was only provided on delivery lines.
171 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
173 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
174 authenticator client_send option. Previously the next char, after a pair
175 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
176 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
177 documentation. There is still no way to get a leading ^ immediately
178 after a NUL (ie. for the password of a PLAIN method authenticator.
180 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
181 files. For bad sizes read, delete the record and whine to paniclog.
183 JH/38 When logging an AUTH failure, as server, do not include sensitive
184 information. Previously, the credentials would be included if given
185 as part of the AUTH command line and an ACL denied authentidcation.
187 JH/39 Bug 2691: fix $local_part_data. When the matching list element
188 referred to a file, bad data was returned. This likely also affected
196 JH/01 Avoid costly startup code when not strictly needed. This reduces time
197 for some exim process initialisations. It does mean that the logging
198 of TLS configuration problems is only done for the daemon startup.
200 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
202 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
203 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
206 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
207 DKIM signing is being done. Previously a transport_filter would always
208 disable CHUNKING, falling back to traditional DATA.
210 JH/05 Regard command-line recipients as tainted.
212 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
214 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
215 PAM library frees one of the arguments given to it, despite the
216 documentation. Therefore a plain malloc must be used.
218 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
219 on-stack buffers were used, resulting in a taint trap when DSN information
220 copied from a received message was written into the buffer.
222 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
223 the ordering of its ARC headers. This caused a crash.
225 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
226 a new record was being constructed with information from the peer, a trap
229 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
230 installation would get error messages from DMARC verify, when it hit the
231 nonexistent file indicated by the default. Distros wanting DMARC enabled
232 should both provide the file and set the option.
233 Also enforce no DMARC verification for command-line sourced messages.
235 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
236 could, depending on the platform, hang at the STARTTLS response.
238 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
239 message on a connection. Previously if one message had ARC headers and
240 the following one did not, a crash could result when adding an
241 Authentication-Results: header.
243 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
244 Exim main code and Exim-related utities. The introduction of taint
245 tracking also did many adjustments to string handling. Since then, eximon
246 frequently terminated with an assert failure.
248 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
249 check for 452 responses. This slightly helps the inefficieny of doing
250 a large alias-expansion into a recipient-limited target. The max_rcpt
251 transport option still applies (and at the current default, will override
252 the new feature). The check is done for either cause of synch, and forces
253 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
254 connection. The new facility is not tunable at this time.
256 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
257 library live data was being used, so the results became garbage. Make
258 copies while it is still usable.
260 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
261 on delivery deferred (==) and failed (**) lines (if a delivery was
262 attemtped). Previously it was only on completion (=>) lines.
264 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
265 for the expansion of the server_scram_iter and server_scram_salt options.
267 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
268 are now specifically given a NO_DATA response without hitting the system
269 resolver. The library goes on to do the now-standard TXT lookup.
270 Use of dnsdb lookups is not affected.
272 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
273 only retrieve the errormessage once. Previously two calls to dlerror()
274 were used, and the second one (for mainlog/paniclog) retrieved null
277 JH/20 Taint checking: disallow use of tainted data for
278 - the appendfile transport file and directory options
279 - the pipe transport command
280 - the autoreply transport file, log and once options
281 - file names used by the redirect router (including filter files)
283 - paths used by single-key lookups
284 Previously this was permitted.
286 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
287 adjusted the size of a major service buffer; this failed because the
288 buffer was in use at the time. Change to a compile-time increase in the
289 buffer size, when this authenticator is compiled into exim.
291 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
292 previous fast-mode was untenable in the face of glibs using mmap to
293 support larger malloc requests.
295 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
296 New values supported, if defined on system where compiled:
297 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
298 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
300 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
301 running a limited number of proceses in parallel, a benefit is gained. The
302 amount varies with the platform hardware and load. The use of the option
303 queue_run_in_order means we cannot do this, as ordering becomes
306 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
307 had introduced a string-copy (for ensuring NUL-termination) which was not
308 appropriate for that case, which can include embedded NUL bytes in the
309 block of data. Investigation showed the copy to actually be needless, the
310 data being length-specified.
312 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
313 done during a receiving connection, and both used TLS, global info was
314 used rather than per-connection info for tracking the state of data
315 queued for transmission. This could result in a connection hang.
317 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
318 Previously, when delivering serveral messages down a single connection
319 only the first would provide a SIZE. This was due to the size information
320 not being properly tracked.
322 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
323 TAI (at 37 seconds currently), pretend to be in UTC for time-related
324 expansion and logging. Previously, spurious values such as a future
325 minute could be seen.
327 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
328 it could crash from a null-deref. This could also affect the
329 ${addresses: } operator and ${readsock } item.
331 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
332 message following a mime one, the variable was not reset.
334 JH/30 When an pipelined-connect fails at the first response, assume incorrect
335 cached capability (perhaps the peer reneged?) and immediately retry in
338 JH/31 Fix spurious detection of timeout while writing to transport filter.
340 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
341 an attempt to copy the string was made before checking it.
343 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
344 taint of the lookup key was maintained; we now regard the presence in the
345 filesystem as sufficient validation.
347 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
348 argument is supplied.
350 JH/35 The dsearch lookup now requires that the directory is an absolute path.
351 Previously this was not checked, and nonempty relative paths made an
352 access under Exim's current working directory.
354 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
355 Previously no event was raised.
357 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
358 parameter supplied by the sender MAIL FROM command. Previously it was
359 ignored, and only the check_spool_space option value for the required
362 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
363 the size of the signing public-key. Previously it was instead giving
364 the size of the signature hash.
366 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
367 the default. See the (new) dkim_verify_min_keysizes option.
369 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
370 and an ACL use a lookup for checking either the local_part or domain,
371 stale data could be accessed. Ensure that variable references are
372 dropped between messages.
374 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
375 by the client was not checked as pointing within response data before
376 being used. A malicious client could thus cause an out-of-bounds read and
377 possibly gain authentication. Fix by adding the check.
379 JH/42 Internationalisation: change the default for downconversion in the smtp
380 transport to be "if needed". Previously it was "as previously set" for
381 the message, which usually meant "if needed" for message-submission but
382 "no" for everything else. However, MTAs have been seen using SMTPUTF8
383 even when the envelope addresses did not need it, resulting in forwarding
384 failures to non-supporting MTAs. A downconvert in such cases will be
385 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
386 transport. The change does mean that addresses needing conversion will
387 be converted when previously a delivery failure would occur.
389 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
390 response was received it would be used unchecked in a fail-DSN, violating
391 standards on line-length limits. Truncate if needed.
393 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
394 doesn't work. (Additionally add a new main config option to configure the
401 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
402 side, to decode a TLS 1.3 packet capture.
404 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
405 Previously the default library behaviour applied, sending two, each in
408 JH/03 Debug output for ACL now gives the config file name and line number for
411 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
413 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
415 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
416 buffer overrun for (non-chunking) other transports.
418 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
419 TLS1.3, means that a server rejecting a client certificate is not visible
420 to the client until the first read of encrypted data (typically the
421 response to EHLO). Add detection for that case and treat it as a failed
422 TLS connection attempt, so that the normal retry-in-clear can work (if
423 suitably configured).
425 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
426 and/or domain. Found and fixed by Jason Betts.
428 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
429 configuration). If a CNAME target was not a wellformed name pattern, a
432 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
433 the OS reports them interleaved with other addresses.
435 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
436 used both for input and for a verify callout, both encrypted, SMTP
437 responses being sent by the server could be lost. This resulted in
438 dropped connections and sometimes bounces generated by a peer sending
441 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
442 string. Previously it was possible to cause undefined behaviour in a
443 library routine (usually a crash). Found by "zerons".
445 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
448 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
449 API was removed, so update to use the newer ones.
451 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
452 any timeout set, is taking a long time. Previously we would hang on to a
453 rotated logfile "forever" if the input was arriving with long gaps
454 (a previous attempt to fix addressed lack, for a long time, of initial
457 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
458 shared (NFS) environment. The length of the tempfile name is now
459 4 + 16 ("hdr.$message_exim_id") which might break on file
460 systems which restrict the file name length to lower values.
463 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
464 shared (NFS) environment.
466 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
467 did for all versions <4.90). Notably -M, -m, --invert, -I may be
470 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
471 on some platforms for bit 31.
473 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
474 to changes apparently associated with TLS1.3 handling some of the APIs
475 previously used were either nonfunctional or inappropriate. Strings
476 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
477 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
478 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
479 This affects log line X= elements, the $tls_{in,out}_cipher variables,
480 and the use of specific cipher names in the encrypted= ACL condition.
482 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
484 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
485 verification result was not updated unless hosts_require_ocsp applied.
487 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
488 queue_list_requires_admin set to false, non-admin users were denied the
491 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
492 directory-of-certs mode. Previously they were advertised despite the
495 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
496 A single TCP connection by a client will now hold a TLS connection open
497 for multiple message deliveries, by default. Previously the default was to
500 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
501 default. If built with the facility, DANE will be used. The facility
502 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
504 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
505 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
506 must be defined and you must still, unless you define DISABLE_TLS, manage
507 the the include-dir and library-file requirements that go with that
508 choice. Non-TLS builds are still supported.
510 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
513 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
514 default. If the platform supports and has the facility enabled, it will
515 be requested on all coneections.
517 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
518 controlled by the build-time option SUPPORT_PIPE_CONNECT.
520 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
522 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
523 success-DSN messages. Previously the From: header was always the default
524 one for these; the option was ignored.
526 JH/28 Fix the timeout on smtp response to apply to the whole response.
527 Previously it was reset for every read, so a teergrubing peer sending
528 single bytes within the time limit could extend the connection for a
529 long time. Credit to Qualsys Security Advisory Team for the discovery.
531 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
532 delivery address, which leaked information of the results of local
533 forwarding. Change to the original envelope recipient address, per
536 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
537 requested. Previously not bounce was generated and a log entry of
538 error ignored was made.
540 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
542 JH/32 Introduce a general tainting mechanism for values read from the input
543 channel, and values derived from them. Refuse to expand any tainted
544 values, to catch one form of exploit.
546 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
547 was unused and the unexpanded text used for the test. Found and
548 fixed by Ruben Jenster.
550 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
551 an attempt to use a TLS library read routine dereffed a nul pointer,
554 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
555 them in our smtp response.
557 JH/36 Have the general router option retry_use_local_part default to true when
558 any of the restrictive preconditions are set (to anything). Previously it
559 was only for check_local user. The change removes one item of manual
560 configuration which is required for proper retries when a remote router
561 handles a subset of addresses for a domain.
563 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
564 link count into consideration.
566 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
567 caused the extension of big_buffer, the following lines were ignored.
569 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
570 accordance with RFC 2308. Previously there was no expiry, so a longlived
571 receive process (eg. due to ACL delays) versus a short SOA value could
574 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
576 JH/39 Promote DMARC support to mainline.
578 JH/40 Bug 2452: Add a References: header to DSNs.
580 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
581 parameters. The relevant library call is documented as "Deprecated: This
582 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
583 3.6.0, DH parameters are negotiated following RFC7919."
585 HS/06 Change the default of dnssec_request_domains to "*"
587 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
588 carried on and emitted a BDAT command, even when PIPELINING was not
591 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
592 buffer was used for the filename, resulting in a trap when tainted
593 arguments (eg. $domain) were used.
595 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
596 recommended to avoid a possible server-load attack. The feature can be
597 re-enabled via the openssl_options main cofiguration option.
599 JH/45 local_scan API: documented the current smtp_printf() call. This changed
600 for version 4.90 - adding a "more data" boolean to the arguments.
601 Bumped the ABI version number also, this having been missed previously;
602 release versions 4.90 to 4.92.3 inclusive were effectively broken in
603 respect of usage of smtp_printf() by either local_scan code or libraries
604 accessed via the ${dlfunc } expansion item. Both will need coding
605 adjustment for any calls to smtp_printf() to match the new function
606 signature; a FALSE value for the new argument is always safe.
608 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
609 the file-offset (which the Linux syscall does, and exim expects); this
610 resulted in an indefinite loop.
612 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
613 to do ARC verification. The Authentication-Results: header line added
614 by the configuration then had no ARC item.
620 JH/01 Remove code calling the customisable local_scan function, unless a new
621 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
623 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
624 non-signal-safe functions being used.
626 JH/03 Bug 2269: When presented with a received message having a stupidly large
627 number of DKIM-Signature headers, disable DKIM verification to avoid
628 a resource-consumption attack. The limit is set at twenty.
630 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
631 report of oldest_pass in ${authres } in consequence, and separate out
632 some descriptions of reasons for verification fail.
634 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
635 files in the spool were present and unlocked. A queue-runner could spot
636 them, resulting in a duplicate delivery. Fix that by doing the unlock
637 after the unlink. Investigation by Tim Stewart. Take the opportunity to
638 add more error-checking on spoolfile handling while that code is being
641 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
642 No known attacks, no CVE, this is defensive hardening.
644 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
645 a queue-runner could start a delivery while other operations were ongoing.
646 Cutthrough delivery was a common victim, resulting in duplicate delivery.
647 Found and investigated by Tim Stewart. Fix by using the open message data
648 file handle rather than opening another, and not locally closing it (which
649 releases a lock) for that case, while creating the temporary .eml format
650 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
652 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
653 $sender_verify_failure/$recipient_verify_failure to "random".
655 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
658 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
659 Previously this would segfault.
661 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
664 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
665 like zero, since the resolver should be doing this for us, But we need one
666 as a CNAME but no MX presence gets the CNAME returned; we need to check
667 that doesn't point to an MX to declare it "no MX returned" rather than
668 "error, loop". A new main option is added so the older capability of
669 following some limited number of chain links is maintained.
671 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
673 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
674 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
675 now-deprecated earlier definitions used only specified the range up to TLS
676 1.2 (in the older-version library docs).
678 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
680 JH/15 Rework TLS client-side context management. Stop using a global, and
681 explicitly pass a context around. This enables future use of TLS for
682 connections to service-daemons (eg. malware scanning) while a client smtp
683 connection is using TLS; with cutthrough connections this is quite likely.
685 JH/16 Fix ARC verification to do AS checks in reverse order.
687 JH/17 Support a "tls" option on the ${readsocket } expansion item.
689 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
690 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
691 Previously the "utf8" would be re-prepended for every additional message.
693 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
694 Previously thery were accepted, resulting in issues when attempting to
695 forward messages to a non-supporting MTA.
697 PP/02 Let -n work with printing macros too, not just options.
699 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
700 one parent address was copied, and bogus data was used at delivery-logging
701 time. Either a crash (after delivery) or bogus log data could result.
702 Discovery and analysis by Tim Stewart.
704 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
705 Previously if the string ended mid-character, we did not insert the
706 promised '?' replacement.
708 PP/04 Documentation: current string operators work on bytes, not codepoints.
710 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
711 should pack well giving a smaller memory footprint so better caching and
712 therefore performance. Group the declarations where this can't be done so
713 that the byte-sized flag variables are not interspersed among pointer
714 variables, giving a better chance of good packing by the compiler.
716 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
717 non-null, to avoid issues with sites running BATV. Previously reports were
718 sent with an empty envelope sender so looked like bounces.
720 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
721 The ignore_error flag wasn't being returned from the filter subprocess so
722 was not set for later routers. Investigation and fix by Matthias Kurz.
724 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
725 and a msg:complete for the whole, when a message is manually removed using
726 -Mrm. Developement by Matthias Kurz, hacked on by JH.
728 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
729 a "Gnu special" function, asprintf() in the DB utility binary builds; I
730 hope that is portable enough.
732 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
733 requiring a known-CA anchor certificate; make it now rely entirely on the
734 TLSA as an anchor. Checking the name on the leaf cert against the name
735 on the A-record for the host is still done for TA (but not for EE mode).
737 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
738 included in delivery lines for non-proxied connections, when compiled with
739 SUPPORT_SOCKS and running with proxy logging enabled.
741 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
742 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
743 move the existing event to fire before the normal logging of message
744 failure so that custom logging is bracketed by normal logging.
746 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
747 msg:fail:internal event. Developement by Matthias Kurz.
749 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
750 far too small for todays use of crypto signatures stored there. Go all
751 the way to the max DNS message size of 64kB, even though this might be
752 overmuch for IOT constrained device use.
754 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
755 copy a string over itself. The library routine is documented as not
756 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
758 JH/32 For main options check_spool_space and check_inode_space, where the
759 platform supports 64b integers, support more than the previous 2^31 kB
760 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
761 the previous G, M, k.
763 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
764 $authenticated_fail_id variable on authentication failure. Previously
767 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
768 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
769 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
770 GNUTLS_SEC_PARAM_MEDIUM.
772 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
773 side. Previously we would continue as if no SNI had been received.
775 JH/36 Harden the handling of string-lists. When a list consisted of a sole
776 "<" character, which should be a list-separator specification, we walked
777 off past the nul-terimation.
779 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
780 causes) even when the retry time is not yet met. Previously they were
781 not, meaning that when (say) an account was over-quota and temp-rejecting,
782 and multiple senders' messages were queued, only one sender would get
783 notified on each configured delay_warning cycle.
785 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
787 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
788 cipher-suites, an error can be left on the stack even for a succeeding
789 accept; this results in impossible error messages when a later operation
792 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
793 return error codes indicating retry. Under TLS1.3 this becomes required.
795 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
796 it only wrote the new authenticators, resulting in a lack of tracking of
797 peer changes of ESMTP extensions until the next cache flush.
799 JH/41 Fix the loop reading a message header line to check for integer overflow,
800 and more-often against header_maxsize. Previously a crafted message could
801 induce a crash of the recive process; now the message is cleanly rejected.
803 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
804 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
810 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
811 When redis_servers is set to a list of > 1 element, and the Redis servers
812 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
813 case of MOVED into a DEFER case instead, thus moving the query onto the
814 next server in the list. For a cluster of N elements, all N servers must
815 be defined in redis_servers.
817 GF/02 Catch and remove uninitialized value warning in exiqsumm
818 Check for existence of @ARGV before looking at $ARGV[0]
820 JH/01 Replace the store_release() internal interface with store_newblock(),
821 which internalises the check required to safely use the old one, plus
822 the allocate and data copy operations duplicated in both (!) of the
823 extant use locations.
825 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
826 modifier. This matches the restriction on the commandline.
828 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
829 Previously only the last row was returned.
831 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
832 we assumed that tags in the header were well-formed, and parsed the
833 element content after inspecting only the first char of the tag.
834 Assumptions at that stage could crash the receive process on malformed
837 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
838 While running the DKIM ACL we operate on the Permanent memory pool so that
839 variables created with "set" persist to the DATA ACL. Also (at any time)
840 DNS lookups that fail create cache records using the Permanent pool. But
841 expansions release any allocations made on the current pool - so a dnsdb
842 lookup expansion done in the DKIM ACL releases the memory used for the
843 DNS negative-cache, and bad things result. Solution is to switch to the
844 Main pool for expansions.
845 While we're in that code, add checks on the DNS cache during store_reset,
846 active in the testsuite.
847 Problem spotted, and debugging aided, by Wolfgang Breyha.
849 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
850 When none of the hosts presented to a transport match an already-open
851 connection, close it and proceed with the list. Previously we would
852 queue the message. Spotted by Lena with Yahoo, probably involving
855 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
856 Previously a spurious "250 OK id=" response was appended to the proper
859 JH/08 The "support for" informational output now, which built with Content
860 Scanning support, has a line for the malware scanner interfaces compiled
861 in. Interface can be individually included or not at build time.
863 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
864 by the template makefile "src/EDITME". The "STREAM" support for an older
865 ClamAV interface method is removed.
867 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
868 rows affected is given instead).
870 JH/11 The runtime Berkeley DB library version is now additionally output by
871 "exim -d -bV". Previously only the compile-time version was shown.
873 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
874 SMTP connection. Previously, when one had more recipients than the
875 first, an abortive onward connection was made. Move to full support for
876 multiple onward connections in sequence, handling cutthrough connection
877 for all multi-message initiating connections.
879 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
880 routers. Previously, a multi-recipient message would fail to match the
881 onward-connection opened for the first recipient, and cause its closure.
883 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
884 a timeout on read on a GnuTLS initiating connection, resulting in the
885 initiating connection being dropped. This mattered most when the callout
886 was marked defer_ok. Fix to keep the two timeout-detection methods
889 JH/15 Relax results from ACL control request to enable cutthrough, in
890 unsupported situations, from error to silently (except under debug)
891 ignoring. This covers use with PRDR, frozen messages, queue-only and
894 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
896 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
897 metadata, resulting in a crash in free().
899 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
900 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
901 Broken also in d185889f4, with init system revamp.
903 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
904 Previously we abruptly closed the connection after reading a malware-
905 found indication; now we go on to read the "scan ok" response line,
908 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
909 ACL. Previously, a crash would result.
911 JH/19 Speed up macro lookups during configuration file read, by skipping non-
912 macro text after a replacement (previously it was only once per line) and
913 by skipping builtin macros when searching for an uppercase lead character.
915 JH/20 DANE support moved from Experimental to mainline. The Makefile control
916 for the build is renamed.
918 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
919 was allocated for every new TLS startup, meaning one per message. Fix
920 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
922 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
923 reported the original. Fix to report (as far as possible) the ACL
924 result replacing the original.
926 JH/23 Fix memory leak during multi-message connections using STARTTLS under
927 OpenSSL. Certificate information is loaded for every new TLS startup,
928 and the resources needed to be freed.
930 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
932 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
935 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
936 DATA response info to the (existing) per-recipient response info for
937 the "C=" log element. It can have useful tracking info from the
938 destination system. Patch from Simon Arlott.
940 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
941 length value. Previously this would segfault.
943 HS/02 Support Avast multiline protoocol, this allows passing flags to
944 newer versions of the scanner.
946 JH/28 Ensure that variables possibly set during message acceptance are marked
947 dead before release of memory in the daemon loop. This stops complaints
948 about them when the debug_store option is enabled. Discovered specifically
949 for sender_rate_period, but applies to a whole set of variables.
950 Do the same for the queue-runner and queue-list loops, for variables set
951 from spool message files. Do the same for the SMTP per-message loop, for
952 certain variables indirectly set in ACL operations.
954 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
955 as a multi-recipient message from a mailinglist manager). The coding had
956 an arbitrary cutoff number of characters while checking for more input;
957 enforced by writing a NUL into the buffer. This corrupted long / fast
958 input. The problem was exposed more widely when more pipelineing of SMTP
959 responses was introduced, and one Exim system was feeding another.
960 The symptom is log complaints of SMTP syntax error (NUL chars) on the
961 receiving system, and refused recipients seen by the sending system
962 (propating to people being dropped from mailing lists).
963 Discovered and pinpointed by David Carter.
965 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
966 replaced by the ${authresults } expansion.
968 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
970 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
971 allows proper process termination in container environments.
973 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
974 Previously the "final dot" had a newline after it; ensure it is CR,LF.
976 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
977 and "err_perm", deprecated since 4.83 when the RFC-defined words
978 "temperror" and "permerror" were introduced.
980 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
981 transport-filters or DKIM-signing. The restriction was lost in the
982 consolidation of verify-callout and delivery SMTP handling.
983 Extend the restriction to also cover ARC-signing.
985 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
986 in defer=pass mode supply a 450 to the initiator. Previously the message
989 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
990 tls_require_ciphers is used as before.
992 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
993 "pass_unscanned". Only tmpfails from the scanner are written to
994 the paniclog, as they may require admin intervention (permission
995 denied, license issues). Other scanner errors (like decompression
996 bombs) do not cause a paniclog entry.
998 JH/36 Fix reinitialisation of DKIM logging variable between messages.
999 Previously it was possible to log spurious information in receive log
1002 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1003 triggered odd behaviour from Outlook Express clients.
1005 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1008 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1009 since the IETF WG has not yet settled on that versus the original
1010 "bare" representation.
1012 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1013 Previously the millisecond value corrupted the output.
1014 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1015 corrupted the output.
1021 JH/01 Rework error string handling in TLS interface so that the caller in
1022 more cases is responsible for logging. This permits library-sourced
1023 string to be attached to addresses during delivery, and collapses
1024 pairs of long lines into single ones.
1026 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1027 during configuration. Wildcards are allowed and expanded.
1029 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1030 This permits better logging.
1032 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1033 do not close it down and have the child transport start it up again on
1034 the passed-on TCP connection. Instead, proxy the child (and any
1035 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1036 affected: the continued delivery log lines do not have any DNSSEC, TLS
1037 Certificate or OCSP information. TLS cipher information is still logged.
1039 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1040 identical IP addresses on different listening ports. Will also affect
1043 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1044 add noisy ifdef guards to special-case this sillyness.
1045 Patch from Bernd Kuhls.
1047 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1048 than 255 are no longer allowed.
1050 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1051 Disable the session-cache too, which might reduce our load. Since we
1052 currrectly use a new context for every connection, both as server and
1053 client, there is no benefit for these.
1054 GnuTLS appears to not support tickets server-side by default (we don't
1055 call gnutls_session_ticket_enable_server()) but client side is enabled
1056 by default on recent versions (3.1.3 +) unless the PFS priority string
1059 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1060 <https://reproducible-builds.org/specs/source-date-epoch/>.
1062 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1063 the check for any unsuccessful recipients did not notice the limit, and
1064 erroneously found still-pending ones.
1066 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1067 MSG_MORE. Only in-clear (not on TLS connections).
1069 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1070 transport-filter is involved we can buffer the headers in memory for
1071 creating the signature, and read the spool data file once for the
1072 signature and again for transmission.
1074 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1075 4.77 as the kernel support then wasn't solid, having issues in 64bit
1076 mode. Now, it's been long enough. Add support for FreeBSD also.
1078 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1079 case where the routing stage had gathered several addresses to send to
1080 a host before calling the transport for the first, we previously failed
1081 to close down TLS in the old transport process before passing the TCP
1082 connection to the new process. The new one sent a STARTTLS command
1083 which naturally failed, giving a failed delivery and bloating the retry
1084 database. Investigation and fix prototype from Wolfgang Breyha.
1086 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1087 false-negatives in the check that the sender had not preempted a response
1088 or prompt from Exim (running as a server), due to that code's lack of
1089 awareness of the SMTP input buffering.
1091 PP/04 Add commandline_checks_require_admin option.
1092 Exim drops privileges sanely, various checks such as -be aren't a
1093 security problem, as long as you trust local users with access to their
1094 own account. When invoked by services which pass untrusted data to
1095 Exim, this might be an issue. Set this option in main configuration
1096 AND make fixes to the calling application, such as using `--` to stop
1099 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1100 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1101 responses to those, into a single TLS record each way (this usually means
1102 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1105 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1106 affects you only if you're dancing at the edge of the param size limits.
1107 If you are, and this message makes sense to you, then: raise the
1108 configured limit or use OpenSSL 1.1. Nothing we can do for older
1111 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1112 cmdline element to get the documented default one. Previously it was
1115 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1116 are unsuable for non-message-specific reasons - eg. connection timeout,
1117 banner-time rejection.
1119 JH/16 Fix logging of delivery remote port, when specified by router, under
1122 PP/06 Repair manualroute's ability to take options in any order, even if one
1123 is the name of a transport.
1126 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1128 JH/17 Change the list-building routines interface to use the expanding-string
1129 triplet model, for better allocation and copying behaviour.
1131 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1132 Previously it was constructed the first time a possibly-matching string
1133 was met in the configuration file input during startup; now it is done
1136 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1137 compatible one, to avoid the (poorly documented) possibility of a config
1138 file in the working directory redirecting the DB files, possibly correpting
1139 some existing file. CVE-2017-10140 assigned for BDB.
1141 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1142 cache-hot. Previously, although the result was properly cached, the
1143 initial verify call returned a defer.
1145 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1146 the main verify for receipient in uncached-mode.
1148 JH/22 Retire historical build files to an "unsupported" subdir. These are
1149 defined as "ones for which we have no current evidence of testing".
1151 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1152 if present. Previously it was ignored.
1154 JH/24 Start using specified-initialisers in C structure init coding. This is
1155 a C99 feature (it's 2017, so now considered safe).
1157 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1158 if was a fixed-sized field and bitmask ops via macros; it is now more
1161 PP/07 GitHub PR 56: Apply MariaDB build fix.
1162 Patch provided by Jaroslav Škarvada.
1164 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1165 during Coverity cleanups [4.87 JH/47]
1166 Diagnosis and fix provided by Michael Fischer v. Mollard.
1168 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1169 the right size to place the terminating semicolon on its own folded
1170 line, the header hash was calculated to an incorrect value thanks to
1171 the (relaxed) space the fold became.
1173 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1174 and confused the parent.
1176 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1177 which could crash as a result. This could lead to undeliverable messages.
1179 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1182 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1183 matters on fast-turnover and PID-randomising systems, which were getting
1184 out-of-order delivery.
1186 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1187 a possibly-overlapping copy. The symptom was that "Remote host closed
1188 connection in response to HELO" was logged instead of the actual 4xx
1191 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1192 Previously only that bufferd was discarded, resulting in SYMTP command
1195 JH/32 DKIM: when a message has multiple signatures matching an identity given
1196 in dkim_verify_signers, run the dkim acl once for each. Previously only
1197 one run was done. Bug 2189.
1199 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1200 "panic the current process" to "deliberately defer". The panic log is
1201 still written with the problem list name; the mail and reject logs now
1202 get a temp-reject line for the message that was being handled, saying
1203 something like "domains check lookup or other defer". The SMTP 451
1204 message is still "Temporary local problem".
1206 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1207 A crafted sequence of BDAT commands could result in in-use memory beeing
1208 freed. CVE-2017-16943.
1210 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1211 from SMTP input. Previously it was always done; now only done for DATA
1212 and not BDAT commands. CVE-2017-16944.
1214 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1215 to the message (such as an overlong header line). Previously this was
1216 not done and we did not exit BDAT mode. Followon from the previous item
1217 though a different problem.
1223 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1224 than -2003 did; needs libidn2 in addition to libidn.
1226 JH/02 The path option on a pipe transport is now expanded before use.
1228 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1229 Patch provided by "Björn", documentation fix added too.
1231 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1232 missing a wire-to-host endian conversion.
1234 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1235 close after a BDAT command line could be taken as a following command,
1236 giving a synch failure. Fix by only checking for synch immediately
1237 before acknowledging the chunk.
1239 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1240 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1241 macro. Patches provided by Josh Soref.
1243 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1244 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1245 by IANA as a service extension. However, John Klensin suggests that we
1248 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1249 the dkim code may be unix-mode line endings rather than smtp wire-format
1250 CRLF, so prepend a CR to any bare LF.
1252 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1253 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1255 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1256 the first were themselves being wrongly included in the feed into dkim
1257 processing; with most chunk sizes in use this resulted in an incorrect
1258 body hash calculated value.
1260 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1261 DKIM signature block, for verification. Although advised against by
1262 standards it is specifically not ruled illegal.
1264 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1266 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1267 missing a body hash (the bh= tag).
1269 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1270 It seems that HAProxy sends the Proxy Protocol information in clear and
1271 only then does a TLS startup, so do the same.
1273 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1274 TCP connections (such as for Spamd) unless the daemon successfully set
1275 Fast Open mode on its listening sockets. This fixes breakage seen on
1276 too-old kernels or those not configured for Fast Open, at the cost of
1277 requiring both directions being enabled for TFO, and TFO never being used
1278 by non-daemon-related Exim processes.
1280 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1281 endings, at least on the first header line. Try to canonify any that get
1282 past that check, despite the cost.
1284 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1285 now limited to an arbitrary five deep, while parsing addresses with the
1286 strip_excess_angle_brackets option enabled.
1288 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1289 instead leave the unprompted TLS handshake in socket buffer for the
1290 TLS library to consume.
1292 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1294 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1296 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1297 region in one operation, for speed, and this leaves assigned pointers
1298 dangling. Add checks run only under the testsuite which checks all
1299 variables at a store-reset and panics on a dangling pointer; add code
1300 explicitly nulling out all the variables discovered. Fixes one known
1301 bug: a transport crash, where a dangling pointer for $sending_ip_address
1302 originally assigned in a verify callout, is re-used.
1304 PP/06 Drop '.' from @INC in various Perl scripts.
1306 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1308 PP/08 Reduce a number of compilation warnings under clang; building with
1309 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1310 should be warning-free.
1312 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1314 HS/01 Fix portability problems introduced by PP/08 for platforms where
1315 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1317 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1318 chunk. This allows us to accept broken chunked messages. We need a more
1319 general solution here.
1321 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1322 already-broken messages in the queue.
1324 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1326 JH/19 Fix reference counting bug in routing-generated-address tracking.
1332 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1333 supports it and a size is available (ie. the sending peer gave us one).
1335 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1336 years of being deprecated). The replacements are the ACLs
1337 acl_smtp_mime and acl_not_smtp_mime.
1339 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1340 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1341 or even an in-clear connection were permitted. Now, if the host lookup
1342 was dnssec and dane was requested then the host is only used if the
1343 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1344 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1345 if one fails this test.
1346 This means that a poorly-configured remote DNS will make it incommunicado;
1347 but it protects against a DNS-interception attack on it.
1349 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1350 non-noisy when a race steals the message being considered.
1352 JH/05 If main configuration option tls_certificate is unset, generate a
1353 self-signed certificate for inbound TLS connections.
1355 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1356 in rewrites and routers.
1358 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1359 and logged a warning sing 4.83; now they are a configuration file error.
1361 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1362 (lacking @domain). Apply the same qualification processing as RCPT.
1364 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1366 JH/10 Support ${sha256:} applied to a string (as well as the previous
1369 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1370 a cutthrough deliver is pending, as we always want to make a connection.
1371 This also avoids re-routing the message when later placing the cutthrough
1372 connection after a verify cache hit.
1373 Do not update it with the verify result either.
1375 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1376 when routing results in more than one destination address.
1378 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1379 signing (which inhibits the cutthrough capability). Previously only
1380 the presence of an option was tested; now an expansion evaluating as
1381 empty is permissible (obviously it should depend only on data available
1382 when the cutthrough connection is made).
1384 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1385 the relevant preceding SMTP command did not note the pipelining mode.
1387 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1388 Previously they were not counted.
1390 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1391 as one having no matching records. Previously we deferred the message
1392 that needed the lookup.
1394 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1395 distinguished as "(=".
1397 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1398 for missing MX records. Previously it only worked for missing A records.
1400 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1402 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1403 after the data-go-ahead and data-ack. Patch from Jason Betts.
1405 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1406 even for a "none" policy. Patch from Tony Meyer.
1408 JH/22 Fix continued use of a connection for further deliveries. If a port was
1409 specified by a router, it must also match for the delivery to be
1412 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1413 When one of the recipients of a message was unsuitable for the connection
1414 (has no matching addresses), we lost track of needing to mark it
1415 deferred. As a result mail would be lost.
1417 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1419 JH/25 Decoding ACL controls is now done using a binary search; the source code
1420 takes up less space and should be simpler to maintain. Merge the ACL
1421 condition decode tables also, with similar effect.
1423 JH/26 Fix problem with one_time used on a redirect router which returned the
1424 parent address unchanged. A retry would see the parent address marked as
1425 delivered, so not attempt the (identical) child. As a result mail would
1428 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1429 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1430 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1433 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1434 Main config options check_{log,spool}_{inodes,space} are now
1435 100 inodes, 10MB unless set otherwise in the configuration.
1437 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1438 Previously it only applied to the main-section connection policy
1441 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1443 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1444 by me. Added RFC7919 DH primes as an alternative.
1446 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1447 are not in the system include path.
1449 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1450 GnuTLS, when a session startup failed (eg because the client disconnected)
1451 Exim did stdio operations after fclose. This was exposed by a recent
1452 change which nulled out the file handle after the fclose.
1454 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1455 signed directly by the cert-signing cert, rather than an intermediate
1456 OCSP-signing cert. This is the model used by LetsEncrypt.
1458 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1460 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1461 an incoming connection.
1463 HS/02 Bug 1802: Do not half-close the connection after sending a request
1466 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1467 fallback to "prime256v1".
1469 JH/34 SECURITY: Use proper copy of DATA command in error message.
1470 Could leak key material. Remotely exploitable. CVE-2016-9963.
1476 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1477 and 3.4.4 - once the server is enabled to respond to an OCSP request
1478 it does even when not requested, resulting in a stapling non-aware
1479 client dropping the TLS connection.
1481 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1482 support variable-length bit vectors. No functional change.
1484 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1485 The I= interface field on outgoing lines is now after the H= remote
1486 host field, same as incoming lines. There is a separate
1487 outgoing_interface log selector which allows you to disable the
1490 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1491 If not running log_selector +smtp_connection the mainlog would be held
1492 open indefinitely after a "too many connections" event, including to a
1493 deleted file after a log rotate. Leave the per net connection logging
1494 leaving it open for efficiency as that will be quickly detected by the
1495 check on the next write.
1497 HS/01 Bug 1671: Fix post transport crash.
1498 Processing the wait-<transport> messages could crash the delivery
1499 process if the message IDs didn't exist for some reason. When
1500 using 'split_spool_directory=yes' the construction of the spool
1501 file name failed already, exposing the same netto behaviour.
1503 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1504 mime_regex ACL conditions.
1506 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1507 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1508 to HELO, local diagnostic string.
1510 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1511 log line to debug. Even when configured with a tls authenticator many
1512 client connections are expected to not authenticate in this way, so
1513 an authenticate fail is not an error.
1515 HS/02 Add the Exim version string to the process info. This way exiwhat
1516 gives some more detail about the running daemon.
1518 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1519 matter for fast-change records such as DNSBLs.
1521 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1522 retry record, even if constant. There may be multiple transports with
1523 different interface settings and the retry behaviour needs to be kept
1526 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1528 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1530 JH/10 Bug 840: fix log_defer_output option of pipe transport
1532 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1533 hosts_randomize. This matters a lot when combined with mua_wrapper.
1535 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1536 ${quote_pgsql:<string>} operator.
1538 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1540 JH/14 Change header file name for freeradius-client. Relevant if compiling
1541 with Radius support; from the Gentoo tree and checked under Fedora.
1543 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1545 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1546 option transport_filter, meaning no filtering.
1548 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1550 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1551 defaults to "*" (all hosts). The variable is now available when not built
1552 with TLS, default unset, mainly to enable keeping the testsuite sane.
1553 If a server certificate is not supplied (via tls_certificate) an error is
1554 logged, and clients will find TLS connections fail on startup. Presumably
1555 they will retry in-clear.
1556 Packagers of Exim are strongly encouraged to create a server certificate
1557 at installation time.
1559 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1560 with the $config_file variable.
1562 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1563 in transport context, after the attempt, and per-recipient. The latter type
1564 is per host attempted. The event data is the error message, and the errno
1565 information encodes the lookup type (A vs. MX) used for the (first) host,
1566 and the trailing two digits of the smtp 4xx response.
1568 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1569 to write to mainlog (or rejectlog, paniclog) in the window between file
1570 creation and permissions/ownership being changed. Particularly affects
1571 installations where exicyclog is run as root, rather than exim user;
1572 result is that the running daemon panics and dies.
1574 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1576 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1577 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1578 "pri" and "weight". Note that the previous implicit priority given by the
1579 list order is no longer honoured.
1581 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1582 for DKIM processing.
1584 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1585 by defining SUPPORT_SOCKS.
1587 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1588 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1589 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1590 variables are renamed to proxy_{local,external}_{address,port}.
1592 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1593 for a build by defining SUPPORT_I18N
1595 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1596 of the query string, and make ${quote_redis:} do that quoting.
1598 JH/29 Move Events support from Experimental to mainline, enabled by default
1599 and removable for a build by defining DISABLE_EVENT.
1601 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1603 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1604 cached by the daemon.
1606 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1607 by defining LOOKUP_REDIS. The libhiredis library is required.
1609 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1610 keys are given for lookup.
1612 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1613 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1614 only supported when built with TLS support. The PolarSSL SHA routines
1615 are still used when the TLS library is too old for convenient support.
1617 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1618 openssl_options), for security. OpenSSL forces this from version 1.1.0
1619 server-side so match that on older versions.
1621 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1622 allocation for $value could be released as the expansion processing
1623 concluded, but leaving the global pointer active for it.
1625 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1626 and to use the domains and local_parts ACL conditions.
1628 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1629 incorrectly not doubled on cutthrough transmission, hence seen as a
1630 body-termination at the receiving system - resulting in truncated mails.
1631 Commonly the sender saw a TCP-level error, and retransmitted the message
1632 via the normal store-and-forward channel. This could result in duplicates
1633 received - but deduplicating mailstores were liable to retain only the
1634 initial truncated version.
1636 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1638 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1640 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1641 we're in there, support oversigning also; bug 1309.
1643 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1645 HS/04 Add support for keep_environment and add_environment options.
1647 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1648 either intentional arithmetic overflow during PRNG, or testing config-
1651 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1652 delivery resulted in actual delivery. Cancel cutthrough before DATA
1655 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1656 rejects a recipient: pass the reject to the originator.
1658 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1659 Many were false-positives and ignorable, but it's worth fixing the
1662 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1663 for the new environment-manipulation done at startup. Move the routines
1664 from being local to tls.c to being global via the os.c file.
1666 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1667 an extract embedded as result-arg for a map, the first arg for extract
1668 is unavailable so we cannot tell if this is a numbered or keyed
1669 extraction. Accept either.
1675 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1678 JH/02 The smtp transport option "multi_domain" is now expanded.
1680 JH/03 The smtp transport now requests PRDR by default, if the server offers
1683 JH/04 Certificate name checking on server certificates, when exim is a client,
1684 is now done by default. The transport option tls_verify_cert_hostnames
1685 can be used to disable this per-host. The build option
1686 EXPERIMENTAL_CERTNAMES is withdrawn.
1688 JH/05 The value of the tls_verify_certificates smtp transport and main options
1689 default to the word "system" to access the system default CA bundle.
1690 For GnuTLS, only version 3.0.20 or later.
1692 JH/06 Verification of the server certificate for a TLS connection is now tried
1693 (but not required) by default. The verification status is now logged by
1694 default, for both outbound TLS and client-certificate supplying inbound
1697 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1700 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1701 Status Notification (bounce) messages are now MIME format per RFC 3464.
1702 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1703 under the control of the dsn_advertise_hosts option, and routers may
1704 have a dsn_lasthop option.
1706 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1707 default, modifiable by a malware= option. The list separator for
1708 the options can now be changed in the usual way. Bug 68.
1710 JH/10 The smtp_receive_timeout main option is now expanded before use.
1712 JH/11 The incoming_interface log option now also enables logging of the
1713 local interface on delivery outgoing connections.
1715 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1716 if the interface and destination host and port all match.
1718 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1721 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1722 Patch from Andrew Lewis.
1724 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1725 now supports optional time-restrictions, weighting, and priority
1726 modifiers per server. Patch originally by <rommer@active.by>.
1728 JH/16 The spamd_address main option now supports a mixed list of local
1729 and remote servers. Remote servers can be IPv6 addresses, and
1730 specify a port-range.
1732 JH/17 Bug 68: The spamd_address main option now supports an optional
1733 timeout value per server.
1735 JH/18 Bug 1581: Router and transport options headers_add/remove can
1736 now have the list separator specified.
1738 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1741 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1744 JH/21 Support for the A6 type of dns record is withdrawn.
1746 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1747 rather than the verbs used.
1749 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1750 from 255 to 1024 chars.
1752 JH/24 Verification callouts now attempt to use TLS by default.
1754 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1755 are generic router options now. The defaults didn't change.
1757 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1758 Original patch from Alexander Shikoff, worked over by JH.
1760 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1761 files if ZCAT_COMMAND is not executable.
1763 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1765 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1767 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1768 Normally benign, it bites when the pair was led to by a CNAME;
1769 modern usage is to not canonicalize the domain to a CNAME target
1770 (and we were inconsistent anyway for A-only vs AAAA+A).
1772 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1774 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1775 when evaluating $sender_host_dnssec.
1777 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1778 $sender_helo_dnssec variable.
1780 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1782 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1784 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1786 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1787 documented as working, but never had. Support all but $spam_report.
1789 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1790 added for tls authenticator.
1792 HS/03 Add perl_taintmode main config option
1798 TL/01 When running the test suite, the README says that variables such as
1799 no_msglog_check are global and can be placed anywhere in a specific
1800 test's script, however it was observed that placement needed to be near
1801 the beginning for it to behave that way. Changed the runtest perl
1802 script to read through the entire script once to detect and set these
1803 variables, reset to the beginning of the script, and then run through
1804 the script parsing/test process like normal.
1806 TL/02 The BSD's have an arc4random API. One of the functions to induce
1807 adding randomness was arc4random_stir(), but it has been removed in
1808 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1809 function when detected.
1811 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1812 cause callback expansion.
1814 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1815 syntax errors in an expansion can be treated as a string instead of
1816 logging or causing an error, due to the internal use of bool_lax
1817 instead of bool when processing it.
1819 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1820 server certificates when making smtp deliveries.
1822 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1824 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1826 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1828 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1829 Merged patch from Sebastian Wiedenroth.
1831 JH/05 Fix results-pipe from transport process. Several recipients, combined
1832 with certificate use, exposed issues where response data items split
1833 over buffer boundaries were not parsed properly. This eventually
1834 resulted in duplicates being sent. This issue only became common enough
1835 to notice due to the introduction of connection certificate information,
1836 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1838 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1839 size buffer was used, resulting in syntax errors when an expansion
1842 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1843 version 3.3.6 or later.
1845 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1846 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1847 both become "event_action", the variables become $event_name, $event_data
1848 and $event_defer_errno. There is a new variable $verify_mode, usable in
1849 routers, transports and related events. The tls:cert event is now also
1850 raised for inbound connections, if the main configuration event_action
1853 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1854 early OCSP support, but no stapling (appears to be less than 1.0.0).
1856 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1857 server certificate names available under the smtp transport option
1858 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1861 JH/10 Time-related extraction expansions from certificates now use the main
1862 option "timezone" setting for output formatting, and are consistent
1863 between OpenSSL and GnuTLS compilations. Bug 1541.
1865 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1866 encoded parameter in the incoming message. Bug 1558.
1868 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1869 include certificate info, eximon was claiming there were spoolfile
1872 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1874 JH/14 Log delivery-related information more consistently, using the sequence
1875 "H=<name> [<ip>]" wherever possible.
1877 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1878 are problematic for Debian distribution, omit them from the release
1881 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1883 JH/16 Fix string representation of time values on 64bit time_t architectures.
1886 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1887 output list separator was used.
1892 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1893 checkers that were complaining about end of non-void function with no
1896 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1897 This was a regression introduced in 4.83 by another bugfix.
1899 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1901 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1902 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1908 TF/01 Correctly close the server side of TLS when forking for delivery.
1910 When a message was received over SMTP with TLS, Exim failed to clear up
1911 the incoming connection properly after forking off the child process to
1912 deliver the message. In some situations the subsequent outgoing
1913 delivery connection happened to have the same fd number as the incoming
1914 connection previously had. Exim would try to use TLS and fail, logging
1915 a "Bad file descriptor" error.
1917 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1918 utilities have not been installed.
1920 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1921 temporary space as the ACL may create new global variables.
1923 TL/01 LDAP support uses per connection or global context settings, depending
1924 upon the detected version of the libraries at build time.
1926 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1927 to extract and use the src ip:port in logging and expansions as if it
1928 were a direct connection from the outside internet. PPv2 support was
1929 updated based on HAProxy spec change in May 2014.
1931 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1933 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1934 Properly escape header and check for NULL return.
1936 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1939 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1941 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1942 characters in header names, implemented as a verify condition.
1943 Contributed by Michael Fischer v. Mollard.
1945 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1946 results permerror and temperror. Previous values are deprecated but
1947 still accepted. In a future release, err_perm and err_temp will be
1948 completely removed, which will be a backward incompatibility if the
1949 ACL tests for either of these two old results. Patch contributed by
1950 user bes-internal on the mailing list.
1952 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1954 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1955 selectors, in both main and reject logs.
1957 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1960 JH/07 Add malware type "sock" for talking to simple daemon.
1962 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1964 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1965 routers/transports under cutthrough routing.
1967 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1968 numbers. Touch up "bool" conditional to keep the same definition.
1970 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1972 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1974 JH/12 Expand items in router/transport headers_add or headers_remove lists
1975 individually rather than the list as a whole. Bug 1452.
1977 Required for reasonable handling of multiple headers_ options when
1978 they may be empty; requires that headers_remove items with embedded
1979 colons must have them doubled (or the list-separator changed).
1981 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1982 view the policy declared in the DMARC record. Currently, $dmarc_status
1983 is a combined value of both the record presence and the result of the
1986 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1988 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1989 dnslookup router and the smtp transport (applying to the forward
1992 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1993 of ldap servers used for a specific lookup. Patch provided by Heiko
1996 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1997 New variable $lookup_dnssec_authenticated for observability.
1999 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2000 Patch submitted by Lars Timman.
2002 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2004 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2005 Requires trusted mode and valid format message id, aborts otherwise.
2006 Patch contributed by Heiko Schlichting.
2008 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2009 certextract with support for various fields. Bug 1358.
2011 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2012 is requested by default, modifiable by smtp transport option
2015 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2016 operate on certificate variables to give certificate fingerprints
2017 Also new ${sha256:cert_variable}.
2019 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2021 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2022 Christian Aistleitner.
2024 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2026 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2027 file. Patch from Wolfgang Breyha.
2029 JH/25 Expand the coverage of the delivery $host and $host_address to
2030 client authenticators run in verify callout. Bug 1476.
2032 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2033 align with daemon_smtp_ports. Bug 72.
2035 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2036 support and error reporting did not work properly.
2038 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2039 and is readable. Patch from Andrew Colin Kissa.
2041 TL/14 Enhance documentation of ${run expansion and how it parses the
2042 commandline after expansion, particularly in the case when an
2043 unquoted variable expansion results in an empty value.
2045 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2047 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2048 by renaming away from it. Addresses GH issue 15, reported by
2051 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2053 TL/15 SECURITY: prevent double expansion in math comparison functions
2054 (can expand unsanitized data). Not remotely exploitable.
2061 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2063 PP/02 Make -n do something, by making it not do something.
2064 When combined with -bP, the name of an option is not output.
2066 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2069 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2070 $sender_host_name and config options to manage this, and basic check
2073 PP/05 DSCP support for outbound connections and control modifier for inbound.
2075 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2076 (Only plugin which currently uses this is kerberos4, which nobody should
2077 be using, but we should make it available and other future plugins might
2078 conceivably use it, even though it would break NAT; stuff *should* be
2079 using channel bindings instead).
2081 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2082 name; added for Sendmail compatibility; requires admin caller.
2083 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2084 just ignore it); requires trusted caller.
2085 Also parse but ignore: -Ac -Am -X<logfile>
2088 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2090 TL/02 Add +smtp_confirmation as a default logging option.
2092 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2093 Patch by Magnus Holmgren from 2007-02-20.
2095 TL/04 Bugzilla 1281 - Spec typo.
2096 Bugzilla 1283 - Spec typo.
2097 Bugzilla 1290 - Spec grammar fixes.
2099 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2101 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2103 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2104 mailing list by by Dmitry Isaikin.
2106 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2108 JH/02 Support "G" suffix to numbers in ${if comparisons.
2110 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2112 NM/01 Bugzilla 1197 - Spec typo
2113 Bugzilla 1196 - Spec examples corrections
2115 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2117 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2118 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2121 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2122 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2124 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2125 "acl {{name}{arg}...}", and optional args on acl condition
2128 JH/05 Permit multiple router/transport headers_add/remove lines.
2130 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2132 JH/07 Avoid using a waiting database for a single-message-only transport.
2133 Performance patch from Paul Fisher. Bugzilla 1262.
2135 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2138 JH/09 Add $headers_added variable, with content from use of ACL modifier
2139 add_header (but not yet added to the message). Bugzilla 199.
2141 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2142 Pulled from Bugzilla 817 by Wolfgang Breyha.
2144 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2146 (nb: this is the same fix as in Exim 4.80.1)
2148 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2151 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2154 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2156 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2157 advertise SMTP AUTH mechanism to us, instead of a generic
2158 protocol violation error. Also, make Exim more robust to bad
2159 data from the Dovecot auth socket.
2161 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2163 When a queue runner is handling a message, Exim first routes the
2164 recipient addresses, during which it prunes them based on the retry
2165 hints database. After that it attempts to deliver the message to
2166 any remaining recipients. It then updates the hints database using
2169 So if a recipient address works intermittently, it can get repeatedly
2170 deferred at routing time. The retry hints record remains fresh so the
2171 address never reaches the final cutoff time.
2173 This is a fairly common occurrence when a user is bumping up against
2174 their storage quota. Exim had some logic in its local delivery code
2175 to deal with this. However it did not apply to per-recipient defers
2176 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2178 This change adds a proper retry rule check during routing so that the
2179 final cutoff time is checked against the message's age. We only do
2180 this check if there is an address retry record and there is not a
2181 domain retry record; this implies that previous attempts to handle
2182 the address had the retry_use_local_parts option turned on. We use
2183 this as an approximation for the destination being like a local
2184 delivery, as in LMTP.
2186 I suspect this new check makes the old local delivery cutoff check
2187 redundant, but I have not verified this so I left the code in place.
2189 TF/02 Correct gecos expansion when From: is a prefix of the username.
2191 Test 0254 submits a message to Exim with the header
2195 When I ran the test suite under the user fanf2, Exim expanded
2196 the header to contain my full name, whereas it should have added
2197 a Resent-Sender: header. It erroneously treats any prefix of the
2198 username as equal to the username.
2200 This change corrects that bug.
2202 GF/01 DCC debug and logging tidyup
2203 Error conditions log to paniclog rather than rejectlog.
2204 Debug lines prefixed by "DCC: " to remove any ambiguity.
2206 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2208 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2209 Bug spotted by Jeremy Harris; was flawed since initial commit.
2210 Would have resulted in OCSP responses post-SNI triggering an Exim
2211 NULL dereference and crash.
2213 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2215 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2216 Bug detection, analysis and fix by Samuel Thibault.
2217 Bugzilla 1331, Debian bug #698092.
2219 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2221 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2222 Server implementation by Todd Lyons, client by JH.
2223 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2224 config variable "prdr_enable" controls whether the server
2225 advertises the facility. If the client requests PRDR a new
2226 acl_data_smtp_prdr ACL is called once for each recipient, after
2227 the body content is received and before the acl_smtp_data ACL.
2228 The client is controlled by both of: a hosts_try_prdr option
2229 on the smtp transport, and the server advertisement.
2230 Default client logging of deliveries and rejections involving
2231 PRDR are flagged with the string "PRDR".
2233 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2234 fclose(). Diagnosis by Todd Lyons.
2236 PP/17 Update configure.default to handle IPv6 localhost better.
2237 Patch by Alain Williams (plus minor tweaks).
2240 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2241 This is now consistent with GnuTLS, and is now documented: the
2242 previous undocumented portable approach to treating the option as
2243 unset was to force an expansion failure. That still works, and
2244 an empty string is now equivalent.
2246 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2247 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2248 not performing validation itself.
2250 PP/20 Added force_command boolean option to pipe transport.
2251 Patch from Nick Koston, of cPanel Inc.
2253 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2256 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2258 PP/21 Fix eximon continuous updating with timestamped log-files.
2259 Broken in a format-string cleanup in 4.80, missed when I repaired the
2260 other false fix of the same issue.
2261 Report and fix from Heiko Schlichting.
2264 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2265 Report from Prashanth Katuri.
2267 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2268 It's SecureTransport, so affects any MacOS clients which use the
2269 system-integrated TLS libraries, including email clients.
2271 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2272 using a MIME ACL for non-SMTP local injection.
2273 Report and assistance in diagnosis by Warren Baker.
2275 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2277 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2279 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2280 last id that failed so it may be referenced in subsequent ACL's.
2282 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2285 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2286 ldap library initialization, allowing self-signed CA's to be
2287 used. Also properly sets require_cert option later in code by
2288 using NULL (global ldap config) instead of ldap handle (per
2289 session). Bug diagnosis and testing by alxgomz.
2291 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2292 the src/util/ subdirectory.
2294 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2295 renamed to Transport Post Delivery Action by Jeremy Harris, as
2298 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2299 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2300 redis_servers = needs to be configured which will be used by the redis
2301 lookup. Patch from Warren Baker, of The Packet Hub.
2303 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2305 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2306 hostname or reverse DNS when processing a host list. Used suggestions
2307 from multiple comments on this bug.
2309 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2311 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2312 Missed a few lines, added it to make the runtest require no keyboard
2315 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2316 contains upper case chars. Make router use caseful_local_part.
2318 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2319 support when GnuTLS has been built with p11-kit.
2325 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2327 This, or similar/improved, will also be change PP/11 of 4.82.
2333 PP/01 Handle short writes when writing local log-files.
2334 In practice, only affects FreeBSD (8 onwards).
2335 Bugzilla 1053, with thanks to Dmitry Isaikin.
2337 NM/01 Bugzilla 949 - Documentation tweak
2339 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2342 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2344 PP/02 Implemented gsasl authenticator.
2346 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2348 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2349 `pkg-config foo` for cflags/libs.
2351 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2352 with rest of GSASL and with heimdal_gssapi.
2354 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2355 `pkg-config foo` for cflags/libs for the TLS implementation.
2357 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2358 properties get this fed in as external SSF. A number of robustness
2359 and debugging improvements to the cyrus_sasl authenticator.
2361 PP/08 cyrus_sasl server now expands the server_realm option.
2363 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2364 Patch by Jeremy Harris.
2366 PP/10 Added dbmjz lookup type.
2368 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2370 PP/12 MAIL args handles TAB as well as SP, for better interop with
2371 non-compliant senders.
2372 Analysis and variant patch by Todd Lyons.
2374 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2375 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2376 Patch from Dirk Mueller <dmueller@suse.com>
2378 PP/13 tls_peerdn now print-escaped for spool files.
2379 Observed some $tls_peerdn in wild which contained \n, which resulted
2380 in spool file corruption.
2382 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2383 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2384 or write after TLS renegotiation, which otherwise led to messages
2387 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2388 as a tracking header (ie: a signed header comes before the signature).
2389 Patch from Wolfgang Breyha.
2391 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2392 comma-sep list; embedded commas doubled.
2394 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2396 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2398 Report and patch from Dmitry Banschikov.
2400 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2401 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2402 failures. We appear to now support TLS1.1+ with Exim.
2404 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2405 lets Exim select keys and certificates based upon TLS SNI from client.
2406 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2407 before an outbound SMTP session. New log_selector, +tls_sni.
2409 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2410 NULL dereference. Report and patch from Alun Jones.
2412 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2413 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2414 Not seeing resolver debug output on NetBSD, but suspect this is a
2415 resolver implementation change.
2417 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2418 Left warnings. Added "eximon gdb" invocation mode.
2420 PP/21 Defaulting "accept_8bitmime" to true, not false.
2422 PP/22 Added -bw for inetd wait mode support.
2424 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2425 locate the relevant includes and libraries. Made this the default.
2427 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2428 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2430 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2431 This may cause build issues on older platforms.
2433 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2434 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2435 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2436 Added SNI support via GnuTLS too.
2437 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2439 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2441 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2442 Applied second patch from Janne, implementing suggestion to default
2443 multiple-strings-in-record handling to match SPF spec.
2445 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2447 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2448 read-only, out of scope).
2449 Patch from Wolfgang Breyha, report from Stuart Northfield.
2451 PP/29 Fix three issues highlighted by clang analyser static analysis.
2452 Only crash-plausible issue would require the Cambridge-specific
2453 iplookup router and a misconfiguration.
2454 Report from Marcin Mirosław.
2456 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2458 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2459 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2460 As part of this, removing so much warning spew let me fix some minor
2461 real issues in debug logging.
2463 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2464 assignment on my part. Fixed.
2466 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2467 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2468 Janne Snabb (who went above and beyond: thank you).
2470 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2471 string otherwise requires a connection and a bunch more work and it's
2472 relatively easy to get wrong. Should also expose TLS library linkage
2475 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2476 64-bit ${eval} (JH/03).
2478 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2479 GNU libc to support some of the 64-bit stuff, should not lead to
2480 conflicts. Defined before os.h is pulled in, so if a given platform
2481 needs to override this, it can.
2483 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2484 protection layer was required, which is not implemented.
2485 Bugzilla 1254, patch from Wolfgang Breyha.
2487 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2488 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2489 tls_dhparam take prime identifiers. Also unbreak combination of
2490 OpenSSL+DH_params+TLSSNI.
2492 PP/39 Disable SSLv2 by default in OpenSSL support.
2498 PP/01 Solaris build fix for Oracle's LDAP libraries.
2499 Bugzilla 1109, patch from Stephen Usher.
2501 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2503 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2506 TF/02 Fix a couple more cases where we did not log the error message
2507 when unlink() failed. See also change 4.74-TF/03.
2509 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2510 lock up or crash if it happened to be inside a call to libc when it
2511 got a SIGUSR1 from exiwhat.
2513 The SIGUSR1 handler appends the current process status to the process
2514 log which is later printed by exiwhat. It used to use the general
2515 purpose logging code to do this, but several functions it calls are
2516 not safe for signals.
2518 The new output code in the SIGUSR1 handler is specific to the process
2519 log, and simple enough that it's easy to inspect for signal safety.
2520 Removing some special cases also simplifies the general logging code.
2521 Removing the spurious timestamps from the process log simplifies
2524 TF/04 Improved ratelimit ACL condition.
2526 The /noupdate option has been deprecated in favour of /readonly which
2527 has clearer semantics. The /leaky, /strict, and /readonly update modes
2528 are mutually exclusive. The update mode is no longer included in the
2529 database key; it just determines when the database is updated. (This
2530 means that when you upgrade Exim will forget old rate measurements.)
2532 Exim now checks that the per_* options are used with an update mode that
2533 makes sense for the current ACL. For example, when Exim is processing a
2534 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2535 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2536 must specify per_mail/readonly. If you omit the update mode it defaults to
2537 /leaky where that makes sense (as before) or /readonly where required.
2539 The /noupdate option is now undocumented but still supported for
2540 backwards compatibility. It is equivalent to /readonly except that in
2541 ACLs where /readonly is required you may specify /leaky/noupdate or
2542 /strict/noupdate which are treated the same as /readonly.
2544 A useful new feature is the /count= option. This is a generalization
2545 of the per_byte option, so that you can measure the throughput of other
2546 aggregate values. For example, the per_byte option is now equivalent
2547 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2549 The per_rcpt option has been generalized using the /count= mechanism
2550 (though it's more complicated than the per_byte equivalence). When it is
2551 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2552 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2553 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2554 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2555 Note that using per_rcpt with a non-readonly update mode in more than
2556 one ACL will cause the recipients to be double-counted. (The per_mail
2557 and per_byte options don't have this problem.)
2559 The handling of very low rates has changed slightly. If the computed rate
2560 is less than the event's count (usually one) then this event is the first
2561 after a long gap. In this case the rate is set to the same as this event's
2562 count, so that the first message of a spam run is counted properly.
2564 The major new feature is a mechanism for counting the rate of unique
2565 events. The new per_addr option counts the number of different
2566 recipients that someone has sent messages to in the last time period. It
2567 behaves like per_rcpt if all the recipient addresses are different, but
2568 duplicate recipient addresses do not increase the measured rate. Like
2569 the /count= option this is a general mechanism, so the per_addr option
2570 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2571 example, measure the rate that a client uses different sender addresses
2572 with the options per_mail/unique=$sender_address. There are further
2573 details in the main documentation.
2575 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2577 TF/06 Removed a few PCRE remnants.
2579 TF/07 Automatically extract Exim's version number from tags in the git
2580 repository when doing development or release builds.
2582 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2583 Bugzilla 879. Patch from Paul Fisher.
2585 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2586 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2589 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2591 PP/05 Variable $av_failed, true if the AV scanner deferred.
2592 Bugzilla 1078. Patch from John Horne.
2594 PP/06 Stop make process more reliably on build failure.
2595 Bugzilla 1087. Patch from Heiko Schlittermann.
2597 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2598 Bugzilla 1089. Patch from Heiko Schlittermann.
2600 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2601 Bugzilla 1131. Patch from Holger Weiß.
2603 PP/09 Handle IPv6 addresses with SPF.
2604 Bugzilla 860. Patch from Wolfgang Breyha.
2606 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2608 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2611 PP/11 match_* no longer expand right-hand-side by default.
2612 New compile-time build option, EXPAND_LISTMATCH_RHS.
2613 New expansion conditions, "inlist", "inlisti".
2615 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2617 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2619 PP/14 fix log_write() format string regression from TF/03.
2620 Bugzilla 1152. Patch from Dmitry Isaikin.
2626 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2628 PP/02 Harmonised TLS library version reporting; only show if debugging.
2629 Layout now matches that introduced for other libraries in 4.74 PP/03.
2631 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2633 PP/04 New "dns_use_edns0" global option.
2635 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2638 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2639 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2641 TK/01 Updated PolarSSL code to 0.14.2.
2642 Bugzilla 1097. Patch from Andreas Metzler.
2644 PP/07 Catch divide-by-zero in ${eval:...}.
2645 Fixes bugzilla 1102.
2647 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2650 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2651 format-string attack -- SECURITY: remote arbitrary code execution.
2653 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2654 time unintentionally subject to list matching rules, letting the header
2655 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2656 arbitrary string expansion). This allowed for information disclosure.
2658 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2659 INT_MIN/-1 -- value coerced to INT_MAX.
2665 NM/01 Workaround for PCRE version dependency in version reporting
2668 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2669 This fixes portability to compilers other than gcc, notably
2670 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2672 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2673 makefiles for portability to HP-UX and POSIX correctness.
2675 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2676 Also via indented variable definition in the Makefile.
2677 (Debugging by Oliver Heesakkers).
2679 PP/02 Restore caching of spamd results with expanded spamd_address.
2680 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2682 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2683 Improves build reliability. Fix from: Frank Elsner
2685 NM/02 Fix wide character breakage in the rfc2047 coding
2686 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2688 NM/03 Allow underscore in dnslist lookups
2689 Fixes bug 1026. Patch from Graeme Fowler
2691 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2692 Code patches from Adam Ciarcinski of NetBSD.
2694 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2697 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2698 is logged, to avoid truncation. Patch from John Horne.
2700 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2701 Patch from Jakob Hirsch.
2703 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2704 SQL string expansion failure details.
2705 Patch from Andrey Oktyabrski.
2707 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2708 Patch from Simon Arlott.
2710 PP/09 New lookups functionality failed to compile on old gcc which rejects
2711 extern declarations in function scope.
2712 Patch from Oliver Fleischmann
2714 PP/10 Use sig_atomic_t for flags set from signal handlers.
2715 Check getgroups() return and improve debugging.
2716 Fixed developed for diagnosis in bug 927 (which turned out to be
2719 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2720 Patch from Mark Zealey.
2722 PP/12 Bugzilla 1056: Improved spamd server selection.
2723 Patch from Mark Zealey.
2725 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2726 Based on patch from Heiko Schlittermann.
2728 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2729 Patch from Uwe Doering, sign-off by Michael Haardt.
2731 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2732 variable declaration deep within a block. Bug and patch from
2735 PP/15 lookups-Makefile IRIX compatibility coercion.
2737 PP/16 Make DISABLE_DKIM build knob functional.
2739 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2740 Patch by Simon Arlott
2742 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2743 variable argument macros. Our copy now differs from upstream.
2749 TF/01 Failure to get a lock on a hints database can have serious
2750 consequences so log it to the panic log.
2752 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2753 controlled using the smtp_confirmation log selector.
2755 TF/03 Include the error message when we fail to unlink a spool file.
2757 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2758 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2759 for maintaining out-of-tree patches for some time.
2761 PP/01 Bugzilla 139: Documentation and portability issues.
2762 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2763 Handle per-OS dynamic-module compilation flags.
2765 PP/02 Let /dev/null have normal permissions.
2766 The 4.73 fixes were a little too stringent and complained about the
2767 permissions on /dev/null. Exempt it from some checks.
2768 Reported by Andreas M. Kirchwitz.
2770 PP/03 Report version information for many libraries, including
2771 Exim version information for dynamically loaded libraries. Created
2772 version.h, now support a version extension string for distributors
2773 who patch heavily. Dynamic module ABI change.
2775 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2776 privilege escalation vulnerability whereby the Exim run-time user
2777 can cause root to append content of the attacker's choosing to
2780 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2783 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2784 If dropping privileges for untrusted macros, we disabled normal logging
2785 on the basis that it would fail; for the Exim run-time user, this is not
2786 the case, and it resulted in successful deliveries going unlogged.
2787 Fixed. Reported by Andreas Metzler.
2793 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2794 only prepend for the Resent-* case. Fixes regression introduced in
2795 Exim 4.70 by NM/22 for Bugzilla 607.
2797 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2798 increasing numbers of administrators be bitten by this.
2800 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2802 PP/03 Bugzilla 994: added openssl_options main configuration option.
2804 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2806 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2808 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2810 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2811 without a peer certificate, leading to a segfault because of an
2812 assumption that peers always have certificates. Be a little more
2813 paranoid. Problem reported by Martin Tscholak.
2815 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2816 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2817 NB: ClamAV planning to remove STREAM in "middle of 2010".
2818 CL also introduces -bmalware, various -d+acl logging additions and
2819 more caution in buffer sizes.
2821 PP/09 Implemented reverse_ip expansion operator.
2823 PP/10 Bugzilla 937: provide a "debug" ACL control.
2825 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2827 PP/12 Bugzilla 973: Implement --version.
2829 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2831 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2833 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2835 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2836 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2837 ignore trailing whitespace.
2839 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2841 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2844 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2845 Notification from Dr Andrew Aitchison.
2847 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2848 ExtendedDetectionInfo response format.
2849 Notification from John Horne.
2851 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2854 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2855 XSL and documented dependency on system catalogs, with examples of how
2858 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2861 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2862 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2863 configuration file which is writeable by the Exim user or group.
2865 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2866 of configuration files to cover files specified with the -C option if
2867 they are going to be used with root privileges, not just the default
2870 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2871 option (effectively making it always true).
2873 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2874 files to be used while preserving root privileges.
2876 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2877 that rogue child processes cannot use them.
2879 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2880 run-time user, instead of root.
2882 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2883 Exim run-time user without dropping privileges.
2885 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2886 result string, instead of calling string_vformat() twice with the same
2889 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2890 for other users. Others should always drop root privileges if they use
2891 -C on the command line, even for a whitelisted configure file.
2893 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2895 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2901 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2902 $data_path, and $header_path variables; fixed documentation bugs and
2905 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2906 exipick to access non-standard spools, including the "frozen" queue
2909 NM/01 Bugzilla 965: Support mysql stored procedures.
2910 Patch from Alain Williams
2912 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2914 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2915 Patch from Andreas Metzler
2917 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2918 Patch from Kirill Miazine
2920 NM/05 Bugzilla 671: Added umask to procmail example.
2922 JJ/03 installed exipick 20100323.0, fixing doc bug
2924 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2925 directory. Notification and patch from Dan Rosenberg.
2927 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2929 TK/02 Improve log output when DKIM signing operation fails.
2931 MH/01 Treat the transport option dkim_domain as a colon separated
2932 list, not as a single string, and sign the message with each element,
2933 omitting multiple occurences of the same signer.
2935 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2936 Bugzilla 985, 986. Patch by Simon Arlott
2938 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2939 Patch by Simon Arlott
2941 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2942 MBX locking. Notification from Dan Rosenberg.
2948 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2950 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2952 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2954 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2956 NM/04 Bugzilla 915: Flag broken perl installation during build.
2962 TK/01 Added patch by Johannes Berg that expands the main option
2963 "spamd_address" if it starts with a dollar sign.
2965 TK/02 Write list of recipients to X-Envelope-Sender header when building
2966 the mbox-format spool file for content scanning (suggested by Jakob
2969 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2970 (http://www.dcc-servers.net/) support via dccifd. Activated by
2971 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2973 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2974 by Mark Daniel Reidel <mr@df.eu>.
2976 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2977 When building exim an external PCRE library is now needed -
2978 PCRE is a system library on the majority of modern systems.
2979 See entry on PCRE_LIBS in EDITME file.
2981 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2982 conversation. Added nologin parameter to request.
2983 Patch contributed by Kirill Miazine.
2985 TF/01 Do not log submission mode rewrites if they do not change the address.
2987 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2989 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2990 log files in place. Contributed by Roberto Lima.
2992 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2994 TF/03 Bugzilla 615: When checking the local_parts router precondition
2995 after a local_part_suffix or local_part_prefix option, Exim now
2996 does not use the address's named list lookup cache, since this
2997 contains cached lookups for the whole local part.
2999 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3000 Robert Millan. Documentation is in experimental-spec.txt.
3002 TF/04 Bugzilla 668: Fix parallel build (make -j).
3004 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3006 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3007 Patch provided by Jan Srzednicki.
3009 TF/05 Leading white space used to be stripped from $spam_report which
3010 wrecked the formatting. Now it is preserved.
3012 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3013 that they are available at delivery time.
3015 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3017 TF/08 TLS error reporting now respects the incoming_interface and
3018 incoming_port log selectors.
3020 TF/09 Produce a more useful error message if an SMTP transport's hosts
3021 setting expands to an empty string.
3023 NM/06 Bugzilla 744: EXPN did not work under TLS.
3024 Patch provided by Phil Pennock.
3026 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3027 Patch provided by Richard Godbee.
3029 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3030 acl_smtp_notquit, added index entry.
3032 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3033 Patch provided by Eugene Bujak.
3035 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3036 accept(). Patch provided by Maxim Dounin.
3038 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3039 Patch provided by Phil Pennock.
3041 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3043 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3044 Patch provided by Brad "anomie" Jorsch.
3046 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3047 Patch provided by Dean Brooks.
3049 TK/05 Add native DKIM support (does not depend on external libraries).
3051 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3052 Patch provided by Graeme Fowler.
3054 NM/16 Bugzilla 851: Documentation example syntax fix.
3056 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3058 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3061 NM/19 Bugzilla 745: TLS version reporting.
3062 Patch provided by Phil Pennock.
3064 NM/20 Bugzilla 167: bool: condition support.
3065 Patch provided by Phil Pennock.
3067 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3068 clients. Patch provided by Phil Pennock.
3070 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3071 Patch provided by Brad "anomie" Jorsch.
3073 NM/23 Bugzilla 687: Fix misparses in eximstats.
3074 Patch provided by Heiko Schlittermann.
3076 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3077 Patch provided by Heiko Schlittermann.
3079 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3080 plus update to original patch.
3082 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3084 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3085 Patch provided by David Brownlee.
3087 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3089 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3091 NM/30 Bugzilla 888: TLS documentation bugfixes.
3093 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3095 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3096 Unlike the original bugzilla I have changed all shell scripts in src tree.
3098 NM/33 Bugzilla 898: Transport filter timeout fix.
3099 Patch by Todd Rinaldo.
3101 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3102 Patch by Serge Demonchaux.
3104 NM/35 Bugzilla 39: Base64 decode bug fixes.
3105 Patch by Jakob Hirsch.
3107 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3109 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3111 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3113 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3119 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3120 ALT-N's libdkim that I have put here:
3121 http://duncanthrax.net/exim-experimental/
3123 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3124 were called 'true' and it seems that C99 defines that as a reserved
3125 keyword to be used with 'bool' variable types. That means you could
3126 not include C99-style headers which use bools without triggering
3127 build errors in sieve.c.
3129 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3130 as mailq or other aliases. Changed the --help handling significantly
3131 to do whats expected. exim_usage() emits usage/help information.
3133 SC/01 Added the -bylocaldomain option to eximstats.
3135 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3137 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3139 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3145 PH/01 Another patch from the Sieve maintainer.
3147 PH/02 When an IPv6 address is converted to a string for single-key lookup
3148 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3149 dots are used instead of colons so that keys in lsearch files need not
3150 contain colons. This was done some time before quoting was made available
3151 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3152 (notated using the quote facility) so as to distinguish them from IPv4
3153 keys. This meant that lookups for IP addresses in host lists did not work
3154 for iplsearch lookups.
3156 This has been fixed by arranging for IPv6 addresses to be expressed with
3157 colons if the lookup type is iplsearch. This is not incompatible, because
3158 previously such lookups could never work.
3160 The situation is now rather anomalous, since one *can* have colons in
3161 ordinary lsearch keys. However, making the change in all cases is
3162 incompatible and would probably break a number of configurations.
3164 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3167 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3168 conversion specification without a maximum field width, thereby enabling
3169 a rogue spamd server to cause a buffer overflow. While nobody in their
3170 right mind would setup Exim to query an untrusted spamd server, an
3171 attacker that gains access to a server running spamd could potentially
3172 exploit this vulnerability to run arbitrary code as the Exim user.
3174 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3175 $primary_hostname instead of what libspf2 thinks the hosts name is.
3177 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3178 a directory entry by the name of the lookup key. Previously, if a
3179 symlink pointed to a non-existing file or a file in a directory that
3180 Exim lacked permissions to read, a lookup for a key matching that
3181 symlink would fail. Now it is enough that a matching directory entry
3182 exists, symlink or not. (Bugzilla 503.)
3184 PH/03 The body_linecount and body_zerocount variables are now exported in the
3187 PH/04 Added the $dnslist_matched variable.
3189 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3190 This means they are set thereafter only if the connection becomes
3193 PH/06 Added the client_condition to authenticators so that some can be skipped
3194 by clients under certain conditions.
3196 PH/07 The error message for a badly-placed control=no_multiline_responses left
3197 "_responses" off the end of the name.
3199 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3201 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3202 (without spaces) instead of just copying the configuration text.
3204 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3206 PH/11 Added $max_received_linelength.
3208 PH/12 Added +ignore_defer and +include_defer to host lists.
3210 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3211 way in which PCRE > 7.0 is built.
3213 PH/14 Implemented queue_only_load_latch.
3215 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3216 MAIL command. The effect was to mangle the value on 64-bit systems.
3218 PH/16 Another patch from the Sieve maintainer.
3220 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3222 PH/18 If a system quota error occurred while trying to create the file for
3223 a maildir delivery, the message "Mailbox is full" was not appended to the
3224 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3225 only to a quota excession during the actual writing of the file.
3227 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3228 characters?) which causes problems in log lines. The DN values are now
3229 passed through string_printing() before being added to log lines.
3231 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3232 and InterBase are left for another time.)
3234 PH/21 Added message_body_newlines option.
3236 PH/22 Guard against possible overflow in moan_check_errorcopy().
3238 PH/23 POSIX allows open() to be a macro; guard against that.
3240 PH/24 If the recipient of an error message contained an @ in the local part
3241 (suitably quoted, of course), incorrect values were put in $domain and
3242 $local_part during the evaluation of errors_copy.
3248 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3249 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3252 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3253 issue a MAIL command.
3255 PH/02 In an ACL statement such as
3257 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3259 if a client was not listed at all, or was listed with a value other than
3260 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3261 the condition was not true (as it should be), so access was not denied.
3262 The bug was that the ! inversion was incorrectly passed on to the second
3263 item. This has been fixed.
3265 PH/03 Added additional dnslists conditions == and =& which are different from
3266 = and & when the dns lookup returns more than one IP address.
3268 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3269 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3271 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3272 FSYNC, which compiles an option called disable_fsync that allows for
3273 bypassing fsync(). The documentation is heavily laced with warnings.
3275 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3277 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3278 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3279 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3280 including adding "make clean"; (3) Added -fPIC when compiling the test
3281 dynamically loaded module, to get rid of a warning.
3283 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3284 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3285 = 0s. The bug is otherwise harmless.
3287 PH/07 There was a bug in the dovecot authenticator such that the value of
3288 $auth1 could be overwritten, and so not correctly preserved, after a
3289 successful authentication. This usually meant that the value preserved by
3290 the server_setid option was incorrect.
3292 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3294 PH/09 Installed PCRE release 7.0.
3296 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3297 run for batched SMTP input. It is now run at the start of every message
3298 in the batch. While fixing this I discovered that the process information
3299 (output by running exiwhat) was not always getting set for -bs and -bS
3300 input. This is fixed, and it now also says "batched" for BSMTP.
3302 PH/11 Added control=no_pipelining.
3304 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3305 patch, slightly modified), and move the expansion of helo_data till after
3306 the connection is made in the smtp transport (so it can use these
3309 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3311 PH/14 Added log_selector = +pid.
3313 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3315 PH/16 Add ${if forany and ${if forall.
3317 PH/17 Added dsn_from option to vary the From: line in DSNs.
3319 PH/18 Flush SMTP output before performing a callout, unless control =
3320 no_callout_flush is set.
3322 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3323 was true (the default) a successful delivery failed to delete the retry
3324 item, thus causing premature timeout of the address. The bug is now
3327 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3329 PH/21 Long custom messages for fakedefer and fakereject are now split up
3330 into multiline responses in the same way that messages for "deny" and
3331 other ACL rejections are.
3333 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3334 with slight modification.
3336 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3337 draft, changing the syntax and factoring some duplicate code".
3339 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3340 for deliveries of the second and subsequent messages over the same SMTP
3343 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3344 ${reduce, with only minor "tidies".
3346 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3348 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3349 expansion side effects.
3351 PH/27 When a message times out after an over-quota error from an Exim-imposed
3352 quota, the bounce message says "mailbox is full". This message was not
3353 being given when it was a system quota that was exceeded. It now should
3356 MH/03 Made $recipients available in local_scan(). local_scan() already has
3357 better access to the recipient list through recipients_list[], but
3358 $recipients can be useful in postmaster-provided expansion strings.
3360 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3361 in the case of a MAIL command with additional options following the
3362 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3363 were accidentally chopped off.
3365 PH/29 SMTP synchronization checks are implemented when a command is read -
3366 there is a check that no more input is waiting when there shouldn't be
3367 any. However, for some commands, a delay in an ACL can mean that it is
3368 some time before the response is written. In this time, more input might
3369 arrive, invalidly. So now there are extra checks after an ACL has run for
3370 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3371 pipelining has not been advertised.
3373 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3375 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3376 returned in the SMTP response when smtp_return_error_details was false.
3377 This has been fixed.
3379 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3380 instead of the C I/O that was originally supplied, because problems were
3381 reported on Solaris.
3383 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3384 Exim which did not show up earlier: it was assuming that a call to
3385 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3386 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3387 was a macro that became an assignment, so it seemed to work. This has
3388 changed to a proper function call with a void return, hence the compile
3389 error. Exim's code has been fixed.
3391 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3394 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3396 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3398 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3399 because it turns out to be too big or there is a timeout) there is no
3400 "Completed" line in the log. When some messages of this type were
3401 selected by exigrep, they were listed as "not completed". Others were
3402 picked up by some special patterns. I have improved the selection
3403 criteria to be more general.
3405 PH/38 The host_find_failed option in the manualroute router can now be set
3406 to "ignore", to completely ignore a host whose IP address cannot be
3407 found. If all hosts are ignored, the behaviour is controlled by the new
3408 host_all_ignored option.
3410 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3411 homing or because of multiple MX records with /mx) generated more than
3412 one IP address, and the following item turned out to be the local host,
3413 all the secondary addresses of the first item were incorrectly removed
3414 from the list, along with the local host and any following hosts (which
3415 is what is supposed to happen).
3417 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3418 whoever called Exim into the -H file. In the case of the daemon it was
3419 behaving confusingly. When first started, it used values for whoever
3420 started the daemon, but after a SIGHUP it used the Exim user (because it
3421 calls itself on a restart). I have changed the code so that it now always
3424 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3425 message are rejected with the same error (e.g. no authentication or bad
3426 sender address), and a DATA command is nevertheless sent (as can happen
3427 with PIPELINING or a stupid MUA), the error message that was given to the
3428 RCPT commands is included in the rejection of the DATA command. This is
3429 intended to be helpful for MUAs that show only the final error to their
3432 PH/42 Another patch from the Sieve maintainer.
3434 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3435 Eximstats - Fixed some broken HTML links and added missing column headers
3437 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3438 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3440 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3442 PH/43 Yet another patch from the Sieve maintainer.
3444 PH/44 I found a way to check for a TCP/IP connection going away before sending
3445 the response to the final '.' that terminates a message, but only in the
3446 case where the client has not sent further data following the '.'
3447 (unfortunately, this is allowed). However, in many cases there won't be
3448 any further data because there won't be any more messages to send. A call
3449 to select() can be used: if it shows that the input is "ready", there is
3450 either input waiting, or the socket has been closed. An attempt to read
3451 the next input character can distinguish the two cases. Previously, Exim
3452 would have sent an OK response which the client would never have see.
3453 This could lead to message repetition. This fix should cure that, at
3454 least in a lot of common cases.
3456 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3457 advertised in response to EHLO.
3463 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3464 fixed by 4.65/MH/01 (is this a record?) are fixed:
3466 (i) An empty string was always treated as zero by the numeric comparison
3467 operators. This behaviour has been restored.
3469 (ii) It is documented that the numeric comparison operators always treat
3470 their arguments as decimal numbers. This was broken in that numbers
3471 starting with 0 were being interpreted as octal.
3473 While fixing these problems I realized that there was another issue that
3474 hadn't been noticed. Values of message_size_limit (both the global option
3475 and the transport option) were treated as octal if they started with 0.
3476 The documentation was vague. These values are now always treated as
3477 decimal, and I will make that clear in the documentation.
3483 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3484 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3487 MH/01 Don't check that the operands of numeric comparison operators are
3488 integers when their expansion is in "skipping" mode (fixes bug
3489 introduced by 4.64-PH/07).
3491 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3492 child addresses, Exim now panics and dies. Previously, because the count
3493 is held in a short int, deliveries were likely to be lost. As such a
3494 large number of recipients for a single message is ridiculous
3495 (performance will be very, very poor), I have chosen to impose a limit
3496 rather than extend the field.
3502 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3503 leftover -K file (the existence of which was triggered by #402).
3504 While we were at it, introduced process PID as part of the -K
3505 filename. This should rule out race conditions when creating
3508 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3509 processing considerably. Previous code took too long for large mails,
3510 triggering a timeout which in turn triggers #401.
3512 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3513 in the DK code in transports.c. sendfile() is not really portable,
3514 hence the _LINUX specificness.
3516 TF/01 In the add_headers option to the mail command in an Exim filter,
3517 there was a bug that Exim would claim a syntax error in any
3518 header after the first one which had an odd number of characters
3521 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3522 callout verification, Exim cached a "reject" for the entire domain. This
3523 is correct for most verifications, but it is not correct for a recipient
3524 verification with use_sender or use_postmaster set, because in that case
3525 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3526 case of MAIL FROM:<> rejection from other early rejections (e.g.
3527 rejection of HELO). When verifying a recipient using a non-null MAIL
3528 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3529 Whatever the result of the callout, the value of the domain cache is
3530 left unchanged (for any other kind of callout, getting as far as trying
3531 RCPT means that the domain itself is ok).
3533 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3536 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3537 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3538 session, the daemon ignores ECONNECT errors and logs others; it now
3539 ignores EPIPE as well.
3541 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3542 (quoted-printable decoding).
3544 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3545 later the small subsequent patch to fix an introduced bug.
3547 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3549 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3551 PH/08 An error is now given if message_size_limit is specified negative.
3553 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3554 to be given (somewhat) arbitrary names.
3556 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3559 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3560 miscellaneous code fixes
3562 PH/10 Added the log_reject_target ACL modifier to specify where to log
3565 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3566 hostname. This is wrong, because it relates to the incoming message (and
3567 probably the interface on which it is arriving) and not to the outgoing
3568 callout (which could be using a different interface). This has been
3569 changed to use the value of the helo_data option from the smtp transport
3570 instead - this is what is used when a message is actually being sent. If
3571 there is no remote transport (possible with a router that sets up host
3572 addresses), $smtp_active_hostname is used.
3574 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3575 tweaks were necessary in order to get it to work (see also 21 below):
3576 (a) The code assumed that strncpy() returns a negative number on buffer
3577 overflow, which isn't the case. Replaced with Exim's string_format()
3579 (b) There were several signed/unsigned issues. I just did the minimum
3580 hacking in of casts. There is scope for a larger refactoring.
3581 (c) The code used strcasecmp() which is not a standard C function.
3582 Replaced with Exim's strcmpic() function.
3583 (d) The code set only $1; it now sets $auth1 as well.
3584 (e) A simple test gave the error "authentication client didn't specify
3585 service in request". It would seem that Dovecot has changed its
3586 interface. Fortunately there's a specification; I followed it and
3587 changed what the client sends and it appears to be working now.
3589 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3592 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3593 address A is aliased to B and C, where B exists and C does not. Without
3594 -v the output is "A verified" because verification stops after a
3595 successful redirection if more than one address is generated. However,
3596 with -v the child addresses are also verified. Exim was outputting "A
3597 failed to verify" and then showing the successful verification for C,
3598 with its parentage. It now outputs "B failed to verify", showing B's
3599 parentage before showing the successful verification of C.
3601 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3602 look up a TXT record in a specific list after matching in a combined
3605 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3606 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3607 they consult the DNS. I had assumed they would set it the way they
3608 wanted; and indeed my experiments on Linux seem to show that in some
3609 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3610 To be on the safe side, however, I have now made the interface to
3611 host_find_byname() similar to host_find_bydns(), with an argument
3612 containing the DNS resolver options. The host_find_byname() function now
3613 sets these options at its start, just as host_find_bydns() does. The smtp
3614 transport options dns_qualify_single and dns_search_parents are passed to
3615 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3616 of host_find_byname() use the default settings of RES_DEFNAMES
3617 (qualify_single) but not RES_DNSRCH (search_parents).
3619 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3620 spool_read_header() do less string testing, by means of a preliminary
3621 switch on the second character of optional "-foo" lines. (This is
3622 overdue, caused by the large number of possibilities that now exist.
3623 Originally there were few.) While I was there, I also converted the
3624 str(n)cmp tests so they don't re-test the leading "-" and the first
3625 character, in the hope this might squeeze out yet more improvement.
3627 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3628 flag allowing group syntax was set by the header_syntax check but not
3629 turned off, possible causing trouble later; (2) The flag was not being
3630 set at all for the header_verify test, causing "group"-style headers to
3631 be rejected. I have now set it in this case, and also caused header_
3632 verify to ignore an empty address taken from a group. While doing this, I
3633 came across some other cases where the code for allowing group syntax
3634 while scanning a header line wasn't quite right (mostly, not resetting
3635 the flag correctly in the right place). These bugs could have caused
3636 trouble for malformed header lines. I hope it is now all correct.
3638 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3639 with the "reply" argument non-NULL. The code, however (which originally
3640 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3641 but it didn't always do it. This confused somebody who was copying the
3642 code for some other use. I have removed all the tests.
3644 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3645 feature that was used to support insecure browsers during the U.S. crypto
3646 embargo. It requires special client support, and Exim is probably the
3647 only MTA that supported it -- and would never use it because real RSA is
3648 always available. This code has been removed, because it had the bad
3649 effect of slowing Exim down by computing (never used) parameters for the
3650 RSA_EXPORT functionality.
3652 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3653 authenticator to fail if there's a tab character in the incoming data
3654 (there should never be unless someone is messing about, as it's supposed
3655 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3656 if the connection is using TLS or if the remote IP is the same as the
3657 local IP, and the "valid-client-cert option" if a client certificate has
3660 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3661 authenticators. This can be used for authorization after authentication
3662 succeeds. (In the case of plaintext, it servers for both authentication
3665 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3666 if any retry times were supplied.
3668 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3669 connection, where there is no client IP address to check. In this
3670 situation, the verify now always succeeds.
3672 PH/25 Applied John Jetmore's -Mset patch.
3674 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3676 PH/27 In a string expansion for a processed (not raw) header when multiple
3677 headers of the same name were present, leading whitespace was being
3678 removed from all of them, but trailing whitespace was being removed only
3679 from the last one. Now trailing whitespace is removed from each header
3680 before concatenation. Completely empty headers in a concatenation (as
3681 before) are ignored.
3683 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3684 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3686 PH/29 [Removed. This was a change that I later backed out, and forgot to
3687 correct the ChangeLog entry (that I had efficiently created) before
3688 committing the later change.]
3690 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3691 address errors (4xx response to RCPT) over the same connection as other
3692 messages routed to the same hosts. Such deliveries are always "forced",
3693 so retry times are not inspected. This resulted in far too many retries
3694 for the affected addresses. The effect occurred only when there were more
3695 hosts than the hosts_max_try setting in the smtp transport when it had
3696 the 4xx errors. Those hosts that it had tried were not added to the list
3697 of hosts for which the message was waiting, so if all were tried, there
3698 was no problem. Two fixes have been applied:
3700 (i) If there are any address or message errors in an SMTP delivery, none
3701 of the hosts (tried or untried) are now added to the list of hosts
3702 for which the message is waiting, so the message should not be a
3703 candidate for sending over the same connection that was used for a
3704 successful delivery of some other message. This seems entirely
3705 reasonable: after all the message is NOT "waiting for some host".
3706 This is so "obvious" that I'm not sure why it wasn't done
3707 previously. Hope I haven't missed anything, but it can't do any
3708 harm, as the worst effect is to miss an optimization.
3710 (ii) If, despite (i), such a delivery is accidentally attempted, the
3711 routing retry time is respected, so at least it doesn't keep
3712 hammering the server.
3714 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3715 in ${readsocket because some servers need this prod.
3717 PH/32 Added some extra debug output when updating a wait-xxx database.
3719 PH/33 The hint "could be header name not terminated by colon", which has been
3720 given for certain expansion errors for a long time, was not being given
3721 for the ${if def:h_colon_omitted{... case.
3723 PH/34 The spec says: "With one important exception, whenever a domain list is
3724 being scanned, $domain contains the subject domain." There was at least
3725 one case where this was not true.
3727 PH/35 The error "getsockname() failed: connection reset by peer" was being
3728 written to the panic log as well as the main log, but it isn't really
3729 panic-worthy as it just means the connection died rather early on. I have
3730 removed the panic log writing for the ECONNRESET error when getsockname()
3733 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3734 runs only) independently of the message's sender address. This meant
3735 that, if the 4xx error was in fact related to the sender, a different
3736 message to the same recipient with a different sender could confuse
3737 things. In particular, this can happen when sending to a greylisting
3738 server, but other circumstances could also provoke similar problems.
3739 I have changed the default so that the retry time for these errors is now
3740 based a combination of the sender and recipient addresses. This change
3741 can be overridden by setting address_retry_include_sender=false in the
3744 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3745 remote server are returned as part of bounce messages. This was not
3746 happening for LMTP over a pipe (the lmtp transport), but now it is the
3747 same for both kinds of LMTP.
3749 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3750 in header lines that were in fact CNAMEs. This is no longer the case.
3752 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3753 and queue runs started by the daemon processed all messages. This has
3754 been fixed so that -R and -S can now usefully be given with -q<time>.
3756 PH/40 Import PCRE release 6.7 (fixes some bugs).
3758 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3760 PH/42 Give an error if -q is specified more than once.
3762 PH/43 Renamed the variables $interface_address and $interface_port as
3763 $received_ip_address and $received_port, to make it clear that these
3764 values apply to message reception, and not to the outgoing interface when
3765 a message is delivered. (The old names remain recognized, of course.)
3767 PH/44 There was no timeout on the connect() call when using a Unix domain
3768 socket in the ${readsocket expansion. There now is.
3770 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3771 be meaningful with "accept".
3773 SC/01 Eximstats V1.43
3774 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3776 SC/02 Eximstats V1.44
3777 Use a glob alias rather than an array ref in the generated
3778 parser. This improves both readability and performance.
3780 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3781 Collect SpamAssassin and rejection statistics.
3782 Don't display local sender or destination tables unless
3783 there is data to show.
3784 Added average volumes into the top table text output.
3786 SC/04 Eximstats V1.46
3787 Collect data on the number of addresses (recipients)
3788 as well as the number of messages.
3790 SC/05 Eximstats V1.47
3791 Added 'Message too big' to the list of mail rejection
3792 reasons (thanks to Marco Gaiarin).
3794 SC/06 Eximstats V1.48
3795 Mainlog lines which have GMT offsets and are too short to
3796 have a flag are now skipped.
3798 SC/07 Eximstats V1.49 (Alain Williams)
3799 Added the -emptyok flag.
3801 SC/08 Eximstats V1.50
3802 Fixes for obtaining the IP address from reject messages.
3804 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3805 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3806 whitespace changes from 4.64-PH/27
3808 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3811 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3812 are found, allow negative numbers in numeric criteria)
3814 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3816 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3817 to match changes made in 4.64-PH/43
3819 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3821 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3822 "domain" is actually an IP address, and force a failure. This locks out
3823 those revolvers/nameservers that support "A-for-A" lookups, in
3824 contravention of the specifications.
3826 PH/48 When a host name was looked up from an IP address, and the subsequent
3827 forward lookup of the name timed out, the host name was left in
3828 $sender_host_name, contrary to the specification.
3830 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3831 restricted to single-key lookups, Exim was not diagnosing an error if
3832 * or *@ was used with a query-style lookup.
3834 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3836 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3837 long ago, but noone interested enough thought of it. Let's just say that
3838 the "1.1" means that there are some new functions that weren't there at
3839 some point in the past.
3841 PH/51 Error processing for expansion failure of helo_data from an smtp
3842 transport during callout processing was broken.
3844 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3845 tested/used via the -bh/-bhc/-bs options.
3847 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3848 bug, fixed in subsequent PCRE releases).
3850 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3851 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3853 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3859 SC/01 Use a glob alias rather than an array ref in eximstats generated
3860 parser. This improves both readability and performance.
3862 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3863 Don't display local sender or destination tables in eximstats unless
3864 there is data to show.
3865 Added average volumes into the eximstats top table text output.
3867 SC/03 Collect data on the number of addresses (recipients) as well
3868 as the number of messages in eximstats.
3870 TF/01 Correct an error in the documentation for the redirect router. Exim
3871 does (usually) call initgroups() when daemonizing.
3873 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3874 with consistent privilege compared to when running as a daemon.
3876 TF/03 Note in the spec that $authenticated_id is not set for local
3877 submissions from trusted users.
3879 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3880 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3882 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3883 by adding some example configuration directives to the default
3884 configuration file. A little bit of work is required to uncomment the
3885 directives and define how usernames and passwords are checked, but
3886 there is now a framework to start from.
3888 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3889 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3890 without this. I don't know how relevant this is to other LDAP libraries.
3892 PH/02 Add the verb name to the "unknown ACL verb" error.
3894 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3896 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3898 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3899 directory not expanded when it should be if an expanded home directory
3900 was set for the address (which is overridden by the transport).
3902 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3905 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3906 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3907 because it is too late at that time, and has no effect.
3909 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3910 security issue with \' (bugzilla #107). I could not use the
3911 PQescapeStringConn() function, because it needs a PGconn value as one of
3914 PH/08 When testing addresses using -bt, indicate those final addresses that
3915 are duplicates that would not cause an additional delivery. At least one
3916 person was confused, thinking that -bt output corresponded to deliveries.
3917 (Suppressing duplicates isn't a good idea as you lose the information
3918 about possibly different redirections that led to the duplicates.)
3920 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3921 systems where poll() doesn't work, in particular OS X.
3923 PH/10 Added more information to debugging output for retry time not reached.
3925 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3926 operations in malware.c.
3928 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3931 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3932 syslog_duplication set false, log lines that would normally be written
3933 both the the main log and to the reject log were not written to syslog at
3936 PH/14 In the default configuration, change the use of "message" in ACL warn
3937 statements to "add_header".
3939 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3940 not followed by a command (e.g. "seen endif").
3942 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3943 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3946 PH/17 Added extra conditions to the default value of delay_warning_condition
3950 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3951 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3952 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3955 The Auto-Submitted: and various List- headers are standardised, whereas I
3956 don't think Precedence: ever was.
3958 PH/18 Refactored debugging code in route_finduser() to show more information,
3959 in particular, the error code if getpwnam() issues one.
3961 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3962 This is apparently needed in addition to the PH/07 change above to avoid
3963 any possible encoding problems.
3965 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3966 but not after initializing Perl.
3968 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3969 output them only if debugging. By default they are written stderr,
3970 apparently, which is not desirable.
3972 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3975 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3978 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3980 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3981 authenticated or an ident call has been made. Suppress the default
3982 values for $authenticated_id and $authenticated_sender (but permit -oMai
3983 and -oMas) when testing with -bh.
3985 PH/24 Re-jigged the order of the tests in the default configuration so that the
3986 tests for valid domains and recipients precede the DNS black list and CSA
3987 tests, on the grounds that those ones are more expensive.
3989 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3990 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3991 This bug exists in every version of Exim that I still have, right back to
3994 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3995 However, an attempt to turn on case-sensitivity in a regex key by
3996 including (?-i) didn't work because the subject string was already
3997 lowercased, and the effects were non-intuitive. It turns out that a
3998 one-line patch can be used to allow (?-i) to work as expected.
4004 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4005 other effects) broke the use of negated acl sub-conditions.
4007 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4010 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4011 "Deny" causes Exim to reject the incoming connection with a 554 error.
4012 Unfortunately, if there is a major crisis, such as a disk failure,
4013 tcp-wrappers gives "deny", whereas what one would like would be some
4014 kind of temporary error. A kludge has been added to help with this.
4015 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4016 554 error is used if errno is still zero or contains ENOENT (which occurs
4017 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4020 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4022 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4023 errors. Otherwise a message that provokes a temporary error (when other
4024 messages do not) can cause a whole host to time out.
4026 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4027 addresses were routed directly to files or pipes from a redirect router.
4028 File deliveries just didn't batch; pipe deliveries might have suffered
4031 PH/06 A failure to get a lock for a hints database would erroneously always say
4032 "Failed to get write lock", even when it was really a read lock.
4034 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4035 of 0600. This has been changed to use the value of the lockfile_mode
4036 option (which defaults to 0600).
4038 PH/08 Applied small patch from the Sieve maintainer.
4040 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4041 folder from quota calculations, a direct delivery into this folder messed
4042 up the contents of the maildirsize file. This was because the regex was
4043 used only to exclude .Trash (or whatever) when the size of the mailbox
4044 was calculated. There was no check that a delivery was happening into an
4045 excluded directory. This bug has been fixed by ignoring all quota
4046 processing for deliveries into excluded directories.
4048 PH/10 Added the maildirfolder_create_regex option to appendfile.
4054 PH/01 The code for finding all the local interface addresses on a FreeBSD
4055 system running IPv6 was broken. This may well have applied to all BSD
4056 systems, as well as to others that have similar system calls. The broken
4057 code found IPv4 interfaces correctly, but gave incorrect values for the
4058 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4059 that it would not match correctly against @[] and not recognize the IPv6
4062 PH/02 The ipliteral router was not recognizing addresses of the form user@
4063 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4065 PH/03 Added disable_ipv6.
4067 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4068 decoded form, because it is most often used to construct To: headers
4069 lines in autoreplies, and the decoded form may well be syntactically
4070 invalid. However, $reply_address has leading white space removed, and all
4071 newlines turned into spaces so that the autoreply transport does not
4074 PH/05 If group was specified without a user on a router, and no group or user
4075 was specified on a transport, the group from the router was ignored.
4077 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4078 for visible compile-time settings that can be used to change these
4079 numbers, for those that want even more. Backwards compatibility with old
4080 spool files has been maintained. However, going back to a previous Exim
4081 release will lost any variables that are in spool files.
4083 PH/07 Two small changes when running in the test harness: increase delay when
4084 passing a TCP/IP connection to a new process, in case the original
4085 process has to generate a bounce, and remove special handling of
4086 127.0.0.2 (sic), which is no longer necessary.
4088 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4089 be the same on different OS.
4091 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4094 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4095 whether --show-vars was specified or not
4097 JJ/02 exipick: Added support for new ACL variable spool format introduced
4100 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4101 syntactically invalid From: or Reply-to: line, and a filter used this to
4102 generate an autoreply, and therefore failed to obtain an address for the
4103 autoreply, Exim could try to deliver to a non-existent relative file
4104 name, causing unrelated and misleading errors. What now happens is that
4105 it logs this as a hard delivery error, but does not attempt to create a
4108 PH/11 The exinext utility has a -C option for testing purposes, but although
4109 the given file was scanned by exinext itself; it wasn't being passed on
4110 when Exim was called.
4112 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4113 an end-of-file indication when reading a command response.
4115 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4116 compiled. In many other places in Exim, IPv6 addresses are always
4117 recognized, so I have changed this. It also means that IPv4 domain
4118 literals of the form [IPV4:n.n.n.n] are now always recognized.
4120 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4121 used if the router is not running as root, for example, when verifying at
4122 ACL time, or when using -bh. The debugging output from this situation was
4123 non-existent - all you got was a failure to exec. I have made two
4126 (a) Failures to set uid/gid, the current directory, or a process leader
4127 in a subprocess such as that created by queryprogram now generate
4128 suitable debugging output when -d is set.
4130 (b) The queryprogram router detects when it is not running as root,
4131 outputs suitable debugging information if -d is set, and then runs
4132 the subprocess without attempting to change uid/gid.
4134 PH/15 Minor change to Makefile for building test_host (undocumented testing
4137 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4138 additional section of a DNS packet that returns MX or SRV records.
4139 Instead, it always explicitly searches for A/AAAA records. This avoids
4140 major problems that occur when a DNS server includes only records of one
4141 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4142 fixed another bug: if SRV records were looked up and the corresponding
4143 address records were *not* found in the additional section, the port
4144 values from the SRV records were lost.
4146 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4147 using the correct key (the original address) when searching the retry
4148 rules in order to find which one to use for generating the retry hint.
4150 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4151 adding the default one. Similarly, if it contains a Reply-To: header, the
4152 errors_reply_to option, if set, is not used.
4154 PH/19 When calculating a retry time, Exim used to measure the "time since
4155 failure" by looking at the "first failed" field in the retry record. Now
4156 it does not use this if it is later than than the arrival time of the
4157 message. Instead it uses the arrival time. This makes for better
4158 behaviour in cases where some deliveries succeed, thus re-setting the
4159 "first failed" field. An example is a quota failure for a huge message
4160 when small messages continue to be delivered. Without this change, the
4161 "time since failure" will always be short, possible causing more frequent
4162 delivery attempts for the huge message than are intended.
4163 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4165 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4166 $1, $2, $3) because the numerical variables can be reset during some
4167 expansion items (e.g. "match"), thereby losing the authentication data.
4169 PH/21 Make -bV show the size of off_t variables so that the test suite can
4170 decide whether to run tests for quotas > 2G.
4172 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4173 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4174 filecount value is greater than 2G or if a quota value is greater than 2G
4175 on a system where the size of off_t is not greater than 4, a panic error
4178 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4179 never match. The debug and -bh output now contains an explicit error
4180 message indicating a malformed IPv4 address or mask.
4182 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4183 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4184 PH/23 above applies.
4186 PH/25 Do not write to syslog when running in the test harness. The only
4187 occasion when this arises is a failure to open the main or panic logs
4188 (for which there is an explicit test).
4190 PH/26 Added the /no_tell option to "control=freeze".
4192 PH/27 If a host name lookup failed very early in a connection, for example, if
4193 the IP address matched host_lookup and the reverse lookup yielded a name
4194 that did not have a forward lookup, an error message of the form "no IP
4195 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4196 could be logged. Now it outputs the IP address instead of "NULL".
4198 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4199 allows the sender and the authenticated sender to be set when
4200 submitting a message from within Exim. Since child_open_exim() is
4201 documented for local_scan(), the new function should be too.
4203 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4204 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4205 results in an empty string is now treated as unset.
4207 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4209 PH/31 Added sender_verify_fail logging option.
4211 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4212 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4213 tidied the source and removed it altogether.
4215 PH/33 When a queue run was abandoned because the load average was too high, a
4216 log line was always written; now it is written only if the queue_run log
4217 selector is set. In addition, the log line for abandonment now contains
4218 information about the queue run such as the pid. This is always present
4219 in "start" and "stop" lines but was omitted from the "abandon" line.
4221 PH/34 Omit spaces between a header name and the colon in the error message that
4222 is given when verify = headers_syntax fails (if there are lots of them,
4223 the message gets confusing).
4225 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4226 names, as there are now some PTR records that contain slashes. This check
4227 is only to protect against broken name servers that fall over on strange
4228 characters, so the fact that it applies to all lookups doesn't matter.
4230 PH/36 Now that the new test suite is complete, we can remove some of the
4231 special code in Exim that was needed for the old test suite. For example,
4232 sorting DNS records because real resolvers return them in an arbitrary
4233 order. The new test suite's fake resolver always returns records in the
4236 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4237 bounces) except when queue_only is set, to avoid logging races between
4238 the different processes.
4240 PH/38 Panic-die if .include specifies a non-absolute path.
4242 PH/39 A tweak to the "H" retry rule from its user.
4244 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4245 a label. They prevented compilation on older perls.
4247 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4248 a warning to be raised on newish perls.
4250 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4251 on queue. Changes to match documented behaviour of showing count of
4252 messages matching specified criteria.
4254 PH/40 Changed the default ident timeout from 30s to 5s.
4256 PH/41 Added support for the use of login_cap features, on those BSD systems
4257 that have them, for controlling the resources used by pipe deliveries.
4259 PH/42 The content-scanning code uses fopen() to create files in which to put
4260 message data. Previously it was not paying any attention to the mode of
4261 the files. Exim runs with umask(0) because the rest of the code creates
4262 files with open(), and sets the required mode explicitly. Thus, these
4263 files were ending up world-writeable. This was not a big issue, because,
4264 being within the spool directory, they were not world-accessible. I have
4265 created a function called modefopen, which takes an additional mode
4266 argument. It sets umask(777), creates the file, chmods it to the required
4267 mode, then resets the umask. All the relevant calls to fopen() in the
4268 content scanning code have been changed to use this function.
4270 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4271 to 24 hours. This avoids potential overflow problems when processing G
4272 and H retry rules. I suspect nobody ever tinkers with this value.
4274 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4276 PH/45 When the plaintext authenticator is running as a client, the server's
4277 challenges are checked to ensure they are valid base64 strings. By
4278 default, the authentication attempt is cancelled if an invalid string is
4279 received. Setting client_ignore_invalid_base64 true ignores these errors.
4280 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4281 they are received. Thus, the responses can be made to depend on the
4282 challenges. If an invalid string is ignored, an empty string is placed in
4285 PH/46 Messages that are created by the autoreply transport now contains a
4286 References: header, in accordance with RFCs 2822 and 3834.
4288 PH/47 Added authenticated_sender_force to the smtp transport.
4290 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4292 PH/49 Installed latest patch from the Sieve maintainer.
4294 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4295 was also set, the appendfile transport was unnecessarily scanning a
4296 directory of message files (e.g. for maildir delivery) to find the count
4297 of files (along with the size), even though it did not need this
4298 information. It now does the scan only if it needs to find either the
4299 size of the count of files.
4301 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4303 PH/52 Two bugs concerned with error handling when the smtp transport is
4306 (i) Exim was not creating retry information for temporary errors given
4307 for individual recipients after the DATA command when the smtp transport
4308 was used in LMTP mode. This meant that they could be retried too
4309 frequently, and not timed out correctly.
4311 (ii) Exim was setting the flag that allows error details to be returned
4312 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4313 recipients that were returned after the DATA command.
4315 PH/53 This is related to PH/52, but is more general: for any failing address,
4316 when detailed error information was permitted to be returned to the
4317 sender, but the error was temporary, then after the final timeout, only
4318 "retry timeout exceeded" was returned. Now it returns the full error as
4319 well as "retry timeout exceeded".
4321 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4322 do this, and (what is worse) MTAs that accept it.
4324 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4325 will now be deprecated.
4327 PH/56 New os.c-cygwin from the Cygwin maintainer.
4329 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4330 formats (previously only available in exim formats via -bpr, -bpru,
4331 and -bpra. Now also available in native and exiqgrep formats)
4333 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4334 with very large, slow to parse queues
4336 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4338 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4340 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4341 responses to authentication challenges, though it was showing the
4342 challenges; (ii) I've removed the CR characters from the debug output for
4345 PH/58 Allow for the insertion of a newline as well as a space when a string
4346 is turned into more than one encoded-word during RFC 2047 encoding. The
4347 Sieve code now uses this.
4349 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4350 data_4xx, lost_connection, tls_required.
4352 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4353 message was being sent as an SMTP response.
4355 PH/61 Add -l and -k options to exicyclog.
4357 PH/62 When verifying, if an address was redirected to one new address, so that
4358 verification continued, and the new address failed or deferred after
4359 having set something in $address_data, the value of $address_data was not
4360 passed back to the ACL. This was different to the case when no
4361 redirection occurred. The value is now passed back in both cases.
4363 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4364 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4365 use login_cap.h, so on its own it isn't the distinguishing feature. The
4366 new name refers directly to the setclassresources() function.
4368 PH/65 Added configuration files for NetBSD3.
4370 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4372 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4373 is preferred over IPv4.
4375 PH/68 The bounce_return_message and bounce_return_body options were not being
4376 honoured for bounces generated during the reception of non-SMTP messages.
4377 In particular, this applied to messages rejected by the ACL. This bug has
4378 been fixed. However, if bounce_return_message is true and bounce_return_
4379 body is false, the headers that are returned for a non-SMTP message
4380 include only those that have been read before the error was detected.
4381 (In the case of an ACL rejection, they have all been read.)
4383 PH/69 The HTML version of the specification is now built in a directory called
4384 spec_html instead of spec.html, because the latter looks like a path with
4385 a MIME-type, and this confuses some software.
4387 PH/70 Catch two compiler warnings in sieve.c.
4389 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4390 function verify_get_ident() calls ip_connect() to connect a socket, but
4391 if the "connect()" function timed out, ip_connect() used to close the
4392 socket. However, verify_get_ident() also closes the socket later, and in
4393 between Exim writes to the log, which may get opened at this point. When
4394 the socket was closed in ip_connect(), the log could get the same file
4395 descriptor number as the socket. This naturally causes chaos. The fix is
4396 not to close the socket in ip_connect(); the socket should be closed by
4397 the function that creates it. There was only one place in the code where
4398 this was missing, in the iplookup router, which I don't think anybody now
4399 uses, but I've fixed it anyway.
4401 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4402 well as to direct DNS lookups. Otherwise the handling of names in host
4403 lists is inconsistent and therefore confusing.
4409 PH/01 Two changes to the default runtime configuration:
4411 (1) Move the checks for relay_from_hosts and authenticated clients from
4412 after to before the (commented out) DNS black list checks.
4414 (2) Add control=submission to the relay_from_hosts and authenticated
4415 clients checks, on the grounds that messages accepted by these
4416 statements are most likely to be submissions.
4418 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4420 (1) Generate an error if the third argument for the ${prvs expansion is
4423 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4426 (3) Reset the variables that are obtained from the first argument of
4427 ${prvscheck and used in the second argument before leaving the code,
4428 because their memory is reclaimed, so using them afterwards may do
4431 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4432 one (it's much easier than Tom thought :-).
4434 (5) Because of (4), we can now allow for the use of $prvscheck_result
4435 inside the third argument.
4437 PH/03 For some reason, the default setting of PATH when running a command from
4438 a pipe transport was just "/usr/bin". I have changed it to
4441 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4442 anything to be listed in the output from -bV.
4444 PH/05 When a filter generated an autoreply, the entire To: header line was
4445 quoted in the delivery log line, like this:
4447 => >A.N.Other <ano@some.domain> <original@ddress> ...
4449 This has been changed so that it extracts the operative address. There
4450 may be more than one such address. If so, they are comma-separated, like
4453 => >ano@some.domain,ona@other.domain <original@ddress> ...
4455 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4456 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4457 not being looked up in the rDNS to get a host name, Exim was showing the
4458 IP address twice in Received: lines, even though the IP addresses were
4459 identical. For example:
4461 Received: from [1.2.3.4] (helo=[1.2.3.4])
4463 However, if the real host name was known, it was omitting the HELO data
4464 if it matched the actual IP address. This has been tidied up so that it
4465 doesn't show the same IP address twice.
4467 PH/07 When both +timestamp and +memory debugging was on, the value given by
4468 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4469 called by the debug printing, thereby overwriting the timestamp buffer.
4470 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4472 PH/08 When the original message was included in an autoreply transport, it
4473 always said "this is a copy of the message, including all the headers",
4474 even if body_only or headers_only was set. It now gives an appropriate
4477 PH/09 Applied a patch from the Sieve maintainer which:
4479 o fixes some comments
4480 o adds the (disabled) notify extension core
4481 o adds some debug output for the result of if/elsif tests
4482 o points to the current vacation draft in the documentation
4483 and documents the missing references header update
4487 o fixes a bug in processing the envelope test (when testing
4488 multiple envelope elements, the last element determined the
4491 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4492 Electronic Mail") by including:
4494 Auto-submitted: auto-generated
4496 in the messages that it generates (bounce messages and others, such as
4497 warnings). In the case of bounce messages for non-SMTP messages, there was
4498 also a typo: it was using "Auto_submitted" (underscore instead of
4499 hyphen). Since every message generated by Exim is necessarily in response
4500 to another message, thes have all been changed to:
4502 Auto-Submitted: auto-replied
4504 in accordance with these statements in the RFC:
4506 The auto-replied keyword:
4508 - SHOULD be used on messages sent in direct response to another
4509 message by an automatic process,
4511 - MUST NOT be used on manually-generated messages,
4513 - MAY be used on Delivery Status Notifications (DSNs) and Message
4514 Disposition Notifications (MDNs),
4516 - MUST NOT be used on messages generated by automatic or periodic
4517 processes, except for messages which are automatic responses to
4520 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4521 to the default Received: header definition.
4523 PH/12 Added log selector acl_warn_skipped (default on).
4525 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4526 variables because (a) they are in the wrong storage pool and (b) even if
4527 they were copied, it wouldn't work properly because of the caching.
4529 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4530 checking when decoding. Apparently there are clients that generate
4531 overlong encoded strings. Why am I not surprised?
4533 PH/15 If the first argument of "${if match_address" was not empty, but did not
4534 contain an "@" character, Exim crashed. Now it writes a panic log message
4535 and treats the condition as false.
4537 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4539 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4540 extension "envelope-auth". The code is finished and in agreement with
4541 other implementations, but there is no documentation so far and in fact,
4542 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4543 not changing the active code.
4545 Print executed "if" and "elsif" statements when debugging is used. This
4546 helps a great deal to understand what a filter does.
4548 Document more things not specified clearly in RFC3028. I had all this
4549 sorted out, when out of a sudden new issues came to my mind. Oops."
4551 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4554 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4555 canonical form (as documented). However, after a host name lookup from
4556 the IP address, check_host() was doing a simple string comparison with
4557 addresses acquired from the DNS when checking that the found name did
4558 have the original IP as one of its addresses. Since any found IPv6
4559 addresses are likely to be in abbreviated form, the comparison could
4560 fail. Luckily, there already exists a function for doing the comparison
4561 by converting both addresses to binary, so now that is used instead of
4562 the text comparison.
4564 PH/20 There was another similar case to PH/19, when a complete host name was
4565 given in a host list; looking up its IP address could give an abbreviated
4566 form, whereas the current host's name might or might not be abbreviated.
4567 The same fix has been applied.
4573 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4574 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4577 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4578 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4580 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4582 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4583 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4584 Also, for "Auto-Submitted", treat anything other than "no" as
4585 non-personal, in accordance with RFC 3834. (Previously it treated
4586 anything starting "auto-" as non-personal.)
4588 TF/01 The control=submission/name=... option had a problem with syntax
4589 errors if the name included a slash character. The /name= option
4590 now slurps the rest of the string, so it can include any characters
4591 but it must come last in the list of options (after /sender_retain
4594 PH/05 Some modifications to the interface to the fake nameserver for the new
4602 TK/01 Added the "success_on_redirect" address verification option. See
4603 NewStuff for rationale and an example.
4605 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4607 PH/02 Patch to exigrep to allow it to work on syslog lines.
4609 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4610 fread() to skip over the body file's header line, because in Cygwin the
4611 header line is locked and is inaccessible.
4613 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4614 co-exist for some time) to make it clear that it is the Exim ID that is
4615 referenced, not the Message-ID: header line.
4617 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4618 string_format() function, because snprintf() does not exist on all
4621 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4622 ${stat: expansion item.
4624 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4625 protocol synchronization error", to keep the pedants happy.
4627 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4628 well as for IRIX systems, when gcc is being used. See the host.c source
4631 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4633 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4636 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4637 transport and to the smtp transport in LMTP mode.
4639 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4641 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4642 run a filter in a subprocess. This could lead to confusion in subsequent
4643 lookups in the parent process. There should also be a search_tidyup() at
4644 the end of the subprocess.
4646 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4647 only if the host matched helo_try_verify_hosts, which caused the
4648 verification to occur when the EHLO/HELO command was issued. The ACL just
4649 tested the remembered result. Now, if a previous verification attempt has
4650 not happened, "verify = helo" does it there and then.
4652 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4654 TK/03 Fix log output including CR from clamd.
4656 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4657 exist provoked a memory error which could cause a segfault.
4659 PH/15 Installed PCRE 6.2
4661 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4663 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4664 of the problem. Specifically, suggested +O2 rather than +O1 for the
4667 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4669 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4670 record, the port was not correctly logged when the outgoing_port log
4671 selector was set (it logged the transort's default port).
4673 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4674 fallback_hosts, and "hosts" in the smtp transport.
4676 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4677 host errors such as "Connection refused".
4679 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4680 authentication with radiusclient 0.4.9:
4682 - Error returned from rc_read_config was caught wrongly
4683 - Username/password not passed on to radius server due to wrong length.
4685 The presumption is that some radiusclient API changes for 4.51/PH/17
4686 were not taken care of correctly. The code is still untested by me (my
4687 Linux distribution still has 0.3.2 of radiusclient), but it was
4688 contributed by a Radius user.
4690 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4691 expanding the "port" option of the smtp transport.
4693 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4694 while reading a MIME header. Thanks to Tom Hughes for a patch.
4696 PH/24 Include config.h inside local_scan.h so that configuration settings are
4699 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4700 that in an ACL for RCPT (for example), you can examine exactly what was
4703 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4704 commands, but it was not correctly comparing the address with the actual
4705 client host address. Thus, it would show the EHLO address in Received:
4706 header lines when this was not necessary.
4708 PH/27 Added the % operator to ${eval:}.
4710 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4711 it should be ignoring failures (because with -C, for example, it has lost
4712 privilege). It wasn't ignoring creation failures other than "already
4715 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4718 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4719 because an input line was too long, either on its own, or by virtue of
4720 too many continuations, the temporary file was not being removed, and the
4721 return code was incorrect.
4723 PH/31 Missing "BOOL" in function definition in filtertest.c.
4725 PH/32 Applied Sieve patches from the maintainer.
4727 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4729 PH/33 Added "verify = not_blind".
4731 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4732 Local/Makefile (with some defaults set). These are used in built scripts
4733 such as exicyclog, but they have never been used in the exim_install
4734 script (though there are many overriding facilities there). I have
4735 arranged that the exim_install script now takes note of these two
4738 PH/35 Installed configuration files for Dragonfly.
4740 PH/36 When a locally submitted message by a trusted user did not contain a
4741 From: header, and the sender address was obtained from -f or from an SMTP
4742 MAIL command, and the trusted user did not use -F to supply a sender
4743 name, $originator_name was incorrectly used when constructing a From:
4744 header. Furthermore, $originator_name was used for submission mode
4745 messages from external hosts without From: headers in a similar way,
4746 which is clearly wrong.
4748 PH/37 Added control=suppress_local_fixups.
4750 PH/38 When log_selector = +received_sender was set, and the addition of the
4751 sender made the log line's construction buffer exactly full, or one byte
4752 less than full, an overflow happened when the terminating "\n" was
4755 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4756 when the result of a list match is failure because a DNS lookup failed.
4758 PH/40 RM_COMMAND is now used in the building process.
4760 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4761 the "build-* directories that it finds.
4763 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4764 address in a domain literal was a prefix of an interface address.
4766 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4767 when verifying a sender address, unless rewrite_headers is false.
4769 PH/44 Wrote a long comment about why errors_to addresses are verified as
4770 recipients, not senders.
4772 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4773 the ratelimit ACL was added.
4775 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4777 PH/46 Added extra information about PostgreSQL errors to the error string.
4779 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4780 avoiding the need to install special zones in a real server. This is
4781 backwards compatible; if it can't find the fake resolver, it drops back.
4782 Thus, both old and new test suites can be run.
4784 TF/02 Added util/ratelimit.pl
4786 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4787 clock is set back in time.
4789 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4790 Candler <B.Candler@pobox.com>.
4792 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4793 OK for addresses that are the result of redirections.
4795 PH/48 A number of further additions for the benefit of the new test suite,
4796 including a fake gethostbyname() that interfaces to the fake DNS resolver
4799 TF/06 The fix for widen_domains has also been applied to qualify_single and
4800 search_parents which are the other dnslookup options that can cause
4803 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4806 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4808 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4809 service" if it encountered an MX record with an empty target hostname.
4810 The message is now "an MX or SRV record indicated no SMTP service".
4812 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4813 if the DNS of the sending site is misconfigured. This is quite a
4814 common situation. This change restores the behaviour of treating a
4815 helo verification defer as a failure.
4817 PH/51 If self=fail was set on a router, the bounce message did not include the
4818 actual error message.
4824 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4826 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4827 command itself ended in error, the underlying message about the transport
4828 filter timeout was being overwritten with the pipe command error. Now the
4829 underlying error message should be appended to the second error message.
4831 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4833 PH/02 Reduce the amount of output that "make" produces by default. Full output
4834 can still be requested.
4836 PH/03 The warning log line about a condition test deferring for a "warn" verb
4837 was being output only once per connection, rather than after each
4838 occurrence (because it was using the same function as for successful
4839 "warn" verbs). This seems wrong, so I have changed it.
4841 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4842 it should not have, which might have caused a crash in the right
4843 circumstances, but probably never did.
4845 PH/04 Installed a modified version of Tony Finch's patch to make submission
4846 mode fix the return path as well as the Sender: header line, and to
4847 add a /name= option so that you can make the user's friendly name appear
4850 TF/03 Added the control = fakedefer ACL modifier.
4852 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4853 Mark Lowes for thorough testing.
4855 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4857 TK/03 Merged latest SRS patch from Miles Wilton.
4859 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4860 with the definition in sysexits.h (which is #included earlier).
4861 Fortunately, Exim does not actually use EX_OK. The code used to try to
4862 preserve the sysexits.h value, by assuming that macro definitions were
4863 scanned for macro replacements. I have been disabused of this notion,
4864 so now the code just undefines EX_OK before #including unistd.h.
4866 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4867 in the smtp transport. When a block could not be written in a single
4868 write() function, the timeout was being re-applied to each part-write.
4869 This seems wrong - if the receiver was accepting one byte at a time it
4870 would take for ever. The timeout is now adjusted when this happens. It
4871 doesn't have to be particularly precise.
4873 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4874 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4876 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4877 without a domain if the check to <postmaster@domain> fails.
4879 SC/01 Eximstats: added -xls and the ability to specify output files
4880 (patch written by Frank Heydlauf).
4882 SC/02 Eximstats: use FileHandles for outputting results.
4884 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4886 SC/04 Eximstats: fixed display of large numbers with -nvr option
4888 SC/05 Eximstats: fixed merging of reports with empty tables.
4890 SC/06 Eximstats: added the -include_original_destination flag
4892 SC/07 Eximstats: removed tabs and trailing whitespace.
4894 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4896 TK/06 MBOX spool code: Add real "From " MBOX separator line
4897 so the .eml file is really in mbox format (even though
4898 most programs do not really care). Patch from Alex Miller.
4900 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4901 The latter is generated from $received_to and is only set if the
4902 message has one envelope recipient. SA can use these headers,
4903 obviously out-of-the-box. Patch from Alex Miller.
4905 PH/08 The ${def test on a variable was returning false if the variable's
4906 value was "0", contrary to what the specification has always said!
4907 The result should be true unless the variable is empty.
4909 PH/09 The syntax error of a character other than { following "${if
4910 def:variable_name" (after optional whitespace) was not being diagnosed.
4911 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4912 accidental colon was present, for example, could give incorrect results.
4914 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4915 result is used (not including appendfile, where other changes are about
4918 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4919 This involved changing a lot of size variables from int to off_t. It
4920 should work with maildirs and everything.
4922 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4923 spamd dying while we are connected to it.
4925 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4928 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4929 function for BDB 4.3.
4931 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4933 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4934 the sources for PCRE have been re-arranged and more files are now
4937 PH/15 The code I had for printing potentially long long variables in PH/11
4938 above was not the best (it lost precision). The length of off_t variables
4939 is now inspected at build time, and an appropriate printing format (%ld
4940 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4941 to be "long long int" or "long int". This is needed for the internal
4942 formatting function string_vformat().
4944 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4945 the configuration file to be ":syslog", then the script "guesses" where
4946 the logs files are, rather than using the compiled in default. In our
4947 case the guess is not the same as the compiled default, so the script
4948 suddenly stopped working when I started to use syslog. The patch checks
4949 to see if log_file_path is "". If so, it attempts to read it from exim
4950 with no configuration file to get the compiled in version, before it
4951 falls back to the previous guessing code."
4953 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4954 implementing BATV in an Exim configuration. See NewStuff for the gory
4957 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4958 Makefile that are specific to HP-UX.
4960 PH/18 If the "use_postmaster" option was set for a recipient callout together
4961 with the "random" option, the postmaster address was used as the MAIL
4962 FROM address for the random test, but not for the subsequent recipient
4963 test. It is now used for both.
4965 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4966 patch removes a few documentation additions to RFC 3028, because the
4967 latest draft now contains them. It adds the new en;ascii-case comparator
4968 and a new error check for 8bit text in MIME parts. Comparator and
4969 require names are now matched exactly. I enabled the subaddress
4970 extension, but it is not well tested yet (read: it works for me)."
4972 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4973 rework some of the code of TK/09 above to avoid the hardwired use of
4974 "%lld" and "long long". Replaced the call to snprintf() with a call to
4977 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4978 records point to non-existent hosts", "retry timeout exceeded", and
4979 "retry time not reached for any host after a long failure period".
4981 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4982 experimental DomainKeys support:
4984 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4985 (2) On an error such as an illegally used "control", the wrong name for
4986 the control was given.
4988 These problems did NOT occur unless DomainKeys support was compiled.
4990 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4992 PH/24 Added ${if match_ip condition.
4994 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4995 databases so that it will be absolutely obvious if a crash occurs in the
4996 DB library. This is a regular occurrence (often caused by mis-matched
4999 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5000 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5001 picked up on a user's system that detects such things. There doesn't seem
5002 to be a gcc warning option for this - only an attribute that has to be
5003 put on the function's prototype. It seems that in Fedora Core 4 they have
5004 set this on a number of new functions. No doubt there will be more in due
5007 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5008 specify a transport. However, if an address that was verified by such a
5009 router was the subject of a callout, Exim crashed because it tried to
5010 read the rcpt_include_affixes from the non-existent transport. Now it
5011 just assumes that the setting of that option is false. This bug was
5012 introduced by 4.51/PH/31.
5014 PH/28 Changed -d+all to exclude +memory, because that information is very
5015 rarely of interest, but it makes the output a lot bigger. People tend to
5016 do -d+all out of habit.
5018 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5019 code in os-type was giving problems when libc.so lives in lib64, like on
5022 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5023 aren't the modern standard, and it seems that some systems' include files
5024 don't always have them. Exim was already checking for some of the newer
5025 ones like T_AAAA, and defining it itself. I've added checks for all the
5026 record types that Exim uses.
5028 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5029 not automatically generating a new one, as it is supposed to. This
5030 prevented TLS from working. If the file did exist, but contained invalid
5031 data, a new version was generated, as expected. It was only the case of a
5032 non-existent file that was broken.
5034 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5035 with a change in libdomainkeys > 0.64.
5037 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5038 from DNS. If the selector record carries the flag, it now has
5039 precedence over the domain-wide flag.
5041 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5043 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5044 the use of a port name that isn't defined in /etc/services) occurred, the
5045 message was deferred as in a normal delivery, and thus remained on the
5046 spool, instead of being failed because of the mua_wrapper setting. This
5047 is now fixed, and I tidied up some of the mua_wrapper messages at the
5050 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5051 the messages in a hash of arrays rather than using individual hashes.
5052 This is a bit cleaner and results in dramatic memory savings, albeit
5053 at a slight CPU cost.
5055 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5056 as requested by Marc Sherman.
5058 SC/10 Eximstats: added histograms for user specified patterns as requested
5061 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5063 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5064 fopen() in the content-scanning modules that did not already have it.
5070 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5071 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5073 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5075 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5077 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5078 to test Sieve filters that use "vacation".
5080 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5081 that changes the way the GnuTLS parameters are stored in the cache file.
5082 The new format can be generated externally. For backward compatibility,
5083 if the data in the cache doesn't make sense, Exim assumes it has read an
5084 old-format file, and it generates new data and writes a new file. This
5085 means that you can't go back to an older release without removing the
5088 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5089 work if there are any delivery delays because "one_time" forces the
5090 parent to be marked "delivered", so its unseen clone is never tried
5091 again. For this reason, Exim now forbids the simultaneous setting of
5094 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5095 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5096 entry if you want to know the details. The fix, however, neglected to
5097 consider the case when local delivery batching is involved. The test for
5098 "previously delivered" was not happening when checking to see if an
5099 address could be batched with a previous (undelivered) one; under
5100 certain circumstances this could lead to multiple deliveries to the same
5103 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5104 in its include files, and this causes problems building Exim.
5106 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5107 header_syntax) but Exim was just ignoring anything given after a slash.
5108 In particular, this caused confusion with an attempt to use "verify =
5109 reverse_host_lookup/defer_ok". An error is now given when options are
5110 supplied for verify items that do not have them. (Maybe reverse_host_
5111 lookup should have a defer_ok option, but that's a different point.)
5113 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5114 defined by RFC 821) to 2048, because there were problems with some AUTH
5115 commands, and RFC 1869 says the size should be increased for extended
5116 SMTP commands that take arguments.
5118 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5121 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5122 "unknown" error; now it says that the functionality isn't in the binary.
5124 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5125 an address' error message when a string expansion fails (syntax or
5126 whatever). Otherwise the password may appear in the log. Following change
5127 PH/42 below, there is no longer a chance of it appearing in a bounce
5130 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5132 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5133 of its addresses was ever tried. (Bugzilla bug #2.)
5135 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5136 the result incorrectly in the debug output. (It correctly added a newline
5137 to what was transported.)
5139 TF/01 Added $received_time.
5141 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5142 commented out examples of how to interface to a virus scanner and to
5143 SpamAssassin. Also added commented examples of av_scanner and
5144 spamd_address settings.
5146 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5147 and controls are allowed in which ACLs. There were a couple of minor
5148 errors. Some of the entries in the conditions table (which is a table of
5149 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5150 negation of where the condition IS allowed.
5152 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5154 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5155 header file does not have a version number, so I've had to invent a new
5156 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5157 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5158 radiusclient), but it was contributed by a Radius user.
5160 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5161 files or directories, for OpenSSL.
5163 PH/19 When an Exim process that is running as root has to create an Exim log
5164 file, it does so in a subprocess that runs as exim:exim so as to get the
5165 ownership right at creation (otherwise, other Exim processes might see
5166 the file with the wrong ownership). There was no test for failure of this
5167 fork() call, which would lead to the process getting stuck as it waited
5168 for a non-existent subprocess. Forks do occasionally fail when resources
5169 run out. I reviewed all the other calls to fork(); they all seem to check
5172 PH/20 When checking for unexpected SMTP input at connect time (before writing
5173 the banner), Exim was not dealing correctly with a non-positive return
5174 from the read() function. If the client had disconnected by this time,
5175 the result was a log entry for a synchronization error with an empty
5176 string after "input=" when read() returned zero. If read() returned -1
5177 (an event I could not check), uninitialized data bytes were printed.
5178 There were reports of junk text (parts of files, etc) appearing after
5181 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5183 PH/22 Added support for macro redefinition, and (re)definition in between
5184 driver and ACL definitions.
5186 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5187 forgetting to use the resulting value; it was using the unexpanded value.
5189 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5190 hadn't been configured. The fix is from Juergen Kreileder, who
5191 understands it better than I do:
5193 "Here's what I see happening with three configured cyrus_sasl
5194 authenticators configured (plain, login, cram-md5):
5196 On startup auth_cyrus_sasl_init() gets called for each of these.
5197 This means three calls to sasl_listmech() without a specified mech_list.
5198 => SASL tests which mechs of all available mechs actually work
5199 => three warnings about OTP not working
5200 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5202 With the patch, sasl_listmech() also gets called three times. But now
5203 SASL's mech_list option is set to the server_mech specified in the the
5204 authenticator. Or in other words, the answer from sasl_listmech()
5205 gets limited to just the mech you're testing for (which is different
5207 => the return list contains just 'plain' or 'login', 'cram-md5' or
5208 nothing depending on the value of ob->server_mech.
5210 I've just tested the patch: Authentication still works fine,
5211 unavailable mechs specified in the exim configuration are still
5212 caught, and the auth.log warnings about OTP are gone."
5214 PH/25 When debugging is enabled, the contents of the command line are added
5215 to the debugging output, even when log_selector=+arguments is not
5218 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5219 answer is "GNU", and only if the return is "GNU/something" is the answer
5222 PH/27 $acl_verify_message is now set immediately after the failure of a
5223 verification in an ACL, and so is available in subsequent modifiers. In
5224 particular, the message can be preserved by coding like this:
5226 warn !verify = sender
5227 set acl_m0 = $acl_verify_message
5229 Previously, $acl_verify_message was set only while expanding "message"
5230 and "log_message" when a very denied access.
5232 PH/28 Modified OS/os.c-Linux with
5234 -#ifndef OS_LOAD_AVERAGE
5235 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5237 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5238 nomenclature these days.)
5240 PH/29 Installed patch from the Sieve maintainer that adds the options
5241 sieve_useraddress and sieve_subaddress to the redirect router.
5243 PH/30 In these circumstances:
5244 . Two addresses routed to the same list of hosts;
5245 . First host does not offer TLS;
5246 . First host accepts first address;
5247 . First host gives temporary error to second address;
5248 . Second host offers TLS and a TLS session is established;
5249 . Second host accepts second address.
5250 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5251 and peerdn, if requested) that were in fact used only for the second
5254 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5255 attention to any local part prefix or suffix that was matched by the
5256 router that accepted the address. It now behaves in the same way as it
5257 does for delivery: the affixes are removed from the local part unless
5258 rcpt_include_affixes is set on the transport.
5260 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5261 timeout during the DATA phase of an incoming message.
5263 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5264 applied a patch sanctioned by the Sieve maintainer.
5266 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5267 the uid or gid is negative. A case of a negative gid caused this to be
5268 noticed. The fix allows for either to be negative.
5270 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5271 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5274 PH/36 The change PH/12 above was broken. Fixed it.
5276 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5277 the grounds that routing the same address twice would always produce the
5278 same answer. This might have been true once, but it is certainly no
5279 longer true now. Routing a child address may depend on the previous
5280 routing that produced that child. Some complicated redirection strategies
5281 went wrong when messages had multiple recipients, and made Exim's
5282 behaviour dependent on the order in which the addresses were given.
5284 I have moved the duplicate checking until after the routing is complete.
5285 Exim scans the addresses that are assigned to local and remote
5286 transports, and removes any duplicates. This means that more work will be
5287 done, as duplicates will always all be routed, but duplicates are
5288 presumably rare, so I don't expect this is of any significance.
5290 For deliveries to pipes, files, and autoreplies, the duplicate checking
5291 still happens during the routing process, since they are not going to be
5294 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5295 It corrects a timeout issue with spamd. This is Ian's comment: "The
5296 background is that sometimes spamd either never reads data from a
5297 connection it has accepted, or it never writes response data. The exiscan
5298 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5299 blindly assumes that writes won't block so it may never time out."
5301 PH/39 Allow G after quota size as well as K and M.
5303 PH/40 The value set for $authenticated_id in an authenticator may not contain
5304 binary zeroes or newlines because the value is written to log lines and
5305 to spool files. There was no check on this. Now the value is run through
5306 the string_printing() function so that such characters are converted to
5307 printable escape sequences.
5309 PH/41 $message_linecount is a new variable that contains the total number of
5310 lines in the message. Compare $body_linecount, which is the count for the
5313 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5314 bounce and delay warning messages, except in certain special cases, which
5317 (a) An SMTP error message from a remote host;
5318 (b) A message specified in a :fail: redirection;
5319 (c) A message specified in a "fail" command in a system filter;
5320 (d) A message specified in a FAIL return from the queryprogram router;
5321 (e) A message specified by the cannot_route_message router option.
5323 In these cases only, Exim does include the error details in bounce and
5324 warning messages. There are also a few cases where bland messages such
5325 as "unrouteable address" or "local delivery error" are given.
5327 PH/43 $value is now also set for the "else" part of a ${run expansion.
5329 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5330 being worked on, but at least Exim now implements the latest version to
5333 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5334 process to complete was treated as a delivery failure, a timeout while
5335 writing the message to the pipe was logged, but erroneously treated as a
5336 successful delivery. Such timeouts include transport filter timeouts. For
5337 consistency with the overall process timeout, these timeouts are now
5338 treated as errors, giving rise to delivery failures by default. However,
5339 there is now a new Boolean option for the pipe transport called
5340 timeout_defer, which, if set TRUE, converts the failures into defers for
5341 both kinds of timeout. A transport filter timeout is now identified in
5344 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5345 systems where "make" and "gmake" are different, calling "gmake" at top
5346 level broke things. I've arranged for the value of $(MAKE) to be passed
5347 from the Makefile to this script so that it can call the same version of
5351 A note about Exim versions 4.44 and 4.50
5352 ----------------------------------------
5354 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5355 changes of various kinds. As a consequence, a big documentation update was
5356 needed. This delayed the release for rather longer than seemed good, especially
5357 in the light of a couple of (minor) security issues. Therefore, the changes
5358 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5359 release. So 4.44 and 4.50 are in effect two different branches that both start
5362 I have left the 4.50 change log unchanged; it contains all the changes since
5363 4.43. The change log for 4.44 is below; many of its items are identical to
5364 those for 4.50. This seems to be the most sensible way to preserve the
5365 historical information.
5371 1. Minor wording change to the doc/README.SIEVE file.
5373 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5374 computation of the current number of files was incorrect.
5376 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5377 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5380 4. Give more explanation in the error message when the command for a transport
5381 filter fails to execute.
5383 5. There are several places where Exim runs a non-Exim command in a
5384 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5385 was being done only for the command run by the queryprogram router. It is
5386 now done for all such subprocesses. The other cases are: ${run, transport
5387 filters, and the commands run by the lmtp and pipe transports.
5389 6. Added CONFIGURE_GROUP build-time option.
5391 7. Some older OS have a limit of 256 on the maximum number of file
5392 descriptors. Exim was using setrlimit() to set 1000 as a large value
5393 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5394 systems. I've change it so that if it can't get 1000, it tries for 256.
5396 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5397 was an oversight, and furthermore, ever since the addition of extra
5398 controls (e.g. 4.43/32), the checks on when to allow different forms of
5399 "control" were broken. There should now be diagnostics for all cases when a
5400 control that does not make sense is encountered.
5402 9. Added the /retain_sender option to "control=submission".
5404 10. $recipients is now available in the predata ACL (oversight).
5406 11. Tidy the search cache before the fork to do a delivery from a message
5407 received from the command line. Otherwise the child will trigger a lookup
5408 failure and thereby defer the delivery if it tries to use (for example) a
5409 cached ldap connection that the parent has called unbind on.
5411 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5412 of $address_data from the recipient verification was clobbered by the
5413 sender verification.
5415 13. The value of address_data from a sender verification is now available in
5416 $sender_address_data in subsequent conditions in the ACL statement.
5418 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5420 15. Added a new option "connect=<time>" to callout options, to set a different
5423 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5424 was its contents. (It was OK if the option was not defined at all.)
5426 17. A "Completed" log line is now written for messages that are removed from
5427 the spool by the -Mrm option.
5429 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5430 information about exactly what failed.
5432 19. Added -dd to debug only the daemon process.
5434 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5435 handles timeouts, both on the server side and network timeouts. Renamed the
5436 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5438 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5439 It is now set to "smtps".
5441 22. $host_address is now set to the target address during the checking of
5442 ignore_target_hosts.
5444 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5445 being passed; this would have caused $sender_host_name to have been used if
5446 matching the list had actually called for a host name (not very likely,
5447 since this list is usually IP addresses). A host name is now passed as
5450 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5451 code that specifies a non-restarting handler (typically sigaction() in
5452 modern systems) in an attempt to fix a rare and obscure crash bug.
5454 25. Narrowed the window for a race in the daemon that could cause it to ignore
5455 SIGCHLD signals. This is not a major problem, because they are used only to
5456 wake it up if nothing else does.
5458 26. A malformed maildirsize file could cause Exim to calculate negative values
5459 for the mailbox size or file count. Odd effects could occur as a result.
5460 The maildirsize information is now recalculated if the size or filecount
5463 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5464 support for a long time. Removed HAVE_SYS_VFS_H.
5466 28. Installed the latest version of exipick from John Jetmore.
5468 29. In an address list, if the pattern was not a regular expression, an empty
5469 subject address (from a bounce message) matched only if the pattern was an
5470 empty string. Non-empty patterns were not even tested. This was the wrong
5471 because it is perfectly reasonable to use an empty address as part of a
5472 database query. An empty address is now tested by patterns that are
5473 lookups. However, all the other forms of pattern expect the subject to
5474 contain a local part and a domain, and therefore, for them, an empty
5475 address still always fails if the pattern is not itself empty.
5477 30. Exim went into a mad DNS loop when attempting to do a callout where the
5478 host was specified on an smtp transport, and looking it up yielded more
5479 than one IP address.
5481 31. Re-factored the code for checking spool and log partition space into a
5482 function that finds that data and another that does the check. The former
5483 is then used to implement four new variables: $spool_space, $log_space,
5484 $spool_inodes, and $log_inodes.
5486 32. The RFC2047 encoding function was originally intended for short strings
5487 such as real names; it was not keeping to the 75-character limit for
5488 encoded words that the RFC imposes. It now respects the limit, and
5489 generates multiple encoded words if necessary. To be on the safe side, I
5490 have increased the buffer size for the ${rfc2047: expansion operator from
5493 33. It is now permitted to omit both strings after an "if" condition; if the
5494 condition is true, the result is "true". As before, when the second string
5495 is omitted, a false condition yields an empty string. This makes it less
5496 cumbersome to write custom ACL and router conditions.
5498 34. Failure to deliver a bounce message always caused it to be frozen, even if
5499 there was an errors_to setting on the router. The errors_to setting is now
5502 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5503 canonical form (fully expanded) before being placed in
5504 $sender_host_address.
5506 36. The table in the code that translates DNS record types into text (T_A to
5507 "A" for instance) was missing entries for NS and CNAME. It is just possible
5508 that this could have caused confusion if both these types were looked up
5509 for the same domain, because the text type is used as part of Exim's
5510 per-process caching. But the chance of anyone hitting this buglet seems
5513 37. The dnsdb lookup has been extended in a number of ways.
5515 (1) There is a new type, "zns", which walks up the domain tree until it
5516 finds some nameserver records. It should be used with care.
5518 (2) There is a new type, "mxh", which is like "mx" except that it returns
5519 just the host names, not the priorities.
5521 (3) It is now possible to give a list of domains (or IP addresses) to be
5522 looked up. The behaviour when one of the lookups defers can be
5523 controlled by a keyword.
5525 (4) It is now possible to specify the separator character for use when
5526 multiple records are returned.
5528 38. The dnslists ACL condition has been extended: it is now possible to supply
5529 a list of IP addresses and/or domains to be looked up in a particular DNS
5532 39. Added log_selector=+queue_time_overall.
5534 40. When running the queue in the test harness, wait just a tad after forking a
5535 delivery process, to get repeatability of debugging output.
5537 41. Include certificate and key file names in error message when GnuTLS fails
5538 to set them up, because the GnuTLS error message doesn't include the name
5539 of the failing file when there is a problem reading it.
5541 42. Allow both -bf and -bF in the same test run.
5543 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5545 44. The "Exiscan patch" is now merged into the mainline Exim source.
5547 45. Sometimes the final signoff response after QUIT could fail to get
5548 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5549 < 0 before doing a fflush(). This bug looks as though it goes back to the
5550 introduction of TLS in release 3.20, but "sometimes" must have been rare
5551 because the tests only now provoked it.
5553 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5554 (this can affect the format of dates).
5556 47. exim_tidydb, when checking for the continued existence of a message for
5557 which it has found a message-specific retry record, was not finding
5558 messages that were in split spool directories. Consequently, it was
5559 deleting retry records that should have stayed in existence.
5561 48. Steve fixed some bugs in eximstats.
5563 49. The SPA authentication driver was not abandoning authentication and moving
5564 on to the next authenticator when an expansion was forced to fail,
5565 contradicting the general specification for all authenticators. Instead it
5566 was generating a temporary error. It now behaves as specified.
5568 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5569 (the order specifies the preference for clients). The order is now AES256,
5570 AES128, 3DES, ARCFOUR128.
5572 51. Small patch to Sieve code - explicitly set From: when generating an
5575 52. Exim crashed if a remote delivery caused a very long error message to be
5576 recorded - for instance if somebody sent an entire SpamAssassin report back
5577 as a large number of 550 error lines. This bug was coincidentally fixed by
5578 increasing the size of one of Exim's internal buffers (big_buffer) that
5579 happened as part of the Exiscan merge. However, to be on the safe side, I
5580 have made the code more robust (and fixed the comments that describe what
5583 53. Now that there can be additional text after "Completed" in log lines (if
5584 the queue_time_overall log selector is set), a one-byte patch to exigrep
5585 was needed to allow it to recognize "Completed" as not the last thing in
5588 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5589 patch that reportedly fixes this has been added. I am not expert enough to
5590 create a test for it. This is what the patch creator wrote:
5592 "I found a little strange behaviour of ldap code when working with
5593 Windows 2003 AD Domain, where users was placed in more than one
5594 Organization Units. When I tried to give exim partial DN, the exit code
5595 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5596 But simultaneously result of request was absolutely normal ldap result,
5597 so I produce this patch..."
5599 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5600 REFERENCE, so I have modified the code to exclude the patch when that macro
5603 55. Some experimental protocols are using DNS PTR records for new purposes. The
5604 keys for these records are domain names, not reversed IP addresses. The
5605 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5606 leaves it alone. Component reversal etc. now happens only for IP addresses.
5609 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5611 57. Double the size of the debug message buffer (to 2048) so that more of very
5612 long debug lines gets shown.
5614 58. The exicyclog utility now does better if the number of log files to keep
5615 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5617 59. Two changes related to the smtp_active_hostname option:
5619 (1) $smtp_active_hostname is now available as a variable.
5620 (2) The default for smtp_banner uses $smtp_active_hostname instead
5621 of $primary_hostname.
5623 60. The host_aton() function is supposed to be passed a string that is known
5624 to be a valid IP address. However, in the case of IPv6 addresses, it was
5625 not checking this. This is a hostage to fortune. Exim now panics and dies
5626 if the condition is not met. A case was found where this could be provoked
5627 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5628 components; fortuitously, this particular loophole had already been fixed
5629 by change 4.50/55 above.
5631 If there are any other similar loopholes, the new check in host_aton()
5632 itself should stop them being exploited. The report I received stated that
5633 data on the command line could provoke the exploit when Exim was running as
5634 exim, but did not say which command line option was involved. All I could
5635 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5636 running as the user.
5639 61. There was a buffer overflow vulnerability in the SPA authentication code
5640 (which came originally from the Samba project). I have added a test to the
5641 spa_base64_to_bits() function which I hope fixes it.
5644 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5645 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5647 63. The daemon start-up calls getloadavg() while still root for those OS that
5648 need the first call to be done as root, but it missed one case: when
5649 deliver_queue_load_max is set with deliver_drop_privilege. This is
5650 necessary for the benefit of the queue runner, because there is no re-exec
5651 when deliver_drop_privilege is set.
5653 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5654 This has been fixed.
5656 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5657 was in use, was not putting the data itself into the right store pool;
5658 consequently, it could be overwritten for a subsequent message in the same
5659 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5662 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5664 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5665 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5666 the function were treating the return as a boolean value, which happened to
5667 work because 0=false and not-0=true, but is not correct code.
5669 68. The host_aton() function was not handling scoped IPv6 addresses (those
5670 with, for example, "%eth0" on the end) correctly.
5672 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5673 negated items (that is, ~something) in unsigned ints. Some compilers
5674 apparently mutter when there is no cast.
5676 70. If an address verification called from an ACL failed, and did not produce a
5677 user-specific message (i.e. there was only a "system" message), nothing was
5678 put in $acl_verify_message. In this situation, it now puts the system
5681 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5682 session; change 4.31/43 added the unwanted input to the log line - except
5683 that it did not do this in the start of session case. It now does.
5685 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5686 This is wrong and can cause the other end to generate a synchronization
5687 error if it is another Exim or anything else that does the synchronization
5688 check. A QUIT command is no longer sent after a timeout.
5690 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5691 during host lookups.
5693 74. The defer_ok option of callout verification was not working if it was used
5694 when verifying addresses in header lines, that is, for this case:
5696 verify = header_sender/callout=defer_ok
5698 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5699 those file descriptors could be used for SMTP connections. If anything
5700 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5701 could be sent to the SMTP client, causing chaos. The daemon now opens
5702 stdin, stdout, and stderr to /dev/null when it puts itself into the
5705 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5706 log by default. The user can override this with suitable Perl magic.
5708 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5709 the log message when discard triggers, was not working for the DATA ACL or
5710 for the non-SMTP ACL.
5712 78. Error message wording change in sieve.c.
5714 79. If smtp_accept_max_per_host was set, the number of connections could be
5715 restricted to fewer than expected, because the daemon was trying to set up
5716 a new connection before checking whether the processes handling previous
5717 connections had finished. The check for completed processes is now done
5718 earlier. On busy systems, this bug wouldn't be noticed because something
5719 else would have woken the daemon, and it would have reaped the completed
5722 80. If a message was submitted locally by a user whose login name contained one
5723 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5724 It caused a spool format error. I have fixed the spool reading code. A
5725 related problem was that the "from" clause in the Received: line became
5726 illegal because of the space(s). It is now covered by ${quote_local_part.
5728 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5731 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5733 83. Patch from Sieve maintainer to fix unterminated string problem in
5734 "vacation" handling.
5736 84. Some minor changes to the Linux configuration files to help with other
5737 OS variants using glibc.
5739 85. One more patch for Sieve to update vacation handling to latest spec.
5742 ----------------------------------------------------
5743 See the note above about the 4.44 and 4.50 releases.
5744 ----------------------------------------------------
5750 1. Change 4.43/35 introduced a bug that caused file counts to be
5751 incorrectly computed when quota_filecount was set in an appendfile
5754 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5755 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5758 3. Give more explanation in the error message when the command for a transport
5759 filter fails to execute.
5761 4. There are several places where Exim runs a non-Exim command in a
5762 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5763 was being done only for the command run by the queryprogram router. It is
5764 now done for all such subprocesses. The other cases are: ${run, transport
5765 filters, and the commands run by the lmtp and pipe transports.
5767 5. Some older OS have a limit of 256 on the maximum number of file
5768 descriptors. Exim was using setrlimit() to set 1000 as a large value
5769 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5770 systems. I've change it so that if it can't get 1000, it tries for 256.
5772 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5773 was an oversight, and furthermore, ever since the addition of extra
5774 controls (e.g. 4.43/32), the checks on when to allow different forms of
5775 "control" were broken. There should now be diagnostics for all cases when a
5776 control that does not make sense is encountered.
5778 7. $recipients is now available in the predata ACL (oversight).
5780 8. Tidy the search cache before the fork to do a delivery from a message
5781 received from the command line. Otherwise the child will trigger a lookup
5782 failure and thereby defer the delivery if it tries to use (for example) a
5783 cached ldap connection that the parent has called unbind on.
5785 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5786 of $address_data from the recipient verification was clobbered by the
5787 sender verification.
5789 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5790 was its contents. (It was OK if the option was not defined at all.)
5792 11. A "Completed" log line is now written for messages that are removed from
5793 the spool by the -Mrm option.
5795 12. $host_address is now set to the target address during the checking of
5796 ignore_target_hosts.
5798 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5799 being passed; this would have caused $sender_host_name to have been used if
5800 matching the list had actually called for a host name (not very likely,
5801 since this list is usually IP addresses). A host name is now passed as
5804 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5805 code that specifies a non-restarting handler (typically sigaction() in
5806 modern systems) in an attempt to fix a rare and obscure crash bug.
5808 15. Narrowed the window for a race in the daemon that could cause it to ignore
5809 SIGCHLD signals. This is not a major problem, because they are used only to
5810 wake it up if nothing else does.
5812 16. A malformed maildirsize file could cause Exim to calculate negative values
5813 for the mailbox size or file count. Odd effects could occur as a result.
5814 The maildirsize information is now recalculated if the size or filecount
5817 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5818 support for a long time. Removed HAVE_SYS_VFS_H.
5820 18. Updated exipick to current release from John Jetmore.
5822 19. Allow an empty sender to be matched against a lookup in an address list.
5823 Previously the only cases considered were a regular expression, or an
5826 20. Exim went into a mad DNS lookup loop when doing a callout where the
5827 host was specified on the transport, if the DNS lookup yielded more than
5830 21. The RFC2047 encoding function was originally intended for short strings
5831 such as real names; it was not keeping to the 75-character limit for
5832 encoded words that the RFC imposes. It now respects the limit, and
5833 generates multiple encoded words if necessary. To be on the safe side, I
5834 have increased the buffer size for the ${rfc2047: expansion operator from
5837 22. Failure to deliver a bounce message always caused it to be frozen, even if
5838 there was an errors_to setting on the router. The errors_to setting is now
5841 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5842 canonical form (fully expanded) before being placed in
5843 $sender_host_address.
5845 24. Updated eximstats to version 1.33
5847 25. Include certificate and key file names in error message when GnuTLS fails
5848 to set them up, because the GnuTLS error message doesn't include the name
5849 of the failing file when there is a problem reading it.
5851 26. Expand error message when OpenSSL has problems setting up cert/key files.
5854 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5855 (this can affect the format of dates).
5857 28. exim_tidydb, when checking for the continued existence of a message for
5858 which it has found a message-specific retry record, was not finding
5859 messages that were in split spool directories. Consequently, it was
5860 deleting retry records that should have stayed in existence.
5862 29. eximstats updated to version 1.35
5863 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5864 1.35 - bugfix such that pie charts by volume are generated correctly
5866 30. The SPA authentication driver was not abandoning authentication and moving
5867 on to the next authenticator when an expansion was forced to fail,
5868 contradicting the general specification for all authenticators. Instead it
5869 was generating a temporary error. It now behaves as specified.
5871 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5872 (the order specifies the preference for clients). The order is now AES256,
5873 AES128, 3DES, ARCFOUR128.
5875 31. Small patch to Sieve code - explicitly set From: when generating an
5878 32. Exim crashed if a remote delivery caused a very long error message to be
5879 recorded - for instance if somebody sent an entire SpamAssassin report back
5880 as a large number of 550 error lines. This bug was coincidentally fixed by
5881 increasing the size of one of Exim's internal buffers (big_buffer) that
5882 happened as part of the Exiscan merge. However, to be on the safe side, I
5883 have made the code more robust (and fixed the comments that describe what
5886 33. Some experimental protocols are using DNS PTR records for new purposes. The
5887 keys for these records are domain names, not reversed IP addresses. The
5888 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5889 leaves it alone. Component reversal etc. now happens only for IP addresses.
5892 34. The host_aton() function is supposed to be passed a string that is known
5893 to be a valid IP address. However, in the case of IPv6 addresses, it was
5894 not checking this. This is a hostage to fortune. Exim now panics and dies
5895 if the condition is not met. A case was found where this could be provoked
5896 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5897 components; fortuitously, this particular loophole had already been fixed
5898 by change 4.50/55 or 4.44/33 above.
5900 If there are any other similar loopholes, the new check in host_aton()
5901 itself should stop them being exploited. The report I received stated that
5902 data on the command line could provoke the exploit when Exim was running as
5903 exim, but did not say which command line option was involved. All I could
5904 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5905 running as the user.
5908 35. There was a buffer overflow vulnerability in the SPA authentication code
5909 (which came originally from the Samba project). I have added a test to the
5910 spa_base64_to_bits() function which I hope fixes it.
5913 36. The daemon start-up calls getloadavg() while still root for those OS that
5914 need the first call to be done as root, but it missed one case: when
5915 deliver_queue_load_max is set with deliver_drop_privilege. This is
5916 necessary for the benefit of the queue runner, because there is no re-exec
5917 when deliver_drop_privilege is set.
5919 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5920 was in use, was not putting the data itself into the right store pool;
5921 consequently, it could be overwritten for a subsequent message in the same
5922 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5925 38. Sometimes the final signoff response after QUIT could fail to get
5926 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5927 < 0 before doing a fflush(). This bug looks as though it goes back to the
5928 introduction of TLS in release 3.20, but "sometimes" must have been rare
5929 because the tests only now provoked it.
5935 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5936 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5937 Now it returns an integer. A number of calls were still expecting a T/F
5938 return. Fortuitously, in all cases, the tests worked in OK situations,
5939 which is the norm. However, things would have gone wrong on any write
5940 failures on the smtp file descriptor. This function is used when sending
5941 messages over SMTP and also when doing verify callouts.
5943 2. When Exim is called to do synchronous delivery of a locally submitted
5944 message (the -odf or -odi options), it no longer closes stderr before doing
5947 3. Implemented the mua_wrapper option.
5949 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5951 5. Implemented the functions header_remove(), header_testname(),
5952 header_add_at_position(), and receive_remove_recipient(), and exported them
5955 6. If an ACL "warn" statement specified the addition of headers, Exim already
5956 inserted X-ACL-Warn: at the start if there was no header name. However, it
5957 was not making this test for the second and subsequent header lines if
5958 there were newlines in the string. This meant that an invalid header could
5959 be inserted if Exim was badly configured.
5961 7. Allow an ACL "warn" statement to add header lines at the start or after all
5962 the Received: headers, as well as at the end.
5964 8. Added the rcpt_4xx retry error code.
5966 9. Added postmaster_mailfrom=xxx to callout verification option.
5968 10. Added mailfrom=xxxx to the callout verification option, for verify=
5971 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5972 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5974 12. Inserted some casts to stop certain compilers warning when using pointer
5975 differences as field lengths or precisions in printf-type calls (mostly
5976 affecting debugging statements).
5978 13. Added optional readline() support for -be (dynamically loaded).
5980 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5981 same clock tick as a message's arrival, so that its received time was the
5982 same as the "first fail" time on the retry record, and that message
5983 remained on the queue past the ultimate address timeout, every queue runner
5984 would try a delivery (because it was past the ultimate address timeout) but
5985 after another failure, the ultimate address timeout, which should have then
5986 bounced the address, did not kick in. This was a "< instead of <=" error;
5987 in most cases the first failure would have been in the next clock tick
5988 after the received time, and all would be well.
5990 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5991 being recognized when the domain list was tested by the match_domain
5992 condition in an expansion string.
5994 16. Added the ${str2b64: operator.
5996 17. Exim was always calling setrlimit() to set a large limit for the number of
5997 processes, without checking whether the existing limit was already
5998 adequate. (It did check for the limit on file descriptors.) Furthermore,
5999 errors from getrlimit() and setrlimit() were being ignored. Now they are
6000 logged to the main and panic logs, but Exim does carry on, to try to do its
6001 job under whatever limits there are.
6003 18. Imported PCRE 5.0.
6005 19. Trivial typo in log message " temporarily refused connection" (the leading
6008 20. If the log selector return_path_on_delivery was set and an address was
6009 redirected to /dev/null, the delivery process crashed because it assumed
6010 that a return path would always be set for a "successful" delivery. In this
6011 case, the whole delivery is bypassed as an optimization, and therefore no
6014 21. Internal re-arrangement: the function for sending a challenge and reading
6015 a response while authentication was assuming a zero-terminated challenge
6016 string. It's now changed to take a pointer and a length, to allow for
6017 binary data in such strings.
6019 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6021 23. Exim was not respecting finduser_retries when seeking the login of the
6022 uid under which it was called; it was always trying 10 times. (The default
6023 setting of finduser_retries is zero.) Also, it was sleeping after the final
6024 failure, which is pointless.
6026 24. Implemented tls_on_connect_ports.
6028 25. Implemented acl_smtp_predata.
6030 26. If the domain in control=submission is set empty, Exim assumes that the
6031 authenticated id is a complete email address when it generates From: or
6032 Sender: header lines.
6034 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6035 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6036 chown and chgrp in /bin and hostname in /usr/bin.
6038 28. Exim was keeping the "process log" file open after each use, just as it
6039 does for the main log. This opens the possibility of it remaining open for
6040 long periods when the USR1 signal hits a daemon. Occasional processlog
6041 errors were reported, that could have been caused by this. Anyway, it seems
6042 much more sensible not to leave this file open at all, so that is what now
6045 29. The long-running daemon process does not normally write to the log once it
6046 has entered its main loop, and it closes the log before doing so. This is
6047 so that log files can straightforwardly be renamed and moved. However,
6048 there are a couple of unusual error situations where the daemon does write
6049 log entries, and I had neglected to close the log afterwards.
6051 30. The text of an SMTP error response that was received during a remote
6052 delivery was being truncated at 512 bytes. This is too short for some of
6053 the long messages that one sometimes sees. I've increased the limit to
6056 31. It is now possible to make retry rules that apply only when a message has a
6057 specific sender, in particular, an empty sender.
6059 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6060 it possible to be selective about when SMTP synchronization is enforced.
6062 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6064 32. Implemented hosts_connection_nolog.
6066 33. Added an ACL for QUIT.
6068 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6071 35. Added mailbox_size and mailbox_filecount to appendfile.
6073 36. Added control = no_multiline_responses to ACLs.
6075 37. There was a bug in the logic of the code that waits for the clock to tick
6076 in the case where the clock went backwards by a substantial amount such
6077 that the microsecond fraction of "now" was more than the microsecond
6078 fraction of "then" (but the whole seconds number was less).
6080 38. Added support for the libradius Radius client library this is found on
6081 FreeBSD (previously only the radiusclient library was supported).
6087 1. When certain lookups returned multiple values in the form name=value, the
6088 quoting of the values was not always being done properly. Specifically:
6089 (a) If the value started with a double quote, but contained no whitespace,
6091 (b) If the value contained whitespace other than a space character (i.e.
6092 tabs or newlines or carriage returns) it was not quoted.
6093 This fix has been applied to the mysql and pgsql lookups by writing a
6094 separate quoting function and calling it from the lookup code. The fix
6095 should probably also be applied to nisplus, ibase and oracle lookups, but
6096 since I cannot test any of those, I have not disturbed their existing code.
6098 2. A hit in the callout cache for a specific address caused a log line with no
6099 reason for rejecting RCPT. Now it says "Previous (cached) callout
6100 verification failure".
6102 3. There was an off-by-one bug in the queryprogram router. An over-long
6103 return line was truncated at 256 instead of 255 characters, thereby
6104 overflowing its buffer with the terminating zero. As well as fixing this, I
6105 have increased the buffer size to 1024 (and made a note to document this).
6107 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6108 when Exim is waiting for an SMTP response from a remote server, Exim
6109 restarts its select() call on the socket, thereby resetting its timeout.
6110 This is not a problem when such interrupts are rare. Somebody set up a cron
6111 job to run exiwhat every 2 minutes, which is less than the normal select()
6112 timeout (5 or 10 minutes). This meant that the select() timeout never
6113 kicked in because it was always reset. I have fixed this by comparing the
6114 time when an interrupt arrives with the time at the start of the first call
6115 to select(). If more time than the timeout has elapsed, the interrupt is
6116 treated as a timeout.
6118 5. Some internal re-factoring in preparation for the addition of Sieve
6119 extensions (by MH). In particular, the "personal" test is moved to a
6120 separate function, and given an option for scanning Cc: and Bcc: (which is
6121 not set for Exim filters).
6123 6. When Exim created an email address using the login of the caller as the
6124 local part (e.g. when creating a From: or Sender: header line), it was not
6125 quoting the local part when it contained special characters such as @.
6127 7. Installed new OpenBSD configuration files.
6129 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6130 try to make them clearer.
6132 9. Callout options, other than the timeout value, were being ignored when
6133 verifying sender addresses in header lines. For example, when using
6135 verify = header_sender/callout=no_cache
6137 the cache was (incorrectly) being used.
6139 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6140 only the Cygwin environment.
6142 11. When return_path_on_delivery was set as a log selector, if different remote
6143 addresses in the same message used different return paths and parallel
6144 remote delivery occurred, the wrong values would sometimes be logged.
6145 (Whenever a remote delivery process finished, the return path value from
6146 the most recently started remote delivery process was logged.)
6148 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6149 lines when AUTH and/or TLS are in use. This is the "received protocol"
6150 field. Exim used to use "asmtp" for authenticated SMTP, without any
6151 indication (in the protocol name) for TLS use. Now it follows the RFC and
6152 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6153 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6154 names appear in log lines as well as in Received: header lines.
6156 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6157 extensions, and comparison tests, and to fix some bugs.
6159 14. Changes to the "personal" filter test:
6161 (1) The test was buggy in that it was just doing the equivalent of
6162 "contains" tests on header lines. For example, if a user's address was
6163 anne@some.where, the "personal" test would incorrectly be true for
6165 To: susanne@some.where
6167 This test is now done by extracting each address from the header in turn,
6168 and checking the entire address. Other tests that are part of "personal"
6169 are now done using regular expressions (for example, to check local parts
6170 of addresses in From: header lines).
6172 (2) The list of non-personal local parts in From: addresses has been
6173 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6174 taken from the Sieve specification recommendations.
6176 (3) If the message contains any header line starting with "List-" it is
6177 treated as non-personal.
6179 (4) The test for "circular" in the Subject: header line has been removed
6180 because it now seems ill-conceived.
6182 15. Minor typos in src/EDITME comments corrected.
6184 16. Installed latest exipick from John Jetmore.
6186 17. If headers_add on a router specified a text string that was too long for
6187 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6188 of string_sprintf() is now avoided.
6190 18. $message_body_size was not set (it was always zero) when running the DATA
6191 ACL and the local_scan() function.
6193 19. For the "mail" command in an Exim filter, no default was being set for
6194 the once_repeat time, causing a random time value to be used if "once" was
6195 specified. (If the value happened to be <= 0, no repeat happened.) The
6196 default is now 0s, meaning "never repeat". The "vacation" command was OK
6197 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6198 (I found it when inspecting the code).
6200 20. There is now an overall timeout for performing a callout verification. It
6201 defaults to 4 times the callout timeout, which applies to individual SMTP
6202 commands during the callout. The overall timeout applies when there is more
6203 than one host that can be tried. The timeout is checked before trying the
6204 next host. This prevents very long delays if there are a large number of
6205 hosts and all are timing out (e.g. when the network connections are timing
6206 out). The value of the overall timeout can be changed by specifying an
6207 additional sub-option for "callout", called "maxwait". For example:
6209 verify = sender/callout=5s,maxwait=20s
6211 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6212 to the end before writing, but this should make it even safer).
6214 22. Exim was forgetting that it had advertised PIPELINING for the second and
6215 subsequent messages on an SMTP connection. It was also not resetting its
6216 memory on STARTTLS and an internal HELO.
6218 23. When Exim logs an SMTP synchronization error within a session, it now
6219 records whether PIPELINING has been advertised or not.
6221 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6222 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6223 rather than long int.
6225 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6227 26. Added the never_mail option to autoreply.
6233 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6234 crash if the getsockname() call failed; this can happen if a connection is
6235 closed very soon after it is established. The problem was simply in the
6236 order in which certain operations were done, causing Exim to try to write
6237 to the SMTP stream before it had set up the file descriptor. The bug has
6238 been fixed by making things happen in the correct order.
6244 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6245 before the connection was closed, thus losing the rejection response.
6247 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6248 some early Solaris releases, but causes trouble in current releases where
6249 socklen_t is defined.
6251 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6254 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6257 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6258 that releasing the top of it at the end releases what was used for sub-
6259 expansions (unless the block got too big). However, discard this block if
6260 the first thing is a variable or header, so that we can use its block when
6261 it is dynamic (useful for very large $message_headers, for example).
6263 6. Lookups now cache *every* query, not just the most recent. A new, separate
6264 store pool is used for this. It can be recovered when all lookup caches are
6265 flushed. Lookups now release memory at the end of their result strings.
6266 This has involved some general refactoring of the lookup sources.
6268 7. Some code has been added to the store_xxx() functions to reduce the amount
6269 of flapping under certain conditions.
6271 8. log_incoming_interface used to affect only the <= reception log lines. Now
6272 it causes the local interface and port to be added to several more SMTP log
6273 lines, for example "SMTP connection from", and rejection lines.
6275 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6277 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6279 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6280 could be overwritten at the end of the current message (or the start of a
6281 new message if it was set in a HELO ACL). The value is now preserved for
6282 the duration of the SMTP connection.
6284 12. If a transport had a headers_rewrite setting, and a matching header line
6285 contained an unqualified address, that address was qualified, even if it
6286 did not match any rewriting rules. The underlying bug was that the values
6287 of the flags that permit the existence of unqualified sender and recipient
6288 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6289 non-local messages, and by -bnq for local messages) were not being
6290 preserved with the message after it was received.
6292 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6293 "next input=" as part of the text comprising the host identity instead of
6294 the correct text. The code was using the same buffer for two different
6295 strings. However, depending on which order the printing function evaluated
6296 its arguments, the bug did not always show up. Under Linux, for example, my
6297 test suite worked just fine.
6299 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6300 doesn't work with some older versions of Perl. It has been changed to "my",
6301 which in any case is probably the better facility to use.
6303 15. A really picky compiler found some instances of statements for creating
6304 error messages that either had too many or two few arguments for the format
6307 16. The size of the buffer for calls to the DNS resolver has been increased
6308 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6309 for addresses that have a lot of PTR records. This alleviates a problem; it
6310 does not fully solve it.
6312 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6313 buffer now truncates the list and logs the incident, which is the same
6314 action as happens when Exim is looking up a host name and its aliases.
6315 Previously in this situation something unpredictable would happen;
6316 sometimes it was "internal error: store_reset failed".
6318 18. If a server dropped the connection unexpectedly when an Exim client was
6319 using GnuTLS and trying to read a response, the client delivery process
6320 crashed while trying to generate an error log message.
6322 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6323 string, for example:
6325 warn message = H1: something\nH2: something
6327 the text was added as a single header line from Exim's point of view
6328 though it ended up OK in the delivered message. However, searching for the
6329 second and subsequent header lines using $h_h2: did not work. This has been
6330 fixed. Similarly, if a system filter added multiple headers in this way,
6331 the routers could not see them.
6333 20. Expanded the error message when iplsearch is called with an invalid key to
6334 suggest using net-iplsearch in a host list.
6336 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6337 ACLs are no longer actually imposed (and a message to that effect is
6340 22. If a "gecos" field in a passwd entry contained escaped characters, in
6341 particular, if it contained a \" sequence, Exim got it wrong when building
6342 a From: or a Sender: header from that name. A second bug also caused
6343 incorrect handling when an unquoted " was present following a character
6344 that needed quoting.
6346 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6347 was not being matched caselessly.
6349 24. Arranged for all hyphens in the exim.8 source to be escaped with
6352 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6353 properly. Recipient callout cache records were still being keyed to include
6354 the sender, even when use_sender was set false. This led to far more
6355 callouts that were necessary. The sender is no longer included in the key
6356 when use_sender is false.
6358 26. Added "control = submission" modifier to ACLs.
6360 27. Added the ${base62d: operator to decode base 62 numbers.
6362 28. dnsdb lookups can now access SRV records.
6364 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6365 the configuration file.
6367 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6368 (-v) mode. This makes the output for a verbose queue run more intelligible.
6370 31. Added a use_postmaster feature to recipient callouts.
6372 32. Added the $body_zerocount variable, containing the number of binary zero
6373 bytes in the message body.
6375 33. The time of last modification of the "new" subdirectory is now used as the
6376 "mailbox time last read" when there is a quota error for a maildir
6379 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6381 35. Added +ignore_unknown as a special item in host lists.
6383 36. Code for decoding IPv6 addresses in host lists is now included, even if
6384 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6385 address was recognized as an IP address, but was then not correctly decoded
6386 into binary, causing unexpected and incorrect effects when compared with
6393 1. Very minor rewording of debugging text in manualroute to say "list of
6394 hosts" instead of "hostlist".
6396 2. If verify=header_syntax was set, and a header line with an unqualified
6397 address (no domain) and a large number of spaces between the end of the
6398 name and the colon was received, the reception process suffered a buffer
6399 overflow, and (when I tested it) crashed. This was caused by some obsolete
6400 code that should have been removed. The fix is to remove it!
6402 3. When running in the test harness, delay a bit after writing a bounce
6403 message to get a bit more predictability in the log output.
6405 4. Added a call to search_tidyup() just before forking a reception process. In
6406 theory, someone could use a lookup in the expansion of smtp_accept_max_
6407 per_host which, without the tidyup, could leave open a database connection.
6409 5. Added the variables $recipient_data and $sender_data which get set from a
6410 lookup success in an ACL "recipients" or "senders" condition, or a router
6411 "senders" option, similar to $domain_data and $local_part_data.
6413 6. Moved the writing of debug_print from before to after the "senders" test
6416 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6417 problems for message scanning, either using a data ACL, or using
6418 local_scan() because the Received: header was not generated till after they
6419 were called (in order to set the time as the time of reception completion).
6420 I have revised the way this works. The header is now generated after the
6421 body is received, but before the ACL or local_scan() are called. After they
6422 are run, the timestamp in the header is updated.
6428 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6429 before starting a queue runner without re-exec. This happened only when
6430 deliver_drop_privilege was set or when the Exim user was set to root. The
6431 effect of the bug was that timeouts during subsequent deliveries caused
6432 crashes instead of being properly handled. The handler is now left at its
6433 default (and expected) setting.
6435 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6436 message, again when deliver_drop_privilege is set or Exim is run as root.
6437 The bug described in (1) was not present in this case, but the tidying up
6438 of the other signals was missing. I have made the two cases consistent.
6440 3. The ignore_target_hosts setting on a manualroute router was being ignored
6441 for hosts that were looked up using the /MX notation.
6443 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6446 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6447 operated on the sender address. After changing the $sender_address to <>
6448 for the sender address verify, Exim was re-instated it as the original
6449 (before rewriting) address, but remembering that it had rewritten it, so it
6450 wasn't rewriting it again. This bug also had the effect of breaking the
6451 sender address verification caching when the sender address was rewritten.
6453 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6454 This has been changed so that if the ip literal address matches
6455 ignore_target_hosts, the router declines.
6457 7. Added expansion conditions match_domain, match_address, and match_local_
6458 part (NOT match_host).
6460 8. The placeholder for the Received: header didn't have a length field set.
6462 9. Added code to Exim itself and to exim_lock to test for a specific race
6463 condition that could lead to file corruption when using MBX delivery. The
6464 issue is with the lockfile that is created in /tmp. If this file is removed
6465 after a process has opened it but before that process has acquired a lock,
6466 there is the potential for a second process to recreate the file and also
6467 acquire a lock. This could lead to two Exim processes writing to the file
6468 at the same time. The added code performs the same test as UW imapd; it
6469 checks after acquiring the lock that its file descriptor still refers to
6470 the same named file.
6472 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6473 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6476 11. Added the smtp_active_hostname option. If used, this will typically be made
6477 to depend on the incoming interface address. Because $interface_address is
6478 not set up until the daemon has forked a reception process, error responses
6479 that can happen earlier (such as "too many connections") no longer contain
6482 12. If an expansion in a condition on a "warn" statement fails because a lookup
6483 defers, the "warn" statement is abandoned, and the next ACL statement is
6484 processed. Previously this caused the whole ACL to be aborted.
6486 13. Added the iplsearch lookup type.
6488 14. Added ident_timeout as a log selector.
6490 15. Added tls_certificate_verified as a log selector.
6492 16. Added a global option tls_require_ciphers (compare the smtp transport
6493 option of the same name). This controls incoming TLS connections.
6495 17. I finally figured out how to make tls_require_ciphers do a similar thing
6496 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6497 before starting the TLS session.
6499 18. Tabs are now shown as \t in -bP output.
6501 19. If the log selector return_path_on_delivery was set, Exim crashed when
6502 bouncing a message because it had too many Received: header lines.
6504 20. If two routers both had headers_remove settings, and the first one included
6505 a superfluous trailing colon, the final name in the first list and the
6506 first name in the second list were incorrectly joined into one item (with a
6507 colon in the middle).
6513 1. Added -C and -D options to the exinext utility, mainly to make it easier
6514 to include in the automated testing, but these could be helpful when
6515 multiple configurations are in use.
6517 2. The exinext utility was not formatting the output nicely when there was
6518 an alternate port involved in the retry record key, nor when there was a
6519 message id as well (for retries that were specific to a specific message
6520 and a specific host). It was also confused by IPv6 addresses, because of
6521 the additional colons they contain. I have fixed the IPv4 problem, and
6522 patched it up to do a reasonable job for IPv6.
6524 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6525 delivery, the log line now contains "pipelined" if PIPELINING was used.
6527 4. An SMTP transport process used to panic and die if the bind() call to set
6528 an explicit outgoing interface failed. This has been changed; it is now
6529 treated in the same way as a connect() failure.
6531 5. A reference to $sender_host_name in the part of a conditional expansion
6532 that was being skipped was still causing a DNS lookup. This no longer
6535 6. The def: expansion condition was not recognizing references to header lines
6536 that used bh_ and bheader_.
6538 7. Added the _cache feature to named lists.
6540 8. The code for checking quota_filecount in the appendfile transport was
6541 allowing one more file than it should have been.
6543 9. For compatibility with Sendmail, the command line option
6551 and sets the incoming protocol and host name (for trusted callers). The
6552 host name and its colon can be omitted when only the protocol is to be set.
6553 Note the Exim already has two private options, -pd and -ps, that refer to
6554 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6555 "s", but I don't think that's a major issue.
6557 10. A number of refactoring changes to the code, none of which should affect
6560 (a) The number of logging options was getting close to filling up the
6561 32-bit word that was used as a bit map. I have split them into two classes:
6562 those that are passed in the argument to log_write(), and those that are
6563 only ever tested independently outside of that function. These are now in
6564 separate 32-bit words, so there is plenty of room for expansion again.
6565 There is no change in the user interface or the logging behaviour.
6567 (b) When building, for example, log lines, the code previously used a
6568 macro that called string_cat() twice, in order to add two strings. This is
6569 not really sufficiently general. Furthermore, there was one instance where
6570 it was actually wrong because one of the argument was used twice, and in
6571 one call a function was used. (As it happened, calling the function twice
6572 did not affect the overall behaviour.) The macro has been replaced by a
6573 function that can join an arbitrary number of extra strings onto a growing
6576 (c) The code for expansion conditions now uses a table and a binary chop
6577 instead of a serial search (which was left over from when there were very
6578 few conditions). Also, it now recognizes conditions like "pam" even when
6579 the relevant support is not compiled in: a suitably worded error message is
6580 given if an attempt is made to use such a condition.
6582 11. Added ${time_interval:xxxxx}.
6584 12. A bug was causing one of the ddress fields not to be passed back correctly
6585 from remote delivery subprocesses. The field in question was not being
6586 subsequently used, so this caused to problems in practice.
6588 13. Added new log selectors queue_time and deliver_time.
6590 14. Might have fixed a bug in maildirsizefile handling that threw up
6591 "unexpected character" debug warnings, and recalculated the data
6592 unnecessarily. In any case, I expanded the warning message to give more
6595 15. Added the message "Restricted characters in address" to the statements in
6596 the default ACL that block characters like @ and % in local parts.
6598 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6599 Three changes have been made:
6601 (a) There was a serious bug; a negative response to MAIL caused the whole
6602 recipient domain to be cached as invalid, thereby blocking all messages
6603 to all local parts at the same domain, from all senders. This bug has
6604 been fixed. The domain is no longer cached after a negative response to
6605 MAIL if the sender used is not empty.
6607 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6610 (c) A new callout option, "use_sender" has been added for people who want
6611 the modified behaviour.
6617 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6620 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6621 indeed breaks things for older releases.
6623 3. Added additional logging to the case where there is a problem reading data
6624 from a filter that is running in a subprocess using a pipe, in order to
6625 try to track down a specific problem.
6627 4. Testing facility fudge: when running in the test harness and attempting
6628 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6629 getting "No route to host". Convert this to a timeout.
6631 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6634 6. Some OS don't have socklen_t but use size_t instead. This affects the
6635 fifth argument of getsockopt() amongst other things. This is now
6636 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6637 can be set for individual OS. I have set it for SunOS5, OSF1, and
6638 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6639 some earlier ones do not.
6641 7. Change 4.30/15 was not doing the test caselessly.
6643 8. The standard form for an IPv6 address literal was being rejected by address
6644 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6645 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6646 this, as well as the form without the "IPv6" on the front (but only when
6647 address literals are enabled, of course).
6649 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6651 10. Exim crashed if a message with an empty sender address specified by -f
6652 encountered a router with an errors_to setting. This could be provoked only
6653 by a command such as
6657 where an empty string was supplied; "<>" did not hit this bug.
6659 11. Installed PCRE release 4.5.
6661 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6662 remained set. It is now erased.
6664 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6665 times from message ids (which are base 36 rather than the normal 62).
6667 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6668 were being counted as actual protocol errors, and logged if the log
6669 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6670 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6671 and DATA following a set of rejected RCPTs do not count as protocol errors.
6672 In other words, Exim assumes they were pipelined, though this may not
6673 actually be the case. Of course, in all cases the client gets an
6674 appropriate error code.
6676 15. If a lookup fails in an ACL condition, a message about the failure may
6677 be available; it is used if testing the ACL cannot continue, because most
6678 such messages specify what the cause of the deferral is. However, some
6679 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6680 that caused an old message to be retained and used if a later statement
6681 caused a defer, replacing the real cause of the deferral.
6683 16. If an IP address had so many PTR records that the DNS lookup buffer
6684 was not large enough to hold them, Exim could crash while trying to process
6685 the truncated data. It now detects and logs this case.
6687 17. Further to 4.21/58, another change has been made: if (and only if) the
6688 first line of a message (the first header line) ends with CRLF, a bare LF
6689 in a subsequent header line has a space inserted after it, so as not to
6690 terminate the header.
6692 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6693 unnecessarily, used atoi() instead of strtol(), and didn't check the
6694 termination when getting file sizes from file names by regex.
6696 19. Completely re-implemented the support for maildirsize files, in the light
6697 of a number of problems with the previous contributed implementation
6698 (4.30/29). In particular:
6700 . If the quota is zero, the maildirsize file is maintained, but no quota is
6703 . If the maildir directory does not exist, it is created before any attempt
6704 to write a maildirsize file.
6706 . The quota value in the file is just a cache; if the quota is changed in
6707 the transport, the new value overrides.
6709 . A regular expression is available for excluding directories from the
6712 20. The autoreply transport checks the characters in options that define the
6713 message's headers; it allows continued headers, but it was checking with
6714 isspace() after an embedded newline instead of explicitly looking for a
6717 21. If all the "regular" hosts to which an address was routed had passed their
6718 expiry times, and had not reached their retry times, the address was
6719 bounced, even if fallback hosts were defined. Now Exim should go on to try
6722 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6723 equivalent code in the SMTP transport. Some hosts send humungous responses
6724 to HELO/EHLO, more than 1024 it seems.
6726 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6727 gives compiler warnings in some environments. I've now done it "properly",
6730 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6731 (because of problems with the built-in one) was declared to return uschar *
6732 instead of char *, causing compiler failure.
6734 25. Fixed a file descriptor leak when processing alias/forward files.
6736 26. Fixed a minor format string issue in dbfn.c.
6738 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6740 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6741 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6744 29. When a sender address is verified, it is cached, to save repeating the test
6745 when there is more than one recipient in a message. However, when the
6746 verification involves a callout, it is possible for different callout
6747 options to be set for different recipients. It is too complicated to keep
6748 track of this in the cache, so now Exim always runs a verification when a
6749 callout is required, relying on the callout cache for the optimization.
6750 The overhead is duplication of the address routing, but this should not be
6753 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6754 to be verified with callout=postmaster, and the main callout worked but the
6755 postmaster check failed, the verification correctly failed. However, if a
6756 subsequent RCPT command asked for sender verification *without* the
6757 postmaster check, incorrect caching caused this verification also to fail,
6760 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6761 it was not caching the DNS options (qualify_single, search_parents) that
6762 were used when the lookup failed. A subsequent lookup with different
6763 options therefore always gave the same answer, though there were cases
6764 where it should not have. (Example: a "domains = !$mx_any" option on a
6765 dnslookup router: the "domains" option is always processed without any
6766 widening, but the router might have qualify_single set.) Now Exim uses the
6767 cached value only when the same options are set.
6769 32. Added John Jetmore's "exipick" utility to the distribution.
6771 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6772 than a timeout (e.g. a certificate is required, and is not provided), an
6773 Exim server now closes the connection immediately. Previously it waited for
6774 the client to close - but if the client is SSL, it seems that they each
6775 wait for each other, leading to a delay before one of them times out.
6777 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6778 maintained 0.8.x compatibility because I don't think many are using it, and
6779 it is clearly obsolete.
6781 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6784 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6785 client certificate was expired. A simple patch fixes this, though I don't
6786 understand the full logic of why the verify callback is called multiple
6789 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6790 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6791 which causes problems with some clients (such as the Certicom SSL Plus
6792 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6793 disables the coutermeasure allowing Eudora to connect."
6795 38. Exim was not checking that a write() to a log file succeeded. This could
6796 lead to Bad Things if a log got too big, in particular if it hit a file
6797 size limit. Exim now panics and dies if it cannot write to a log file, just
6798 as it does if it cannot open a log file.
6800 39. Modified OS/Makefile-Linux so that it now contains
6802 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6804 The two -D definitions ensure that Exim is compiled with large file
6805 support, which makes it possible to handle log files that are bigger than
6808 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6809 instance) a domain was checked against a named list that involved a lookup,
6810 causing $domain_data to be set, then another domain was checked against the
6811 same list, then the first domain was re-checked, the value of $domain_data
6812 after the final check could be wrong. In particular, if the second check
6813 failed, it could be set empty. This bug probably also applied to
6816 41. The strip_trailing_dot option was not being applied to the address given
6817 with the -f command-line option.
6819 42. The code for reading a message's header from the spool was incrementing
6820 $received_count, but never initializing it. This meant that the value was
6821 incorrect (doubled) while delivering a message in the same process in which
6822 it was received. In the most common configuration of Exim, this never
6823 happens - a fresh exec is done - but it can happen when
6824 deliver_drop_privilege is set.
6826 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6827 it now includes up to 150 characters of the unexpected data in the log
6830 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6831 and building data strings. The size of both of these buffers was 10 000
6832 bytes - far larger than anybody would *ever* want, thought I. Needless to
6833 say, somebody hit the limit. I have increased the maximum line length to
6834 20 000 and the maximum data length of concatenated lines to 100 000. I have
6835 also fixed two bugs, because there was no checking on these buffers. Tsk,
6836 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6837 buffer is too small.
6839 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6840 lsearch lookups. Now it does.
6842 46. When parsing a route_list item in a manualroute router, a fixed-length
6843 buffer was used for the list of hosts. I made this 1024 bytes long,
6844 thinking that nobody would ever have a list of hosts that long. Wrong.
6845 Somebody had a whole pile of complicated expansion conditions, and the
6846 string was silently truncated, leading to an expansion error. It turns out
6847 that it is easier to change to an unlimited length (owing to other changes
6848 that have happened since this code was originally written) than to build
6849 structure for giving a limitation error. The length of the item that
6850 expands into the list of hosts is now unlimited.
6852 47. The lsearch lookup could not handle data where the length of text line was
6853 more than 4095 characters. Such lines were truncated, leading to shortened
6854 data being returned. It should now handle lines of any length.
6856 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6857 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6860 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6861 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6862 no longer happens in comment lines. A list of replacements is now placed
6863 at the head of all of the source files, except those whose only change is
6864 to replace PERL_COMMAND in the very first #! line.
6866 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6867 messages on the queue, with a bottom-up merge sort, using code contributed
6868 by Michael Haardt. This should make operations like -bp somewhat faster on
6869 large queues. It won't affect queue runners, except when queue_run_in_order
6872 51. Installed eximstats 1.31 in the distribution.
6874 52. Added support for SRV lookups to the dnslookup router.
6876 53. If an ACL referred to $message_body or $message_body_end, the value was not
6877 reset for any messages that followed in the same SMTP session.
6879 54. The store-handling optimization for building very long strings was not
6880 differentiating between the different store pools. I don't think this
6881 actually made any difference in practice, but I've tidied it.
6883 55. While running the routers to verify a sender address, $sender_address
6884 was still set to the sender address. This is wrong, because when routing to
6885 send a bounce to the sender, it would be empty. Therefore, I have changed
6886 it so that, while verifying a sender address, $sender_address is set to <>.
6887 (There is no change to what happens when verifying a recipient address.)
6889 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6890 target A or AAAA records (if not already returned) without resetting the
6891 qualify_single or search_parents options of the DNS resolver. These are
6892 inappropriate in this case because the targets of MX and SRV records must
6893 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6894 target that, when qualified, matched something in the local domain. These
6895 two options are now turned off when doing these lookups.
6897 57. It seems that at least some releases of Reiserfs (which does not have the
6898 concept of a fixed number of inodes) returns zero and not -1 for the
6899 number of available inodes. This interacted badly with check_spool_inodes,
6900 which assumed that -1 was the "no such thing" setting. What I have done is
6901 to check that the total number of inodes is greater than zero before doing
6902 the test of how many are available.
6904 58. When a "warn" ACL statement has a log_message modifier, the message is
6905 remembered, and not repeated. This is to avoid a lot of repetition when a
6906 message has many recipients that cause the same warning to be written.
6907 However, Exim was preserving the list of already written lines for an
6908 entire SMTP session, which doesn't seem right. The memory is now reset if a
6909 new message is started.
6911 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6912 debugging output unless log_rewrite was also set.
6914 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6915 of (int)(handle) when we know that handle contains (void *)(-1).
6917 61. The Exim daemon panic-logs an error return when it closes the incoming
6918 connection. However "connection reset by peer" seems to be common, and
6919 isn't really an error worthy of noting specially, so that particular error
6922 62. When Exim is trying to find all the local interfaces, it used to panic and
6923 die if the ioctl to get the interface flags failed. However, it seems that
6924 on at least one OS (Solaris 9) it is possible to have an interface that is
6925 included in the list of interfaces, but for which you get a failure error
6926 for this call. This happens when the interface is not "plumbed" into a
6927 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6928 failure of the "get flags" call assumes that the interface is down.
6930 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6931 makes life easier for people who are doing arithmetic on fields extracted
6932 from dates, where you often get leading zeros that should not be
6933 interpreted as octal.
6935 64. Added qualify_domain to the redirect router, to override the global
6938 65. If a pathologically long header line contained very many addresses (the
6939 report of this problem mentioned 10 000) and each of them was rewritten,
6940 Exim could use up a very large amount of memory. (It kept on making new
6941 copies of the header line as it rewrote, and never released the old ones.)
6942 At the expense of a bit more processing, the header rewriting function has
6943 been changed so that it no longer eats memory in this way.
6945 66. The generation of the Received: header has been moved from the time that a
6946 message starts to be received, to the time that it finishes. The timestamp
6947 in the Received: header should now be very close to that of the <= log
6948 line. There are two side-effects of this change:
6950 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6951 logged header lines no longer include the local Received: line, because
6952 it has not yet been created. The same applies to a copy of the message
6953 that is returned to a non-SMTP sender when a message is rejected.
6955 (b) When a filter file is tested using -bf, no additional Received: header
6956 is added to the test message. After some thought, I decided that this
6959 This change does not affect the value of $received_for. It is still set
6960 after address rewriting, but before local_scan() is called.
6962 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6964 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6965 gave an unhelpful panic error message, and a defer error. I have managed to
6966 change this behaviour so that it now rejects any supplied certificate,
6967 which seems right, as the list of acceptable certificates is empty.
6969 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6970 gave an unhelpful defer error. I have not managed to make this reject any
6971 supplied certificates, but the error message it gives is "no certificate
6972 supplied", which is not helpful.
6974 70. exigrep's output now also includes lines that are not associated with any
6975 message, but which match the given pattern. Implemented by a patch from
6976 Martin Sluka, which also tidied up the Perl a bit.
6978 71. Recipient callout verification, like sender verification, was using <> in
6979 the MAIL FROM command. This isn't really the right thing, since the actual
6980 sender may affect whether the remote host accepts the recipient or not. I
6981 have changed it to use the actual sender in the callout; this means that
6982 the cache record is now keyed on a recipient/sender pair, not just the
6983 recipient address. There doesn't seem to be a real danger of callout loops,
6984 since a callout by the remote host to check the sender would use <>.
6985 [SEE ABOVE: changed after hitting problems.]
6987 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6988 temporary errors. However, in the case of such a code being given after
6989 the end of a data transmission (i.e. after ".") Exim was failing to write
6990 a retry record for the message. (Yes, there was some broken host that was
6991 actually sending 8xx at this point.)
6993 73. An unknown lookup type in a host list could cause Exim to panic-die when
6994 the list was checked. (An example that provoked this was putting <; in the
6995 middle of a list instead of at the start.) If this happened during a DATA
6996 ACL check, a -D file could be left lying around. This kind of configuration
6997 error no longer causes Exim to die; instead it causes a defer error. The
6998 incident is still logged to the main and panic logs.
7000 74. Buglet left over from Exim 3 conversion. The message "too many messages
7001 in one connection" was written to the rejectlog but not the mainlog, except
7002 when address rewriting (yes!) was being logged.
7004 75. Added write_rejectlog option.
7006 76. When a system filter was run not as root (that is, when system_filter_user
7007 was set), the values of the $n variables were not being returned to the
7008 main process; thus, they were not subsequently available in the $sn
7011 77. Added +return_path_on_delivery log selector.
7013 78. A connection timeout was being treated differently from recipients deferred
7014 when testing hosts_max_try with a message that was older than the host's
7015 retry timeout. (The host should not be counted, thus allowing all hosts to
7016 be tried at least once before bouncing.) This may have been the cause of an
7017 occasionally reported bug whereby a message would remain on the queue
7018 longer than the retry timeout, but would be bounced if a delivery was
7019 forced. I say "may" because I never totally pinned down the problem;
7020 setting up timeout/retry tests is difficult. See also the next item.
7022 79. The ultimate address timeout was not being applied to errors that involved
7023 a combination of host plus message (for example, a timeout on a MAIL
7024 command). When an address resolved to a number of possible hosts, and they
7025 were not all tried for each delivery (e.g. because of hosts_max_try), a
7026 message could remain on the queue longer than the retry timeout.
7028 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7031 81. Fixed an obscure SMTP outgoing bug which required at least the following
7032 conditions: (a) there was another message waiting for the same server;
7033 (b) the server returned 5xx to all RCPT commands in the first message so
7034 that the message was not completed; (c) the server dropped the connection
7035 or gave a negative response to the RSET that Exim sends to abort the
7036 transaction. The observed case was a dropped connection after DATA that had
7037 been sent in pipelining mode. That is, the server had advertised PIPELINING
7038 but was not implementing it correctly. The effect of the bug was incorrect
7039 behaviour, such as trying another host, and this could lead to a crash.
7045 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7046 and daemon.c were passed as pointers to ints; they should have been
7047 pointers to socklen_t variables (which are typically unsigned ints).
7049 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7052 3. Fixed a really odd bug that affected only the testing scheme; patching a
7053 certain fixed string in the binary changed the value of another string that
7054 happened to be identical to the end of the original first string.
7056 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7057 name", it returns that address as the IP address. On some operating
7058 systems (e.g. Solaris), it also passes back the IP address string as the
7059 "host name". However, on others (e.g. Linux), it passes back an empty
7060 string. Exim wasn't checking for this, and was changing the host name to an
7061 empty string, assuming it had been canonicalized.
7063 5. Although rare, it is permitted to have more than one PTR record for a given
7064 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7065 all the names associated with an address, because they do in Solaris.
7066 However, it seems that they do not in Linux for data that comes from the
7067 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7068 I found this out when I moved to a new Linux workstation and tried to run
7069 the Exim test suite.
7071 To get round this problem I have changed the code so that it now does its
7072 own call to the DNS to look up PTR records when searching for a host name.
7073 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7074 addresses that are only in /etc/hosts are still found.
7076 This behaviour is, however, controlled by an option called host_lookup_
7077 order, which defaults to "bydns:byaddr". If people want to use the other
7078 order, or indeed, just use one or the other means of lookup, they can
7079 specify it in this variable.
7081 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7082 some operating systems, this comes back from gethostbyaddr() as an empty
7083 string, and this is what Exim used to test for. However, it seems that in
7084 other systems, "." is yielded. Exim now tests for this case too.
7086 7. The values of check_spool_space and check_log_space are now held internally
7087 as a number of kilobytes instead of an absolute number of bytes. If a
7088 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7089 kilobyte. This means that much larger values can be stored.
7091 8. Exim monitor: an attempt to get the action menu when not actually pointing
7092 at a message produces an empty menu entitled "No message selected". This
7093 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7094 no entries in it ("Shell widget menu has zero width and/or height"). So I
7095 have added a single, blank menu entry in this case.
7097 9. Added ${quote_local_part.
7099 10. MIME decoding is now applied to the contents of Subject: header lines when
7102 11. Now that a reference to $sender_host_address automatically causes a reverse
7103 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7104 host lookup before query-style lookups in lists that might use this
7105 variable. This has therefore been abolished, and the "net-" prefix is no
7106 longer necessary for query-style lookups.
7108 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7109 have been a typo for LFLAGS, so it has been changed.
7111 13. The install script calls Exim with "-C /dev/null" in order to find the
7112 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7113 to be output. However, since Exim outputs its version number before the
7114 error, it didn't break the script. It just looked ugly. I fixed this by
7115 always allowing "-C /dev/null" if the caller is root.
7117 14. Ignore overlarge ACL variable number when reading spool file - insurance
7118 against a later release with more variables having written the file.
7120 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7121 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7122 this, as well as the form without the "IPv6" on the front.
7124 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7125 OS/Makefile-Darwin file.
7127 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7128 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7130 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7131 variable. However, some OS use ints for inodes. I've added cast to long int
7132 to get rid of the compiler warning.
7134 19. I had forgotten to lock out "/../" in configuration file names when
7135 ALT_CONFIG_PREFIX was set.
7137 20. Routers used for verification do not need to specify transports. However,
7138 if such a router generated a host list, and callout was configured, Exim
7139 crashed, because it could not find a port number from the (non-existent)
7140 transport. It now assumes port 25 in this circumstance.
7142 21. Added the -t option to exigrep.
7144 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7145 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7146 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7147 them is set, LOOKUP_LSEARCH is forced.
7149 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7151 24. Added sender and host information to the "rejected by local_scan()" log
7152 line; previously there was no indication of these.
7154 25. Added .include_if_exists.
7156 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7157 Linux. It turns out that not all file systems support this. Apparently some
7158 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7159 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7160 not supported on the file descriptor, is now ignored when Exim is trying to
7161 sync a directory. This applies only to Linux.
7163 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7165 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7166 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7169 29. Added support for maildirsize files from supplied patch (modified a bit).
7171 30. The use of :fail: followed by an empty string could lead Exim to respond to
7172 sender verification failures with (e.g.):
7174 550 Verification failed for <xxx>
7175 550 Sender verify failed
7177 where the first response line was missing the '-' that indicates it is not
7178 the final line of the response.
7180 31. The loop for finding the name of the user that called Exim had a hardwired
7181 limit of 10; it now uses the value of finduser_retries, which is used for
7182 all other user lookups.
7184 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7187 33. Exim was neglecting to zero errno before one call of strtol() when
7188 expanding a string and expecting an integer value. On some systems this
7189 resulted in spurious "integer overflow" errors. Also, it was casting the
7190 result into an int without checking.
7192 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7193 did not work. The code looks as if it has *never* worked, though it appears
7194 to have been documented since at least release 1.62. I have made it work.
7196 35. The "timeout_DNS" error in retry rules, also documented since at least
7197 1.62, also never worked. As it isn't clear exactly what this means, and
7198 clearly it isn't a major issue, I have abolished the feature by treating it
7199 as "timeout", and writing a warning to the main and panic logs.
7201 36. The display of retry rules for -brt wasn't always showing the error code
7204 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7205 timeout_connect_A, timeout_connect_MX.
7207 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7208 to the empty sender.
7210 39. The daemon was not analysing the content of -oX till after it had closed
7211 stderr and disconnected from the controlling terminal. This meant that any
7212 syntax errors were only noted on the panic log, and the return code from
7213 the command was 0. By re-arranging the code a little, I've made the
7214 decoding happen first, so such errors now appear on stderr, and the return
7215 code is 1. However, the actual setting up of the sockets still happens in
7216 the disconnected process, so errors there are still only recorded on the
7219 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7220 connections (as happens on some IP stacks) was logged at start up time as
7221 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7222 it from "IPv6 and IPv4", which means that two separate sockets are being
7225 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7226 says whether AF_INET or AF_INET6 was passed as an argument.
7228 42. Exiwhat output was messed up when time zones were included in log
7231 43. Exiwhat now gives more information about the daemon's listening ports,
7232 and whether -tls-on-connect was used.
7234 44. The "port" option of the smtp transport is now expanded.
7236 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7237 silently ignored. Now an error message is written to the main and panic
7240 46. There's a new ACL modifier called "logwrite" which writes to a log file
7241 as soon as it is encountered.
7243 47. Added $local_user_uid and $local_user_gid at routing time.
7245 48. Exim crashed when trying to verify a sender address that was being
7248 49. Exim was recognizing only a space character after ".include". It now also
7249 recognizes a tab character.
7251 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7252 extracting the relevant information from the specification. The man page no
7253 longer contains scrambled data for the -d option, and I've added a section
7254 at the front about calling Exim under different names.
7256 51. Added "extra_headers" argument to the "mail" command in filter files.
7258 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7261 53. Installed eximstats 1.29.
7263 54. Added transport_filter_timeout as a generic transport option.
7265 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7266 Cc: header lines. This was required by RFC 822, but it not required by RFC
7269 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7270 incoming messages that did not have them. Now it does so only if the
7271 message originates locally, that is, if there is no associated remote host
7272 address. When Resent- header lines are present, this applies to the Resent-
7273 lines rather than the non-Resent- lines.
7275 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7276 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7278 58. Messages for configuration errors now include the name of the main
7279 configuration file - useful now that there may be more than one file in a
7280 list (.included file names were always shown).
7282 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7283 for those rare installations that do not start the daemon as root or run it
7284 setuid root. I've cut out the call to initgroups() if the daemon is not
7287 60. The Exim user and group can now be bound into the binary as text strings
7288 that are looked up at the start of Exim's processing.
7290 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7292 62. Added $mailstore_basename variable.
7294 63. Installed patch to sieve.c from Michael Haardt.
7296 64. When Exim failed to open the panic log after failing to open the main log,
7297 the original message it was trying to log was written to stderr and debug
7298 output, but if they were not available (the usual case in production), it
7299 was lost. Now it is written to syslog before the two lines that record the
7300 failures to open the logs.
7302 65. Users' Exim filters run in subprocesses under the user's uid. It is
7303 possible for a "deliver" command or an alias in a "personal" command to
7304 provoke an address rewrite. If logging of address rewriting is configured,
7305 this fails because the process is not running as root or exim. There may be
7306 a better way of dealing with this, but for the moment (because 4.30 needs
7307 to be released), I have disabled address rewrite logging when running a
7308 filter in a non-root, non-exim process.
7314 1. The buildconfig auxiliary program wasn't quoting the value set for
7315 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7316 not defined. This bug was masked in 4.22 by the effect that was fixed in
7319 2. Some messages that were rejected after a message id was allocated were
7320 shown as "incomplete" by exigrep. It no longer does this for messages that
7321 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7323 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7324 have allow_domain_literals set, the ID did not get logged in the <= line.
7325 Domain literals are now always recognized in Message-ID: header lines.
7327 4. The first argument for a ${extract expansion item is the key name or field
7328 number. Leading and trailing spaces in this item were not being ignored,
7329 causing some misleading effects.
7331 5. When deliver_drop_privilege was set, single queue runner processes started
7332 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7333 same command in the process it spins off) were not dropping privilege.
7335 6. When the daemon running as "exim" started a queue runner, it always
7336 re-executed Exim in the spun-off process. This is a waste of effort when
7337 deliver_drop_privilege is set. The new process now just calls the
7338 queue-runner function directly.
7344 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7347 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7348 ignored. Though the use of -oP was forcing the writing of a pid file, it
7349 was always written to the default place.
7351 3. If the message "no IP address found for host xxxx" is generated during
7352 incoming verification, it is now followed by identification of the incoming
7353 connection (so you can more easily find what provoked it).
7355 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7357 5. Added some features to "harden" Exim a bit more against certain attacks:
7359 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7360 be put in Local/Makefile. This is like the never_users runtime option,
7361 but it cannot be overridden. The default setting is "root".
7363 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7364 prefix string with which any file named in a -C command line option
7367 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7368 is retained for -C and -D only if the caller of Exim is root. Without
7369 it, the exim user may also use -C and -D and retain privilege.
7371 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7372 command line option is disabled.
7374 6. Macro names set by the -D option must start with an upper case letter, just
7375 like macro names defined in the configuration file.
7377 7. Added "dereference=" facility to LDAP.
7379 8. Two instances of the typo "uknown" in the source files are fixed.
7381 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7382 the Configure-Makefile script screwed up while processing it.
7384 10. Incorporated PCRE 4.4.
7386 11. The SMTP synchronization check was not operating right at the start of an
7387 SMTP session. For example, it could not catch a HELO sent before the client
7388 waited for the greeting. There is now a check for outstanding input at the
7389 point when the greeting is written. Because of the duplex, asynchronous
7390 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7391 way, but not yet received, when the check is performed.
7393 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7394 on TCP/IP sockets, because this apparently causes some broken clients to
7397 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7398 unchanged) from the Cygwin maintainer.
7400 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7401 was supported instead of testing for mbx. Effectively a typo.
7403 15. The spa authenticator server code was not checking that the input it
7404 received was valid base64.
7406 16. The debug output line for the "set" modifier in ACLs was not showing the
7407 name of the variable that was being set.
7409 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7411 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7412 lookup on its own. Something else was needed to trigger the lookup. For
7413 example, a match in host_lookup or the need for a host name in a host list.
7414 Now, if $sender_host_name is referenced and the host name has not yet been
7415 looked up, a lookup is performed. If the lookup fails, the variable remains
7416 empty, and $host_lookup_failed is set to "1".
7418 19. Added "eqi" as a case-independent comparison operator.
7420 20. The saslauthd authentication condition could segfault if neither service
7421 nor realm was specified.
7423 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7424 error message that was logged was misleading, and incoming SMTP
7425 connections were dropped. The message is now more accurate, and temporary
7426 errors are given to SMTP connections.
7428 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7429 (except RSET) with a 421 error, until QUIT is received. However, it was
7430 failing to send a response to QUIT.
7432 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7433 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7434 ACL is now run after the helo_try_verify_hosts code.
7436 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7437 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7438 case-independent, so other case variants are also recognized). Apparently
7439 some systems use these upper case variants.
7441 25. If more than two messages were waiting for the same host, and a transport
7442 filter was specified for the transport, Exim sent two messages over the
7443 same TCP/IP connection, and then failed with "socket operation on non-
7444 socket" when it tried to send the third.
7446 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7448 27. The extern definition of crypt16() in expand.c was not being excluded when
7449 the OS had its own crypt16() function.
7451 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7452 as a preferred synonym for return_size_limit, both as an option and as an
7455 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7457 30. Changed the default configuration ACL to relax the local part checking rule
7458 for addresses that are not in any local domains. For these addresses,
7459 slashes and pipe symbols are allowed within local parts, but the sequence
7460 /../ is explicitly forbidden.
7462 31. SPA server authentication was not clearing the challenge buffer before
7465 32. log_message in a "warn" ACL statement was writing to the reject log as
7466 well as to the main log, which contradicts the documentation and doesn't
7467 seem right (because no rejection is happening). So I have stopped it.
7469 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7470 However, I am unable to do any testing of this.
7472 34. Fixed an infelicity in the appendfile transport. When checking directories
7473 for a mailbox, to see if any needed to be created, it was accidentally
7474 using path names with one or more superfluous leading slashes; tracing
7475 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7477 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7478 contents are added to the log line that is written for every discarded
7479 recipient. (Previously a log_message setting was ignored.)
7481 36. The ${quote: operator now quotes the string if it is empty.
7483 37. The install script runs exim in order to find its version number. If for
7484 some reason other than non-existence or emptiness, which it checks, it
7485 could not run './exim', it was installing it with an empty version number,
7486 i.e. as "exim-". This error state is now caught, and the installation is
7489 38. An argument was missing from the function that creates an error message
7490 when Exim fails to connect to the socket for saslauthd authentication.
7491 This could cause Exim to crash, or give a corrupted message.
7493 39. Added isip, isip4, and isip6 to ${if conditions.
7495 40. The ACL variables $acl_xx are now saved with the message, and can be
7496 accessed later in routers, transports, and filters.
7498 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7499 strings in the file are not string-expanded.
7501 42. If a MAIL command specified a SIZE value that was too large to fit into an
7502 int variable, the check against message_size_limit failed. Such values are
7503 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7504 day this will have to be increased, but I don't think I want to be around
7505 when emails are that large.
7512 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7513 iconv() is not standard in FreeBSD.
7515 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7516 IPv6 enabled. The observed symptom was a segmentation fault on return from
7517 the function os_common_find_running_interfaces() in src/os.c.
7519 3. In the check_special_case() function in daemon.c I had used "errno" as an
7520 argument name, which causes warnings on some systems. This was basically a
7521 typo, since it was named "eno" in the comments!
7523 4. The code that waits for the clock to tick (at a resolution of some fraction
7524 of a second) so as to ensure message-id uniqueness was always waiting for
7525 at least one whole tick, when it could have waited for less. [This is
7526 almost certainly not relevant at current processor speeds, where it is
7527 unlikely to ever wait at all. But we try to future-proof.]
7529 5. The function that sleeps for a time interval that includes fractions of a
7530 second contained a race. It did not block SIGALRM between setting the
7531 timer, and suspending (a couple of lines later). If the interval was short
7532 and the sigsuspend() was delayed until after it had expired, the suspension
7533 never ended. On busy systems this could lead to processes getting stuck for
7536 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7537 process, before it forks any delivery processes. The open lookup caching
7538 mechanism meant that the open file or database connection was passed into
7539 the delivery process. The problem was that delivery processes always tidy
7540 up cached lookup data. This could cause a problem for the next delivery
7541 process started by the queue runner, because the external queue runner
7542 process does not know about the closure. So the next delivery process
7543 still has data in the lookup cache. In the case of a file lookup, there was
7544 no problem because closing a file descriptor in a subprocess doesn't affect
7545 the parent. However, if the lookup was caching a connection to a database,
7546 the connection was closed, and the second delivery process was likely to
7547 see errors such as "PGSQL: query failed: server closed the connection
7548 unexpectedly". The problem has been fixed by closing all cached lookups
7549 in a queue runner before running a delivery process.
7551 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7552 seem to have the "const" qualifier which it has on other OS. I've
7555 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7556 *of the same type* (client or server) with the same public name that an
7557 error should be diagnosed.
7559 9. When Exim looked up a host name for an IP address, but failed to find the
7560 original IP address when looking up the host name (a safety check), it
7561 output the message "<ip address> does not match any IP for NULL", which was
7562 confusing, to say the least. The bug was that the host name should have
7563 appeared instead of "NULL".
7565 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7566 user that is built into the binary, and the -C or -D options is used, root
7567 privilege is dropped before the configuration file is read. In addition,
7568 logging is switched to stderr instead of the normal log files. If the
7569 configuration then re-defines the Exim user, the unprivileged environment
7570 is probably not what is expected, so Exim logs a panic warning message (but
7573 However, if deliver_drop_privilege is set, the unprivileged state may well
7574 be exactly what is intended, so the warning has been cut out in that case,
7575 and Exim is allowed to try to write to its normal log files.
7581 1. smtp_return_error_details was not giving details for temporary sender
7582 or receiver verification errors.
7584 2. Diagnose a configuration error if two authenticators have the same public
7587 3. Exim used not to create the message log file for a message until the first
7588 delivery attempt. This could be confusing when incoming messages were held
7589 for policy or load reasons. The message log file is now created at the time
7590 the message is received, and an initial "Received" line is written to it.
7592 4. The automatically generated man page for command line options had a minor
7593 bug that caused no ill effects; however, a more serious problem was that
7594 the procedure for building the man page automatically didn't always
7595 operate. Consequently, release 4.20 contains an out-of-date version. This
7596 shouldn't happen again.
7598 5. When building Exim with embedded Perl support, the script that builds the
7599 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7600 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7602 6. The freeze_tell option was not being used for messages that were frozen on
7603 arrival, either by an ACL or by local_scan().
7605 7. Added the smtp_incomplete_transaction log selector.
7607 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7608 was accepting AUTH without a new EHLO.
7610 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7611 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7614 10. Logging of TCP/IP connections (when configured) now happens in the main
7615 daemon process instead of the child process, so that the TCP/IP connection
7616 count is more accurate (but it can never be perfect).
7618 11. The use of "drop" in a nested ACL was not being handled correctly in the
7619 outer ACL. Now, if condition failure induced by the nested "drop" causes
7620 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7621 or "require"), the connection is dropped.
7623 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7624 that yield "discard" can now be used with an "accept" or a "discard" verb,
7625 but an error is generated for any others (because I can't see a useful way
7626 to define what should happen).
7628 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7629 are now processed in the same way as lines in the Exim configuration file.
7630 In particular, continuation lines are supported.
7632 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7634 15. Added -ti meaning -t -i.
7636 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7637 domains, and warn by logging if others are found.
7639 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7640 returned from GIFCONF when Exim is trying to find the list of interfaces on
7641 a host. The code in os.c has been modified to copy each ifreq to an aligned
7642 structure in all cases.
7644 Also, in some cases, the returned ifreq's were being copied to a 'struct
7645 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7646 means the last couple of bytes of an IPv6 address could be chopped if the
7647 ifreq contained only a normal sockaddr (14 bytes storage).
7649 18. Named domain lists were not supported in the hosts_treat_as_local option.
7650 An entry such as +xxxx was not recognized, and was treated as a literal
7653 19. Ensure that header lines added by a DATA ACL are included in the reject log
7654 if the ACL subsequently rejects the message.
7656 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7657 MD5 (which is deprecated).
7659 21. When testing a filter file using -bf, Exim was writing a message when it
7660 took the sender from a "From " line in the message, but it was not doing so
7661 when it took $return_path from a Return-Path: header line. It now does.
7663 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7664 with a valid header line field name (a series of printing characters
7665 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7667 23. Changed "disc" in the source to "disk" to conform to the documentation and
7668 the book and for uniformity.
7670 24. Ignore Sendmail's -Ooption=value command line item.
7672 25. When execve() failed while trying to run a command in a pipe transport,
7673 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7674 could be confused with a return value of 69 from the command itself. This
7675 has been changed to 127, the value the shell returns if it is asked to run
7676 a non-existent command. The wording for the related log line suggests a
7677 non-existent command as the problem.
7679 26. If received_header_text expands to an empty string, do not add a Received:
7680 header line to the message. (Well, it adds a token one on the spool, but
7681 marks it "old" so that it doesn't get used or transmitted.)
7683 27. Installed eximstats 1.28 (addition of -nt option).
7685 28. There was no check for failure on the call to getsockname() in the daemon
7686 code. This can fail if there is a shortage of resources on the system, with
7687 ENOMEM, for example. A temporary error is now given on failure.
7689 29. Contrary to the C standard, it seems that in some environments, the
7690 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7691 program. Exim now does this explicitly; it affects the formatting of
7692 timestamps using strftime().
7694 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7695 complaints. I've now initialized all the variables, to avoid this.
7697 32. Header lines added by a system filter were not being "seen" during
7698 transport-time rewrites.
7700 33. The info_callback() function passed to OpenSSL is set up with type void
7701 (*)(SSL *, int, int), as described somewhere. However, when calling the
7702 function (actually a macro) that sets it up, the type void(*)() is
7703 expected. I've put in a cast to prevent warnings from picky compilers.
7705 34. If a DNS black list lookup found a CNAME record, but there were no A
7706 records associated with the domain it pointed at, Exim crashed.
7708 35. If a DNS black list lookup returned more than one A record, Exim ignored
7709 all but the first. It now scans all returned addresses if a particular IP
7710 value is being sought. In this situation, the contents of the
7711 $dnslist_value variable are a list of all the addresses, separated by a
7714 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7715 to accept a host name and all its aliases if the forward lookup for any of
7716 them yielded the IP address of the incoming connection. Now it accepts only
7717 those names whose forward lookup yields the correct IP address. Any other
7718 names are discarded. This closes a loophole whereby a rogue DNS
7719 administrator could create reverse DNS records to break through a
7720 wildcarded host restriction in an ACL.
7722 37. If a user filter or a system filter that ran in a subprocess used any of
7723 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7724 the wrong values were passed to the pipe command ($thisaddress had the
7725 value of $0, $0 had the value of $1, etc). This bug was introduced by
7726 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7728 38. Improved the line breaking for long SMTP error messages from ACLs.
7729 Previously, if there was no break point between 40 and 75 characters, Exim
7730 left the rest of the message alone. Two changes have been made: (a) I've
7731 reduced the minimum length to 35 characters; (b) if it can't find a break
7732 point between 35 and 75 characters, it looks ahead and uses the first one
7733 that it finds. This may give the occasional overlong line, but at least the
7734 remaining text gets split now.
7736 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7737 file descriptors might be low, and that setting 1000 would always raise it.
7738 It turns out that in some environments, the limit is already over 1000 and
7739 that lowering it causes trouble. So now Exim takes care not to decrease it.
7741 40. When delivering a message, the value of $return_path is set to $sender_
7742 address at the start of routing (routers may change the value). By an
7743 oversight, this default was not being set up when an address was tested by
7744 -bt or -bv, which affected the outcome if any router or filter referred to
7747 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7748 log only when "message" or "log_message" are set. However, if one of the
7749 conditions was an address verification, or a call to a nested ACL, the
7750 messages generated by the underlying test were being passed through. This
7751 no longer happens. The underlying message is available in $acl_verify_
7752 message for both "message" and "log_message" expansions, so it can be
7753 passed through if needed.
7755 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7756 new expansion $bh_ to give the encoded byte string without charset
7757 translation. Translation happens only if iconv() is available; HAVE_ICONV
7758 indicates this at build time. HEADERS_CHARSET gives the charset to
7759 translate to; headers_charset can change it in the configuration, and
7760 "headers charset" can change it in an individual filter file.
7762 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7763 that creates RFC 2047 encoded "words" labels them as that charset instead
7764 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7765 expansion operator; (ii) when Exim creates a From: line for a local
7766 message; (iii) when a header line is rewritten to include a "phrase" part.
7768 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7769 buggy, causing it to skip the first lines of messages whose message ID
7770 ended in 'D'. This would not have bitten before Exim release 4.14, because
7771 message IDs were unlikely to end in 'D' before then. The effect was to have
7772 incorrect size information for certain domains.
7774 45. #include "config.h" was missing at the start of the crypt16.c module. This
7775 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7778 46. If there was a timeout during a "random" callout check, Exim treated it as
7779 a failure of the random address, and carried on sending RSET and the real
7780 address. If the delay was just some slowness somewhere, the response to the
7781 original RCPT would be taken as a response to RSET and so on, causing
7782 mayhem of various kinds.
7784 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7785 when I implemented it. It didn't allow for the fact that some option values
7786 may legitimately be negative (e.g. size_addition), and it didn't even do
7787 the right test for positive values.
7789 48. Domain names in DNS records are case-independent. Exim always looks them up
7790 in lower case. Some resolvers return domain names in exactly the case they
7791 appear in the zone file, that is, they may contain uppercase letters. Not
7792 all resolvers do this - some return always lower case. Exim was treating a
7793 change of case by a resolver as a change of domain, similar to a widening
7794 of a domain abbreviation. This triggered its re-routing code and so it was
7795 trying to route what was effectively the same domain again. This normally
7796 caused routing to fail (because the router wouldn't handle the domain
7797 twice). Now Exim checks for this case specially, and just changes the
7798 casing of the domain that it ultimately uses when it transmits the message
7801 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7804 50. If a filter generated a file delivery with a non-absolute name (possible if
7805 no home directory exists for the router), the forbid_file option was not
7808 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7809 the existing equality matching.
7811 52. Exim was using ints instead of ino_t variables in some places where it was
7812 dealing with inode numbers.
7814 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7815 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7816 called TMPDIR, and if it finds it is different, it changes its value.
7818 54. The smtp_printf() function is now made available to local_scan() so
7819 additional output lines can be written before returning. There is also an
7820 smtp_fflush() function to enable the detection of a dropped connection.
7821 The variables smtp_input and smtp_batched_input are exported to
7824 55. Changed the default runtime configuration: the message "Unknown user"
7825 has been removed from the ACL, and instead placed on the localuser router,
7826 using the cannot_route_message feature. This means that any verification
7827 failures that generate their own messages won't get overridden. Similarly,
7828 the "Unrouteable address" message that was in the ACL for unverifiable
7829 relay addresses has also been removed.
7831 56. Added hosts_avoid_esmtp to the smtp transport.
7833 57. The exicyclog script was not checking for the esoteric option
7834 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7835 will work only if exicyclog is run under the appropriate euid.
7837 58. Following a discussion on the list, the rules by which Exim recognises line
7838 endings on incoming messages have been changed. The -dropcr and drop_cr
7839 options are now no-ops, retained only for backwards compatibility. The
7840 following line terminators are recognized: LF CRLF CR. However, special
7841 processing applies to CR:
7843 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7844 nor a local message in the state where . is a terminator.
7846 (ii) If a bare CR is encountered in a header line, an extra space is added
7847 after the line terminator so as not to end the header. The reasoning
7848 behind this is that bare CRs in header lines are most likely either
7849 to be mistakes, or people trying to play silly games.
7851 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7852 was being incorrectly given as 18 bytes larger than it should have been.
7853 This is a VOB (very old bug).
7855 60. This may never have affected anything current, but just in case it has:
7856 When the local host is found other than at the start of a list of hosts,
7857 the local host, those with the same MX, and any that follow, are discarded.
7858 When the list in question was part of a longer list of hosts, the following
7859 hosts (not currently being processed) were also being discarded. This no
7860 longer happens. I'm not sure if this situation could ever has previously
7863 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7866 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7867 header. This is something that is recommended in a new Internet Draft, and
7868 is something that is documented as being done by Sendmail. There are two
7869 possible values. For messages generated by the autoreply transport, Exim
7872 Auto-Submitted: auto-replied
7874 whereas for all other generated messages (e.g. bounces) it adds
7876 Auto-Submitted: auto-generated
7878 63. The "personal" condition in filters now includes a test for the
7879 Auto-Submitted: header. If it contains the string "auto-" the message it
7880 not considered personal.
7882 64. Added rcpt_include_affixes as a generic transport option.
7884 65. Added queue_only_override (default true).
7886 66. Added the syslog_duplication option.
7888 67. If what should have been the first header line of a message consisted of
7889 a space followed by a colon, Exim was mis-interpreting it as a header line.
7890 It isn't of course - it is syntactically invalid and should therefore be
7891 treated as the start of the message body. The misbehaviour could have
7892 caused a number of strange effects, including loss of data in subsequent
7893 header lines, and spool format errors.
7895 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7896 client host had authenticated. This control can now be exercised by an ACL
7897 for more flexibility.
7899 69. By default, callouts do not happen when testing with -bh. There is now a
7900 variant, -bhc, which does actually run the callout code, including
7901 consulting and updating the callout cache.
7903 70. Added support for saslauthd authentication, courtesy of Alexander
7906 71. If statvfs() failed on the spool or log directories while checking their
7907 size for availability, Exim confusingly gave the error "space shortage".
7908 Furthermore, in debugging mode it crashed with a floating point exception.
7909 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7910 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7911 serious problem, Exim now writes to the main and panic logs when this
7912 happens, with details of the failure. It then refuses to accept the
7913 incoming message, giving the message "spool directory problem" or "log
7914 directory problem" with a 421 code for SMTP messages.
7916 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7917 0, 1, and 2 exist, because some OS complain for execs without them (see
7918 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7919 descriptors. However, the code omitted to check that the open succeeded,
7920 causing mysterious errors if for some reason the permissions on /dev/null
7921 got screwed. Now Exim writes a message to the main and panic logs, and
7922 bombs out if it can't open /dev/null.
7924 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7925 interact so that it is all more flexible. It is supposed to remain
7926 backwards compatible. Also added extra_local_interfaces.
7928 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7929 to bomb out with an assertion failure - to the client this appears as a
7930 connection drop. This problem occurs in the part of the code that was taken
7931 from the Samba project. Fortunately, the assertion is in a very simple
7932 function, so I have fixed this by reproducing the function inline in the
7933 one place where it is called, and arranging for authentication to fail
7934 instead of killing the process with assert().
7936 75. The SPA client code was not working when the server requested OEM rather
7937 than Unicode encoding.
7939 76. Added code to make require_files with a specific uid setting more usable in
7940 the case where statting the file as root fails - usually a non-root-mounted
7941 NFS file system. When this happens and the failure is EACCES, Exim now
7942 forks a subprocess and does the per-uid checking as the relevant uid.
7944 77. Added process_log_path.
7946 78. If log_file_path was not explicitly set, a setting of check_log_space or
7947 check_log_inodes was ignored.
7949 79. If a space check for the spool or log partitions fails, the incident is now
7950 logged. Of course, in the latter case the data may get lost...
7952 80. Added the %p formatting code to string_format() so that it can be used to
7953 print addresses in debug_print(). Adjusted all the address printing in the
7954 debugging in store.c to use %p rather than %d.
7956 81. There was a concern that a line of code in smtp_in.c could overflow a
7957 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7958 initially expressed, the concern was not well-founded, because trailing
7959 spaces are removed early. However, if the trailing spaces were followed by
7960 a NULL, they did not get removed, so the overflow was possible. Two fixes
7963 (a) I re-wrote the offending code in a cleaner fashion.
7964 (b) If an incoming SMTP command contains a NULL character, it is rejected
7967 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7968 runs initgroups(), so that if the Exim user is in any additional groups,
7969 they will be used during message reception.
7975 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob