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
191 jh/40 The gsasl authenticator now supports caching of the salted password
192 generated by the client-side implementation. This required the addition
193 of a new variable: $auth4.
200 JH/01 Avoid costly startup code when not strictly needed. This reduces time
201 for some exim process initialisations. It does mean that the logging
202 of TLS configuration problems is only done for the daemon startup.
204 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
206 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
207 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
210 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
211 DKIM signing is being done. Previously a transport_filter would always
212 disable CHUNKING, falling back to traditional DATA.
214 JH/05 Regard command-line recipients as tainted.
216 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
218 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
219 PAM library frees one of the arguments given to it, despite the
220 documentation. Therefore a plain malloc must be used.
222 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
223 on-stack buffers were used, resulting in a taint trap when DSN information
224 copied from a received message was written into the buffer.
226 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
227 the ordering of its ARC headers. This caused a crash.
229 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
230 a new record was being constructed with information from the peer, a trap
233 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
234 installation would get error messages from DMARC verify, when it hit the
235 nonexistent file indicated by the default. Distros wanting DMARC enabled
236 should both provide the file and set the option.
237 Also enforce no DMARC verification for command-line sourced messages.
239 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
240 could, depending on the platform, hang at the STARTTLS response.
242 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
243 message on a connection. Previously if one message had ARC headers and
244 the following one did not, a crash could result when adding an
245 Authentication-Results: header.
247 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
248 Exim main code and Exim-related utities. The introduction of taint
249 tracking also did many adjustments to string handling. Since then, eximon
250 frequently terminated with an assert failure.
252 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
253 check for 452 responses. This slightly helps the inefficieny of doing
254 a large alias-expansion into a recipient-limited target. The max_rcpt
255 transport option still applies (and at the current default, will override
256 the new feature). The check is done for either cause of synch, and forces
257 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
258 connection. The new facility is not tunable at this time.
260 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
261 library live data was being used, so the results became garbage. Make
262 copies while it is still usable.
264 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
265 on delivery deferred (==) and failed (**) lines (if a delivery was
266 attemtped). Previously it was only on completion (=>) lines.
268 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
269 for the expansion of the server_scram_iter and server_scram_salt options.
271 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
272 are now specifically given a NO_DATA response without hitting the system
273 resolver. The library goes on to do the now-standard TXT lookup.
274 Use of dnsdb lookups is not affected.
276 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
277 only retrieve the errormessage once. Previously two calls to dlerror()
278 were used, and the second one (for mainlog/paniclog) retrieved null
281 JH/20 Taint checking: disallow use of tainted data for
282 - the appendfile transport file and directory options
283 - the pipe transport command
284 - the autoreply transport file, log and once options
285 - file names used by the redirect router (including filter files)
287 - paths used by single-key lookups
288 Previously this was permitted.
290 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
291 adjusted the size of a major service buffer; this failed because the
292 buffer was in use at the time. Change to a compile-time increase in the
293 buffer size, when this authenticator is compiled into exim.
295 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
296 previous fast-mode was untenable in the face of glibs using mmap to
297 support larger malloc requests.
299 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
300 New values supported, if defined on system where compiled:
301 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
302 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
304 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
305 running a limited number of proceses in parallel, a benefit is gained. The
306 amount varies with the platform hardware and load. The use of the option
307 queue_run_in_order means we cannot do this, as ordering becomes
310 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
311 had introduced a string-copy (for ensuring NUL-termination) which was not
312 appropriate for that case, which can include embedded NUL bytes in the
313 block of data. Investigation showed the copy to actually be needless, the
314 data being length-specified.
316 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
317 done during a receiving connection, and both used TLS, global info was
318 used rather than per-connection info for tracking the state of data
319 queued for transmission. This could result in a connection hang.
321 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
322 Previously, when delivering serveral messages down a single connection
323 only the first would provide a SIZE. This was due to the size information
324 not being properly tracked.
326 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
327 TAI (at 37 seconds currently), pretend to be in UTC for time-related
328 expansion and logging. Previously, spurious values such as a future
329 minute could be seen.
331 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
332 it could crash from a null-deref. This could also affect the
333 ${addresses: } operator and ${readsock } item.
335 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
336 message following a mime one, the variable was not reset.
338 JH/30 When an pipelined-connect fails at the first response, assume incorrect
339 cached capability (perhaps the peer reneged?) and immediately retry in
342 JH/31 Fix spurious detection of timeout while writing to transport filter.
344 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
345 an attempt to copy the string was made before checking it.
347 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
348 taint of the lookup key was maintained; we now regard the presence in the
349 filesystem as sufficient validation.
351 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
352 argument is supplied.
354 JH/35 The dsearch lookup now requires that the directory is an absolute path.
355 Previously this was not checked, and nonempty relative paths made an
356 access under Exim's current working directory.
358 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
359 Previously no event was raised.
361 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
362 parameter supplied by the sender MAIL FROM command. Previously it was
363 ignored, and only the check_spool_space option value for the required
366 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
367 the size of the signing public-key. Previously it was instead giving
368 the size of the signature hash.
370 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
371 the default. See the (new) dkim_verify_min_keysizes option.
373 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
374 and an ACL use a lookup for checking either the local_part or domain,
375 stale data could be accessed. Ensure that variable references are
376 dropped between messages.
378 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
379 by the client was not checked as pointing within response data before
380 being used. A malicious client could thus cause an out-of-bounds read and
381 possibly gain authentication. Fix by adding the check.
383 JH/42 Internationalisation: change the default for downconversion in the smtp
384 transport to be "if needed". Previously it was "as previously set" for
385 the message, which usually meant "if needed" for message-submission but
386 "no" for everything else. However, MTAs have been seen using SMTPUTF8
387 even when the envelope addresses did not need it, resulting in forwarding
388 failures to non-supporting MTAs. A downconvert in such cases will be
389 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
390 transport. The change does mean that addresses needing conversion will
391 be converted when previously a delivery failure would occur.
393 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
394 response was received it would be used unchecked in a fail-DSN, violating
395 standards on line-length limits. Truncate if needed.
397 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
398 doesn't work. (Additionally add a new main config option to configure the
405 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
406 side, to decode a TLS 1.3 packet capture.
408 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
409 Previously the default library behaviour applied, sending two, each in
412 JH/03 Debug output for ACL now gives the config file name and line number for
415 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
417 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
419 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
420 buffer overrun for (non-chunking) other transports.
422 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
423 TLS1.3, means that a server rejecting a client certificate is not visible
424 to the client until the first read of encrypted data (typically the
425 response to EHLO). Add detection for that case and treat it as a failed
426 TLS connection attempt, so that the normal retry-in-clear can work (if
427 suitably configured).
429 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
430 and/or domain. Found and fixed by Jason Betts.
432 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
433 configuration). If a CNAME target was not a wellformed name pattern, a
436 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
437 the OS reports them interleaved with other addresses.
439 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
440 used both for input and for a verify callout, both encrypted, SMTP
441 responses being sent by the server could be lost. This resulted in
442 dropped connections and sometimes bounces generated by a peer sending
445 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
446 string. Previously it was possible to cause undefined behaviour in a
447 library routine (usually a crash). Found by "zerons".
449 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
452 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
453 API was removed, so update to use the newer ones.
455 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
456 any timeout set, is taking a long time. Previously we would hang on to a
457 rotated logfile "forever" if the input was arriving with long gaps
458 (a previous attempt to fix addressed lack, for a long time, of initial
461 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
462 shared (NFS) environment. The length of the tempfile name is now
463 4 + 16 ("hdr.$message_exim_id") which might break on file
464 systems which restrict the file name length to lower values.
467 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
468 shared (NFS) environment.
470 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
471 did for all versions <4.90). Notably -M, -m, --invert, -I may be
474 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
475 on some platforms for bit 31.
477 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
478 to changes apparently associated with TLS1.3 handling some of the APIs
479 previously used were either nonfunctional or inappropriate. Strings
480 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
481 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
482 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
483 This affects log line X= elements, the $tls_{in,out}_cipher variables,
484 and the use of specific cipher names in the encrypted= ACL condition.
486 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
488 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
489 verification result was not updated unless hosts_require_ocsp applied.
491 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
492 queue_list_requires_admin set to false, non-admin users were denied the
495 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
496 directory-of-certs mode. Previously they were advertised despite the
499 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
500 A single TCP connection by a client will now hold a TLS connection open
501 for multiple message deliveries, by default. Previously the default was to
504 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
505 default. If built with the facility, DANE will be used. The facility
506 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
508 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
509 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
510 must be defined and you must still, unless you define DISABLE_TLS, manage
511 the the include-dir and library-file requirements that go with that
512 choice. Non-TLS builds are still supported.
514 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
517 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
518 default. If the platform supports and has the facility enabled, it will
519 be requested on all coneections.
521 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
522 controlled by the build-time option SUPPORT_PIPE_CONNECT.
524 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
526 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
527 success-DSN messages. Previously the From: header was always the default
528 one for these; the option was ignored.
530 JH/28 Fix the timeout on smtp response to apply to the whole response.
531 Previously it was reset for every read, so a teergrubing peer sending
532 single bytes within the time limit could extend the connection for a
533 long time. Credit to Qualsys Security Advisory Team for the discovery.
535 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
536 delivery address, which leaked information of the results of local
537 forwarding. Change to the original envelope recipient address, per
540 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
541 requested. Previously not bounce was generated and a log entry of
542 error ignored was made.
544 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
546 JH/32 Introduce a general tainting mechanism for values read from the input
547 channel, and values derived from them. Refuse to expand any tainted
548 values, to catch one form of exploit.
550 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
551 was unused and the unexpanded text used for the test. Found and
552 fixed by Ruben Jenster.
554 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
555 an attempt to use a TLS library read routine dereffed a nul pointer,
558 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
559 them in our smtp response.
561 JH/36 Have the general router option retry_use_local_part default to true when
562 any of the restrictive preconditions are set (to anything). Previously it
563 was only for check_local user. The change removes one item of manual
564 configuration which is required for proper retries when a remote router
565 handles a subset of addresses for a domain.
567 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
568 link count into consideration.
570 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
571 caused the extension of big_buffer, the following lines were ignored.
573 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
574 accordance with RFC 2308. Previously there was no expiry, so a longlived
575 receive process (eg. due to ACL delays) versus a short SOA value could
578 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
580 JH/39 Promote DMARC support to mainline.
582 JH/40 Bug 2452: Add a References: header to DSNs.
584 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
585 parameters. The relevant library call is documented as "Deprecated: This
586 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
587 3.6.0, DH parameters are negotiated following RFC7919."
589 HS/06 Change the default of dnssec_request_domains to "*"
591 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
592 carried on and emitted a BDAT command, even when PIPELINING was not
595 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
596 buffer was used for the filename, resulting in a trap when tainted
597 arguments (eg. $domain) were used.
599 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
600 recommended to avoid a possible server-load attack. The feature can be
601 re-enabled via the openssl_options main cofiguration option.
603 JH/45 local_scan API: documented the current smtp_printf() call. This changed
604 for version 4.90 - adding a "more data" boolean to the arguments.
605 Bumped the ABI version number also, this having been missed previously;
606 release versions 4.90 to 4.92.3 inclusive were effectively broken in
607 respect of usage of smtp_printf() by either local_scan code or libraries
608 accessed via the ${dlfunc } expansion item. Both will need coding
609 adjustment for any calls to smtp_printf() to match the new function
610 signature; a FALSE value for the new argument is always safe.
612 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
613 the file-offset (which the Linux syscall does, and exim expects); this
614 resulted in an indefinite loop.
616 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
617 to do ARC verification. The Authentication-Results: header line added
618 by the configuration then had no ARC item.
624 JH/01 Remove code calling the customisable local_scan function, unless a new
625 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
627 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
628 non-signal-safe functions being used.
630 JH/03 Bug 2269: When presented with a received message having a stupidly large
631 number of DKIM-Signature headers, disable DKIM verification to avoid
632 a resource-consumption attack. The limit is set at twenty.
634 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
635 report of oldest_pass in ${authres } in consequence, and separate out
636 some descriptions of reasons for verification fail.
638 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
639 files in the spool were present and unlocked. A queue-runner could spot
640 them, resulting in a duplicate delivery. Fix that by doing the unlock
641 after the unlink. Investigation by Tim Stewart. Take the opportunity to
642 add more error-checking on spoolfile handling while that code is being
645 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
646 No known attacks, no CVE, this is defensive hardening.
648 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
649 a queue-runner could start a delivery while other operations were ongoing.
650 Cutthrough delivery was a common victim, resulting in duplicate delivery.
651 Found and investigated by Tim Stewart. Fix by using the open message data
652 file handle rather than opening another, and not locally closing it (which
653 releases a lock) for that case, while creating the temporary .eml format
654 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
656 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
657 $sender_verify_failure/$recipient_verify_failure to "random".
659 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
662 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
663 Previously this would segfault.
665 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
668 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
669 like zero, since the resolver should be doing this for us, But we need one
670 as a CNAME but no MX presence gets the CNAME returned; we need to check
671 that doesn't point to an MX to declare it "no MX returned" rather than
672 "error, loop". A new main option is added so the older capability of
673 following some limited number of chain links is maintained.
675 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
677 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
678 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
679 now-deprecated earlier definitions used only specified the range up to TLS
680 1.2 (in the older-version library docs).
682 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
684 JH/15 Rework TLS client-side context management. Stop using a global, and
685 explicitly pass a context around. This enables future use of TLS for
686 connections to service-daemons (eg. malware scanning) while a client smtp
687 connection is using TLS; with cutthrough connections this is quite likely.
689 JH/16 Fix ARC verification to do AS checks in reverse order.
691 JH/17 Support a "tls" option on the ${readsocket } expansion item.
693 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
694 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
695 Previously the "utf8" would be re-prepended for every additional message.
697 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
698 Previously thery were accepted, resulting in issues when attempting to
699 forward messages to a non-supporting MTA.
701 PP/02 Let -n work with printing macros too, not just options.
703 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
704 one parent address was copied, and bogus data was used at delivery-logging
705 time. Either a crash (after delivery) or bogus log data could result.
706 Discovery and analysis by Tim Stewart.
708 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
709 Previously if the string ended mid-character, we did not insert the
710 promised '?' replacement.
712 PP/04 Documentation: current string operators work on bytes, not codepoints.
714 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
715 should pack well giving a smaller memory footprint so better caching and
716 therefore performance. Group the declarations where this can't be done so
717 that the byte-sized flag variables are not interspersed among pointer
718 variables, giving a better chance of good packing by the compiler.
720 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
721 non-null, to avoid issues with sites running BATV. Previously reports were
722 sent with an empty envelope sender so looked like bounces.
724 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
725 The ignore_error flag wasn't being returned from the filter subprocess so
726 was not set for later routers. Investigation and fix by Matthias Kurz.
728 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
729 and a msg:complete for the whole, when a message is manually removed using
730 -Mrm. Developement by Matthias Kurz, hacked on by JH.
732 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
733 a "Gnu special" function, asprintf() in the DB utility binary builds; I
734 hope that is portable enough.
736 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
737 requiring a known-CA anchor certificate; make it now rely entirely on the
738 TLSA as an anchor. Checking the name on the leaf cert against the name
739 on the A-record for the host is still done for TA (but not for EE mode).
741 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
742 included in delivery lines for non-proxied connections, when compiled with
743 SUPPORT_SOCKS and running with proxy logging enabled.
745 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
746 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
747 move the existing event to fire before the normal logging of message
748 failure so that custom logging is bracketed by normal logging.
750 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
751 msg:fail:internal event. Developement by Matthias Kurz.
753 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
754 far too small for todays use of crypto signatures stored there. Go all
755 the way to the max DNS message size of 64kB, even though this might be
756 overmuch for IOT constrained device use.
758 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
759 copy a string over itself. The library routine is documented as not
760 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
762 JH/32 For main options check_spool_space and check_inode_space, where the
763 platform supports 64b integers, support more than the previous 2^31 kB
764 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
765 the previous G, M, k.
767 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
768 $authenticated_fail_id variable on authentication failure. Previously
771 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
772 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
773 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
774 GNUTLS_SEC_PARAM_MEDIUM.
776 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
777 side. Previously we would continue as if no SNI had been received.
779 JH/36 Harden the handling of string-lists. When a list consisted of a sole
780 "<" character, which should be a list-separator specification, we walked
781 off past the nul-terimation.
783 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
784 causes) even when the retry time is not yet met. Previously they were
785 not, meaning that when (say) an account was over-quota and temp-rejecting,
786 and multiple senders' messages were queued, only one sender would get
787 notified on each configured delay_warning cycle.
789 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
791 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
792 cipher-suites, an error can be left on the stack even for a succeeding
793 accept; this results in impossible error messages when a later operation
796 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
797 return error codes indicating retry. Under TLS1.3 this becomes required.
799 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
800 it only wrote the new authenticators, resulting in a lack of tracking of
801 peer changes of ESMTP extensions until the next cache flush.
803 JH/41 Fix the loop reading a message header line to check for integer overflow,
804 and more-often against header_maxsize. Previously a crafted message could
805 induce a crash of the recive process; now the message is cleanly rejected.
807 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
808 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
814 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
815 When redis_servers is set to a list of > 1 element, and the Redis servers
816 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
817 case of MOVED into a DEFER case instead, thus moving the query onto the
818 next server in the list. For a cluster of N elements, all N servers must
819 be defined in redis_servers.
821 GF/02 Catch and remove uninitialized value warning in exiqsumm
822 Check for existence of @ARGV before looking at $ARGV[0]
824 JH/01 Replace the store_release() internal interface with store_newblock(),
825 which internalises the check required to safely use the old one, plus
826 the allocate and data copy operations duplicated in both (!) of the
827 extant use locations.
829 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
830 modifier. This matches the restriction on the commandline.
832 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
833 Previously only the last row was returned.
835 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
836 we assumed that tags in the header were well-formed, and parsed the
837 element content after inspecting only the first char of the tag.
838 Assumptions at that stage could crash the receive process on malformed
841 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
842 While running the DKIM ACL we operate on the Permanent memory pool so that
843 variables created with "set" persist to the DATA ACL. Also (at any time)
844 DNS lookups that fail create cache records using the Permanent pool. But
845 expansions release any allocations made on the current pool - so a dnsdb
846 lookup expansion done in the DKIM ACL releases the memory used for the
847 DNS negative-cache, and bad things result. Solution is to switch to the
848 Main pool for expansions.
849 While we're in that code, add checks on the DNS cache during store_reset,
850 active in the testsuite.
851 Problem spotted, and debugging aided, by Wolfgang Breyha.
853 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
854 When none of the hosts presented to a transport match an already-open
855 connection, close it and proceed with the list. Previously we would
856 queue the message. Spotted by Lena with Yahoo, probably involving
859 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
860 Previously a spurious "250 OK id=" response was appended to the proper
863 JH/08 The "support for" informational output now, which built with Content
864 Scanning support, has a line for the malware scanner interfaces compiled
865 in. Interface can be individually included or not at build time.
867 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
868 by the template makefile "src/EDITME". The "STREAM" support for an older
869 ClamAV interface method is removed.
871 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
872 rows affected is given instead).
874 JH/11 The runtime Berkeley DB library version is now additionally output by
875 "exim -d -bV". Previously only the compile-time version was shown.
877 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
878 SMTP connection. Previously, when one had more recipients than the
879 first, an abortive onward connection was made. Move to full support for
880 multiple onward connections in sequence, handling cutthrough connection
881 for all multi-message initiating connections.
883 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
884 routers. Previously, a multi-recipient message would fail to match the
885 onward-connection opened for the first recipient, and cause its closure.
887 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
888 a timeout on read on a GnuTLS initiating connection, resulting in the
889 initiating connection being dropped. This mattered most when the callout
890 was marked defer_ok. Fix to keep the two timeout-detection methods
893 JH/15 Relax results from ACL control request to enable cutthrough, in
894 unsupported situations, from error to silently (except under debug)
895 ignoring. This covers use with PRDR, frozen messages, queue-only and
898 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
900 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
901 metadata, resulting in a crash in free().
903 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
904 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
905 Broken also in d185889f4, with init system revamp.
907 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
908 Previously we abruptly closed the connection after reading a malware-
909 found indication; now we go on to read the "scan ok" response line,
912 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
913 ACL. Previously, a crash would result.
915 JH/19 Speed up macro lookups during configuration file read, by skipping non-
916 macro text after a replacement (previously it was only once per line) and
917 by skipping builtin macros when searching for an uppercase lead character.
919 JH/20 DANE support moved from Experimental to mainline. The Makefile control
920 for the build is renamed.
922 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
923 was allocated for every new TLS startup, meaning one per message. Fix
924 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
926 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
927 reported the original. Fix to report (as far as possible) the ACL
928 result replacing the original.
930 JH/23 Fix memory leak during multi-message connections using STARTTLS under
931 OpenSSL. Certificate information is loaded for every new TLS startup,
932 and the resources needed to be freed.
934 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
936 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
939 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
940 DATA response info to the (existing) per-recipient response info for
941 the "C=" log element. It can have useful tracking info from the
942 destination system. Patch from Simon Arlott.
944 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
945 length value. Previously this would segfault.
947 HS/02 Support Avast multiline protoocol, this allows passing flags to
948 newer versions of the scanner.
950 JH/28 Ensure that variables possibly set during message acceptance are marked
951 dead before release of memory in the daemon loop. This stops complaints
952 about them when the debug_store option is enabled. Discovered specifically
953 for sender_rate_period, but applies to a whole set of variables.
954 Do the same for the queue-runner and queue-list loops, for variables set
955 from spool message files. Do the same for the SMTP per-message loop, for
956 certain variables indirectly set in ACL operations.
958 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
959 as a multi-recipient message from a mailinglist manager). The coding had
960 an arbitrary cutoff number of characters while checking for more input;
961 enforced by writing a NUL into the buffer. This corrupted long / fast
962 input. The problem was exposed more widely when more pipelineing of SMTP
963 responses was introduced, and one Exim system was feeding another.
964 The symptom is log complaints of SMTP syntax error (NUL chars) on the
965 receiving system, and refused recipients seen by the sending system
966 (propating to people being dropped from mailing lists).
967 Discovered and pinpointed by David Carter.
969 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
970 replaced by the ${authresults } expansion.
972 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
974 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
975 allows proper process termination in container environments.
977 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
978 Previously the "final dot" had a newline after it; ensure it is CR,LF.
980 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
981 and "err_perm", deprecated since 4.83 when the RFC-defined words
982 "temperror" and "permerror" were introduced.
984 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
985 transport-filters or DKIM-signing. The restriction was lost in the
986 consolidation of verify-callout and delivery SMTP handling.
987 Extend the restriction to also cover ARC-signing.
989 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
990 in defer=pass mode supply a 450 to the initiator. Previously the message
993 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
994 tls_require_ciphers is used as before.
996 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
997 "pass_unscanned". Only tmpfails from the scanner are written to
998 the paniclog, as they may require admin intervention (permission
999 denied, license issues). Other scanner errors (like decompression
1000 bombs) do not cause a paniclog entry.
1002 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1003 Previously it was possible to log spurious information in receive log
1006 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1007 triggered odd behaviour from Outlook Express clients.
1009 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1012 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1013 since the IETF WG has not yet settled on that versus the original
1014 "bare" representation.
1016 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1017 Previously the millisecond value corrupted the output.
1018 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1019 corrupted the output.
1025 JH/01 Rework error string handling in TLS interface so that the caller in
1026 more cases is responsible for logging. This permits library-sourced
1027 string to be attached to addresses during delivery, and collapses
1028 pairs of long lines into single ones.
1030 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1031 during configuration. Wildcards are allowed and expanded.
1033 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1034 This permits better logging.
1036 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1037 do not close it down and have the child transport start it up again on
1038 the passed-on TCP connection. Instead, proxy the child (and any
1039 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1040 affected: the continued delivery log lines do not have any DNSSEC, TLS
1041 Certificate or OCSP information. TLS cipher information is still logged.
1043 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1044 identical IP addresses on different listening ports. Will also affect
1047 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1048 add noisy ifdef guards to special-case this sillyness.
1049 Patch from Bernd Kuhls.
1051 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1052 than 255 are no longer allowed.
1054 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1055 Disable the session-cache too, which might reduce our load. Since we
1056 currrectly use a new context for every connection, both as server and
1057 client, there is no benefit for these.
1058 GnuTLS appears to not support tickets server-side by default (we don't
1059 call gnutls_session_ticket_enable_server()) but client side is enabled
1060 by default on recent versions (3.1.3 +) unless the PFS priority string
1063 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1064 <https://reproducible-builds.org/specs/source-date-epoch/>.
1066 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1067 the check for any unsuccessful recipients did not notice the limit, and
1068 erroneously found still-pending ones.
1070 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1071 MSG_MORE. Only in-clear (not on TLS connections).
1073 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1074 transport-filter is involved we can buffer the headers in memory for
1075 creating the signature, and read the spool data file once for the
1076 signature and again for transmission.
1078 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1079 4.77 as the kernel support then wasn't solid, having issues in 64bit
1080 mode. Now, it's been long enough. Add support for FreeBSD also.
1082 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1083 case where the routing stage had gathered several addresses to send to
1084 a host before calling the transport for the first, we previously failed
1085 to close down TLS in the old transport process before passing the TCP
1086 connection to the new process. The new one sent a STARTTLS command
1087 which naturally failed, giving a failed delivery and bloating the retry
1088 database. Investigation and fix prototype from Wolfgang Breyha.
1090 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1091 false-negatives in the check that the sender had not preempted a response
1092 or prompt from Exim (running as a server), due to that code's lack of
1093 awareness of the SMTP input buffering.
1095 PP/04 Add commandline_checks_require_admin option.
1096 Exim drops privileges sanely, various checks such as -be aren't a
1097 security problem, as long as you trust local users with access to their
1098 own account. When invoked by services which pass untrusted data to
1099 Exim, this might be an issue. Set this option in main configuration
1100 AND make fixes to the calling application, such as using `--` to stop
1103 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1104 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1105 responses to those, into a single TLS record each way (this usually means
1106 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1109 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1110 affects you only if you're dancing at the edge of the param size limits.
1111 If you are, and this message makes sense to you, then: raise the
1112 configured limit or use OpenSSL 1.1. Nothing we can do for older
1115 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1116 cmdline element to get the documented default one. Previously it was
1119 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1120 are unsuable for non-message-specific reasons - eg. connection timeout,
1121 banner-time rejection.
1123 JH/16 Fix logging of delivery remote port, when specified by router, under
1126 PP/06 Repair manualroute's ability to take options in any order, even if one
1127 is the name of a transport.
1130 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1132 JH/17 Change the list-building routines interface to use the expanding-string
1133 triplet model, for better allocation and copying behaviour.
1135 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1136 Previously it was constructed the first time a possibly-matching string
1137 was met in the configuration file input during startup; now it is done
1140 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1141 compatible one, to avoid the (poorly documented) possibility of a config
1142 file in the working directory redirecting the DB files, possibly correpting
1143 some existing file. CVE-2017-10140 assigned for BDB.
1145 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1146 cache-hot. Previously, although the result was properly cached, the
1147 initial verify call returned a defer.
1149 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1150 the main verify for receipient in uncached-mode.
1152 JH/22 Retire historical build files to an "unsupported" subdir. These are
1153 defined as "ones for which we have no current evidence of testing".
1155 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1156 if present. Previously it was ignored.
1158 JH/24 Start using specified-initialisers in C structure init coding. This is
1159 a C99 feature (it's 2017, so now considered safe).
1161 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1162 if was a fixed-sized field and bitmask ops via macros; it is now more
1165 PP/07 GitHub PR 56: Apply MariaDB build fix.
1166 Patch provided by Jaroslav Škarvada.
1168 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1169 during Coverity cleanups [4.87 JH/47]
1170 Diagnosis and fix provided by Michael Fischer v. Mollard.
1172 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1173 the right size to place the terminating semicolon on its own folded
1174 line, the header hash was calculated to an incorrect value thanks to
1175 the (relaxed) space the fold became.
1177 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1178 and confused the parent.
1180 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1181 which could crash as a result. This could lead to undeliverable messages.
1183 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1186 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1187 matters on fast-turnover and PID-randomising systems, which were getting
1188 out-of-order delivery.
1190 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1191 a possibly-overlapping copy. The symptom was that "Remote host closed
1192 connection in response to HELO" was logged instead of the actual 4xx
1195 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1196 Previously only that bufferd was discarded, resulting in SYMTP command
1199 JH/32 DKIM: when a message has multiple signatures matching an identity given
1200 in dkim_verify_signers, run the dkim acl once for each. Previously only
1201 one run was done. Bug 2189.
1203 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1204 "panic the current process" to "deliberately defer". The panic log is
1205 still written with the problem list name; the mail and reject logs now
1206 get a temp-reject line for the message that was being handled, saying
1207 something like "domains check lookup or other defer". The SMTP 451
1208 message is still "Temporary local problem".
1210 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1211 A crafted sequence of BDAT commands could result in in-use memory beeing
1212 freed. CVE-2017-16943.
1214 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1215 from SMTP input. Previously it was always done; now only done for DATA
1216 and not BDAT commands. CVE-2017-16944.
1218 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1219 to the message (such as an overlong header line). Previously this was
1220 not done and we did not exit BDAT mode. Followon from the previous item
1221 though a different problem.
1227 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1228 than -2003 did; needs libidn2 in addition to libidn.
1230 JH/02 The path option on a pipe transport is now expanded before use.
1232 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1233 Patch provided by "Björn", documentation fix added too.
1235 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1236 missing a wire-to-host endian conversion.
1238 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1239 close after a BDAT command line could be taken as a following command,
1240 giving a synch failure. Fix by only checking for synch immediately
1241 before acknowledging the chunk.
1243 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1244 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1245 macro. Patches provided by Josh Soref.
1247 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1248 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1249 by IANA as a service extension. However, John Klensin suggests that we
1252 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1253 the dkim code may be unix-mode line endings rather than smtp wire-format
1254 CRLF, so prepend a CR to any bare LF.
1256 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1257 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1259 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1260 the first were themselves being wrongly included in the feed into dkim
1261 processing; with most chunk sizes in use this resulted in an incorrect
1262 body hash calculated value.
1264 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1265 DKIM signature block, for verification. Although advised against by
1266 standards it is specifically not ruled illegal.
1268 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1270 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1271 missing a body hash (the bh= tag).
1273 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1274 It seems that HAProxy sends the Proxy Protocol information in clear and
1275 only then does a TLS startup, so do the same.
1277 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1278 TCP connections (such as for Spamd) unless the daemon successfully set
1279 Fast Open mode on its listening sockets. This fixes breakage seen on
1280 too-old kernels or those not configured for Fast Open, at the cost of
1281 requiring both directions being enabled for TFO, and TFO never being used
1282 by non-daemon-related Exim processes.
1284 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1285 endings, at least on the first header line. Try to canonify any that get
1286 past that check, despite the cost.
1288 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1289 now limited to an arbitrary five deep, while parsing addresses with the
1290 strip_excess_angle_brackets option enabled.
1292 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1293 instead leave the unprompted TLS handshake in socket buffer for the
1294 TLS library to consume.
1296 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1298 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1300 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1301 region in one operation, for speed, and this leaves assigned pointers
1302 dangling. Add checks run only under the testsuite which checks all
1303 variables at a store-reset and panics on a dangling pointer; add code
1304 explicitly nulling out all the variables discovered. Fixes one known
1305 bug: a transport crash, where a dangling pointer for $sending_ip_address
1306 originally assigned in a verify callout, is re-used.
1308 PP/06 Drop '.' from @INC in various Perl scripts.
1310 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1312 PP/08 Reduce a number of compilation warnings under clang; building with
1313 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1314 should be warning-free.
1316 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1318 HS/01 Fix portability problems introduced by PP/08 for platforms where
1319 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1321 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1322 chunk. This allows us to accept broken chunked messages. We need a more
1323 general solution here.
1325 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1326 already-broken messages in the queue.
1328 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1330 JH/19 Fix reference counting bug in routing-generated-address tracking.
1336 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1337 supports it and a size is available (ie. the sending peer gave us one).
1339 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1340 years of being deprecated). The replacements are the ACLs
1341 acl_smtp_mime and acl_not_smtp_mime.
1343 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1344 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1345 or even an in-clear connection were permitted. Now, if the host lookup
1346 was dnssec and dane was requested then the host is only used if the
1347 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1348 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1349 if one fails this test.
1350 This means that a poorly-configured remote DNS will make it incommunicado;
1351 but it protects against a DNS-interception attack on it.
1353 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1354 non-noisy when a race steals the message being considered.
1356 JH/05 If main configuration option tls_certificate is unset, generate a
1357 self-signed certificate for inbound TLS connections.
1359 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1360 in rewrites and routers.
1362 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1363 and logged a warning sing 4.83; now they are a configuration file error.
1365 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1366 (lacking @domain). Apply the same qualification processing as RCPT.
1368 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1370 JH/10 Support ${sha256:} applied to a string (as well as the previous
1373 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1374 a cutthrough deliver is pending, as we always want to make a connection.
1375 This also avoids re-routing the message when later placing the cutthrough
1376 connection after a verify cache hit.
1377 Do not update it with the verify result either.
1379 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1380 when routing results in more than one destination address.
1382 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1383 signing (which inhibits the cutthrough capability). Previously only
1384 the presence of an option was tested; now an expansion evaluating as
1385 empty is permissible (obviously it should depend only on data available
1386 when the cutthrough connection is made).
1388 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1389 the relevant preceding SMTP command did not note the pipelining mode.
1391 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1392 Previously they were not counted.
1394 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1395 as one having no matching records. Previously we deferred the message
1396 that needed the lookup.
1398 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1399 distinguished as "(=".
1401 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1402 for missing MX records. Previously it only worked for missing A records.
1404 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1406 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1407 after the data-go-ahead and data-ack. Patch from Jason Betts.
1409 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1410 even for a "none" policy. Patch from Tony Meyer.
1412 JH/22 Fix continued use of a connection for further deliveries. If a port was
1413 specified by a router, it must also match for the delivery to be
1416 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1417 When one of the recipients of a message was unsuitable for the connection
1418 (has no matching addresses), we lost track of needing to mark it
1419 deferred. As a result mail would be lost.
1421 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1423 JH/25 Decoding ACL controls is now done using a binary search; the source code
1424 takes up less space and should be simpler to maintain. Merge the ACL
1425 condition decode tables also, with similar effect.
1427 JH/26 Fix problem with one_time used on a redirect router which returned the
1428 parent address unchanged. A retry would see the parent address marked as
1429 delivered, so not attempt the (identical) child. As a result mail would
1432 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1433 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1434 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1437 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1438 Main config options check_{log,spool}_{inodes,space} are now
1439 100 inodes, 10MB unless set otherwise in the configuration.
1441 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1442 Previously it only applied to the main-section connection policy
1445 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1447 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1448 by me. Added RFC7919 DH primes as an alternative.
1450 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1451 are not in the system include path.
1453 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1454 GnuTLS, when a session startup failed (eg because the client disconnected)
1455 Exim did stdio operations after fclose. This was exposed by a recent
1456 change which nulled out the file handle after the fclose.
1458 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1459 signed directly by the cert-signing cert, rather than an intermediate
1460 OCSP-signing cert. This is the model used by LetsEncrypt.
1462 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1464 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1465 an incoming connection.
1467 HS/02 Bug 1802: Do not half-close the connection after sending a request
1470 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1471 fallback to "prime256v1".
1473 JH/34 SECURITY: Use proper copy of DATA command in error message.
1474 Could leak key material. Remotely exploitable. CVE-2016-9963.
1480 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1481 and 3.4.4 - once the server is enabled to respond to an OCSP request
1482 it does even when not requested, resulting in a stapling non-aware
1483 client dropping the TLS connection.
1485 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1486 support variable-length bit vectors. No functional change.
1488 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1489 The I= interface field on outgoing lines is now after the H= remote
1490 host field, same as incoming lines. There is a separate
1491 outgoing_interface log selector which allows you to disable the
1494 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1495 If not running log_selector +smtp_connection the mainlog would be held
1496 open indefinitely after a "too many connections" event, including to a
1497 deleted file after a log rotate. Leave the per net connection logging
1498 leaving it open for efficiency as that will be quickly detected by the
1499 check on the next write.
1501 HS/01 Bug 1671: Fix post transport crash.
1502 Processing the wait-<transport> messages could crash the delivery
1503 process if the message IDs didn't exist for some reason. When
1504 using 'split_spool_directory=yes' the construction of the spool
1505 file name failed already, exposing the same netto behaviour.
1507 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1508 mime_regex ACL conditions.
1510 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1511 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1512 to HELO, local diagnostic string.
1514 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1515 log line to debug. Even when configured with a tls authenticator many
1516 client connections are expected to not authenticate in this way, so
1517 an authenticate fail is not an error.
1519 HS/02 Add the Exim version string to the process info. This way exiwhat
1520 gives some more detail about the running daemon.
1522 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1523 matter for fast-change records such as DNSBLs.
1525 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1526 retry record, even if constant. There may be multiple transports with
1527 different interface settings and the retry behaviour needs to be kept
1530 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1532 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1534 JH/10 Bug 840: fix log_defer_output option of pipe transport
1536 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1537 hosts_randomize. This matters a lot when combined with mua_wrapper.
1539 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1540 ${quote_pgsql:<string>} operator.
1542 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1544 JH/14 Change header file name for freeradius-client. Relevant if compiling
1545 with Radius support; from the Gentoo tree and checked under Fedora.
1547 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1549 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1550 option transport_filter, meaning no filtering.
1552 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1554 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1555 defaults to "*" (all hosts). The variable is now available when not built
1556 with TLS, default unset, mainly to enable keeping the testsuite sane.
1557 If a server certificate is not supplied (via tls_certificate) an error is
1558 logged, and clients will find TLS connections fail on startup. Presumably
1559 they will retry in-clear.
1560 Packagers of Exim are strongly encouraged to create a server certificate
1561 at installation time.
1563 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1564 with the $config_file variable.
1566 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1567 in transport context, after the attempt, and per-recipient. The latter type
1568 is per host attempted. The event data is the error message, and the errno
1569 information encodes the lookup type (A vs. MX) used for the (first) host,
1570 and the trailing two digits of the smtp 4xx response.
1572 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1573 to write to mainlog (or rejectlog, paniclog) in the window between file
1574 creation and permissions/ownership being changed. Particularly affects
1575 installations where exicyclog is run as root, rather than exim user;
1576 result is that the running daemon panics and dies.
1578 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1580 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1581 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1582 "pri" and "weight". Note that the previous implicit priority given by the
1583 list order is no longer honoured.
1585 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1586 for DKIM processing.
1588 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1589 by defining SUPPORT_SOCKS.
1591 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1592 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1593 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1594 variables are renamed to proxy_{local,external}_{address,port}.
1596 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1597 for a build by defining SUPPORT_I18N
1599 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1600 of the query string, and make ${quote_redis:} do that quoting.
1602 JH/29 Move Events support from Experimental to mainline, enabled by default
1603 and removable for a build by defining DISABLE_EVENT.
1605 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1607 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1608 cached by the daemon.
1610 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1611 by defining LOOKUP_REDIS. The libhiredis library is required.
1613 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1614 keys are given for lookup.
1616 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1617 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1618 only supported when built with TLS support. The PolarSSL SHA routines
1619 are still used when the TLS library is too old for convenient support.
1621 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1622 openssl_options), for security. OpenSSL forces this from version 1.1.0
1623 server-side so match that on older versions.
1625 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1626 allocation for $value could be released as the expansion processing
1627 concluded, but leaving the global pointer active for it.
1629 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1630 and to use the domains and local_parts ACL conditions.
1632 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1633 incorrectly not doubled on cutthrough transmission, hence seen as a
1634 body-termination at the receiving system - resulting in truncated mails.
1635 Commonly the sender saw a TCP-level error, and retransmitted the message
1636 via the normal store-and-forward channel. This could result in duplicates
1637 received - but deduplicating mailstores were liable to retain only the
1638 initial truncated version.
1640 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1642 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1644 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1645 we're in there, support oversigning also; bug 1309.
1647 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1649 HS/04 Add support for keep_environment and add_environment options.
1651 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1652 either intentional arithmetic overflow during PRNG, or testing config-
1655 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1656 delivery resulted in actual delivery. Cancel cutthrough before DATA
1659 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1660 rejects a recipient: pass the reject to the originator.
1662 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1663 Many were false-positives and ignorable, but it's worth fixing the
1666 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1667 for the new environment-manipulation done at startup. Move the routines
1668 from being local to tls.c to being global via the os.c file.
1670 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1671 an extract embedded as result-arg for a map, the first arg for extract
1672 is unavailable so we cannot tell if this is a numbered or keyed
1673 extraction. Accept either.
1679 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1682 JH/02 The smtp transport option "multi_domain" is now expanded.
1684 JH/03 The smtp transport now requests PRDR by default, if the server offers
1687 JH/04 Certificate name checking on server certificates, when exim is a client,
1688 is now done by default. The transport option tls_verify_cert_hostnames
1689 can be used to disable this per-host. The build option
1690 EXPERIMENTAL_CERTNAMES is withdrawn.
1692 JH/05 The value of the tls_verify_certificates smtp transport and main options
1693 default to the word "system" to access the system default CA bundle.
1694 For GnuTLS, only version 3.0.20 or later.
1696 JH/06 Verification of the server certificate for a TLS connection is now tried
1697 (but not required) by default. The verification status is now logged by
1698 default, for both outbound TLS and client-certificate supplying inbound
1701 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1704 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1705 Status Notification (bounce) messages are now MIME format per RFC 3464.
1706 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1707 under the control of the dsn_advertise_hosts option, and routers may
1708 have a dsn_lasthop option.
1710 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1711 default, modifiable by a malware= option. The list separator for
1712 the options can now be changed in the usual way. Bug 68.
1714 JH/10 The smtp_receive_timeout main option is now expanded before use.
1716 JH/11 The incoming_interface log option now also enables logging of the
1717 local interface on delivery outgoing connections.
1719 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1720 if the interface and destination host and port all match.
1722 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1725 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1726 Patch from Andrew Lewis.
1728 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1729 now supports optional time-restrictions, weighting, and priority
1730 modifiers per server. Patch originally by <rommer@active.by>.
1732 JH/16 The spamd_address main option now supports a mixed list of local
1733 and remote servers. Remote servers can be IPv6 addresses, and
1734 specify a port-range.
1736 JH/17 Bug 68: The spamd_address main option now supports an optional
1737 timeout value per server.
1739 JH/18 Bug 1581: Router and transport options headers_add/remove can
1740 now have the list separator specified.
1742 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1745 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1748 JH/21 Support for the A6 type of dns record is withdrawn.
1750 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1751 rather than the verbs used.
1753 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1754 from 255 to 1024 chars.
1756 JH/24 Verification callouts now attempt to use TLS by default.
1758 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1759 are generic router options now. The defaults didn't change.
1761 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1762 Original patch from Alexander Shikoff, worked over by JH.
1764 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1765 files if ZCAT_COMMAND is not executable.
1767 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1769 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1771 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1772 Normally benign, it bites when the pair was led to by a CNAME;
1773 modern usage is to not canonicalize the domain to a CNAME target
1774 (and we were inconsistent anyway for A-only vs AAAA+A).
1776 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1778 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1779 when evaluating $sender_host_dnssec.
1781 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1782 $sender_helo_dnssec variable.
1784 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1786 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1788 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1790 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1791 documented as working, but never had. Support all but $spam_report.
1793 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1794 added for tls authenticator.
1796 HS/03 Add perl_taintmode main config option
1802 TL/01 When running the test suite, the README says that variables such as
1803 no_msglog_check are global and can be placed anywhere in a specific
1804 test's script, however it was observed that placement needed to be near
1805 the beginning for it to behave that way. Changed the runtest perl
1806 script to read through the entire script once to detect and set these
1807 variables, reset to the beginning of the script, and then run through
1808 the script parsing/test process like normal.
1810 TL/02 The BSD's have an arc4random API. One of the functions to induce
1811 adding randomness was arc4random_stir(), but it has been removed in
1812 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1813 function when detected.
1815 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1816 cause callback expansion.
1818 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1819 syntax errors in an expansion can be treated as a string instead of
1820 logging or causing an error, due to the internal use of bool_lax
1821 instead of bool when processing it.
1823 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1824 server certificates when making smtp deliveries.
1826 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1828 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1830 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1832 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1833 Merged patch from Sebastian Wiedenroth.
1835 JH/05 Fix results-pipe from transport process. Several recipients, combined
1836 with certificate use, exposed issues where response data items split
1837 over buffer boundaries were not parsed properly. This eventually
1838 resulted in duplicates being sent. This issue only became common enough
1839 to notice due to the introduction of connection certificate information,
1840 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1842 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1843 size buffer was used, resulting in syntax errors when an expansion
1846 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1847 version 3.3.6 or later.
1849 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1850 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1851 both become "event_action", the variables become $event_name, $event_data
1852 and $event_defer_errno. There is a new variable $verify_mode, usable in
1853 routers, transports and related events. The tls:cert event is now also
1854 raised for inbound connections, if the main configuration event_action
1857 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1858 early OCSP support, but no stapling (appears to be less than 1.0.0).
1860 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1861 server certificate names available under the smtp transport option
1862 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1865 JH/10 Time-related extraction expansions from certificates now use the main
1866 option "timezone" setting for output formatting, and are consistent
1867 between OpenSSL and GnuTLS compilations. Bug 1541.
1869 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1870 encoded parameter in the incoming message. Bug 1558.
1872 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1873 include certificate info, eximon was claiming there were spoolfile
1876 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1878 JH/14 Log delivery-related information more consistently, using the sequence
1879 "H=<name> [<ip>]" wherever possible.
1881 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1882 are problematic for Debian distribution, omit them from the release
1885 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1887 JH/16 Fix string representation of time values on 64bit time_t architectures.
1890 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1891 output list separator was used.
1896 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1897 checkers that were complaining about end of non-void function with no
1900 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1901 This was a regression introduced in 4.83 by another bugfix.
1903 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1905 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1906 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1912 TF/01 Correctly close the server side of TLS when forking for delivery.
1914 When a message was received over SMTP with TLS, Exim failed to clear up
1915 the incoming connection properly after forking off the child process to
1916 deliver the message. In some situations the subsequent outgoing
1917 delivery connection happened to have the same fd number as the incoming
1918 connection previously had. Exim would try to use TLS and fail, logging
1919 a "Bad file descriptor" error.
1921 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1922 utilities have not been installed.
1924 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1925 temporary space as the ACL may create new global variables.
1927 TL/01 LDAP support uses per connection or global context settings, depending
1928 upon the detected version of the libraries at build time.
1930 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1931 to extract and use the src ip:port in logging and expansions as if it
1932 were a direct connection from the outside internet. PPv2 support was
1933 updated based on HAProxy spec change in May 2014.
1935 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1937 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1938 Properly escape header and check for NULL return.
1940 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1943 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1945 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1946 characters in header names, implemented as a verify condition.
1947 Contributed by Michael Fischer v. Mollard.
1949 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1950 results permerror and temperror. Previous values are deprecated but
1951 still accepted. In a future release, err_perm and err_temp will be
1952 completely removed, which will be a backward incompatibility if the
1953 ACL tests for either of these two old results. Patch contributed by
1954 user bes-internal on the mailing list.
1956 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1958 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1959 selectors, in both main and reject logs.
1961 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1964 JH/07 Add malware type "sock" for talking to simple daemon.
1966 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1968 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1969 routers/transports under cutthrough routing.
1971 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1972 numbers. Touch up "bool" conditional to keep the same definition.
1974 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1976 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1978 JH/12 Expand items in router/transport headers_add or headers_remove lists
1979 individually rather than the list as a whole. Bug 1452.
1981 Required for reasonable handling of multiple headers_ options when
1982 they may be empty; requires that headers_remove items with embedded
1983 colons must have them doubled (or the list-separator changed).
1985 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1986 view the policy declared in the DMARC record. Currently, $dmarc_status
1987 is a combined value of both the record presence and the result of the
1990 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1992 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1993 dnslookup router and the smtp transport (applying to the forward
1996 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1997 of ldap servers used for a specific lookup. Patch provided by Heiko
2000 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2001 New variable $lookup_dnssec_authenticated for observability.
2003 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2004 Patch submitted by Lars Timman.
2006 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2008 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2009 Requires trusted mode and valid format message id, aborts otherwise.
2010 Patch contributed by Heiko Schlichting.
2012 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2013 certextract with support for various fields. Bug 1358.
2015 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2016 is requested by default, modifiable by smtp transport option
2019 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2020 operate on certificate variables to give certificate fingerprints
2021 Also new ${sha256:cert_variable}.
2023 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2025 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2026 Christian Aistleitner.
2028 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2030 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2031 file. Patch from Wolfgang Breyha.
2033 JH/25 Expand the coverage of the delivery $host and $host_address to
2034 client authenticators run in verify callout. Bug 1476.
2036 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2037 align with daemon_smtp_ports. Bug 72.
2039 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2040 support and error reporting did not work properly.
2042 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2043 and is readable. Patch from Andrew Colin Kissa.
2045 TL/14 Enhance documentation of ${run expansion and how it parses the
2046 commandline after expansion, particularly in the case when an
2047 unquoted variable expansion results in an empty value.
2049 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2051 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2052 by renaming away from it. Addresses GH issue 15, reported by
2055 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2057 TL/15 SECURITY: prevent double expansion in math comparison functions
2058 (can expand unsanitized data). Not remotely exploitable.
2065 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2067 PP/02 Make -n do something, by making it not do something.
2068 When combined with -bP, the name of an option is not output.
2070 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2073 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2074 $sender_host_name and config options to manage this, and basic check
2077 PP/05 DSCP support for outbound connections and control modifier for inbound.
2079 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2080 (Only plugin which currently uses this is kerberos4, which nobody should
2081 be using, but we should make it available and other future plugins might
2082 conceivably use it, even though it would break NAT; stuff *should* be
2083 using channel bindings instead).
2085 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2086 name; added for Sendmail compatibility; requires admin caller.
2087 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2088 just ignore it); requires trusted caller.
2089 Also parse but ignore: -Ac -Am -X<logfile>
2092 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2094 TL/02 Add +smtp_confirmation as a default logging option.
2096 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2097 Patch by Magnus Holmgren from 2007-02-20.
2099 TL/04 Bugzilla 1281 - Spec typo.
2100 Bugzilla 1283 - Spec typo.
2101 Bugzilla 1290 - Spec grammar fixes.
2103 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2105 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2107 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2108 mailing list by by Dmitry Isaikin.
2110 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2112 JH/02 Support "G" suffix to numbers in ${if comparisons.
2114 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2116 NM/01 Bugzilla 1197 - Spec typo
2117 Bugzilla 1196 - Spec examples corrections
2119 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2121 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2122 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2125 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2126 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2128 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2129 "acl {{name}{arg}...}", and optional args on acl condition
2132 JH/05 Permit multiple router/transport headers_add/remove lines.
2134 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2136 JH/07 Avoid using a waiting database for a single-message-only transport.
2137 Performance patch from Paul Fisher. Bugzilla 1262.
2139 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2142 JH/09 Add $headers_added variable, with content from use of ACL modifier
2143 add_header (but not yet added to the message). Bugzilla 199.
2145 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2146 Pulled from Bugzilla 817 by Wolfgang Breyha.
2148 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2150 (nb: this is the same fix as in Exim 4.80.1)
2152 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2155 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2158 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2160 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2161 advertise SMTP AUTH mechanism to us, instead of a generic
2162 protocol violation error. Also, make Exim more robust to bad
2163 data from the Dovecot auth socket.
2165 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2167 When a queue runner is handling a message, Exim first routes the
2168 recipient addresses, during which it prunes them based on the retry
2169 hints database. After that it attempts to deliver the message to
2170 any remaining recipients. It then updates the hints database using
2173 So if a recipient address works intermittently, it can get repeatedly
2174 deferred at routing time. The retry hints record remains fresh so the
2175 address never reaches the final cutoff time.
2177 This is a fairly common occurrence when a user is bumping up against
2178 their storage quota. Exim had some logic in its local delivery code
2179 to deal with this. However it did not apply to per-recipient defers
2180 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2182 This change adds a proper retry rule check during routing so that the
2183 final cutoff time is checked against the message's age. We only do
2184 this check if there is an address retry record and there is not a
2185 domain retry record; this implies that previous attempts to handle
2186 the address had the retry_use_local_parts option turned on. We use
2187 this as an approximation for the destination being like a local
2188 delivery, as in LMTP.
2190 I suspect this new check makes the old local delivery cutoff check
2191 redundant, but I have not verified this so I left the code in place.
2193 TF/02 Correct gecos expansion when From: is a prefix of the username.
2195 Test 0254 submits a message to Exim with the header
2199 When I ran the test suite under the user fanf2, Exim expanded
2200 the header to contain my full name, whereas it should have added
2201 a Resent-Sender: header. It erroneously treats any prefix of the
2202 username as equal to the username.
2204 This change corrects that bug.
2206 GF/01 DCC debug and logging tidyup
2207 Error conditions log to paniclog rather than rejectlog.
2208 Debug lines prefixed by "DCC: " to remove any ambiguity.
2210 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2212 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2213 Bug spotted by Jeremy Harris; was flawed since initial commit.
2214 Would have resulted in OCSP responses post-SNI triggering an Exim
2215 NULL dereference and crash.
2217 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2219 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2220 Bug detection, analysis and fix by Samuel Thibault.
2221 Bugzilla 1331, Debian bug #698092.
2223 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2225 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2226 Server implementation by Todd Lyons, client by JH.
2227 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2228 config variable "prdr_enable" controls whether the server
2229 advertises the facility. If the client requests PRDR a new
2230 acl_data_smtp_prdr ACL is called once for each recipient, after
2231 the body content is received and before the acl_smtp_data ACL.
2232 The client is controlled by both of: a hosts_try_prdr option
2233 on the smtp transport, and the server advertisement.
2234 Default client logging of deliveries and rejections involving
2235 PRDR are flagged with the string "PRDR".
2237 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2238 fclose(). Diagnosis by Todd Lyons.
2240 PP/17 Update configure.default to handle IPv6 localhost better.
2241 Patch by Alain Williams (plus minor tweaks).
2244 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2245 This is now consistent with GnuTLS, and is now documented: the
2246 previous undocumented portable approach to treating the option as
2247 unset was to force an expansion failure. That still works, and
2248 an empty string is now equivalent.
2250 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2251 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2252 not performing validation itself.
2254 PP/20 Added force_command boolean option to pipe transport.
2255 Patch from Nick Koston, of cPanel Inc.
2257 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2260 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2262 PP/21 Fix eximon continuous updating with timestamped log-files.
2263 Broken in a format-string cleanup in 4.80, missed when I repaired the
2264 other false fix of the same issue.
2265 Report and fix from Heiko Schlichting.
2268 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2269 Report from Prashanth Katuri.
2271 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2272 It's SecureTransport, so affects any MacOS clients which use the
2273 system-integrated TLS libraries, including email clients.
2275 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2276 using a MIME ACL for non-SMTP local injection.
2277 Report and assistance in diagnosis by Warren Baker.
2279 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2281 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2283 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2284 last id that failed so it may be referenced in subsequent ACL's.
2286 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2289 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2290 ldap library initialization, allowing self-signed CA's to be
2291 used. Also properly sets require_cert option later in code by
2292 using NULL (global ldap config) instead of ldap handle (per
2293 session). Bug diagnosis and testing by alxgomz.
2295 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2296 the src/util/ subdirectory.
2298 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2299 renamed to Transport Post Delivery Action by Jeremy Harris, as
2302 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2303 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2304 redis_servers = needs to be configured which will be used by the redis
2305 lookup. Patch from Warren Baker, of The Packet Hub.
2307 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2309 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2310 hostname or reverse DNS when processing a host list. Used suggestions
2311 from multiple comments on this bug.
2313 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2315 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2316 Missed a few lines, added it to make the runtest require no keyboard
2319 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2320 contains upper case chars. Make router use caseful_local_part.
2322 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2323 support when GnuTLS has been built with p11-kit.
2329 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2331 This, or similar/improved, will also be change PP/11 of 4.82.
2337 PP/01 Handle short writes when writing local log-files.
2338 In practice, only affects FreeBSD (8 onwards).
2339 Bugzilla 1053, with thanks to Dmitry Isaikin.
2341 NM/01 Bugzilla 949 - Documentation tweak
2343 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2346 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2348 PP/02 Implemented gsasl authenticator.
2350 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2352 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2353 `pkg-config foo` for cflags/libs.
2355 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2356 with rest of GSASL and with heimdal_gssapi.
2358 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2359 `pkg-config foo` for cflags/libs for the TLS implementation.
2361 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2362 properties get this fed in as external SSF. A number of robustness
2363 and debugging improvements to the cyrus_sasl authenticator.
2365 PP/08 cyrus_sasl server now expands the server_realm option.
2367 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2368 Patch by Jeremy Harris.
2370 PP/10 Added dbmjz lookup type.
2372 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2374 PP/12 MAIL args handles TAB as well as SP, for better interop with
2375 non-compliant senders.
2376 Analysis and variant patch by Todd Lyons.
2378 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2379 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2380 Patch from Dirk Mueller <dmueller@suse.com>
2382 PP/13 tls_peerdn now print-escaped for spool files.
2383 Observed some $tls_peerdn in wild which contained \n, which resulted
2384 in spool file corruption.
2386 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2387 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2388 or write after TLS renegotiation, which otherwise led to messages
2391 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2392 as a tracking header (ie: a signed header comes before the signature).
2393 Patch from Wolfgang Breyha.
2395 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2396 comma-sep list; embedded commas doubled.
2398 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2400 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2402 Report and patch from Dmitry Banschikov.
2404 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2405 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2406 failures. We appear to now support TLS1.1+ with Exim.
2408 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2409 lets Exim select keys and certificates based upon TLS SNI from client.
2410 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2411 before an outbound SMTP session. New log_selector, +tls_sni.
2413 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2414 NULL dereference. Report and patch from Alun Jones.
2416 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2417 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2418 Not seeing resolver debug output on NetBSD, but suspect this is a
2419 resolver implementation change.
2421 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2422 Left warnings. Added "eximon gdb" invocation mode.
2424 PP/21 Defaulting "accept_8bitmime" to true, not false.
2426 PP/22 Added -bw for inetd wait mode support.
2428 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2429 locate the relevant includes and libraries. Made this the default.
2431 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2432 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2434 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2435 This may cause build issues on older platforms.
2437 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2438 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2439 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2440 Added SNI support via GnuTLS too.
2441 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2443 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2445 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2446 Applied second patch from Janne, implementing suggestion to default
2447 multiple-strings-in-record handling to match SPF spec.
2449 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2451 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2452 read-only, out of scope).
2453 Patch from Wolfgang Breyha, report from Stuart Northfield.
2455 PP/29 Fix three issues highlighted by clang analyser static analysis.
2456 Only crash-plausible issue would require the Cambridge-specific
2457 iplookup router and a misconfiguration.
2458 Report from Marcin Mirosław.
2460 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2462 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2463 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2464 As part of this, removing so much warning spew let me fix some minor
2465 real issues in debug logging.
2467 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2468 assignment on my part. Fixed.
2470 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2471 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2472 Janne Snabb (who went above and beyond: thank you).
2474 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2475 string otherwise requires a connection and a bunch more work and it's
2476 relatively easy to get wrong. Should also expose TLS library linkage
2479 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2480 64-bit ${eval} (JH/03).
2482 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2483 GNU libc to support some of the 64-bit stuff, should not lead to
2484 conflicts. Defined before os.h is pulled in, so if a given platform
2485 needs to override this, it can.
2487 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2488 protection layer was required, which is not implemented.
2489 Bugzilla 1254, patch from Wolfgang Breyha.
2491 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2492 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2493 tls_dhparam take prime identifiers. Also unbreak combination of
2494 OpenSSL+DH_params+TLSSNI.
2496 PP/39 Disable SSLv2 by default in OpenSSL support.
2502 PP/01 Solaris build fix for Oracle's LDAP libraries.
2503 Bugzilla 1109, patch from Stephen Usher.
2505 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2507 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2510 TF/02 Fix a couple more cases where we did not log the error message
2511 when unlink() failed. See also change 4.74-TF/03.
2513 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2514 lock up or crash if it happened to be inside a call to libc when it
2515 got a SIGUSR1 from exiwhat.
2517 The SIGUSR1 handler appends the current process status to the process
2518 log which is later printed by exiwhat. It used to use the general
2519 purpose logging code to do this, but several functions it calls are
2520 not safe for signals.
2522 The new output code in the SIGUSR1 handler is specific to the process
2523 log, and simple enough that it's easy to inspect for signal safety.
2524 Removing some special cases also simplifies the general logging code.
2525 Removing the spurious timestamps from the process log simplifies
2528 TF/04 Improved ratelimit ACL condition.
2530 The /noupdate option has been deprecated in favour of /readonly which
2531 has clearer semantics. The /leaky, /strict, and /readonly update modes
2532 are mutually exclusive. The update mode is no longer included in the
2533 database key; it just determines when the database is updated. (This
2534 means that when you upgrade Exim will forget old rate measurements.)
2536 Exim now checks that the per_* options are used with an update mode that
2537 makes sense for the current ACL. For example, when Exim is processing a
2538 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2539 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2540 must specify per_mail/readonly. If you omit the update mode it defaults to
2541 /leaky where that makes sense (as before) or /readonly where required.
2543 The /noupdate option is now undocumented but still supported for
2544 backwards compatibility. It is equivalent to /readonly except that in
2545 ACLs where /readonly is required you may specify /leaky/noupdate or
2546 /strict/noupdate which are treated the same as /readonly.
2548 A useful new feature is the /count= option. This is a generalization
2549 of the per_byte option, so that you can measure the throughput of other
2550 aggregate values. For example, the per_byte option is now equivalent
2551 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2553 The per_rcpt option has been generalized using the /count= mechanism
2554 (though it's more complicated than the per_byte equivalence). When it is
2555 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2556 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2557 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2558 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2559 Note that using per_rcpt with a non-readonly update mode in more than
2560 one ACL will cause the recipients to be double-counted. (The per_mail
2561 and per_byte options don't have this problem.)
2563 The handling of very low rates has changed slightly. If the computed rate
2564 is less than the event's count (usually one) then this event is the first
2565 after a long gap. In this case the rate is set to the same as this event's
2566 count, so that the first message of a spam run is counted properly.
2568 The major new feature is a mechanism for counting the rate of unique
2569 events. The new per_addr option counts the number of different
2570 recipients that someone has sent messages to in the last time period. It
2571 behaves like per_rcpt if all the recipient addresses are different, but
2572 duplicate recipient addresses do not increase the measured rate. Like
2573 the /count= option this is a general mechanism, so the per_addr option
2574 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2575 example, measure the rate that a client uses different sender addresses
2576 with the options per_mail/unique=$sender_address. There are further
2577 details in the main documentation.
2579 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2581 TF/06 Removed a few PCRE remnants.
2583 TF/07 Automatically extract Exim's version number from tags in the git
2584 repository when doing development or release builds.
2586 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2587 Bugzilla 879. Patch from Paul Fisher.
2589 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2590 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2593 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2595 PP/05 Variable $av_failed, true if the AV scanner deferred.
2596 Bugzilla 1078. Patch from John Horne.
2598 PP/06 Stop make process more reliably on build failure.
2599 Bugzilla 1087. Patch from Heiko Schlittermann.
2601 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2602 Bugzilla 1089. Patch from Heiko Schlittermann.
2604 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2605 Bugzilla 1131. Patch from Holger Weiß.
2607 PP/09 Handle IPv6 addresses with SPF.
2608 Bugzilla 860. Patch from Wolfgang Breyha.
2610 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2612 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2615 PP/11 match_* no longer expand right-hand-side by default.
2616 New compile-time build option, EXPAND_LISTMATCH_RHS.
2617 New expansion conditions, "inlist", "inlisti".
2619 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2621 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2623 PP/14 fix log_write() format string regression from TF/03.
2624 Bugzilla 1152. Patch from Dmitry Isaikin.
2630 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2632 PP/02 Harmonised TLS library version reporting; only show if debugging.
2633 Layout now matches that introduced for other libraries in 4.74 PP/03.
2635 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2637 PP/04 New "dns_use_edns0" global option.
2639 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2642 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2643 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2645 TK/01 Updated PolarSSL code to 0.14.2.
2646 Bugzilla 1097. Patch from Andreas Metzler.
2648 PP/07 Catch divide-by-zero in ${eval:...}.
2649 Fixes bugzilla 1102.
2651 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2654 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2655 format-string attack -- SECURITY: remote arbitrary code execution.
2657 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2658 time unintentionally subject to list matching rules, letting the header
2659 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2660 arbitrary string expansion). This allowed for information disclosure.
2662 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2663 INT_MIN/-1 -- value coerced to INT_MAX.
2669 NM/01 Workaround for PCRE version dependency in version reporting
2672 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2673 This fixes portability to compilers other than gcc, notably
2674 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2676 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2677 makefiles for portability to HP-UX and POSIX correctness.
2679 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2680 Also via indented variable definition in the Makefile.
2681 (Debugging by Oliver Heesakkers).
2683 PP/02 Restore caching of spamd results with expanded spamd_address.
2684 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2686 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2687 Improves build reliability. Fix from: Frank Elsner
2689 NM/02 Fix wide character breakage in the rfc2047 coding
2690 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2692 NM/03 Allow underscore in dnslist lookups
2693 Fixes bug 1026. Patch from Graeme Fowler
2695 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2696 Code patches from Adam Ciarcinski of NetBSD.
2698 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2701 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2702 is logged, to avoid truncation. Patch from John Horne.
2704 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2705 Patch from Jakob Hirsch.
2707 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2708 SQL string expansion failure details.
2709 Patch from Andrey Oktyabrski.
2711 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2712 Patch from Simon Arlott.
2714 PP/09 New lookups functionality failed to compile on old gcc which rejects
2715 extern declarations in function scope.
2716 Patch from Oliver Fleischmann
2718 PP/10 Use sig_atomic_t for flags set from signal handlers.
2719 Check getgroups() return and improve debugging.
2720 Fixed developed for diagnosis in bug 927 (which turned out to be
2723 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2724 Patch from Mark Zealey.
2726 PP/12 Bugzilla 1056: Improved spamd server selection.
2727 Patch from Mark Zealey.
2729 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2730 Based on patch from Heiko Schlittermann.
2732 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2733 Patch from Uwe Doering, sign-off by Michael Haardt.
2735 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2736 variable declaration deep within a block. Bug and patch from
2739 PP/15 lookups-Makefile IRIX compatibility coercion.
2741 PP/16 Make DISABLE_DKIM build knob functional.
2743 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2744 Patch by Simon Arlott
2746 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2747 variable argument macros. Our copy now differs from upstream.
2753 TF/01 Failure to get a lock on a hints database can have serious
2754 consequences so log it to the panic log.
2756 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2757 controlled using the smtp_confirmation log selector.
2759 TF/03 Include the error message when we fail to unlink a spool file.
2761 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2762 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2763 for maintaining out-of-tree patches for some time.
2765 PP/01 Bugzilla 139: Documentation and portability issues.
2766 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2767 Handle per-OS dynamic-module compilation flags.
2769 PP/02 Let /dev/null have normal permissions.
2770 The 4.73 fixes were a little too stringent and complained about the
2771 permissions on /dev/null. Exempt it from some checks.
2772 Reported by Andreas M. Kirchwitz.
2774 PP/03 Report version information for many libraries, including
2775 Exim version information for dynamically loaded libraries. Created
2776 version.h, now support a version extension string for distributors
2777 who patch heavily. Dynamic module ABI change.
2779 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2780 privilege escalation vulnerability whereby the Exim run-time user
2781 can cause root to append content of the attacker's choosing to
2784 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2787 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2788 If dropping privileges for untrusted macros, we disabled normal logging
2789 on the basis that it would fail; for the Exim run-time user, this is not
2790 the case, and it resulted in successful deliveries going unlogged.
2791 Fixed. Reported by Andreas Metzler.
2797 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2798 only prepend for the Resent-* case. Fixes regression introduced in
2799 Exim 4.70 by NM/22 for Bugzilla 607.
2801 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2802 increasing numbers of administrators be bitten by this.
2804 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2806 PP/03 Bugzilla 994: added openssl_options main configuration option.
2808 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2810 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2812 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2814 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2815 without a peer certificate, leading to a segfault because of an
2816 assumption that peers always have certificates. Be a little more
2817 paranoid. Problem reported by Martin Tscholak.
2819 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2820 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2821 NB: ClamAV planning to remove STREAM in "middle of 2010".
2822 CL also introduces -bmalware, various -d+acl logging additions and
2823 more caution in buffer sizes.
2825 PP/09 Implemented reverse_ip expansion operator.
2827 PP/10 Bugzilla 937: provide a "debug" ACL control.
2829 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2831 PP/12 Bugzilla 973: Implement --version.
2833 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2835 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2837 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2839 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2840 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2841 ignore trailing whitespace.
2843 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2845 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2848 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2849 Notification from Dr Andrew Aitchison.
2851 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2852 ExtendedDetectionInfo response format.
2853 Notification from John Horne.
2855 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2858 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2859 XSL and documented dependency on system catalogs, with examples of how
2862 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2865 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2866 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2867 configuration file which is writeable by the Exim user or group.
2869 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2870 of configuration files to cover files specified with the -C option if
2871 they are going to be used with root privileges, not just the default
2874 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2875 option (effectively making it always true).
2877 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2878 files to be used while preserving root privileges.
2880 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2881 that rogue child processes cannot use them.
2883 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2884 run-time user, instead of root.
2886 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2887 Exim run-time user without dropping privileges.
2889 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2890 result string, instead of calling string_vformat() twice with the same
2893 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2894 for other users. Others should always drop root privileges if they use
2895 -C on the command line, even for a whitelisted configure file.
2897 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2899 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2905 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2906 $data_path, and $header_path variables; fixed documentation bugs and
2909 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2910 exipick to access non-standard spools, including the "frozen" queue
2913 NM/01 Bugzilla 965: Support mysql stored procedures.
2914 Patch from Alain Williams
2916 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2918 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2919 Patch from Andreas Metzler
2921 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2922 Patch from Kirill Miazine
2924 NM/05 Bugzilla 671: Added umask to procmail example.
2926 JJ/03 installed exipick 20100323.0, fixing doc bug
2928 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2929 directory. Notification and patch from Dan Rosenberg.
2931 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2933 TK/02 Improve log output when DKIM signing operation fails.
2935 MH/01 Treat the transport option dkim_domain as a colon separated
2936 list, not as a single string, and sign the message with each element,
2937 omitting multiple occurences of the same signer.
2939 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2940 Bugzilla 985, 986. Patch by Simon Arlott
2942 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2943 Patch by Simon Arlott
2945 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2946 MBX locking. Notification from Dan Rosenberg.
2952 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2954 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2956 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2958 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2960 NM/04 Bugzilla 915: Flag broken perl installation during build.
2966 TK/01 Added patch by Johannes Berg that expands the main option
2967 "spamd_address" if it starts with a dollar sign.
2969 TK/02 Write list of recipients to X-Envelope-Sender header when building
2970 the mbox-format spool file for content scanning (suggested by Jakob
2973 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2974 (http://www.dcc-servers.net/) support via dccifd. Activated by
2975 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2977 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2978 by Mark Daniel Reidel <mr@df.eu>.
2980 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2981 When building exim an external PCRE library is now needed -
2982 PCRE is a system library on the majority of modern systems.
2983 See entry on PCRE_LIBS in EDITME file.
2985 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2986 conversation. Added nologin parameter to request.
2987 Patch contributed by Kirill Miazine.
2989 TF/01 Do not log submission mode rewrites if they do not change the address.
2991 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2993 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2994 log files in place. Contributed by Roberto Lima.
2996 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2998 TF/03 Bugzilla 615: When checking the local_parts router precondition
2999 after a local_part_suffix or local_part_prefix option, Exim now
3000 does not use the address's named list lookup cache, since this
3001 contains cached lookups for the whole local part.
3003 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3004 Robert Millan. Documentation is in experimental-spec.txt.
3006 TF/04 Bugzilla 668: Fix parallel build (make -j).
3008 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3010 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3011 Patch provided by Jan Srzednicki.
3013 TF/05 Leading white space used to be stripped from $spam_report which
3014 wrecked the formatting. Now it is preserved.
3016 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3017 that they are available at delivery time.
3019 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3021 TF/08 TLS error reporting now respects the incoming_interface and
3022 incoming_port log selectors.
3024 TF/09 Produce a more useful error message if an SMTP transport's hosts
3025 setting expands to an empty string.
3027 NM/06 Bugzilla 744: EXPN did not work under TLS.
3028 Patch provided by Phil Pennock.
3030 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3031 Patch provided by Richard Godbee.
3033 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3034 acl_smtp_notquit, added index entry.
3036 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3037 Patch provided by Eugene Bujak.
3039 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3040 accept(). Patch provided by Maxim Dounin.
3042 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3043 Patch provided by Phil Pennock.
3045 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3047 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3048 Patch provided by Brad "anomie" Jorsch.
3050 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3051 Patch provided by Dean Brooks.
3053 TK/05 Add native DKIM support (does not depend on external libraries).
3055 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3056 Patch provided by Graeme Fowler.
3058 NM/16 Bugzilla 851: Documentation example syntax fix.
3060 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3062 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3065 NM/19 Bugzilla 745: TLS version reporting.
3066 Patch provided by Phil Pennock.
3068 NM/20 Bugzilla 167: bool: condition support.
3069 Patch provided by Phil Pennock.
3071 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3072 clients. Patch provided by Phil Pennock.
3074 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3075 Patch provided by Brad "anomie" Jorsch.
3077 NM/23 Bugzilla 687: Fix misparses in eximstats.
3078 Patch provided by Heiko Schlittermann.
3080 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3081 Patch provided by Heiko Schlittermann.
3083 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3084 plus update to original patch.
3086 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3088 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3089 Patch provided by David Brownlee.
3091 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3093 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3095 NM/30 Bugzilla 888: TLS documentation bugfixes.
3097 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3099 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3100 Unlike the original bugzilla I have changed all shell scripts in src tree.
3102 NM/33 Bugzilla 898: Transport filter timeout fix.
3103 Patch by Todd Rinaldo.
3105 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3106 Patch by Serge Demonchaux.
3108 NM/35 Bugzilla 39: Base64 decode bug fixes.
3109 Patch by Jakob Hirsch.
3111 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3113 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3115 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3117 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3123 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3124 ALT-N's libdkim that I have put here:
3125 http://duncanthrax.net/exim-experimental/
3127 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3128 were called 'true' and it seems that C99 defines that as a reserved
3129 keyword to be used with 'bool' variable types. That means you could
3130 not include C99-style headers which use bools without triggering
3131 build errors in sieve.c.
3133 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3134 as mailq or other aliases. Changed the --help handling significantly
3135 to do whats expected. exim_usage() emits usage/help information.
3137 SC/01 Added the -bylocaldomain option to eximstats.
3139 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3141 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3143 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3149 PH/01 Another patch from the Sieve maintainer.
3151 PH/02 When an IPv6 address is converted to a string for single-key lookup
3152 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3153 dots are used instead of colons so that keys in lsearch files need not
3154 contain colons. This was done some time before quoting was made available
3155 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3156 (notated using the quote facility) so as to distinguish them from IPv4
3157 keys. This meant that lookups for IP addresses in host lists did not work
3158 for iplsearch lookups.
3160 This has been fixed by arranging for IPv6 addresses to be expressed with
3161 colons if the lookup type is iplsearch. This is not incompatible, because
3162 previously such lookups could never work.
3164 The situation is now rather anomalous, since one *can* have colons in
3165 ordinary lsearch keys. However, making the change in all cases is
3166 incompatible and would probably break a number of configurations.
3168 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3171 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3172 conversion specification without a maximum field width, thereby enabling
3173 a rogue spamd server to cause a buffer overflow. While nobody in their
3174 right mind would setup Exim to query an untrusted spamd server, an
3175 attacker that gains access to a server running spamd could potentially
3176 exploit this vulnerability to run arbitrary code as the Exim user.
3178 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3179 $primary_hostname instead of what libspf2 thinks the hosts name is.
3181 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3182 a directory entry by the name of the lookup key. Previously, if a
3183 symlink pointed to a non-existing file or a file in a directory that
3184 Exim lacked permissions to read, a lookup for a key matching that
3185 symlink would fail. Now it is enough that a matching directory entry
3186 exists, symlink or not. (Bugzilla 503.)
3188 PH/03 The body_linecount and body_zerocount variables are now exported in the
3191 PH/04 Added the $dnslist_matched variable.
3193 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3194 This means they are set thereafter only if the connection becomes
3197 PH/06 Added the client_condition to authenticators so that some can be skipped
3198 by clients under certain conditions.
3200 PH/07 The error message for a badly-placed control=no_multiline_responses left
3201 "_responses" off the end of the name.
3203 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3205 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3206 (without spaces) instead of just copying the configuration text.
3208 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3210 PH/11 Added $max_received_linelength.
3212 PH/12 Added +ignore_defer and +include_defer to host lists.
3214 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3215 way in which PCRE > 7.0 is built.
3217 PH/14 Implemented queue_only_load_latch.
3219 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3220 MAIL command. The effect was to mangle the value on 64-bit systems.
3222 PH/16 Another patch from the Sieve maintainer.
3224 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3226 PH/18 If a system quota error occurred while trying to create the file for
3227 a maildir delivery, the message "Mailbox is full" was not appended to the
3228 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3229 only to a quota excession during the actual writing of the file.
3231 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3232 characters?) which causes problems in log lines. The DN values are now
3233 passed through string_printing() before being added to log lines.
3235 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3236 and InterBase are left for another time.)
3238 PH/21 Added message_body_newlines option.
3240 PH/22 Guard against possible overflow in moan_check_errorcopy().
3242 PH/23 POSIX allows open() to be a macro; guard against that.
3244 PH/24 If the recipient of an error message contained an @ in the local part
3245 (suitably quoted, of course), incorrect values were put in $domain and
3246 $local_part during the evaluation of errors_copy.
3252 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3253 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3256 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3257 issue a MAIL command.
3259 PH/02 In an ACL statement such as
3261 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3263 if a client was not listed at all, or was listed with a value other than
3264 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3265 the condition was not true (as it should be), so access was not denied.
3266 The bug was that the ! inversion was incorrectly passed on to the second
3267 item. This has been fixed.
3269 PH/03 Added additional dnslists conditions == and =& which are different from
3270 = and & when the dns lookup returns more than one IP address.
3272 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3273 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3275 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3276 FSYNC, which compiles an option called disable_fsync that allows for
3277 bypassing fsync(). The documentation is heavily laced with warnings.
3279 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3281 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3282 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3283 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3284 including adding "make clean"; (3) Added -fPIC when compiling the test
3285 dynamically loaded module, to get rid of a warning.
3287 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3288 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3289 = 0s. The bug is otherwise harmless.
3291 PH/07 There was a bug in the dovecot authenticator such that the value of
3292 $auth1 could be overwritten, and so not correctly preserved, after a
3293 successful authentication. This usually meant that the value preserved by
3294 the server_setid option was incorrect.
3296 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3298 PH/09 Installed PCRE release 7.0.
3300 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3301 run for batched SMTP input. It is now run at the start of every message
3302 in the batch. While fixing this I discovered that the process information
3303 (output by running exiwhat) was not always getting set for -bs and -bS
3304 input. This is fixed, and it now also says "batched" for BSMTP.
3306 PH/11 Added control=no_pipelining.
3308 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3309 patch, slightly modified), and move the expansion of helo_data till after
3310 the connection is made in the smtp transport (so it can use these
3313 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3315 PH/14 Added log_selector = +pid.
3317 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3319 PH/16 Add ${if forany and ${if forall.
3321 PH/17 Added dsn_from option to vary the From: line in DSNs.
3323 PH/18 Flush SMTP output before performing a callout, unless control =
3324 no_callout_flush is set.
3326 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3327 was true (the default) a successful delivery failed to delete the retry
3328 item, thus causing premature timeout of the address. The bug is now
3331 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3333 PH/21 Long custom messages for fakedefer and fakereject are now split up
3334 into multiline responses in the same way that messages for "deny" and
3335 other ACL rejections are.
3337 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3338 with slight modification.
3340 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3341 draft, changing the syntax and factoring some duplicate code".
3343 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3344 for deliveries of the second and subsequent messages over the same SMTP
3347 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3348 ${reduce, with only minor "tidies".
3350 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3352 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3353 expansion side effects.
3355 PH/27 When a message times out after an over-quota error from an Exim-imposed
3356 quota, the bounce message says "mailbox is full". This message was not
3357 being given when it was a system quota that was exceeded. It now should
3360 MH/03 Made $recipients available in local_scan(). local_scan() already has
3361 better access to the recipient list through recipients_list[], but
3362 $recipients can be useful in postmaster-provided expansion strings.
3364 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3365 in the case of a MAIL command with additional options following the
3366 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3367 were accidentally chopped off.
3369 PH/29 SMTP synchronization checks are implemented when a command is read -
3370 there is a check that no more input is waiting when there shouldn't be
3371 any. However, for some commands, a delay in an ACL can mean that it is
3372 some time before the response is written. In this time, more input might
3373 arrive, invalidly. So now there are extra checks after an ACL has run for
3374 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3375 pipelining has not been advertised.
3377 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3379 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3380 returned in the SMTP response when smtp_return_error_details was false.
3381 This has been fixed.
3383 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3384 instead of the C I/O that was originally supplied, because problems were
3385 reported on Solaris.
3387 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3388 Exim which did not show up earlier: it was assuming that a call to
3389 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3390 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3391 was a macro that became an assignment, so it seemed to work. This has
3392 changed to a proper function call with a void return, hence the compile
3393 error. Exim's code has been fixed.
3395 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3398 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3400 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3402 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3403 because it turns out to be too big or there is a timeout) there is no
3404 "Completed" line in the log. When some messages of this type were
3405 selected by exigrep, they were listed as "not completed". Others were
3406 picked up by some special patterns. I have improved the selection
3407 criteria to be more general.
3409 PH/38 The host_find_failed option in the manualroute router can now be set
3410 to "ignore", to completely ignore a host whose IP address cannot be
3411 found. If all hosts are ignored, the behaviour is controlled by the new
3412 host_all_ignored option.
3414 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3415 homing or because of multiple MX records with /mx) generated more than
3416 one IP address, and the following item turned out to be the local host,
3417 all the secondary addresses of the first item were incorrectly removed
3418 from the list, along with the local host and any following hosts (which
3419 is what is supposed to happen).
3421 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3422 whoever called Exim into the -H file. In the case of the daemon it was
3423 behaving confusingly. When first started, it used values for whoever
3424 started the daemon, but after a SIGHUP it used the Exim user (because it
3425 calls itself on a restart). I have changed the code so that it now always
3428 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3429 message are rejected with the same error (e.g. no authentication or bad
3430 sender address), and a DATA command is nevertheless sent (as can happen
3431 with PIPELINING or a stupid MUA), the error message that was given to the
3432 RCPT commands is included in the rejection of the DATA command. This is
3433 intended to be helpful for MUAs that show only the final error to their
3436 PH/42 Another patch from the Sieve maintainer.
3438 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3439 Eximstats - Fixed some broken HTML links and added missing column headers
3441 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3442 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3444 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3446 PH/43 Yet another patch from the Sieve maintainer.
3448 PH/44 I found a way to check for a TCP/IP connection going away before sending
3449 the response to the final '.' that terminates a message, but only in the
3450 case where the client has not sent further data following the '.'
3451 (unfortunately, this is allowed). However, in many cases there won't be
3452 any further data because there won't be any more messages to send. A call
3453 to select() can be used: if it shows that the input is "ready", there is
3454 either input waiting, or the socket has been closed. An attempt to read
3455 the next input character can distinguish the two cases. Previously, Exim
3456 would have sent an OK response which the client would never have see.
3457 This could lead to message repetition. This fix should cure that, at
3458 least in a lot of common cases.
3460 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3461 advertised in response to EHLO.
3467 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3468 fixed by 4.65/MH/01 (is this a record?) are fixed:
3470 (i) An empty string was always treated as zero by the numeric comparison
3471 operators. This behaviour has been restored.
3473 (ii) It is documented that the numeric comparison operators always treat
3474 their arguments as decimal numbers. This was broken in that numbers
3475 starting with 0 were being interpreted as octal.
3477 While fixing these problems I realized that there was another issue that
3478 hadn't been noticed. Values of message_size_limit (both the global option
3479 and the transport option) were treated as octal if they started with 0.
3480 The documentation was vague. These values are now always treated as
3481 decimal, and I will make that clear in the documentation.
3487 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3488 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3491 MH/01 Don't check that the operands of numeric comparison operators are
3492 integers when their expansion is in "skipping" mode (fixes bug
3493 introduced by 4.64-PH/07).
3495 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3496 child addresses, Exim now panics and dies. Previously, because the count
3497 is held in a short int, deliveries were likely to be lost. As such a
3498 large number of recipients for a single message is ridiculous
3499 (performance will be very, very poor), I have chosen to impose a limit
3500 rather than extend the field.
3506 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3507 leftover -K file (the existence of which was triggered by #402).
3508 While we were at it, introduced process PID as part of the -K
3509 filename. This should rule out race conditions when creating
3512 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3513 processing considerably. Previous code took too long for large mails,
3514 triggering a timeout which in turn triggers #401.
3516 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3517 in the DK code in transports.c. sendfile() is not really portable,
3518 hence the _LINUX specificness.
3520 TF/01 In the add_headers option to the mail command in an Exim filter,
3521 there was a bug that Exim would claim a syntax error in any
3522 header after the first one which had an odd number of characters
3525 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3526 callout verification, Exim cached a "reject" for the entire domain. This
3527 is correct for most verifications, but it is not correct for a recipient
3528 verification with use_sender or use_postmaster set, because in that case
3529 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3530 case of MAIL FROM:<> rejection from other early rejections (e.g.
3531 rejection of HELO). When verifying a recipient using a non-null MAIL
3532 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3533 Whatever the result of the callout, the value of the domain cache is
3534 left unchanged (for any other kind of callout, getting as far as trying
3535 RCPT means that the domain itself is ok).
3537 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3540 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3541 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3542 session, the daemon ignores ECONNECT errors and logs others; it now
3543 ignores EPIPE as well.
3545 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3546 (quoted-printable decoding).
3548 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3549 later the small subsequent patch to fix an introduced bug.
3551 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3553 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3555 PH/08 An error is now given if message_size_limit is specified negative.
3557 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3558 to be given (somewhat) arbitrary names.
3560 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3563 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3564 miscellaneous code fixes
3566 PH/10 Added the log_reject_target ACL modifier to specify where to log
3569 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3570 hostname. This is wrong, because it relates to the incoming message (and
3571 probably the interface on which it is arriving) and not to the outgoing
3572 callout (which could be using a different interface). This has been
3573 changed to use the value of the helo_data option from the smtp transport
3574 instead - this is what is used when a message is actually being sent. If
3575 there is no remote transport (possible with a router that sets up host
3576 addresses), $smtp_active_hostname is used.
3578 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3579 tweaks were necessary in order to get it to work (see also 21 below):
3580 (a) The code assumed that strncpy() returns a negative number on buffer
3581 overflow, which isn't the case. Replaced with Exim's string_format()
3583 (b) There were several signed/unsigned issues. I just did the minimum
3584 hacking in of casts. There is scope for a larger refactoring.
3585 (c) The code used strcasecmp() which is not a standard C function.
3586 Replaced with Exim's strcmpic() function.
3587 (d) The code set only $1; it now sets $auth1 as well.
3588 (e) A simple test gave the error "authentication client didn't specify
3589 service in request". It would seem that Dovecot has changed its
3590 interface. Fortunately there's a specification; I followed it and
3591 changed what the client sends and it appears to be working now.
3593 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3596 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3597 address A is aliased to B and C, where B exists and C does not. Without
3598 -v the output is "A verified" because verification stops after a
3599 successful redirection if more than one address is generated. However,
3600 with -v the child addresses are also verified. Exim was outputting "A
3601 failed to verify" and then showing the successful verification for C,
3602 with its parentage. It now outputs "B failed to verify", showing B's
3603 parentage before showing the successful verification of C.
3605 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3606 look up a TXT record in a specific list after matching in a combined
3609 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3610 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3611 they consult the DNS. I had assumed they would set it the way they
3612 wanted; and indeed my experiments on Linux seem to show that in some
3613 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3614 To be on the safe side, however, I have now made the interface to
3615 host_find_byname() similar to host_find_bydns(), with an argument
3616 containing the DNS resolver options. The host_find_byname() function now
3617 sets these options at its start, just as host_find_bydns() does. The smtp
3618 transport options dns_qualify_single and dns_search_parents are passed to
3619 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3620 of host_find_byname() use the default settings of RES_DEFNAMES
3621 (qualify_single) but not RES_DNSRCH (search_parents).
3623 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3624 spool_read_header() do less string testing, by means of a preliminary
3625 switch on the second character of optional "-foo" lines. (This is
3626 overdue, caused by the large number of possibilities that now exist.
3627 Originally there were few.) While I was there, I also converted the
3628 str(n)cmp tests so they don't re-test the leading "-" and the first
3629 character, in the hope this might squeeze out yet more improvement.
3631 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3632 flag allowing group syntax was set by the header_syntax check but not
3633 turned off, possible causing trouble later; (2) The flag was not being
3634 set at all for the header_verify test, causing "group"-style headers to
3635 be rejected. I have now set it in this case, and also caused header_
3636 verify to ignore an empty address taken from a group. While doing this, I
3637 came across some other cases where the code for allowing group syntax
3638 while scanning a header line wasn't quite right (mostly, not resetting
3639 the flag correctly in the right place). These bugs could have caused
3640 trouble for malformed header lines. I hope it is now all correct.
3642 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3643 with the "reply" argument non-NULL. The code, however (which originally
3644 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3645 but it didn't always do it. This confused somebody who was copying the
3646 code for some other use. I have removed all the tests.
3648 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3649 feature that was used to support insecure browsers during the U.S. crypto
3650 embargo. It requires special client support, and Exim is probably the
3651 only MTA that supported it -- and would never use it because real RSA is
3652 always available. This code has been removed, because it had the bad
3653 effect of slowing Exim down by computing (never used) parameters for the
3654 RSA_EXPORT functionality.
3656 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3657 authenticator to fail if there's a tab character in the incoming data
3658 (there should never be unless someone is messing about, as it's supposed
3659 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3660 if the connection is using TLS or if the remote IP is the same as the
3661 local IP, and the "valid-client-cert option" if a client certificate has
3664 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3665 authenticators. This can be used for authorization after authentication
3666 succeeds. (In the case of plaintext, it servers for both authentication
3669 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3670 if any retry times were supplied.
3672 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3673 connection, where there is no client IP address to check. In this
3674 situation, the verify now always succeeds.
3676 PH/25 Applied John Jetmore's -Mset patch.
3678 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3680 PH/27 In a string expansion for a processed (not raw) header when multiple
3681 headers of the same name were present, leading whitespace was being
3682 removed from all of them, but trailing whitespace was being removed only
3683 from the last one. Now trailing whitespace is removed from each header
3684 before concatenation. Completely empty headers in a concatenation (as
3685 before) are ignored.
3687 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3688 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3690 PH/29 [Removed. This was a change that I later backed out, and forgot to
3691 correct the ChangeLog entry (that I had efficiently created) before
3692 committing the later change.]
3694 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3695 address errors (4xx response to RCPT) over the same connection as other
3696 messages routed to the same hosts. Such deliveries are always "forced",
3697 so retry times are not inspected. This resulted in far too many retries
3698 for the affected addresses. The effect occurred only when there were more
3699 hosts than the hosts_max_try setting in the smtp transport when it had
3700 the 4xx errors. Those hosts that it had tried were not added to the list
3701 of hosts for which the message was waiting, so if all were tried, there
3702 was no problem. Two fixes have been applied:
3704 (i) If there are any address or message errors in an SMTP delivery, none
3705 of the hosts (tried or untried) are now added to the list of hosts
3706 for which the message is waiting, so the message should not be a
3707 candidate for sending over the same connection that was used for a
3708 successful delivery of some other message. This seems entirely
3709 reasonable: after all the message is NOT "waiting for some host".
3710 This is so "obvious" that I'm not sure why it wasn't done
3711 previously. Hope I haven't missed anything, but it can't do any
3712 harm, as the worst effect is to miss an optimization.
3714 (ii) If, despite (i), such a delivery is accidentally attempted, the
3715 routing retry time is respected, so at least it doesn't keep
3716 hammering the server.
3718 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3719 in ${readsocket because some servers need this prod.
3721 PH/32 Added some extra debug output when updating a wait-xxx database.
3723 PH/33 The hint "could be header name not terminated by colon", which has been
3724 given for certain expansion errors for a long time, was not being given
3725 for the ${if def:h_colon_omitted{... case.
3727 PH/34 The spec says: "With one important exception, whenever a domain list is
3728 being scanned, $domain contains the subject domain." There was at least
3729 one case where this was not true.
3731 PH/35 The error "getsockname() failed: connection reset by peer" was being
3732 written to the panic log as well as the main log, but it isn't really
3733 panic-worthy as it just means the connection died rather early on. I have
3734 removed the panic log writing for the ECONNRESET error when getsockname()
3737 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3738 runs only) independently of the message's sender address. This meant
3739 that, if the 4xx error was in fact related to the sender, a different
3740 message to the same recipient with a different sender could confuse
3741 things. In particular, this can happen when sending to a greylisting
3742 server, but other circumstances could also provoke similar problems.
3743 I have changed the default so that the retry time for these errors is now
3744 based a combination of the sender and recipient addresses. This change
3745 can be overridden by setting address_retry_include_sender=false in the
3748 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3749 remote server are returned as part of bounce messages. This was not
3750 happening for LMTP over a pipe (the lmtp transport), but now it is the
3751 same for both kinds of LMTP.
3753 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3754 in header lines that were in fact CNAMEs. This is no longer the case.
3756 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3757 and queue runs started by the daemon processed all messages. This has
3758 been fixed so that -R and -S can now usefully be given with -q<time>.
3760 PH/40 Import PCRE release 6.7 (fixes some bugs).
3762 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3764 PH/42 Give an error if -q is specified more than once.
3766 PH/43 Renamed the variables $interface_address and $interface_port as
3767 $received_ip_address and $received_port, to make it clear that these
3768 values apply to message reception, and not to the outgoing interface when
3769 a message is delivered. (The old names remain recognized, of course.)
3771 PH/44 There was no timeout on the connect() call when using a Unix domain
3772 socket in the ${readsocket expansion. There now is.
3774 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3775 be meaningful with "accept".
3777 SC/01 Eximstats V1.43
3778 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3780 SC/02 Eximstats V1.44
3781 Use a glob alias rather than an array ref in the generated
3782 parser. This improves both readability and performance.
3784 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3785 Collect SpamAssassin and rejection statistics.
3786 Don't display local sender or destination tables unless
3787 there is data to show.
3788 Added average volumes into the top table text output.
3790 SC/04 Eximstats V1.46
3791 Collect data on the number of addresses (recipients)
3792 as well as the number of messages.
3794 SC/05 Eximstats V1.47
3795 Added 'Message too big' to the list of mail rejection
3796 reasons (thanks to Marco Gaiarin).
3798 SC/06 Eximstats V1.48
3799 Mainlog lines which have GMT offsets and are too short to
3800 have a flag are now skipped.
3802 SC/07 Eximstats V1.49 (Alain Williams)
3803 Added the -emptyok flag.
3805 SC/08 Eximstats V1.50
3806 Fixes for obtaining the IP address from reject messages.
3808 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3809 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3810 whitespace changes from 4.64-PH/27
3812 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3815 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3816 are found, allow negative numbers in numeric criteria)
3818 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3820 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3821 to match changes made in 4.64-PH/43
3823 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3825 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3826 "domain" is actually an IP address, and force a failure. This locks out
3827 those revolvers/nameservers that support "A-for-A" lookups, in
3828 contravention of the specifications.
3830 PH/48 When a host name was looked up from an IP address, and the subsequent
3831 forward lookup of the name timed out, the host name was left in
3832 $sender_host_name, contrary to the specification.
3834 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3835 restricted to single-key lookups, Exim was not diagnosing an error if
3836 * or *@ was used with a query-style lookup.
3838 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3840 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3841 long ago, but noone interested enough thought of it. Let's just say that
3842 the "1.1" means that there are some new functions that weren't there at
3843 some point in the past.
3845 PH/51 Error processing for expansion failure of helo_data from an smtp
3846 transport during callout processing was broken.
3848 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3849 tested/used via the -bh/-bhc/-bs options.
3851 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3852 bug, fixed in subsequent PCRE releases).
3854 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3855 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3857 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3863 SC/01 Use a glob alias rather than an array ref in eximstats generated
3864 parser. This improves both readability and performance.
3866 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3867 Don't display local sender or destination tables in eximstats unless
3868 there is data to show.
3869 Added average volumes into the eximstats top table text output.
3871 SC/03 Collect data on the number of addresses (recipients) as well
3872 as the number of messages in eximstats.
3874 TF/01 Correct an error in the documentation for the redirect router. Exim
3875 does (usually) call initgroups() when daemonizing.
3877 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3878 with consistent privilege compared to when running as a daemon.
3880 TF/03 Note in the spec that $authenticated_id is not set for local
3881 submissions from trusted users.
3883 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3884 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3886 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3887 by adding some example configuration directives to the default
3888 configuration file. A little bit of work is required to uncomment the
3889 directives and define how usernames and passwords are checked, but
3890 there is now a framework to start from.
3892 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3893 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3894 without this. I don't know how relevant this is to other LDAP libraries.
3896 PH/02 Add the verb name to the "unknown ACL verb" error.
3898 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3900 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3902 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3903 directory not expanded when it should be if an expanded home directory
3904 was set for the address (which is overridden by the transport).
3906 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3909 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3910 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3911 because it is too late at that time, and has no effect.
3913 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3914 security issue with \' (bugzilla #107). I could not use the
3915 PQescapeStringConn() function, because it needs a PGconn value as one of
3918 PH/08 When testing addresses using -bt, indicate those final addresses that
3919 are duplicates that would not cause an additional delivery. At least one
3920 person was confused, thinking that -bt output corresponded to deliveries.
3921 (Suppressing duplicates isn't a good idea as you lose the information
3922 about possibly different redirections that led to the duplicates.)
3924 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3925 systems where poll() doesn't work, in particular OS X.
3927 PH/10 Added more information to debugging output for retry time not reached.
3929 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3930 operations in malware.c.
3932 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3935 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3936 syslog_duplication set false, log lines that would normally be written
3937 both the the main log and to the reject log were not written to syslog at
3940 PH/14 In the default configuration, change the use of "message" in ACL warn
3941 statements to "add_header".
3943 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3944 not followed by a command (e.g. "seen endif").
3946 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3947 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3950 PH/17 Added extra conditions to the default value of delay_warning_condition
3954 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3955 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3956 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3959 The Auto-Submitted: and various List- headers are standardised, whereas I
3960 don't think Precedence: ever was.
3962 PH/18 Refactored debugging code in route_finduser() to show more information,
3963 in particular, the error code if getpwnam() issues one.
3965 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3966 This is apparently needed in addition to the PH/07 change above to avoid
3967 any possible encoding problems.
3969 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3970 but not after initializing Perl.
3972 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3973 output them only if debugging. By default they are written stderr,
3974 apparently, which is not desirable.
3976 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3979 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3982 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3984 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3985 authenticated or an ident call has been made. Suppress the default
3986 values for $authenticated_id and $authenticated_sender (but permit -oMai
3987 and -oMas) when testing with -bh.
3989 PH/24 Re-jigged the order of the tests in the default configuration so that the
3990 tests for valid domains and recipients precede the DNS black list and CSA
3991 tests, on the grounds that those ones are more expensive.
3993 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3994 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3995 This bug exists in every version of Exim that I still have, right back to
3998 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3999 However, an attempt to turn on case-sensitivity in a regex key by
4000 including (?-i) didn't work because the subject string was already
4001 lowercased, and the effects were non-intuitive. It turns out that a
4002 one-line patch can be used to allow (?-i) to work as expected.
4008 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4009 other effects) broke the use of negated acl sub-conditions.
4011 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4014 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4015 "Deny" causes Exim to reject the incoming connection with a 554 error.
4016 Unfortunately, if there is a major crisis, such as a disk failure,
4017 tcp-wrappers gives "deny", whereas what one would like would be some
4018 kind of temporary error. A kludge has been added to help with this.
4019 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4020 554 error is used if errno is still zero or contains ENOENT (which occurs
4021 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4024 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4026 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4027 errors. Otherwise a message that provokes a temporary error (when other
4028 messages do not) can cause a whole host to time out.
4030 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4031 addresses were routed directly to files or pipes from a redirect router.
4032 File deliveries just didn't batch; pipe deliveries might have suffered
4035 PH/06 A failure to get a lock for a hints database would erroneously always say
4036 "Failed to get write lock", even when it was really a read lock.
4038 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4039 of 0600. This has been changed to use the value of the lockfile_mode
4040 option (which defaults to 0600).
4042 PH/08 Applied small patch from the Sieve maintainer.
4044 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4045 folder from quota calculations, a direct delivery into this folder messed
4046 up the contents of the maildirsize file. This was because the regex was
4047 used only to exclude .Trash (or whatever) when the size of the mailbox
4048 was calculated. There was no check that a delivery was happening into an
4049 excluded directory. This bug has been fixed by ignoring all quota
4050 processing for deliveries into excluded directories.
4052 PH/10 Added the maildirfolder_create_regex option to appendfile.
4058 PH/01 The code for finding all the local interface addresses on a FreeBSD
4059 system running IPv6 was broken. This may well have applied to all BSD
4060 systems, as well as to others that have similar system calls. The broken
4061 code found IPv4 interfaces correctly, but gave incorrect values for the
4062 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4063 that it would not match correctly against @[] and not recognize the IPv6
4066 PH/02 The ipliteral router was not recognizing addresses of the form user@
4067 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4069 PH/03 Added disable_ipv6.
4071 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4072 decoded form, because it is most often used to construct To: headers
4073 lines in autoreplies, and the decoded form may well be syntactically
4074 invalid. However, $reply_address has leading white space removed, and all
4075 newlines turned into spaces so that the autoreply transport does not
4078 PH/05 If group was specified without a user on a router, and no group or user
4079 was specified on a transport, the group from the router was ignored.
4081 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4082 for visible compile-time settings that can be used to change these
4083 numbers, for those that want even more. Backwards compatibility with old
4084 spool files has been maintained. However, going back to a previous Exim
4085 release will lost any variables that are in spool files.
4087 PH/07 Two small changes when running in the test harness: increase delay when
4088 passing a TCP/IP connection to a new process, in case the original
4089 process has to generate a bounce, and remove special handling of
4090 127.0.0.2 (sic), which is no longer necessary.
4092 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4093 be the same on different OS.
4095 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4098 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4099 whether --show-vars was specified or not
4101 JJ/02 exipick: Added support for new ACL variable spool format introduced
4104 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4105 syntactically invalid From: or Reply-to: line, and a filter used this to
4106 generate an autoreply, and therefore failed to obtain an address for the
4107 autoreply, Exim could try to deliver to a non-existent relative file
4108 name, causing unrelated and misleading errors. What now happens is that
4109 it logs this as a hard delivery error, but does not attempt to create a
4112 PH/11 The exinext utility has a -C option for testing purposes, but although
4113 the given file was scanned by exinext itself; it wasn't being passed on
4114 when Exim was called.
4116 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4117 an end-of-file indication when reading a command response.
4119 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4120 compiled. In many other places in Exim, IPv6 addresses are always
4121 recognized, so I have changed this. It also means that IPv4 domain
4122 literals of the form [IPV4:n.n.n.n] are now always recognized.
4124 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4125 used if the router is not running as root, for example, when verifying at
4126 ACL time, or when using -bh. The debugging output from this situation was
4127 non-existent - all you got was a failure to exec. I have made two
4130 (a) Failures to set uid/gid, the current directory, or a process leader
4131 in a subprocess such as that created by queryprogram now generate
4132 suitable debugging output when -d is set.
4134 (b) The queryprogram router detects when it is not running as root,
4135 outputs suitable debugging information if -d is set, and then runs
4136 the subprocess without attempting to change uid/gid.
4138 PH/15 Minor change to Makefile for building test_host (undocumented testing
4141 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4142 additional section of a DNS packet that returns MX or SRV records.
4143 Instead, it always explicitly searches for A/AAAA records. This avoids
4144 major problems that occur when a DNS server includes only records of one
4145 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4146 fixed another bug: if SRV records were looked up and the corresponding
4147 address records were *not* found in the additional section, the port
4148 values from the SRV records were lost.
4150 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4151 using the correct key (the original address) when searching the retry
4152 rules in order to find which one to use for generating the retry hint.
4154 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4155 adding the default one. Similarly, if it contains a Reply-To: header, the
4156 errors_reply_to option, if set, is not used.
4158 PH/19 When calculating a retry time, Exim used to measure the "time since
4159 failure" by looking at the "first failed" field in the retry record. Now
4160 it does not use this if it is later than than the arrival time of the
4161 message. Instead it uses the arrival time. This makes for better
4162 behaviour in cases where some deliveries succeed, thus re-setting the
4163 "first failed" field. An example is a quota failure for a huge message
4164 when small messages continue to be delivered. Without this change, the
4165 "time since failure" will always be short, possible causing more frequent
4166 delivery attempts for the huge message than are intended.
4167 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4169 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4170 $1, $2, $3) because the numerical variables can be reset during some
4171 expansion items (e.g. "match"), thereby losing the authentication data.
4173 PH/21 Make -bV show the size of off_t variables so that the test suite can
4174 decide whether to run tests for quotas > 2G.
4176 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4177 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4178 filecount value is greater than 2G or if a quota value is greater than 2G
4179 on a system where the size of off_t is not greater than 4, a panic error
4182 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4183 never match. The debug and -bh output now contains an explicit error
4184 message indicating a malformed IPv4 address or mask.
4186 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4187 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4188 PH/23 above applies.
4190 PH/25 Do not write to syslog when running in the test harness. The only
4191 occasion when this arises is a failure to open the main or panic logs
4192 (for which there is an explicit test).
4194 PH/26 Added the /no_tell option to "control=freeze".
4196 PH/27 If a host name lookup failed very early in a connection, for example, if
4197 the IP address matched host_lookup and the reverse lookup yielded a name
4198 that did not have a forward lookup, an error message of the form "no IP
4199 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4200 could be logged. Now it outputs the IP address instead of "NULL".
4202 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4203 allows the sender and the authenticated sender to be set when
4204 submitting a message from within Exim. Since child_open_exim() is
4205 documented for local_scan(), the new function should be too.
4207 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4208 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4209 results in an empty string is now treated as unset.
4211 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4213 PH/31 Added sender_verify_fail logging option.
4215 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4216 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4217 tidied the source and removed it altogether.
4219 PH/33 When a queue run was abandoned because the load average was too high, a
4220 log line was always written; now it is written only if the queue_run log
4221 selector is set. In addition, the log line for abandonment now contains
4222 information about the queue run such as the pid. This is always present
4223 in "start" and "stop" lines but was omitted from the "abandon" line.
4225 PH/34 Omit spaces between a header name and the colon in the error message that
4226 is given when verify = headers_syntax fails (if there are lots of them,
4227 the message gets confusing).
4229 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4230 names, as there are now some PTR records that contain slashes. This check
4231 is only to protect against broken name servers that fall over on strange
4232 characters, so the fact that it applies to all lookups doesn't matter.
4234 PH/36 Now that the new test suite is complete, we can remove some of the
4235 special code in Exim that was needed for the old test suite. For example,
4236 sorting DNS records because real resolvers return them in an arbitrary
4237 order. The new test suite's fake resolver always returns records in the
4240 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4241 bounces) except when queue_only is set, to avoid logging races between
4242 the different processes.
4244 PH/38 Panic-die if .include specifies a non-absolute path.
4246 PH/39 A tweak to the "H" retry rule from its user.
4248 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4249 a label. They prevented compilation on older perls.
4251 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4252 a warning to be raised on newish perls.
4254 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4255 on queue. Changes to match documented behaviour of showing count of
4256 messages matching specified criteria.
4258 PH/40 Changed the default ident timeout from 30s to 5s.
4260 PH/41 Added support for the use of login_cap features, on those BSD systems
4261 that have them, for controlling the resources used by pipe deliveries.
4263 PH/42 The content-scanning code uses fopen() to create files in which to put
4264 message data. Previously it was not paying any attention to the mode of
4265 the files. Exim runs with umask(0) because the rest of the code creates
4266 files with open(), and sets the required mode explicitly. Thus, these
4267 files were ending up world-writeable. This was not a big issue, because,
4268 being within the spool directory, they were not world-accessible. I have
4269 created a function called modefopen, which takes an additional mode
4270 argument. It sets umask(777), creates the file, chmods it to the required
4271 mode, then resets the umask. All the relevant calls to fopen() in the
4272 content scanning code have been changed to use this function.
4274 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4275 to 24 hours. This avoids potential overflow problems when processing G
4276 and H retry rules. I suspect nobody ever tinkers with this value.
4278 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4280 PH/45 When the plaintext authenticator is running as a client, the server's
4281 challenges are checked to ensure they are valid base64 strings. By
4282 default, the authentication attempt is cancelled if an invalid string is
4283 received. Setting client_ignore_invalid_base64 true ignores these errors.
4284 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4285 they are received. Thus, the responses can be made to depend on the
4286 challenges. If an invalid string is ignored, an empty string is placed in
4289 PH/46 Messages that are created by the autoreply transport now contains a
4290 References: header, in accordance with RFCs 2822 and 3834.
4292 PH/47 Added authenticated_sender_force to the smtp transport.
4294 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4296 PH/49 Installed latest patch from the Sieve maintainer.
4298 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4299 was also set, the appendfile transport was unnecessarily scanning a
4300 directory of message files (e.g. for maildir delivery) to find the count
4301 of files (along with the size), even though it did not need this
4302 information. It now does the scan only if it needs to find either the
4303 size of the count of files.
4305 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4307 PH/52 Two bugs concerned with error handling when the smtp transport is
4310 (i) Exim was not creating retry information for temporary errors given
4311 for individual recipients after the DATA command when the smtp transport
4312 was used in LMTP mode. This meant that they could be retried too
4313 frequently, and not timed out correctly.
4315 (ii) Exim was setting the flag that allows error details to be returned
4316 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4317 recipients that were returned after the DATA command.
4319 PH/53 This is related to PH/52, but is more general: for any failing address,
4320 when detailed error information was permitted to be returned to the
4321 sender, but the error was temporary, then after the final timeout, only
4322 "retry timeout exceeded" was returned. Now it returns the full error as
4323 well as "retry timeout exceeded".
4325 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4326 do this, and (what is worse) MTAs that accept it.
4328 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4329 will now be deprecated.
4331 PH/56 New os.c-cygwin from the Cygwin maintainer.
4333 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4334 formats (previously only available in exim formats via -bpr, -bpru,
4335 and -bpra. Now also available in native and exiqgrep formats)
4337 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4338 with very large, slow to parse queues
4340 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4342 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4344 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4345 responses to authentication challenges, though it was showing the
4346 challenges; (ii) I've removed the CR characters from the debug output for
4349 PH/58 Allow for the insertion of a newline as well as a space when a string
4350 is turned into more than one encoded-word during RFC 2047 encoding. The
4351 Sieve code now uses this.
4353 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4354 data_4xx, lost_connection, tls_required.
4356 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4357 message was being sent as an SMTP response.
4359 PH/61 Add -l and -k options to exicyclog.
4361 PH/62 When verifying, if an address was redirected to one new address, so that
4362 verification continued, and the new address failed or deferred after
4363 having set something in $address_data, the value of $address_data was not
4364 passed back to the ACL. This was different to the case when no
4365 redirection occurred. The value is now passed back in both cases.
4367 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4368 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4369 use login_cap.h, so on its own it isn't the distinguishing feature. The
4370 new name refers directly to the setclassresources() function.
4372 PH/65 Added configuration files for NetBSD3.
4374 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4376 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4377 is preferred over IPv4.
4379 PH/68 The bounce_return_message and bounce_return_body options were not being
4380 honoured for bounces generated during the reception of non-SMTP messages.
4381 In particular, this applied to messages rejected by the ACL. This bug has
4382 been fixed. However, if bounce_return_message is true and bounce_return_
4383 body is false, the headers that are returned for a non-SMTP message
4384 include only those that have been read before the error was detected.
4385 (In the case of an ACL rejection, they have all been read.)
4387 PH/69 The HTML version of the specification is now built in a directory called
4388 spec_html instead of spec.html, because the latter looks like a path with
4389 a MIME-type, and this confuses some software.
4391 PH/70 Catch two compiler warnings in sieve.c.
4393 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4394 function verify_get_ident() calls ip_connect() to connect a socket, but
4395 if the "connect()" function timed out, ip_connect() used to close the
4396 socket. However, verify_get_ident() also closes the socket later, and in
4397 between Exim writes to the log, which may get opened at this point. When
4398 the socket was closed in ip_connect(), the log could get the same file
4399 descriptor number as the socket. This naturally causes chaos. The fix is
4400 not to close the socket in ip_connect(); the socket should be closed by
4401 the function that creates it. There was only one place in the code where
4402 this was missing, in the iplookup router, which I don't think anybody now
4403 uses, but I've fixed it anyway.
4405 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4406 well as to direct DNS lookups. Otherwise the handling of names in host
4407 lists is inconsistent and therefore confusing.
4413 PH/01 Two changes to the default runtime configuration:
4415 (1) Move the checks for relay_from_hosts and authenticated clients from
4416 after to before the (commented out) DNS black list checks.
4418 (2) Add control=submission to the relay_from_hosts and authenticated
4419 clients checks, on the grounds that messages accepted by these
4420 statements are most likely to be submissions.
4422 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4424 (1) Generate an error if the third argument for the ${prvs expansion is
4427 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4430 (3) Reset the variables that are obtained from the first argument of
4431 ${prvscheck and used in the second argument before leaving the code,
4432 because their memory is reclaimed, so using them afterwards may do
4435 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4436 one (it's much easier than Tom thought :-).
4438 (5) Because of (4), we can now allow for the use of $prvscheck_result
4439 inside the third argument.
4441 PH/03 For some reason, the default setting of PATH when running a command from
4442 a pipe transport was just "/usr/bin". I have changed it to
4445 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4446 anything to be listed in the output from -bV.
4448 PH/05 When a filter generated an autoreply, the entire To: header line was
4449 quoted in the delivery log line, like this:
4451 => >A.N.Other <ano@some.domain> <original@ddress> ...
4453 This has been changed so that it extracts the operative address. There
4454 may be more than one such address. If so, they are comma-separated, like
4457 => >ano@some.domain,ona@other.domain <original@ddress> ...
4459 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4460 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4461 not being looked up in the rDNS to get a host name, Exim was showing the
4462 IP address twice in Received: lines, even though the IP addresses were
4463 identical. For example:
4465 Received: from [1.2.3.4] (helo=[1.2.3.4])
4467 However, if the real host name was known, it was omitting the HELO data
4468 if it matched the actual IP address. This has been tidied up so that it
4469 doesn't show the same IP address twice.
4471 PH/07 When both +timestamp and +memory debugging was on, the value given by
4472 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4473 called by the debug printing, thereby overwriting the timestamp buffer.
4474 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4476 PH/08 When the original message was included in an autoreply transport, it
4477 always said "this is a copy of the message, including all the headers",
4478 even if body_only or headers_only was set. It now gives an appropriate
4481 PH/09 Applied a patch from the Sieve maintainer which:
4483 o fixes some comments
4484 o adds the (disabled) notify extension core
4485 o adds some debug output for the result of if/elsif tests
4486 o points to the current vacation draft in the documentation
4487 and documents the missing references header update
4491 o fixes a bug in processing the envelope test (when testing
4492 multiple envelope elements, the last element determined the
4495 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4496 Electronic Mail") by including:
4498 Auto-submitted: auto-generated
4500 in the messages that it generates (bounce messages and others, such as
4501 warnings). In the case of bounce messages for non-SMTP messages, there was
4502 also a typo: it was using "Auto_submitted" (underscore instead of
4503 hyphen). Since every message generated by Exim is necessarily in response
4504 to another message, thes have all been changed to:
4506 Auto-Submitted: auto-replied
4508 in accordance with these statements in the RFC:
4510 The auto-replied keyword:
4512 - SHOULD be used on messages sent in direct response to another
4513 message by an automatic process,
4515 - MUST NOT be used on manually-generated messages,
4517 - MAY be used on Delivery Status Notifications (DSNs) and Message
4518 Disposition Notifications (MDNs),
4520 - MUST NOT be used on messages generated by automatic or periodic
4521 processes, except for messages which are automatic responses to
4524 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4525 to the default Received: header definition.
4527 PH/12 Added log selector acl_warn_skipped (default on).
4529 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4530 variables because (a) they are in the wrong storage pool and (b) even if
4531 they were copied, it wouldn't work properly because of the caching.
4533 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4534 checking when decoding. Apparently there are clients that generate
4535 overlong encoded strings. Why am I not surprised?
4537 PH/15 If the first argument of "${if match_address" was not empty, but did not
4538 contain an "@" character, Exim crashed. Now it writes a panic log message
4539 and treats the condition as false.
4541 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4543 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4544 extension "envelope-auth". The code is finished and in agreement with
4545 other implementations, but there is no documentation so far and in fact,
4546 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4547 not changing the active code.
4549 Print executed "if" and "elsif" statements when debugging is used. This
4550 helps a great deal to understand what a filter does.
4552 Document more things not specified clearly in RFC3028. I had all this
4553 sorted out, when out of a sudden new issues came to my mind. Oops."
4555 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4558 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4559 canonical form (as documented). However, after a host name lookup from
4560 the IP address, check_host() was doing a simple string comparison with
4561 addresses acquired from the DNS when checking that the found name did
4562 have the original IP as one of its addresses. Since any found IPv6
4563 addresses are likely to be in abbreviated form, the comparison could
4564 fail. Luckily, there already exists a function for doing the comparison
4565 by converting both addresses to binary, so now that is used instead of
4566 the text comparison.
4568 PH/20 There was another similar case to PH/19, when a complete host name was
4569 given in a host list; looking up its IP address could give an abbreviated
4570 form, whereas the current host's name might or might not be abbreviated.
4571 The same fix has been applied.
4577 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4578 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4581 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4582 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4584 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4586 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4587 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4588 Also, for "Auto-Submitted", treat anything other than "no" as
4589 non-personal, in accordance with RFC 3834. (Previously it treated
4590 anything starting "auto-" as non-personal.)
4592 TF/01 The control=submission/name=... option had a problem with syntax
4593 errors if the name included a slash character. The /name= option
4594 now slurps the rest of the string, so it can include any characters
4595 but it must come last in the list of options (after /sender_retain
4598 PH/05 Some modifications to the interface to the fake nameserver for the new
4606 TK/01 Added the "success_on_redirect" address verification option. See
4607 NewStuff for rationale and an example.
4609 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4611 PH/02 Patch to exigrep to allow it to work on syslog lines.
4613 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4614 fread() to skip over the body file's header line, because in Cygwin the
4615 header line is locked and is inaccessible.
4617 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4618 co-exist for some time) to make it clear that it is the Exim ID that is
4619 referenced, not the Message-ID: header line.
4621 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4622 string_format() function, because snprintf() does not exist on all
4625 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4626 ${stat: expansion item.
4628 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4629 protocol synchronization error", to keep the pedants happy.
4631 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4632 well as for IRIX systems, when gcc is being used. See the host.c source
4635 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4637 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4640 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4641 transport and to the smtp transport in LMTP mode.
4643 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4645 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4646 run a filter in a subprocess. This could lead to confusion in subsequent
4647 lookups in the parent process. There should also be a search_tidyup() at
4648 the end of the subprocess.
4650 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4651 only if the host matched helo_try_verify_hosts, which caused the
4652 verification to occur when the EHLO/HELO command was issued. The ACL just
4653 tested the remembered result. Now, if a previous verification attempt has
4654 not happened, "verify = helo" does it there and then.
4656 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4658 TK/03 Fix log output including CR from clamd.
4660 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4661 exist provoked a memory error which could cause a segfault.
4663 PH/15 Installed PCRE 6.2
4665 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4667 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4668 of the problem. Specifically, suggested +O2 rather than +O1 for the
4671 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4673 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4674 record, the port was not correctly logged when the outgoing_port log
4675 selector was set (it logged the transort's default port).
4677 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4678 fallback_hosts, and "hosts" in the smtp transport.
4680 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4681 host errors such as "Connection refused".
4683 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4684 authentication with radiusclient 0.4.9:
4686 - Error returned from rc_read_config was caught wrongly
4687 - Username/password not passed on to radius server due to wrong length.
4689 The presumption is that some radiusclient API changes for 4.51/PH/17
4690 were not taken care of correctly. The code is still untested by me (my
4691 Linux distribution still has 0.3.2 of radiusclient), but it was
4692 contributed by a Radius user.
4694 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4695 expanding the "port" option of the smtp transport.
4697 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4698 while reading a MIME header. Thanks to Tom Hughes for a patch.
4700 PH/24 Include config.h inside local_scan.h so that configuration settings are
4703 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4704 that in an ACL for RCPT (for example), you can examine exactly what was
4707 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4708 commands, but it was not correctly comparing the address with the actual
4709 client host address. Thus, it would show the EHLO address in Received:
4710 header lines when this was not necessary.
4712 PH/27 Added the % operator to ${eval:}.
4714 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4715 it should be ignoring failures (because with -C, for example, it has lost
4716 privilege). It wasn't ignoring creation failures other than "already
4719 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4722 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4723 because an input line was too long, either on its own, or by virtue of
4724 too many continuations, the temporary file was not being removed, and the
4725 return code was incorrect.
4727 PH/31 Missing "BOOL" in function definition in filtertest.c.
4729 PH/32 Applied Sieve patches from the maintainer.
4731 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4733 PH/33 Added "verify = not_blind".
4735 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4736 Local/Makefile (with some defaults set). These are used in built scripts
4737 such as exicyclog, but they have never been used in the exim_install
4738 script (though there are many overriding facilities there). I have
4739 arranged that the exim_install script now takes note of these two
4742 PH/35 Installed configuration files for Dragonfly.
4744 PH/36 When a locally submitted message by a trusted user did not contain a
4745 From: header, and the sender address was obtained from -f or from an SMTP
4746 MAIL command, and the trusted user did not use -F to supply a sender
4747 name, $originator_name was incorrectly used when constructing a From:
4748 header. Furthermore, $originator_name was used for submission mode
4749 messages from external hosts without From: headers in a similar way,
4750 which is clearly wrong.
4752 PH/37 Added control=suppress_local_fixups.
4754 PH/38 When log_selector = +received_sender was set, and the addition of the
4755 sender made the log line's construction buffer exactly full, or one byte
4756 less than full, an overflow happened when the terminating "\n" was
4759 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4760 when the result of a list match is failure because a DNS lookup failed.
4762 PH/40 RM_COMMAND is now used in the building process.
4764 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4765 the "build-* directories that it finds.
4767 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4768 address in a domain literal was a prefix of an interface address.
4770 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4771 when verifying a sender address, unless rewrite_headers is false.
4773 PH/44 Wrote a long comment about why errors_to addresses are verified as
4774 recipients, not senders.
4776 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4777 the ratelimit ACL was added.
4779 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4781 PH/46 Added extra information about PostgreSQL errors to the error string.
4783 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4784 avoiding the need to install special zones in a real server. This is
4785 backwards compatible; if it can't find the fake resolver, it drops back.
4786 Thus, both old and new test suites can be run.
4788 TF/02 Added util/ratelimit.pl
4790 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4791 clock is set back in time.
4793 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4794 Candler <B.Candler@pobox.com>.
4796 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4797 OK for addresses that are the result of redirections.
4799 PH/48 A number of further additions for the benefit of the new test suite,
4800 including a fake gethostbyname() that interfaces to the fake DNS resolver
4803 TF/06 The fix for widen_domains has also been applied to qualify_single and
4804 search_parents which are the other dnslookup options that can cause
4807 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4810 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4812 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4813 service" if it encountered an MX record with an empty target hostname.
4814 The message is now "an MX or SRV record indicated no SMTP service".
4816 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4817 if the DNS of the sending site is misconfigured. This is quite a
4818 common situation. This change restores the behaviour of treating a
4819 helo verification defer as a failure.
4821 PH/51 If self=fail was set on a router, the bounce message did not include the
4822 actual error message.
4828 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4830 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4831 command itself ended in error, the underlying message about the transport
4832 filter timeout was being overwritten with the pipe command error. Now the
4833 underlying error message should be appended to the second error message.
4835 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4837 PH/02 Reduce the amount of output that "make" produces by default. Full output
4838 can still be requested.
4840 PH/03 The warning log line about a condition test deferring for a "warn" verb
4841 was being output only once per connection, rather than after each
4842 occurrence (because it was using the same function as for successful
4843 "warn" verbs). This seems wrong, so I have changed it.
4845 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4846 it should not have, which might have caused a crash in the right
4847 circumstances, but probably never did.
4849 PH/04 Installed a modified version of Tony Finch's patch to make submission
4850 mode fix the return path as well as the Sender: header line, and to
4851 add a /name= option so that you can make the user's friendly name appear
4854 TF/03 Added the control = fakedefer ACL modifier.
4856 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4857 Mark Lowes for thorough testing.
4859 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4861 TK/03 Merged latest SRS patch from Miles Wilton.
4863 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4864 with the definition in sysexits.h (which is #included earlier).
4865 Fortunately, Exim does not actually use EX_OK. The code used to try to
4866 preserve the sysexits.h value, by assuming that macro definitions were
4867 scanned for macro replacements. I have been disabused of this notion,
4868 so now the code just undefines EX_OK before #including unistd.h.
4870 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4871 in the smtp transport. When a block could not be written in a single
4872 write() function, the timeout was being re-applied to each part-write.
4873 This seems wrong - if the receiver was accepting one byte at a time it
4874 would take for ever. The timeout is now adjusted when this happens. It
4875 doesn't have to be particularly precise.
4877 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4878 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4880 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4881 without a domain if the check to <postmaster@domain> fails.
4883 SC/01 Eximstats: added -xls and the ability to specify output files
4884 (patch written by Frank Heydlauf).
4886 SC/02 Eximstats: use FileHandles for outputting results.
4888 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4890 SC/04 Eximstats: fixed display of large numbers with -nvr option
4892 SC/05 Eximstats: fixed merging of reports with empty tables.
4894 SC/06 Eximstats: added the -include_original_destination flag
4896 SC/07 Eximstats: removed tabs and trailing whitespace.
4898 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4900 TK/06 MBOX spool code: Add real "From " MBOX separator line
4901 so the .eml file is really in mbox format (even though
4902 most programs do not really care). Patch from Alex Miller.
4904 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4905 The latter is generated from $received_to and is only set if the
4906 message has one envelope recipient. SA can use these headers,
4907 obviously out-of-the-box. Patch from Alex Miller.
4909 PH/08 The ${def test on a variable was returning false if the variable's
4910 value was "0", contrary to what the specification has always said!
4911 The result should be true unless the variable is empty.
4913 PH/09 The syntax error of a character other than { following "${if
4914 def:variable_name" (after optional whitespace) was not being diagnosed.
4915 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4916 accidental colon was present, for example, could give incorrect results.
4918 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4919 result is used (not including appendfile, where other changes are about
4922 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4923 This involved changing a lot of size variables from int to off_t. It
4924 should work with maildirs and everything.
4926 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4927 spamd dying while we are connected to it.
4929 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4932 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4933 function for BDB 4.3.
4935 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4937 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4938 the sources for PCRE have been re-arranged and more files are now
4941 PH/15 The code I had for printing potentially long long variables in PH/11
4942 above was not the best (it lost precision). The length of off_t variables
4943 is now inspected at build time, and an appropriate printing format (%ld
4944 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4945 to be "long long int" or "long int". This is needed for the internal
4946 formatting function string_vformat().
4948 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4949 the configuration file to be ":syslog", then the script "guesses" where
4950 the logs files are, rather than using the compiled in default. In our
4951 case the guess is not the same as the compiled default, so the script
4952 suddenly stopped working when I started to use syslog. The patch checks
4953 to see if log_file_path is "". If so, it attempts to read it from exim
4954 with no configuration file to get the compiled in version, before it
4955 falls back to the previous guessing code."
4957 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4958 implementing BATV in an Exim configuration. See NewStuff for the gory
4961 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4962 Makefile that are specific to HP-UX.
4964 PH/18 If the "use_postmaster" option was set for a recipient callout together
4965 with the "random" option, the postmaster address was used as the MAIL
4966 FROM address for the random test, but not for the subsequent recipient
4967 test. It is now used for both.
4969 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4970 patch removes a few documentation additions to RFC 3028, because the
4971 latest draft now contains them. It adds the new en;ascii-case comparator
4972 and a new error check for 8bit text in MIME parts. Comparator and
4973 require names are now matched exactly. I enabled the subaddress
4974 extension, but it is not well tested yet (read: it works for me)."
4976 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4977 rework some of the code of TK/09 above to avoid the hardwired use of
4978 "%lld" and "long long". Replaced the call to snprintf() with a call to
4981 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4982 records point to non-existent hosts", "retry timeout exceeded", and
4983 "retry time not reached for any host after a long failure period".
4985 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4986 experimental DomainKeys support:
4988 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4989 (2) On an error such as an illegally used "control", the wrong name for
4990 the control was given.
4992 These problems did NOT occur unless DomainKeys support was compiled.
4994 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4996 PH/24 Added ${if match_ip condition.
4998 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4999 databases so that it will be absolutely obvious if a crash occurs in the
5000 DB library. This is a regular occurrence (often caused by mis-matched
5003 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5004 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5005 picked up on a user's system that detects such things. There doesn't seem
5006 to be a gcc warning option for this - only an attribute that has to be
5007 put on the function's prototype. It seems that in Fedora Core 4 they have
5008 set this on a number of new functions. No doubt there will be more in due
5011 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5012 specify a transport. However, if an address that was verified by such a
5013 router was the subject of a callout, Exim crashed because it tried to
5014 read the rcpt_include_affixes from the non-existent transport. Now it
5015 just assumes that the setting of that option is false. This bug was
5016 introduced by 4.51/PH/31.
5018 PH/28 Changed -d+all to exclude +memory, because that information is very
5019 rarely of interest, but it makes the output a lot bigger. People tend to
5020 do -d+all out of habit.
5022 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5023 code in os-type was giving problems when libc.so lives in lib64, like on
5026 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5027 aren't the modern standard, and it seems that some systems' include files
5028 don't always have them. Exim was already checking for some of the newer
5029 ones like T_AAAA, and defining it itself. I've added checks for all the
5030 record types that Exim uses.
5032 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5033 not automatically generating a new one, as it is supposed to. This
5034 prevented TLS from working. If the file did exist, but contained invalid
5035 data, a new version was generated, as expected. It was only the case of a
5036 non-existent file that was broken.
5038 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5039 with a change in libdomainkeys > 0.64.
5041 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5042 from DNS. If the selector record carries the flag, it now has
5043 precedence over the domain-wide flag.
5045 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5047 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5048 the use of a port name that isn't defined in /etc/services) occurred, the
5049 message was deferred as in a normal delivery, and thus remained on the
5050 spool, instead of being failed because of the mua_wrapper setting. This
5051 is now fixed, and I tidied up some of the mua_wrapper messages at the
5054 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5055 the messages in a hash of arrays rather than using individual hashes.
5056 This is a bit cleaner and results in dramatic memory savings, albeit
5057 at a slight CPU cost.
5059 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5060 as requested by Marc Sherman.
5062 SC/10 Eximstats: added histograms for user specified patterns as requested
5065 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5067 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5068 fopen() in the content-scanning modules that did not already have it.
5074 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5075 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5077 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5079 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5081 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5082 to test Sieve filters that use "vacation".
5084 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5085 that changes the way the GnuTLS parameters are stored in the cache file.
5086 The new format can be generated externally. For backward compatibility,
5087 if the data in the cache doesn't make sense, Exim assumes it has read an
5088 old-format file, and it generates new data and writes a new file. This
5089 means that you can't go back to an older release without removing the
5092 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5093 work if there are any delivery delays because "one_time" forces the
5094 parent to be marked "delivered", so its unseen clone is never tried
5095 again. For this reason, Exim now forbids the simultaneous setting of
5098 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5099 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5100 entry if you want to know the details. The fix, however, neglected to
5101 consider the case when local delivery batching is involved. The test for
5102 "previously delivered" was not happening when checking to see if an
5103 address could be batched with a previous (undelivered) one; under
5104 certain circumstances this could lead to multiple deliveries to the same
5107 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5108 in its include files, and this causes problems building Exim.
5110 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5111 header_syntax) but Exim was just ignoring anything given after a slash.
5112 In particular, this caused confusion with an attempt to use "verify =
5113 reverse_host_lookup/defer_ok". An error is now given when options are
5114 supplied for verify items that do not have them. (Maybe reverse_host_
5115 lookup should have a defer_ok option, but that's a different point.)
5117 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5118 defined by RFC 821) to 2048, because there were problems with some AUTH
5119 commands, and RFC 1869 says the size should be increased for extended
5120 SMTP commands that take arguments.
5122 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5125 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5126 "unknown" error; now it says that the functionality isn't in the binary.
5128 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5129 an address' error message when a string expansion fails (syntax or
5130 whatever). Otherwise the password may appear in the log. Following change
5131 PH/42 below, there is no longer a chance of it appearing in a bounce
5134 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5136 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5137 of its addresses was ever tried. (Bugzilla bug #2.)
5139 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5140 the result incorrectly in the debug output. (It correctly added a newline
5141 to what was transported.)
5143 TF/01 Added $received_time.
5145 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5146 commented out examples of how to interface to a virus scanner and to
5147 SpamAssassin. Also added commented examples of av_scanner and
5148 spamd_address settings.
5150 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5151 and controls are allowed in which ACLs. There were a couple of minor
5152 errors. Some of the entries in the conditions table (which is a table of
5153 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5154 negation of where the condition IS allowed.
5156 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5158 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5159 header file does not have a version number, so I've had to invent a new
5160 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5161 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5162 radiusclient), but it was contributed by a Radius user.
5164 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5165 files or directories, for OpenSSL.
5167 PH/19 When an Exim process that is running as root has to create an Exim log
5168 file, it does so in a subprocess that runs as exim:exim so as to get the
5169 ownership right at creation (otherwise, other Exim processes might see
5170 the file with the wrong ownership). There was no test for failure of this
5171 fork() call, which would lead to the process getting stuck as it waited
5172 for a non-existent subprocess. Forks do occasionally fail when resources
5173 run out. I reviewed all the other calls to fork(); they all seem to check
5176 PH/20 When checking for unexpected SMTP input at connect time (before writing
5177 the banner), Exim was not dealing correctly with a non-positive return
5178 from the read() function. If the client had disconnected by this time,
5179 the result was a log entry for a synchronization error with an empty
5180 string after "input=" when read() returned zero. If read() returned -1
5181 (an event I could not check), uninitialized data bytes were printed.
5182 There were reports of junk text (parts of files, etc) appearing after
5185 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5187 PH/22 Added support for macro redefinition, and (re)definition in between
5188 driver and ACL definitions.
5190 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5191 forgetting to use the resulting value; it was using the unexpanded value.
5193 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5194 hadn't been configured. The fix is from Juergen Kreileder, who
5195 understands it better than I do:
5197 "Here's what I see happening with three configured cyrus_sasl
5198 authenticators configured (plain, login, cram-md5):
5200 On startup auth_cyrus_sasl_init() gets called for each of these.
5201 This means three calls to sasl_listmech() without a specified mech_list.
5202 => SASL tests which mechs of all available mechs actually work
5203 => three warnings about OTP not working
5204 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5206 With the patch, sasl_listmech() also gets called three times. But now
5207 SASL's mech_list option is set to the server_mech specified in the the
5208 authenticator. Or in other words, the answer from sasl_listmech()
5209 gets limited to just the mech you're testing for (which is different
5211 => the return list contains just 'plain' or 'login', 'cram-md5' or
5212 nothing depending on the value of ob->server_mech.
5214 I've just tested the patch: Authentication still works fine,
5215 unavailable mechs specified in the exim configuration are still
5216 caught, and the auth.log warnings about OTP are gone."
5218 PH/25 When debugging is enabled, the contents of the command line are added
5219 to the debugging output, even when log_selector=+arguments is not
5222 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5223 answer is "GNU", and only if the return is "GNU/something" is the answer
5226 PH/27 $acl_verify_message is now set immediately after the failure of a
5227 verification in an ACL, and so is available in subsequent modifiers. In
5228 particular, the message can be preserved by coding like this:
5230 warn !verify = sender
5231 set acl_m0 = $acl_verify_message
5233 Previously, $acl_verify_message was set only while expanding "message"
5234 and "log_message" when a very denied access.
5236 PH/28 Modified OS/os.c-Linux with
5238 -#ifndef OS_LOAD_AVERAGE
5239 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5241 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5242 nomenclature these days.)
5244 PH/29 Installed patch from the Sieve maintainer that adds the options
5245 sieve_useraddress and sieve_subaddress to the redirect router.
5247 PH/30 In these circumstances:
5248 . Two addresses routed to the same list of hosts;
5249 . First host does not offer TLS;
5250 . First host accepts first address;
5251 . First host gives temporary error to second address;
5252 . Second host offers TLS and a TLS session is established;
5253 . Second host accepts second address.
5254 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5255 and peerdn, if requested) that were in fact used only for the second
5258 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5259 attention to any local part prefix or suffix that was matched by the
5260 router that accepted the address. It now behaves in the same way as it
5261 does for delivery: the affixes are removed from the local part unless
5262 rcpt_include_affixes is set on the transport.
5264 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5265 timeout during the DATA phase of an incoming message.
5267 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5268 applied a patch sanctioned by the Sieve maintainer.
5270 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5271 the uid or gid is negative. A case of a negative gid caused this to be
5272 noticed. The fix allows for either to be negative.
5274 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5275 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5278 PH/36 The change PH/12 above was broken. Fixed it.
5280 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5281 the grounds that routing the same address twice would always produce the
5282 same answer. This might have been true once, but it is certainly no
5283 longer true now. Routing a child address may depend on the previous
5284 routing that produced that child. Some complicated redirection strategies
5285 went wrong when messages had multiple recipients, and made Exim's
5286 behaviour dependent on the order in which the addresses were given.
5288 I have moved the duplicate checking until after the routing is complete.
5289 Exim scans the addresses that are assigned to local and remote
5290 transports, and removes any duplicates. This means that more work will be
5291 done, as duplicates will always all be routed, but duplicates are
5292 presumably rare, so I don't expect this is of any significance.
5294 For deliveries to pipes, files, and autoreplies, the duplicate checking
5295 still happens during the routing process, since they are not going to be
5298 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5299 It corrects a timeout issue with spamd. This is Ian's comment: "The
5300 background is that sometimes spamd either never reads data from a
5301 connection it has accepted, or it never writes response data. The exiscan
5302 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5303 blindly assumes that writes won't block so it may never time out."
5305 PH/39 Allow G after quota size as well as K and M.
5307 PH/40 The value set for $authenticated_id in an authenticator may not contain
5308 binary zeroes or newlines because the value is written to log lines and
5309 to spool files. There was no check on this. Now the value is run through
5310 the string_printing() function so that such characters are converted to
5311 printable escape sequences.
5313 PH/41 $message_linecount is a new variable that contains the total number of
5314 lines in the message. Compare $body_linecount, which is the count for the
5317 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5318 bounce and delay warning messages, except in certain special cases, which
5321 (a) An SMTP error message from a remote host;
5322 (b) A message specified in a :fail: redirection;
5323 (c) A message specified in a "fail" command in a system filter;
5324 (d) A message specified in a FAIL return from the queryprogram router;
5325 (e) A message specified by the cannot_route_message router option.
5327 In these cases only, Exim does include the error details in bounce and
5328 warning messages. There are also a few cases where bland messages such
5329 as "unrouteable address" or "local delivery error" are given.
5331 PH/43 $value is now also set for the "else" part of a ${run expansion.
5333 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5334 being worked on, but at least Exim now implements the latest version to
5337 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5338 process to complete was treated as a delivery failure, a timeout while
5339 writing the message to the pipe was logged, but erroneously treated as a
5340 successful delivery. Such timeouts include transport filter timeouts. For
5341 consistency with the overall process timeout, these timeouts are now
5342 treated as errors, giving rise to delivery failures by default. However,
5343 there is now a new Boolean option for the pipe transport called
5344 timeout_defer, which, if set TRUE, converts the failures into defers for
5345 both kinds of timeout. A transport filter timeout is now identified in
5348 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5349 systems where "make" and "gmake" are different, calling "gmake" at top
5350 level broke things. I've arranged for the value of $(MAKE) to be passed
5351 from the Makefile to this script so that it can call the same version of
5355 A note about Exim versions 4.44 and 4.50
5356 ----------------------------------------
5358 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5359 changes of various kinds. As a consequence, a big documentation update was
5360 needed. This delayed the release for rather longer than seemed good, especially
5361 in the light of a couple of (minor) security issues. Therefore, the changes
5362 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5363 release. So 4.44 and 4.50 are in effect two different branches that both start
5366 I have left the 4.50 change log unchanged; it contains all the changes since
5367 4.43. The change log for 4.44 is below; many of its items are identical to
5368 those for 4.50. This seems to be the most sensible way to preserve the
5369 historical information.
5375 1. Minor wording change to the doc/README.SIEVE file.
5377 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5378 computation of the current number of files was incorrect.
5380 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5381 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5384 4. Give more explanation in the error message when the command for a transport
5385 filter fails to execute.
5387 5. There are several places where Exim runs a non-Exim command in a
5388 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5389 was being done only for the command run by the queryprogram router. It is
5390 now done for all such subprocesses. The other cases are: ${run, transport
5391 filters, and the commands run by the lmtp and pipe transports.
5393 6. Added CONFIGURE_GROUP build-time option.
5395 7. Some older OS have a limit of 256 on the maximum number of file
5396 descriptors. Exim was using setrlimit() to set 1000 as a large value
5397 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5398 systems. I've change it so that if it can't get 1000, it tries for 256.
5400 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5401 was an oversight, and furthermore, ever since the addition of extra
5402 controls (e.g. 4.43/32), the checks on when to allow different forms of
5403 "control" were broken. There should now be diagnostics for all cases when a
5404 control that does not make sense is encountered.
5406 9. Added the /retain_sender option to "control=submission".
5408 10. $recipients is now available in the predata ACL (oversight).
5410 11. Tidy the search cache before the fork to do a delivery from a message
5411 received from the command line. Otherwise the child will trigger a lookup
5412 failure and thereby defer the delivery if it tries to use (for example) a
5413 cached ldap connection that the parent has called unbind on.
5415 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5416 of $address_data from the recipient verification was clobbered by the
5417 sender verification.
5419 13. The value of address_data from a sender verification is now available in
5420 $sender_address_data in subsequent conditions in the ACL statement.
5422 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5424 15. Added a new option "connect=<time>" to callout options, to set a different
5427 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5428 was its contents. (It was OK if the option was not defined at all.)
5430 17. A "Completed" log line is now written for messages that are removed from
5431 the spool by the -Mrm option.
5433 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5434 information about exactly what failed.
5436 19. Added -dd to debug only the daemon process.
5438 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5439 handles timeouts, both on the server side and network timeouts. Renamed the
5440 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5442 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5443 It is now set to "smtps".
5445 22. $host_address is now set to the target address during the checking of
5446 ignore_target_hosts.
5448 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5449 being passed; this would have caused $sender_host_name to have been used if
5450 matching the list had actually called for a host name (not very likely,
5451 since this list is usually IP addresses). A host name is now passed as
5454 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5455 code that specifies a non-restarting handler (typically sigaction() in
5456 modern systems) in an attempt to fix a rare and obscure crash bug.
5458 25. Narrowed the window for a race in the daemon that could cause it to ignore
5459 SIGCHLD signals. This is not a major problem, because they are used only to
5460 wake it up if nothing else does.
5462 26. A malformed maildirsize file could cause Exim to calculate negative values
5463 for the mailbox size or file count. Odd effects could occur as a result.
5464 The maildirsize information is now recalculated if the size or filecount
5467 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5468 support for a long time. Removed HAVE_SYS_VFS_H.
5470 28. Installed the latest version of exipick from John Jetmore.
5472 29. In an address list, if the pattern was not a regular expression, an empty
5473 subject address (from a bounce message) matched only if the pattern was an
5474 empty string. Non-empty patterns were not even tested. This was the wrong
5475 because it is perfectly reasonable to use an empty address as part of a
5476 database query. An empty address is now tested by patterns that are
5477 lookups. However, all the other forms of pattern expect the subject to
5478 contain a local part and a domain, and therefore, for them, an empty
5479 address still always fails if the pattern is not itself empty.
5481 30. Exim went into a mad DNS loop when attempting to do a callout where the
5482 host was specified on an smtp transport, and looking it up yielded more
5483 than one IP address.
5485 31. Re-factored the code for checking spool and log partition space into a
5486 function that finds that data and another that does the check. The former
5487 is then used to implement four new variables: $spool_space, $log_space,
5488 $spool_inodes, and $log_inodes.
5490 32. The RFC2047 encoding function was originally intended for short strings
5491 such as real names; it was not keeping to the 75-character limit for
5492 encoded words that the RFC imposes. It now respects the limit, and
5493 generates multiple encoded words if necessary. To be on the safe side, I
5494 have increased the buffer size for the ${rfc2047: expansion operator from
5497 33. It is now permitted to omit both strings after an "if" condition; if the
5498 condition is true, the result is "true". As before, when the second string
5499 is omitted, a false condition yields an empty string. This makes it less
5500 cumbersome to write custom ACL and router conditions.
5502 34. Failure to deliver a bounce message always caused it to be frozen, even if
5503 there was an errors_to setting on the router. The errors_to setting is now
5506 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5507 canonical form (fully expanded) before being placed in
5508 $sender_host_address.
5510 36. The table in the code that translates DNS record types into text (T_A to
5511 "A" for instance) was missing entries for NS and CNAME. It is just possible
5512 that this could have caused confusion if both these types were looked up
5513 for the same domain, because the text type is used as part of Exim's
5514 per-process caching. But the chance of anyone hitting this buglet seems
5517 37. The dnsdb lookup has been extended in a number of ways.
5519 (1) There is a new type, "zns", which walks up the domain tree until it
5520 finds some nameserver records. It should be used with care.
5522 (2) There is a new type, "mxh", which is like "mx" except that it returns
5523 just the host names, not the priorities.
5525 (3) It is now possible to give a list of domains (or IP addresses) to be
5526 looked up. The behaviour when one of the lookups defers can be
5527 controlled by a keyword.
5529 (4) It is now possible to specify the separator character for use when
5530 multiple records are returned.
5532 38. The dnslists ACL condition has been extended: it is now possible to supply
5533 a list of IP addresses and/or domains to be looked up in a particular DNS
5536 39. Added log_selector=+queue_time_overall.
5538 40. When running the queue in the test harness, wait just a tad after forking a
5539 delivery process, to get repeatability of debugging output.
5541 41. Include certificate and key file names in error message when GnuTLS fails
5542 to set them up, because the GnuTLS error message doesn't include the name
5543 of the failing file when there is a problem reading it.
5545 42. Allow both -bf and -bF in the same test run.
5547 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5549 44. The "Exiscan patch" is now merged into the mainline Exim source.
5551 45. Sometimes the final signoff response after QUIT could fail to get
5552 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5553 < 0 before doing a fflush(). This bug looks as though it goes back to the
5554 introduction of TLS in release 3.20, but "sometimes" must have been rare
5555 because the tests only now provoked it.
5557 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5558 (this can affect the format of dates).
5560 47. exim_tidydb, when checking for the continued existence of a message for
5561 which it has found a message-specific retry record, was not finding
5562 messages that were in split spool directories. Consequently, it was
5563 deleting retry records that should have stayed in existence.
5565 48. Steve fixed some bugs in eximstats.
5567 49. The SPA authentication driver was not abandoning authentication and moving
5568 on to the next authenticator when an expansion was forced to fail,
5569 contradicting the general specification for all authenticators. Instead it
5570 was generating a temporary error. It now behaves as specified.
5572 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5573 (the order specifies the preference for clients). The order is now AES256,
5574 AES128, 3DES, ARCFOUR128.
5576 51. Small patch to Sieve code - explicitly set From: when generating an
5579 52. Exim crashed if a remote delivery caused a very long error message to be
5580 recorded - for instance if somebody sent an entire SpamAssassin report back
5581 as a large number of 550 error lines. This bug was coincidentally fixed by
5582 increasing the size of one of Exim's internal buffers (big_buffer) that
5583 happened as part of the Exiscan merge. However, to be on the safe side, I
5584 have made the code more robust (and fixed the comments that describe what
5587 53. Now that there can be additional text after "Completed" in log lines (if
5588 the queue_time_overall log selector is set), a one-byte patch to exigrep
5589 was needed to allow it to recognize "Completed" as not the last thing in
5592 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5593 patch that reportedly fixes this has been added. I am not expert enough to
5594 create a test for it. This is what the patch creator wrote:
5596 "I found a little strange behaviour of ldap code when working with
5597 Windows 2003 AD Domain, where users was placed in more than one
5598 Organization Units. When I tried to give exim partial DN, the exit code
5599 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5600 But simultaneously result of request was absolutely normal ldap result,
5601 so I produce this patch..."
5603 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5604 REFERENCE, so I have modified the code to exclude the patch when that macro
5607 55. Some experimental protocols are using DNS PTR records for new purposes. The
5608 keys for these records are domain names, not reversed IP addresses. The
5609 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5610 leaves it alone. Component reversal etc. now happens only for IP addresses.
5613 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5615 57. Double the size of the debug message buffer (to 2048) so that more of very
5616 long debug lines gets shown.
5618 58. The exicyclog utility now does better if the number of log files to keep
5619 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5621 59. Two changes related to the smtp_active_hostname option:
5623 (1) $smtp_active_hostname is now available as a variable.
5624 (2) The default for smtp_banner uses $smtp_active_hostname instead
5625 of $primary_hostname.
5627 60. The host_aton() function is supposed to be passed a string that is known
5628 to be a valid IP address. However, in the case of IPv6 addresses, it was
5629 not checking this. This is a hostage to fortune. Exim now panics and dies
5630 if the condition is not met. A case was found where this could be provoked
5631 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5632 components; fortuitously, this particular loophole had already been fixed
5633 by change 4.50/55 above.
5635 If there are any other similar loopholes, the new check in host_aton()
5636 itself should stop them being exploited. The report I received stated that
5637 data on the command line could provoke the exploit when Exim was running as
5638 exim, but did not say which command line option was involved. All I could
5639 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5640 running as the user.
5643 61. There was a buffer overflow vulnerability in the SPA authentication code
5644 (which came originally from the Samba project). I have added a test to the
5645 spa_base64_to_bits() function which I hope fixes it.
5648 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5649 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5651 63. The daemon start-up calls getloadavg() while still root for those OS that
5652 need the first call to be done as root, but it missed one case: when
5653 deliver_queue_load_max is set with deliver_drop_privilege. This is
5654 necessary for the benefit of the queue runner, because there is no re-exec
5655 when deliver_drop_privilege is set.
5657 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5658 This has been fixed.
5660 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5661 was in use, was not putting the data itself into the right store pool;
5662 consequently, it could be overwritten for a subsequent message in the same
5663 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5666 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5668 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5669 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5670 the function were treating the return as a boolean value, which happened to
5671 work because 0=false and not-0=true, but is not correct code.
5673 68. The host_aton() function was not handling scoped IPv6 addresses (those
5674 with, for example, "%eth0" on the end) correctly.
5676 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5677 negated items (that is, ~something) in unsigned ints. Some compilers
5678 apparently mutter when there is no cast.
5680 70. If an address verification called from an ACL failed, and did not produce a
5681 user-specific message (i.e. there was only a "system" message), nothing was
5682 put in $acl_verify_message. In this situation, it now puts the system
5685 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5686 session; change 4.31/43 added the unwanted input to the log line - except
5687 that it did not do this in the start of session case. It now does.
5689 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5690 This is wrong and can cause the other end to generate a synchronization
5691 error if it is another Exim or anything else that does the synchronization
5692 check. A QUIT command is no longer sent after a timeout.
5694 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5695 during host lookups.
5697 74. The defer_ok option of callout verification was not working if it was used
5698 when verifying addresses in header lines, that is, for this case:
5700 verify = header_sender/callout=defer_ok
5702 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5703 those file descriptors could be used for SMTP connections. If anything
5704 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5705 could be sent to the SMTP client, causing chaos. The daemon now opens
5706 stdin, stdout, and stderr to /dev/null when it puts itself into the
5709 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5710 log by default. The user can override this with suitable Perl magic.
5712 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5713 the log message when discard triggers, was not working for the DATA ACL or
5714 for the non-SMTP ACL.
5716 78. Error message wording change in sieve.c.
5718 79. If smtp_accept_max_per_host was set, the number of connections could be
5719 restricted to fewer than expected, because the daemon was trying to set up
5720 a new connection before checking whether the processes handling previous
5721 connections had finished. The check for completed processes is now done
5722 earlier. On busy systems, this bug wouldn't be noticed because something
5723 else would have woken the daemon, and it would have reaped the completed
5726 80. If a message was submitted locally by a user whose login name contained one
5727 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5728 It caused a spool format error. I have fixed the spool reading code. A
5729 related problem was that the "from" clause in the Received: line became
5730 illegal because of the space(s). It is now covered by ${quote_local_part.
5732 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5735 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5737 83. Patch from Sieve maintainer to fix unterminated string problem in
5738 "vacation" handling.
5740 84. Some minor changes to the Linux configuration files to help with other
5741 OS variants using glibc.
5743 85. One more patch for Sieve to update vacation handling to latest spec.
5746 ----------------------------------------------------
5747 See the note above about the 4.44 and 4.50 releases.
5748 ----------------------------------------------------
5754 1. Change 4.43/35 introduced a bug that caused file counts to be
5755 incorrectly computed when quota_filecount was set in an appendfile
5758 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5759 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5762 3. Give more explanation in the error message when the command for a transport
5763 filter fails to execute.
5765 4. There are several places where Exim runs a non-Exim command in a
5766 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5767 was being done only for the command run by the queryprogram router. It is
5768 now done for all such subprocesses. The other cases are: ${run, transport
5769 filters, and the commands run by the lmtp and pipe transports.
5771 5. Some older OS have a limit of 256 on the maximum number of file
5772 descriptors. Exim was using setrlimit() to set 1000 as a large value
5773 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5774 systems. I've change it so that if it can't get 1000, it tries for 256.
5776 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5777 was an oversight, and furthermore, ever since the addition of extra
5778 controls (e.g. 4.43/32), the checks on when to allow different forms of
5779 "control" were broken. There should now be diagnostics for all cases when a
5780 control that does not make sense is encountered.
5782 7. $recipients is now available in the predata ACL (oversight).
5784 8. Tidy the search cache before the fork to do a delivery from a message
5785 received from the command line. Otherwise the child will trigger a lookup
5786 failure and thereby defer the delivery if it tries to use (for example) a
5787 cached ldap connection that the parent has called unbind on.
5789 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5790 of $address_data from the recipient verification was clobbered by the
5791 sender verification.
5793 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5794 was its contents. (It was OK if the option was not defined at all.)
5796 11. A "Completed" log line is now written for messages that are removed from
5797 the spool by the -Mrm option.
5799 12. $host_address is now set to the target address during the checking of
5800 ignore_target_hosts.
5802 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5803 being passed; this would have caused $sender_host_name to have been used if
5804 matching the list had actually called for a host name (not very likely,
5805 since this list is usually IP addresses). A host name is now passed as
5808 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5809 code that specifies a non-restarting handler (typically sigaction() in
5810 modern systems) in an attempt to fix a rare and obscure crash bug.
5812 15. Narrowed the window for a race in the daemon that could cause it to ignore
5813 SIGCHLD signals. This is not a major problem, because they are used only to
5814 wake it up if nothing else does.
5816 16. A malformed maildirsize file could cause Exim to calculate negative values
5817 for the mailbox size or file count. Odd effects could occur as a result.
5818 The maildirsize information is now recalculated if the size or filecount
5821 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5822 support for a long time. Removed HAVE_SYS_VFS_H.
5824 18. Updated exipick to current release from John Jetmore.
5826 19. Allow an empty sender to be matched against a lookup in an address list.
5827 Previously the only cases considered were a regular expression, or an
5830 20. Exim went into a mad DNS lookup loop when doing a callout where the
5831 host was specified on the transport, if the DNS lookup yielded more than
5834 21. The RFC2047 encoding function was originally intended for short strings
5835 such as real names; it was not keeping to the 75-character limit for
5836 encoded words that the RFC imposes. It now respects the limit, and
5837 generates multiple encoded words if necessary. To be on the safe side, I
5838 have increased the buffer size for the ${rfc2047: expansion operator from
5841 22. Failure to deliver a bounce message always caused it to be frozen, even if
5842 there was an errors_to setting on the router. The errors_to setting is now
5845 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5846 canonical form (fully expanded) before being placed in
5847 $sender_host_address.
5849 24. Updated eximstats to version 1.33
5851 25. Include certificate and key file names in error message when GnuTLS fails
5852 to set them up, because the GnuTLS error message doesn't include the name
5853 of the failing file when there is a problem reading it.
5855 26. Expand error message when OpenSSL has problems setting up cert/key files.
5858 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5859 (this can affect the format of dates).
5861 28. exim_tidydb, when checking for the continued existence of a message for
5862 which it has found a message-specific retry record, was not finding
5863 messages that were in split spool directories. Consequently, it was
5864 deleting retry records that should have stayed in existence.
5866 29. eximstats updated to version 1.35
5867 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5868 1.35 - bugfix such that pie charts by volume are generated correctly
5870 30. The SPA authentication driver was not abandoning authentication and moving
5871 on to the next authenticator when an expansion was forced to fail,
5872 contradicting the general specification for all authenticators. Instead it
5873 was generating a temporary error. It now behaves as specified.
5875 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5876 (the order specifies the preference for clients). The order is now AES256,
5877 AES128, 3DES, ARCFOUR128.
5879 31. Small patch to Sieve code - explicitly set From: when generating an
5882 32. Exim crashed if a remote delivery caused a very long error message to be
5883 recorded - for instance if somebody sent an entire SpamAssassin report back
5884 as a large number of 550 error lines. This bug was coincidentally fixed by
5885 increasing the size of one of Exim's internal buffers (big_buffer) that
5886 happened as part of the Exiscan merge. However, to be on the safe side, I
5887 have made the code more robust (and fixed the comments that describe what
5890 33. Some experimental protocols are using DNS PTR records for new purposes. The
5891 keys for these records are domain names, not reversed IP addresses. The
5892 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5893 leaves it alone. Component reversal etc. now happens only for IP addresses.
5896 34. The host_aton() function is supposed to be passed a string that is known
5897 to be a valid IP address. However, in the case of IPv6 addresses, it was
5898 not checking this. This is a hostage to fortune. Exim now panics and dies
5899 if the condition is not met. A case was found where this could be provoked
5900 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5901 components; fortuitously, this particular loophole had already been fixed
5902 by change 4.50/55 or 4.44/33 above.
5904 If there are any other similar loopholes, the new check in host_aton()
5905 itself should stop them being exploited. The report I received stated that
5906 data on the command line could provoke the exploit when Exim was running as
5907 exim, but did not say which command line option was involved. All I could
5908 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5909 running as the user.
5912 35. There was a buffer overflow vulnerability in the SPA authentication code
5913 (which came originally from the Samba project). I have added a test to the
5914 spa_base64_to_bits() function which I hope fixes it.
5917 36. The daemon start-up calls getloadavg() while still root for those OS that
5918 need the first call to be done as root, but it missed one case: when
5919 deliver_queue_load_max is set with deliver_drop_privilege. This is
5920 necessary for the benefit of the queue runner, because there is no re-exec
5921 when deliver_drop_privilege is set.
5923 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5924 was in use, was not putting the data itself into the right store pool;
5925 consequently, it could be overwritten for a subsequent message in the same
5926 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5929 38. Sometimes the final signoff response after QUIT could fail to get
5930 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5931 < 0 before doing a fflush(). This bug looks as though it goes back to the
5932 introduction of TLS in release 3.20, but "sometimes" must have been rare
5933 because the tests only now provoked it.
5939 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5940 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5941 Now it returns an integer. A number of calls were still expecting a T/F
5942 return. Fortuitously, in all cases, the tests worked in OK situations,
5943 which is the norm. However, things would have gone wrong on any write
5944 failures on the smtp file descriptor. This function is used when sending
5945 messages over SMTP and also when doing verify callouts.
5947 2. When Exim is called to do synchronous delivery of a locally submitted
5948 message (the -odf or -odi options), it no longer closes stderr before doing
5951 3. Implemented the mua_wrapper option.
5953 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5955 5. Implemented the functions header_remove(), header_testname(),
5956 header_add_at_position(), and receive_remove_recipient(), and exported them
5959 6. If an ACL "warn" statement specified the addition of headers, Exim already
5960 inserted X-ACL-Warn: at the start if there was no header name. However, it
5961 was not making this test for the second and subsequent header lines if
5962 there were newlines in the string. This meant that an invalid header could
5963 be inserted if Exim was badly configured.
5965 7. Allow an ACL "warn" statement to add header lines at the start or after all
5966 the Received: headers, as well as at the end.
5968 8. Added the rcpt_4xx retry error code.
5970 9. Added postmaster_mailfrom=xxx to callout verification option.
5972 10. Added mailfrom=xxxx to the callout verification option, for verify=
5975 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5976 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5978 12. Inserted some casts to stop certain compilers warning when using pointer
5979 differences as field lengths or precisions in printf-type calls (mostly
5980 affecting debugging statements).
5982 13. Added optional readline() support for -be (dynamically loaded).
5984 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5985 same clock tick as a message's arrival, so that its received time was the
5986 same as the "first fail" time on the retry record, and that message
5987 remained on the queue past the ultimate address timeout, every queue runner
5988 would try a delivery (because it was past the ultimate address timeout) but
5989 after another failure, the ultimate address timeout, which should have then
5990 bounced the address, did not kick in. This was a "< instead of <=" error;
5991 in most cases the first failure would have been in the next clock tick
5992 after the received time, and all would be well.
5994 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5995 being recognized when the domain list was tested by the match_domain
5996 condition in an expansion string.
5998 16. Added the ${str2b64: operator.
6000 17. Exim was always calling setrlimit() to set a large limit for the number of
6001 processes, without checking whether the existing limit was already
6002 adequate. (It did check for the limit on file descriptors.) Furthermore,
6003 errors from getrlimit() and setrlimit() were being ignored. Now they are
6004 logged to the main and panic logs, but Exim does carry on, to try to do its
6005 job under whatever limits there are.
6007 18. Imported PCRE 5.0.
6009 19. Trivial typo in log message " temporarily refused connection" (the leading
6012 20. If the log selector return_path_on_delivery was set and an address was
6013 redirected to /dev/null, the delivery process crashed because it assumed
6014 that a return path would always be set for a "successful" delivery. In this
6015 case, the whole delivery is bypassed as an optimization, and therefore no
6018 21. Internal re-arrangement: the function for sending a challenge and reading
6019 a response while authentication was assuming a zero-terminated challenge
6020 string. It's now changed to take a pointer and a length, to allow for
6021 binary data in such strings.
6023 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6025 23. Exim was not respecting finduser_retries when seeking the login of the
6026 uid under which it was called; it was always trying 10 times. (The default
6027 setting of finduser_retries is zero.) Also, it was sleeping after the final
6028 failure, which is pointless.
6030 24. Implemented tls_on_connect_ports.
6032 25. Implemented acl_smtp_predata.
6034 26. If the domain in control=submission is set empty, Exim assumes that the
6035 authenticated id is a complete email address when it generates From: or
6036 Sender: header lines.
6038 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6039 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6040 chown and chgrp in /bin and hostname in /usr/bin.
6042 28. Exim was keeping the "process log" file open after each use, just as it
6043 does for the main log. This opens the possibility of it remaining open for
6044 long periods when the USR1 signal hits a daemon. Occasional processlog
6045 errors were reported, that could have been caused by this. Anyway, it seems
6046 much more sensible not to leave this file open at all, so that is what now
6049 29. The long-running daemon process does not normally write to the log once it
6050 has entered its main loop, and it closes the log before doing so. This is
6051 so that log files can straightforwardly be renamed and moved. However,
6052 there are a couple of unusual error situations where the daemon does write
6053 log entries, and I had neglected to close the log afterwards.
6055 30. The text of an SMTP error response that was received during a remote
6056 delivery was being truncated at 512 bytes. This is too short for some of
6057 the long messages that one sometimes sees. I've increased the limit to
6060 31. It is now possible to make retry rules that apply only when a message has a
6061 specific sender, in particular, an empty sender.
6063 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6064 it possible to be selective about when SMTP synchronization is enforced.
6066 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6068 32. Implemented hosts_connection_nolog.
6070 33. Added an ACL for QUIT.
6072 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6075 35. Added mailbox_size and mailbox_filecount to appendfile.
6077 36. Added control = no_multiline_responses to ACLs.
6079 37. There was a bug in the logic of the code that waits for the clock to tick
6080 in the case where the clock went backwards by a substantial amount such
6081 that the microsecond fraction of "now" was more than the microsecond
6082 fraction of "then" (but the whole seconds number was less).
6084 38. Added support for the libradius Radius client library this is found on
6085 FreeBSD (previously only the radiusclient library was supported).
6091 1. When certain lookups returned multiple values in the form name=value, the
6092 quoting of the values was not always being done properly. Specifically:
6093 (a) If the value started with a double quote, but contained no whitespace,
6095 (b) If the value contained whitespace other than a space character (i.e.
6096 tabs or newlines or carriage returns) it was not quoted.
6097 This fix has been applied to the mysql and pgsql lookups by writing a
6098 separate quoting function and calling it from the lookup code. The fix
6099 should probably also be applied to nisplus, ibase and oracle lookups, but
6100 since I cannot test any of those, I have not disturbed their existing code.
6102 2. A hit in the callout cache for a specific address caused a log line with no
6103 reason for rejecting RCPT. Now it says "Previous (cached) callout
6104 verification failure".
6106 3. There was an off-by-one bug in the queryprogram router. An over-long
6107 return line was truncated at 256 instead of 255 characters, thereby
6108 overflowing its buffer with the terminating zero. As well as fixing this, I
6109 have increased the buffer size to 1024 (and made a note to document this).
6111 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6112 when Exim is waiting for an SMTP response from a remote server, Exim
6113 restarts its select() call on the socket, thereby resetting its timeout.
6114 This is not a problem when such interrupts are rare. Somebody set up a cron
6115 job to run exiwhat every 2 minutes, which is less than the normal select()
6116 timeout (5 or 10 minutes). This meant that the select() timeout never
6117 kicked in because it was always reset. I have fixed this by comparing the
6118 time when an interrupt arrives with the time at the start of the first call
6119 to select(). If more time than the timeout has elapsed, the interrupt is
6120 treated as a timeout.
6122 5. Some internal re-factoring in preparation for the addition of Sieve
6123 extensions (by MH). In particular, the "personal" test is moved to a
6124 separate function, and given an option for scanning Cc: and Bcc: (which is
6125 not set for Exim filters).
6127 6. When Exim created an email address using the login of the caller as the
6128 local part (e.g. when creating a From: or Sender: header line), it was not
6129 quoting the local part when it contained special characters such as @.
6131 7. Installed new OpenBSD configuration files.
6133 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6134 try to make them clearer.
6136 9. Callout options, other than the timeout value, were being ignored when
6137 verifying sender addresses in header lines. For example, when using
6139 verify = header_sender/callout=no_cache
6141 the cache was (incorrectly) being used.
6143 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6144 only the Cygwin environment.
6146 11. When return_path_on_delivery was set as a log selector, if different remote
6147 addresses in the same message used different return paths and parallel
6148 remote delivery occurred, the wrong values would sometimes be logged.
6149 (Whenever a remote delivery process finished, the return path value from
6150 the most recently started remote delivery process was logged.)
6152 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6153 lines when AUTH and/or TLS are in use. This is the "received protocol"
6154 field. Exim used to use "asmtp" for authenticated SMTP, without any
6155 indication (in the protocol name) for TLS use. Now it follows the RFC and
6156 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6157 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6158 names appear in log lines as well as in Received: header lines.
6160 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6161 extensions, and comparison tests, and to fix some bugs.
6163 14. Changes to the "personal" filter test:
6165 (1) The test was buggy in that it was just doing the equivalent of
6166 "contains" tests on header lines. For example, if a user's address was
6167 anne@some.where, the "personal" test would incorrectly be true for
6169 To: susanne@some.where
6171 This test is now done by extracting each address from the header in turn,
6172 and checking the entire address. Other tests that are part of "personal"
6173 are now done using regular expressions (for example, to check local parts
6174 of addresses in From: header lines).
6176 (2) The list of non-personal local parts in From: addresses has been
6177 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6178 taken from the Sieve specification recommendations.
6180 (3) If the message contains any header line starting with "List-" it is
6181 treated as non-personal.
6183 (4) The test for "circular" in the Subject: header line has been removed
6184 because it now seems ill-conceived.
6186 15. Minor typos in src/EDITME comments corrected.
6188 16. Installed latest exipick from John Jetmore.
6190 17. If headers_add on a router specified a text string that was too long for
6191 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6192 of string_sprintf() is now avoided.
6194 18. $message_body_size was not set (it was always zero) when running the DATA
6195 ACL and the local_scan() function.
6197 19. For the "mail" command in an Exim filter, no default was being set for
6198 the once_repeat time, causing a random time value to be used if "once" was
6199 specified. (If the value happened to be <= 0, no repeat happened.) The
6200 default is now 0s, meaning "never repeat". The "vacation" command was OK
6201 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6202 (I found it when inspecting the code).
6204 20. There is now an overall timeout for performing a callout verification. It
6205 defaults to 4 times the callout timeout, which applies to individual SMTP
6206 commands during the callout. The overall timeout applies when there is more
6207 than one host that can be tried. The timeout is checked before trying the
6208 next host. This prevents very long delays if there are a large number of
6209 hosts and all are timing out (e.g. when the network connections are timing
6210 out). The value of the overall timeout can be changed by specifying an
6211 additional sub-option for "callout", called "maxwait". For example:
6213 verify = sender/callout=5s,maxwait=20s
6215 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6216 to the end before writing, but this should make it even safer).
6218 22. Exim was forgetting that it had advertised PIPELINING for the second and
6219 subsequent messages on an SMTP connection. It was also not resetting its
6220 memory on STARTTLS and an internal HELO.
6222 23. When Exim logs an SMTP synchronization error within a session, it now
6223 records whether PIPELINING has been advertised or not.
6225 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6226 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6227 rather than long int.
6229 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6231 26. Added the never_mail option to autoreply.
6237 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6238 crash if the getsockname() call failed; this can happen if a connection is
6239 closed very soon after it is established. The problem was simply in the
6240 order in which certain operations were done, causing Exim to try to write
6241 to the SMTP stream before it had set up the file descriptor. The bug has
6242 been fixed by making things happen in the correct order.
6248 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6249 before the connection was closed, thus losing the rejection response.
6251 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6252 some early Solaris releases, but causes trouble in current releases where
6253 socklen_t is defined.
6255 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6258 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6261 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6262 that releasing the top of it at the end releases what was used for sub-
6263 expansions (unless the block got too big). However, discard this block if
6264 the first thing is a variable or header, so that we can use its block when
6265 it is dynamic (useful for very large $message_headers, for example).
6267 6. Lookups now cache *every* query, not just the most recent. A new, separate
6268 store pool is used for this. It can be recovered when all lookup caches are
6269 flushed. Lookups now release memory at the end of their result strings.
6270 This has involved some general refactoring of the lookup sources.
6272 7. Some code has been added to the store_xxx() functions to reduce the amount
6273 of flapping under certain conditions.
6275 8. log_incoming_interface used to affect only the <= reception log lines. Now
6276 it causes the local interface and port to be added to several more SMTP log
6277 lines, for example "SMTP connection from", and rejection lines.
6279 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6281 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6283 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6284 could be overwritten at the end of the current message (or the start of a
6285 new message if it was set in a HELO ACL). The value is now preserved for
6286 the duration of the SMTP connection.
6288 12. If a transport had a headers_rewrite setting, and a matching header line
6289 contained an unqualified address, that address was qualified, even if it
6290 did not match any rewriting rules. The underlying bug was that the values
6291 of the flags that permit the existence of unqualified sender and recipient
6292 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6293 non-local messages, and by -bnq for local messages) were not being
6294 preserved with the message after it was received.
6296 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6297 "next input=" as part of the text comprising the host identity instead of
6298 the correct text. The code was using the same buffer for two different
6299 strings. However, depending on which order the printing function evaluated
6300 its arguments, the bug did not always show up. Under Linux, for example, my
6301 test suite worked just fine.
6303 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6304 doesn't work with some older versions of Perl. It has been changed to "my",
6305 which in any case is probably the better facility to use.
6307 15. A really picky compiler found some instances of statements for creating
6308 error messages that either had too many or two few arguments for the format
6311 16. The size of the buffer for calls to the DNS resolver has been increased
6312 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6313 for addresses that have a lot of PTR records. This alleviates a problem; it
6314 does not fully solve it.
6316 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6317 buffer now truncates the list and logs the incident, which is the same
6318 action as happens when Exim is looking up a host name and its aliases.
6319 Previously in this situation something unpredictable would happen;
6320 sometimes it was "internal error: store_reset failed".
6322 18. If a server dropped the connection unexpectedly when an Exim client was
6323 using GnuTLS and trying to read a response, the client delivery process
6324 crashed while trying to generate an error log message.
6326 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6327 string, for example:
6329 warn message = H1: something\nH2: something
6331 the text was added as a single header line from Exim's point of view
6332 though it ended up OK in the delivered message. However, searching for the
6333 second and subsequent header lines using $h_h2: did not work. This has been
6334 fixed. Similarly, if a system filter added multiple headers in this way,
6335 the routers could not see them.
6337 20. Expanded the error message when iplsearch is called with an invalid key to
6338 suggest using net-iplsearch in a host list.
6340 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6341 ACLs are no longer actually imposed (and a message to that effect is
6344 22. If a "gecos" field in a passwd entry contained escaped characters, in
6345 particular, if it contained a \" sequence, Exim got it wrong when building
6346 a From: or a Sender: header from that name. A second bug also caused
6347 incorrect handling when an unquoted " was present following a character
6348 that needed quoting.
6350 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6351 was not being matched caselessly.
6353 24. Arranged for all hyphens in the exim.8 source to be escaped with
6356 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6357 properly. Recipient callout cache records were still being keyed to include
6358 the sender, even when use_sender was set false. This led to far more
6359 callouts that were necessary. The sender is no longer included in the key
6360 when use_sender is false.
6362 26. Added "control = submission" modifier to ACLs.
6364 27. Added the ${base62d: operator to decode base 62 numbers.
6366 28. dnsdb lookups can now access SRV records.
6368 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6369 the configuration file.
6371 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6372 (-v) mode. This makes the output for a verbose queue run more intelligible.
6374 31. Added a use_postmaster feature to recipient callouts.
6376 32. Added the $body_zerocount variable, containing the number of binary zero
6377 bytes in the message body.
6379 33. The time of last modification of the "new" subdirectory is now used as the
6380 "mailbox time last read" when there is a quota error for a maildir
6383 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6385 35. Added +ignore_unknown as a special item in host lists.
6387 36. Code for decoding IPv6 addresses in host lists is now included, even if
6388 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6389 address was recognized as an IP address, but was then not correctly decoded
6390 into binary, causing unexpected and incorrect effects when compared with
6397 1. Very minor rewording of debugging text in manualroute to say "list of
6398 hosts" instead of "hostlist".
6400 2. If verify=header_syntax was set, and a header line with an unqualified
6401 address (no domain) and a large number of spaces between the end of the
6402 name and the colon was received, the reception process suffered a buffer
6403 overflow, and (when I tested it) crashed. This was caused by some obsolete
6404 code that should have been removed. The fix is to remove it!
6406 3. When running in the test harness, delay a bit after writing a bounce
6407 message to get a bit more predictability in the log output.
6409 4. Added a call to search_tidyup() just before forking a reception process. In
6410 theory, someone could use a lookup in the expansion of smtp_accept_max_
6411 per_host which, without the tidyup, could leave open a database connection.
6413 5. Added the variables $recipient_data and $sender_data which get set from a
6414 lookup success in an ACL "recipients" or "senders" condition, or a router
6415 "senders" option, similar to $domain_data and $local_part_data.
6417 6. Moved the writing of debug_print from before to after the "senders" test
6420 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6421 problems for message scanning, either using a data ACL, or using
6422 local_scan() because the Received: header was not generated till after they
6423 were called (in order to set the time as the time of reception completion).
6424 I have revised the way this works. The header is now generated after the
6425 body is received, but before the ACL or local_scan() are called. After they
6426 are run, the timestamp in the header is updated.
6432 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6433 before starting a queue runner without re-exec. This happened only when
6434 deliver_drop_privilege was set or when the Exim user was set to root. The
6435 effect of the bug was that timeouts during subsequent deliveries caused
6436 crashes instead of being properly handled. The handler is now left at its
6437 default (and expected) setting.
6439 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6440 message, again when deliver_drop_privilege is set or Exim is run as root.
6441 The bug described in (1) was not present in this case, but the tidying up
6442 of the other signals was missing. I have made the two cases consistent.
6444 3. The ignore_target_hosts setting on a manualroute router was being ignored
6445 for hosts that were looked up using the /MX notation.
6447 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6450 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6451 operated on the sender address. After changing the $sender_address to <>
6452 for the sender address verify, Exim was re-instated it as the original
6453 (before rewriting) address, but remembering that it had rewritten it, so it
6454 wasn't rewriting it again. This bug also had the effect of breaking the
6455 sender address verification caching when the sender address was rewritten.
6457 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6458 This has been changed so that if the ip literal address matches
6459 ignore_target_hosts, the router declines.
6461 7. Added expansion conditions match_domain, match_address, and match_local_
6462 part (NOT match_host).
6464 8. The placeholder for the Received: header didn't have a length field set.
6466 9. Added code to Exim itself and to exim_lock to test for a specific race
6467 condition that could lead to file corruption when using MBX delivery. The
6468 issue is with the lockfile that is created in /tmp. If this file is removed
6469 after a process has opened it but before that process has acquired a lock,
6470 there is the potential for a second process to recreate the file and also
6471 acquire a lock. This could lead to two Exim processes writing to the file
6472 at the same time. The added code performs the same test as UW imapd; it
6473 checks after acquiring the lock that its file descriptor still refers to
6474 the same named file.
6476 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6477 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6480 11. Added the smtp_active_hostname option. If used, this will typically be made
6481 to depend on the incoming interface address. Because $interface_address is
6482 not set up until the daemon has forked a reception process, error responses
6483 that can happen earlier (such as "too many connections") no longer contain
6486 12. If an expansion in a condition on a "warn" statement fails because a lookup
6487 defers, the "warn" statement is abandoned, and the next ACL statement is
6488 processed. Previously this caused the whole ACL to be aborted.
6490 13. Added the iplsearch lookup type.
6492 14. Added ident_timeout as a log selector.
6494 15. Added tls_certificate_verified as a log selector.
6496 16. Added a global option tls_require_ciphers (compare the smtp transport
6497 option of the same name). This controls incoming TLS connections.
6499 17. I finally figured out how to make tls_require_ciphers do a similar thing
6500 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6501 before starting the TLS session.
6503 18. Tabs are now shown as \t in -bP output.
6505 19. If the log selector return_path_on_delivery was set, Exim crashed when
6506 bouncing a message because it had too many Received: header lines.
6508 20. If two routers both had headers_remove settings, and the first one included
6509 a superfluous trailing colon, the final name in the first list and the
6510 first name in the second list were incorrectly joined into one item (with a
6511 colon in the middle).
6517 1. Added -C and -D options to the exinext utility, mainly to make it easier
6518 to include in the automated testing, but these could be helpful when
6519 multiple configurations are in use.
6521 2. The exinext utility was not formatting the output nicely when there was
6522 an alternate port involved in the retry record key, nor when there was a
6523 message id as well (for retries that were specific to a specific message
6524 and a specific host). It was also confused by IPv6 addresses, because of
6525 the additional colons they contain. I have fixed the IPv4 problem, and
6526 patched it up to do a reasonable job for IPv6.
6528 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6529 delivery, the log line now contains "pipelined" if PIPELINING was used.
6531 4. An SMTP transport process used to panic and die if the bind() call to set
6532 an explicit outgoing interface failed. This has been changed; it is now
6533 treated in the same way as a connect() failure.
6535 5. A reference to $sender_host_name in the part of a conditional expansion
6536 that was being skipped was still causing a DNS lookup. This no longer
6539 6. The def: expansion condition was not recognizing references to header lines
6540 that used bh_ and bheader_.
6542 7. Added the _cache feature to named lists.
6544 8. The code for checking quota_filecount in the appendfile transport was
6545 allowing one more file than it should have been.
6547 9. For compatibility with Sendmail, the command line option
6555 and sets the incoming protocol and host name (for trusted callers). The
6556 host name and its colon can be omitted when only the protocol is to be set.
6557 Note the Exim already has two private options, -pd and -ps, that refer to
6558 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6559 "s", but I don't think that's a major issue.
6561 10. A number of refactoring changes to the code, none of which should affect
6564 (a) The number of logging options was getting close to filling up the
6565 32-bit word that was used as a bit map. I have split them into two classes:
6566 those that are passed in the argument to log_write(), and those that are
6567 only ever tested independently outside of that function. These are now in
6568 separate 32-bit words, so there is plenty of room for expansion again.
6569 There is no change in the user interface or the logging behaviour.
6571 (b) When building, for example, log lines, the code previously used a
6572 macro that called string_cat() twice, in order to add two strings. This is
6573 not really sufficiently general. Furthermore, there was one instance where
6574 it was actually wrong because one of the argument was used twice, and in
6575 one call a function was used. (As it happened, calling the function twice
6576 did not affect the overall behaviour.) The macro has been replaced by a
6577 function that can join an arbitrary number of extra strings onto a growing
6580 (c) The code for expansion conditions now uses a table and a binary chop
6581 instead of a serial search (which was left over from when there were very
6582 few conditions). Also, it now recognizes conditions like "pam" even when
6583 the relevant support is not compiled in: a suitably worded error message is
6584 given if an attempt is made to use such a condition.
6586 11. Added ${time_interval:xxxxx}.
6588 12. A bug was causing one of the ddress fields not to be passed back correctly
6589 from remote delivery subprocesses. The field in question was not being
6590 subsequently used, so this caused to problems in practice.
6592 13. Added new log selectors queue_time and deliver_time.
6594 14. Might have fixed a bug in maildirsizefile handling that threw up
6595 "unexpected character" debug warnings, and recalculated the data
6596 unnecessarily. In any case, I expanded the warning message to give more
6599 15. Added the message "Restricted characters in address" to the statements in
6600 the default ACL that block characters like @ and % in local parts.
6602 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6603 Three changes have been made:
6605 (a) There was a serious bug; a negative response to MAIL caused the whole
6606 recipient domain to be cached as invalid, thereby blocking all messages
6607 to all local parts at the same domain, from all senders. This bug has
6608 been fixed. The domain is no longer cached after a negative response to
6609 MAIL if the sender used is not empty.
6611 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6614 (c) A new callout option, "use_sender" has been added for people who want
6615 the modified behaviour.
6621 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6624 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6625 indeed breaks things for older releases.
6627 3. Added additional logging to the case where there is a problem reading data
6628 from a filter that is running in a subprocess using a pipe, in order to
6629 try to track down a specific problem.
6631 4. Testing facility fudge: when running in the test harness and attempting
6632 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6633 getting "No route to host". Convert this to a timeout.
6635 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6638 6. Some OS don't have socklen_t but use size_t instead. This affects the
6639 fifth argument of getsockopt() amongst other things. This is now
6640 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6641 can be set for individual OS. I have set it for SunOS5, OSF1, and
6642 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6643 some earlier ones do not.
6645 7. Change 4.30/15 was not doing the test caselessly.
6647 8. The standard form for an IPv6 address literal was being rejected by address
6648 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6649 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6650 this, as well as the form without the "IPv6" on the front (but only when
6651 address literals are enabled, of course).
6653 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6655 10. Exim crashed if a message with an empty sender address specified by -f
6656 encountered a router with an errors_to setting. This could be provoked only
6657 by a command such as
6661 where an empty string was supplied; "<>" did not hit this bug.
6663 11. Installed PCRE release 4.5.
6665 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6666 remained set. It is now erased.
6668 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6669 times from message ids (which are base 36 rather than the normal 62).
6671 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6672 were being counted as actual protocol errors, and logged if the log
6673 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6674 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6675 and DATA following a set of rejected RCPTs do not count as protocol errors.
6676 In other words, Exim assumes they were pipelined, though this may not
6677 actually be the case. Of course, in all cases the client gets an
6678 appropriate error code.
6680 15. If a lookup fails in an ACL condition, a message about the failure may
6681 be available; it is used if testing the ACL cannot continue, because most
6682 such messages specify what the cause of the deferral is. However, some
6683 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6684 that caused an old message to be retained and used if a later statement
6685 caused a defer, replacing the real cause of the deferral.
6687 16. If an IP address had so many PTR records that the DNS lookup buffer
6688 was not large enough to hold them, Exim could crash while trying to process
6689 the truncated data. It now detects and logs this case.
6691 17. Further to 4.21/58, another change has been made: if (and only if) the
6692 first line of a message (the first header line) ends with CRLF, a bare LF
6693 in a subsequent header line has a space inserted after it, so as not to
6694 terminate the header.
6696 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6697 unnecessarily, used atoi() instead of strtol(), and didn't check the
6698 termination when getting file sizes from file names by regex.
6700 19. Completely re-implemented the support for maildirsize files, in the light
6701 of a number of problems with the previous contributed implementation
6702 (4.30/29). In particular:
6704 . If the quota is zero, the maildirsize file is maintained, but no quota is
6707 . If the maildir directory does not exist, it is created before any attempt
6708 to write a maildirsize file.
6710 . The quota value in the file is just a cache; if the quota is changed in
6711 the transport, the new value overrides.
6713 . A regular expression is available for excluding directories from the
6716 20. The autoreply transport checks the characters in options that define the
6717 message's headers; it allows continued headers, but it was checking with
6718 isspace() after an embedded newline instead of explicitly looking for a
6721 21. If all the "regular" hosts to which an address was routed had passed their
6722 expiry times, and had not reached their retry times, the address was
6723 bounced, even if fallback hosts were defined. Now Exim should go on to try
6726 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6727 equivalent code in the SMTP transport. Some hosts send humungous responses
6728 to HELO/EHLO, more than 1024 it seems.
6730 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6731 gives compiler warnings in some environments. I've now done it "properly",
6734 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6735 (because of problems with the built-in one) was declared to return uschar *
6736 instead of char *, causing compiler failure.
6738 25. Fixed a file descriptor leak when processing alias/forward files.
6740 26. Fixed a minor format string issue in dbfn.c.
6742 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6744 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6745 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6748 29. When a sender address is verified, it is cached, to save repeating the test
6749 when there is more than one recipient in a message. However, when the
6750 verification involves a callout, it is possible for different callout
6751 options to be set for different recipients. It is too complicated to keep
6752 track of this in the cache, so now Exim always runs a verification when a
6753 callout is required, relying on the callout cache for the optimization.
6754 The overhead is duplication of the address routing, but this should not be
6757 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6758 to be verified with callout=postmaster, and the main callout worked but the
6759 postmaster check failed, the verification correctly failed. However, if a
6760 subsequent RCPT command asked for sender verification *without* the
6761 postmaster check, incorrect caching caused this verification also to fail,
6764 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6765 it was not caching the DNS options (qualify_single, search_parents) that
6766 were used when the lookup failed. A subsequent lookup with different
6767 options therefore always gave the same answer, though there were cases
6768 where it should not have. (Example: a "domains = !$mx_any" option on a
6769 dnslookup router: the "domains" option is always processed without any
6770 widening, but the router might have qualify_single set.) Now Exim uses the
6771 cached value only when the same options are set.
6773 32. Added John Jetmore's "exipick" utility to the distribution.
6775 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6776 than a timeout (e.g. a certificate is required, and is not provided), an
6777 Exim server now closes the connection immediately. Previously it waited for
6778 the client to close - but if the client is SSL, it seems that they each
6779 wait for each other, leading to a delay before one of them times out.
6781 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6782 maintained 0.8.x compatibility because I don't think many are using it, and
6783 it is clearly obsolete.
6785 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6788 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6789 client certificate was expired. A simple patch fixes this, though I don't
6790 understand the full logic of why the verify callback is called multiple
6793 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6794 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6795 which causes problems with some clients (such as the Certicom SSL Plus
6796 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6797 disables the coutermeasure allowing Eudora to connect."
6799 38. Exim was not checking that a write() to a log file succeeded. This could
6800 lead to Bad Things if a log got too big, in particular if it hit a file
6801 size limit. Exim now panics and dies if it cannot write to a log file, just
6802 as it does if it cannot open a log file.
6804 39. Modified OS/Makefile-Linux so that it now contains
6806 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6808 The two -D definitions ensure that Exim is compiled with large file
6809 support, which makes it possible to handle log files that are bigger than
6812 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6813 instance) a domain was checked against a named list that involved a lookup,
6814 causing $domain_data to be set, then another domain was checked against the
6815 same list, then the first domain was re-checked, the value of $domain_data
6816 after the final check could be wrong. In particular, if the second check
6817 failed, it could be set empty. This bug probably also applied to
6820 41. The strip_trailing_dot option was not being applied to the address given
6821 with the -f command-line option.
6823 42. The code for reading a message's header from the spool was incrementing
6824 $received_count, but never initializing it. This meant that the value was
6825 incorrect (doubled) while delivering a message in the same process in which
6826 it was received. In the most common configuration of Exim, this never
6827 happens - a fresh exec is done - but it can happen when
6828 deliver_drop_privilege is set.
6830 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6831 it now includes up to 150 characters of the unexpected data in the log
6834 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6835 and building data strings. The size of both of these buffers was 10 000
6836 bytes - far larger than anybody would *ever* want, thought I. Needless to
6837 say, somebody hit the limit. I have increased the maximum line length to
6838 20 000 and the maximum data length of concatenated lines to 100 000. I have
6839 also fixed two bugs, because there was no checking on these buffers. Tsk,
6840 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6841 buffer is too small.
6843 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6844 lsearch lookups. Now it does.
6846 46. When parsing a route_list item in a manualroute router, a fixed-length
6847 buffer was used for the list of hosts. I made this 1024 bytes long,
6848 thinking that nobody would ever have a list of hosts that long. Wrong.
6849 Somebody had a whole pile of complicated expansion conditions, and the
6850 string was silently truncated, leading to an expansion error. It turns out
6851 that it is easier to change to an unlimited length (owing to other changes
6852 that have happened since this code was originally written) than to build
6853 structure for giving a limitation error. The length of the item that
6854 expands into the list of hosts is now unlimited.
6856 47. The lsearch lookup could not handle data where the length of text line was
6857 more than 4095 characters. Such lines were truncated, leading to shortened
6858 data being returned. It should now handle lines of any length.
6860 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6861 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6864 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6865 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6866 no longer happens in comment lines. A list of replacements is now placed
6867 at the head of all of the source files, except those whose only change is
6868 to replace PERL_COMMAND in the very first #! line.
6870 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6871 messages on the queue, with a bottom-up merge sort, using code contributed
6872 by Michael Haardt. This should make operations like -bp somewhat faster on
6873 large queues. It won't affect queue runners, except when queue_run_in_order
6876 51. Installed eximstats 1.31 in the distribution.
6878 52. Added support for SRV lookups to the dnslookup router.
6880 53. If an ACL referred to $message_body or $message_body_end, the value was not
6881 reset for any messages that followed in the same SMTP session.
6883 54. The store-handling optimization for building very long strings was not
6884 differentiating between the different store pools. I don't think this
6885 actually made any difference in practice, but I've tidied it.
6887 55. While running the routers to verify a sender address, $sender_address
6888 was still set to the sender address. This is wrong, because when routing to
6889 send a bounce to the sender, it would be empty. Therefore, I have changed
6890 it so that, while verifying a sender address, $sender_address is set to <>.
6891 (There is no change to what happens when verifying a recipient address.)
6893 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6894 target A or AAAA records (if not already returned) without resetting the
6895 qualify_single or search_parents options of the DNS resolver. These are
6896 inappropriate in this case because the targets of MX and SRV records must
6897 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6898 target that, when qualified, matched something in the local domain. These
6899 two options are now turned off when doing these lookups.
6901 57. It seems that at least some releases of Reiserfs (which does not have the
6902 concept of a fixed number of inodes) returns zero and not -1 for the
6903 number of available inodes. This interacted badly with check_spool_inodes,
6904 which assumed that -1 was the "no such thing" setting. What I have done is
6905 to check that the total number of inodes is greater than zero before doing
6906 the test of how many are available.
6908 58. When a "warn" ACL statement has a log_message modifier, the message is
6909 remembered, and not repeated. This is to avoid a lot of repetition when a
6910 message has many recipients that cause the same warning to be written.
6911 However, Exim was preserving the list of already written lines for an
6912 entire SMTP session, which doesn't seem right. The memory is now reset if a
6913 new message is started.
6915 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6916 debugging output unless log_rewrite was also set.
6918 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6919 of (int)(handle) when we know that handle contains (void *)(-1).
6921 61. The Exim daemon panic-logs an error return when it closes the incoming
6922 connection. However "connection reset by peer" seems to be common, and
6923 isn't really an error worthy of noting specially, so that particular error
6926 62. When Exim is trying to find all the local interfaces, it used to panic and
6927 die if the ioctl to get the interface flags failed. However, it seems that
6928 on at least one OS (Solaris 9) it is possible to have an interface that is
6929 included in the list of interfaces, but for which you get a failure error
6930 for this call. This happens when the interface is not "plumbed" into a
6931 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6932 failure of the "get flags" call assumes that the interface is down.
6934 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6935 makes life easier for people who are doing arithmetic on fields extracted
6936 from dates, where you often get leading zeros that should not be
6937 interpreted as octal.
6939 64. Added qualify_domain to the redirect router, to override the global
6942 65. If a pathologically long header line contained very many addresses (the
6943 report of this problem mentioned 10 000) and each of them was rewritten,
6944 Exim could use up a very large amount of memory. (It kept on making new
6945 copies of the header line as it rewrote, and never released the old ones.)
6946 At the expense of a bit more processing, the header rewriting function has
6947 been changed so that it no longer eats memory in this way.
6949 66. The generation of the Received: header has been moved from the time that a
6950 message starts to be received, to the time that it finishes. The timestamp
6951 in the Received: header should now be very close to that of the <= log
6952 line. There are two side-effects of this change:
6954 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6955 logged header lines no longer include the local Received: line, because
6956 it has not yet been created. The same applies to a copy of the message
6957 that is returned to a non-SMTP sender when a message is rejected.
6959 (b) When a filter file is tested using -bf, no additional Received: header
6960 is added to the test message. After some thought, I decided that this
6963 This change does not affect the value of $received_for. It is still set
6964 after address rewriting, but before local_scan() is called.
6966 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6968 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6969 gave an unhelpful panic error message, and a defer error. I have managed to
6970 change this behaviour so that it now rejects any supplied certificate,
6971 which seems right, as the list of acceptable certificates is empty.
6973 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6974 gave an unhelpful defer error. I have not managed to make this reject any
6975 supplied certificates, but the error message it gives is "no certificate
6976 supplied", which is not helpful.
6978 70. exigrep's output now also includes lines that are not associated with any
6979 message, but which match the given pattern. Implemented by a patch from
6980 Martin Sluka, which also tidied up the Perl a bit.
6982 71. Recipient callout verification, like sender verification, was using <> in
6983 the MAIL FROM command. This isn't really the right thing, since the actual
6984 sender may affect whether the remote host accepts the recipient or not. I
6985 have changed it to use the actual sender in the callout; this means that
6986 the cache record is now keyed on a recipient/sender pair, not just the
6987 recipient address. There doesn't seem to be a real danger of callout loops,
6988 since a callout by the remote host to check the sender would use <>.
6989 [SEE ABOVE: changed after hitting problems.]
6991 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6992 temporary errors. However, in the case of such a code being given after
6993 the end of a data transmission (i.e. after ".") Exim was failing to write
6994 a retry record for the message. (Yes, there was some broken host that was
6995 actually sending 8xx at this point.)
6997 73. An unknown lookup type in a host list could cause Exim to panic-die when
6998 the list was checked. (An example that provoked this was putting <; in the
6999 middle of a list instead of at the start.) If this happened during a DATA
7000 ACL check, a -D file could be left lying around. This kind of configuration
7001 error no longer causes Exim to die; instead it causes a defer error. The
7002 incident is still logged to the main and panic logs.
7004 74. Buglet left over from Exim 3 conversion. The message "too many messages
7005 in one connection" was written to the rejectlog but not the mainlog, except
7006 when address rewriting (yes!) was being logged.
7008 75. Added write_rejectlog option.
7010 76. When a system filter was run not as root (that is, when system_filter_user
7011 was set), the values of the $n variables were not being returned to the
7012 main process; thus, they were not subsequently available in the $sn
7015 77. Added +return_path_on_delivery log selector.
7017 78. A connection timeout was being treated differently from recipients deferred
7018 when testing hosts_max_try with a message that was older than the host's
7019 retry timeout. (The host should not be counted, thus allowing all hosts to
7020 be tried at least once before bouncing.) This may have been the cause of an
7021 occasionally reported bug whereby a message would remain on the queue
7022 longer than the retry timeout, but would be bounced if a delivery was
7023 forced. I say "may" because I never totally pinned down the problem;
7024 setting up timeout/retry tests is difficult. See also the next item.
7026 79. The ultimate address timeout was not being applied to errors that involved
7027 a combination of host plus message (for example, a timeout on a MAIL
7028 command). When an address resolved to a number of possible hosts, and they
7029 were not all tried for each delivery (e.g. because of hosts_max_try), a
7030 message could remain on the queue longer than the retry timeout.
7032 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7035 81. Fixed an obscure SMTP outgoing bug which required at least the following
7036 conditions: (a) there was another message waiting for the same server;
7037 (b) the server returned 5xx to all RCPT commands in the first message so
7038 that the message was not completed; (c) the server dropped the connection
7039 or gave a negative response to the RSET that Exim sends to abort the
7040 transaction. The observed case was a dropped connection after DATA that had
7041 been sent in pipelining mode. That is, the server had advertised PIPELINING
7042 but was not implementing it correctly. The effect of the bug was incorrect
7043 behaviour, such as trying another host, and this could lead to a crash.
7049 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7050 and daemon.c were passed as pointers to ints; they should have been
7051 pointers to socklen_t variables (which are typically unsigned ints).
7053 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7056 3. Fixed a really odd bug that affected only the testing scheme; patching a
7057 certain fixed string in the binary changed the value of another string that
7058 happened to be identical to the end of the original first string.
7060 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7061 name", it returns that address as the IP address. On some operating
7062 systems (e.g. Solaris), it also passes back the IP address string as the
7063 "host name". However, on others (e.g. Linux), it passes back an empty
7064 string. Exim wasn't checking for this, and was changing the host name to an
7065 empty string, assuming it had been canonicalized.
7067 5. Although rare, it is permitted to have more than one PTR record for a given
7068 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7069 all the names associated with an address, because they do in Solaris.
7070 However, it seems that they do not in Linux for data that comes from the
7071 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7072 I found this out when I moved to a new Linux workstation and tried to run
7073 the Exim test suite.
7075 To get round this problem I have changed the code so that it now does its
7076 own call to the DNS to look up PTR records when searching for a host name.
7077 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7078 addresses that are only in /etc/hosts are still found.
7080 This behaviour is, however, controlled by an option called host_lookup_
7081 order, which defaults to "bydns:byaddr". If people want to use the other
7082 order, or indeed, just use one or the other means of lookup, they can
7083 specify it in this variable.
7085 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7086 some operating systems, this comes back from gethostbyaddr() as an empty
7087 string, and this is what Exim used to test for. However, it seems that in
7088 other systems, "." is yielded. Exim now tests for this case too.
7090 7. The values of check_spool_space and check_log_space are now held internally
7091 as a number of kilobytes instead of an absolute number of bytes. If a
7092 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7093 kilobyte. This means that much larger values can be stored.
7095 8. Exim monitor: an attempt to get the action menu when not actually pointing
7096 at a message produces an empty menu entitled "No message selected". This
7097 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7098 no entries in it ("Shell widget menu has zero width and/or height"). So I
7099 have added a single, blank menu entry in this case.
7101 9. Added ${quote_local_part.
7103 10. MIME decoding is now applied to the contents of Subject: header lines when
7106 11. Now that a reference to $sender_host_address automatically causes a reverse
7107 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7108 host lookup before query-style lookups in lists that might use this
7109 variable. This has therefore been abolished, and the "net-" prefix is no
7110 longer necessary for query-style lookups.
7112 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7113 have been a typo for LFLAGS, so it has been changed.
7115 13. The install script calls Exim with "-C /dev/null" in order to find the
7116 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7117 to be output. However, since Exim outputs its version number before the
7118 error, it didn't break the script. It just looked ugly. I fixed this by
7119 always allowing "-C /dev/null" if the caller is root.
7121 14. Ignore overlarge ACL variable number when reading spool file - insurance
7122 against a later release with more variables having written the file.
7124 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7125 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7126 this, as well as the form without the "IPv6" on the front.
7128 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7129 OS/Makefile-Darwin file.
7131 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7132 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7134 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7135 variable. However, some OS use ints for inodes. I've added cast to long int
7136 to get rid of the compiler warning.
7138 19. I had forgotten to lock out "/../" in configuration file names when
7139 ALT_CONFIG_PREFIX was set.
7141 20. Routers used for verification do not need to specify transports. However,
7142 if such a router generated a host list, and callout was configured, Exim
7143 crashed, because it could not find a port number from the (non-existent)
7144 transport. It now assumes port 25 in this circumstance.
7146 21. Added the -t option to exigrep.
7148 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7149 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7150 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7151 them is set, LOOKUP_LSEARCH is forced.
7153 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7155 24. Added sender and host information to the "rejected by local_scan()" log
7156 line; previously there was no indication of these.
7158 25. Added .include_if_exists.
7160 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7161 Linux. It turns out that not all file systems support this. Apparently some
7162 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7163 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7164 not supported on the file descriptor, is now ignored when Exim is trying to
7165 sync a directory. This applies only to Linux.
7167 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7169 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7170 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7173 29. Added support for maildirsize files from supplied patch (modified a bit).
7175 30. The use of :fail: followed by an empty string could lead Exim to respond to
7176 sender verification failures with (e.g.):
7178 550 Verification failed for <xxx>
7179 550 Sender verify failed
7181 where the first response line was missing the '-' that indicates it is not
7182 the final line of the response.
7184 31. The loop for finding the name of the user that called Exim had a hardwired
7185 limit of 10; it now uses the value of finduser_retries, which is used for
7186 all other user lookups.
7188 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7191 33. Exim was neglecting to zero errno before one call of strtol() when
7192 expanding a string and expecting an integer value. On some systems this
7193 resulted in spurious "integer overflow" errors. Also, it was casting the
7194 result into an int without checking.
7196 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7197 did not work. The code looks as if it has *never* worked, though it appears
7198 to have been documented since at least release 1.62. I have made it work.
7200 35. The "timeout_DNS" error in retry rules, also documented since at least
7201 1.62, also never worked. As it isn't clear exactly what this means, and
7202 clearly it isn't a major issue, I have abolished the feature by treating it
7203 as "timeout", and writing a warning to the main and panic logs.
7205 36. The display of retry rules for -brt wasn't always showing the error code
7208 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7209 timeout_connect_A, timeout_connect_MX.
7211 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7212 to the empty sender.
7214 39. The daemon was not analysing the content of -oX till after it had closed
7215 stderr and disconnected from the controlling terminal. This meant that any
7216 syntax errors were only noted on the panic log, and the return code from
7217 the command was 0. By re-arranging the code a little, I've made the
7218 decoding happen first, so such errors now appear on stderr, and the return
7219 code is 1. However, the actual setting up of the sockets still happens in
7220 the disconnected process, so errors there are still only recorded on the
7223 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7224 connections (as happens on some IP stacks) was logged at start up time as
7225 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7226 it from "IPv6 and IPv4", which means that two separate sockets are being
7229 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7230 says whether AF_INET or AF_INET6 was passed as an argument.
7232 42. Exiwhat output was messed up when time zones were included in log
7235 43. Exiwhat now gives more information about the daemon's listening ports,
7236 and whether -tls-on-connect was used.
7238 44. The "port" option of the smtp transport is now expanded.
7240 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7241 silently ignored. Now an error message is written to the main and panic
7244 46. There's a new ACL modifier called "logwrite" which writes to a log file
7245 as soon as it is encountered.
7247 47. Added $local_user_uid and $local_user_gid at routing time.
7249 48. Exim crashed when trying to verify a sender address that was being
7252 49. Exim was recognizing only a space character after ".include". It now also
7253 recognizes a tab character.
7255 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7256 extracting the relevant information from the specification. The man page no
7257 longer contains scrambled data for the -d option, and I've added a section
7258 at the front about calling Exim under different names.
7260 51. Added "extra_headers" argument to the "mail" command in filter files.
7262 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7265 53. Installed eximstats 1.29.
7267 54. Added transport_filter_timeout as a generic transport option.
7269 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7270 Cc: header lines. This was required by RFC 822, but it not required by RFC
7273 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7274 incoming messages that did not have them. Now it does so only if the
7275 message originates locally, that is, if there is no associated remote host
7276 address. When Resent- header lines are present, this applies to the Resent-
7277 lines rather than the non-Resent- lines.
7279 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7280 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7282 58. Messages for configuration errors now include the name of the main
7283 configuration file - useful now that there may be more than one file in a
7284 list (.included file names were always shown).
7286 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7287 for those rare installations that do not start the daemon as root or run it
7288 setuid root. I've cut out the call to initgroups() if the daemon is not
7291 60. The Exim user and group can now be bound into the binary as text strings
7292 that are looked up at the start of Exim's processing.
7294 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7296 62. Added $mailstore_basename variable.
7298 63. Installed patch to sieve.c from Michael Haardt.
7300 64. When Exim failed to open the panic log after failing to open the main log,
7301 the original message it was trying to log was written to stderr and debug
7302 output, but if they were not available (the usual case in production), it
7303 was lost. Now it is written to syslog before the two lines that record the
7304 failures to open the logs.
7306 65. Users' Exim filters run in subprocesses under the user's uid. It is
7307 possible for a "deliver" command or an alias in a "personal" command to
7308 provoke an address rewrite. If logging of address rewriting is configured,
7309 this fails because the process is not running as root or exim. There may be
7310 a better way of dealing with this, but for the moment (because 4.30 needs
7311 to be released), I have disabled address rewrite logging when running a
7312 filter in a non-root, non-exim process.
7318 1. The buildconfig auxiliary program wasn't quoting the value set for
7319 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7320 not defined. This bug was masked in 4.22 by the effect that was fixed in
7323 2. Some messages that were rejected after a message id was allocated were
7324 shown as "incomplete" by exigrep. It no longer does this for messages that
7325 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7327 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7328 have allow_domain_literals set, the ID did not get logged in the <= line.
7329 Domain literals are now always recognized in Message-ID: header lines.
7331 4. The first argument for a ${extract expansion item is the key name or field
7332 number. Leading and trailing spaces in this item were not being ignored,
7333 causing some misleading effects.
7335 5. When deliver_drop_privilege was set, single queue runner processes started
7336 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7337 same command in the process it spins off) were not dropping privilege.
7339 6. When the daemon running as "exim" started a queue runner, it always
7340 re-executed Exim in the spun-off process. This is a waste of effort when
7341 deliver_drop_privilege is set. The new process now just calls the
7342 queue-runner function directly.
7348 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7351 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7352 ignored. Though the use of -oP was forcing the writing of a pid file, it
7353 was always written to the default place.
7355 3. If the message "no IP address found for host xxxx" is generated during
7356 incoming verification, it is now followed by identification of the incoming
7357 connection (so you can more easily find what provoked it).
7359 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7361 5. Added some features to "harden" Exim a bit more against certain attacks:
7363 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7364 be put in Local/Makefile. This is like the never_users runtime option,
7365 but it cannot be overridden. The default setting is "root".
7367 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7368 prefix string with which any file named in a -C command line option
7371 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7372 is retained for -C and -D only if the caller of Exim is root. Without
7373 it, the exim user may also use -C and -D and retain privilege.
7375 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7376 command line option is disabled.
7378 6. Macro names set by the -D option must start with an upper case letter, just
7379 like macro names defined in the configuration file.
7381 7. Added "dereference=" facility to LDAP.
7383 8. Two instances of the typo "uknown" in the source files are fixed.
7385 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7386 the Configure-Makefile script screwed up while processing it.
7388 10. Incorporated PCRE 4.4.
7390 11. The SMTP synchronization check was not operating right at the start of an
7391 SMTP session. For example, it could not catch a HELO sent before the client
7392 waited for the greeting. There is now a check for outstanding input at the
7393 point when the greeting is written. Because of the duplex, asynchronous
7394 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7395 way, but not yet received, when the check is performed.
7397 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7398 on TCP/IP sockets, because this apparently causes some broken clients to
7401 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7402 unchanged) from the Cygwin maintainer.
7404 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7405 was supported instead of testing for mbx. Effectively a typo.
7407 15. The spa authenticator server code was not checking that the input it
7408 received was valid base64.
7410 16. The debug output line for the "set" modifier in ACLs was not showing the
7411 name of the variable that was being set.
7413 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7415 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7416 lookup on its own. Something else was needed to trigger the lookup. For
7417 example, a match in host_lookup or the need for a host name in a host list.
7418 Now, if $sender_host_name is referenced and the host name has not yet been
7419 looked up, a lookup is performed. If the lookup fails, the variable remains
7420 empty, and $host_lookup_failed is set to "1".
7422 19. Added "eqi" as a case-independent comparison operator.
7424 20. The saslauthd authentication condition could segfault if neither service
7425 nor realm was specified.
7427 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7428 error message that was logged was misleading, and incoming SMTP
7429 connections were dropped. The message is now more accurate, and temporary
7430 errors are given to SMTP connections.
7432 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7433 (except RSET) with a 421 error, until QUIT is received. However, it was
7434 failing to send a response to QUIT.
7436 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7437 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7438 ACL is now run after the helo_try_verify_hosts code.
7440 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7441 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7442 case-independent, so other case variants are also recognized). Apparently
7443 some systems use these upper case variants.
7445 25. If more than two messages were waiting for the same host, and a transport
7446 filter was specified for the transport, Exim sent two messages over the
7447 same TCP/IP connection, and then failed with "socket operation on non-
7448 socket" when it tried to send the third.
7450 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7452 27. The extern definition of crypt16() in expand.c was not being excluded when
7453 the OS had its own crypt16() function.
7455 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7456 as a preferred synonym for return_size_limit, both as an option and as an
7459 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7461 30. Changed the default configuration ACL to relax the local part checking rule
7462 for addresses that are not in any local domains. For these addresses,
7463 slashes and pipe symbols are allowed within local parts, but the sequence
7464 /../ is explicitly forbidden.
7466 31. SPA server authentication was not clearing the challenge buffer before
7469 32. log_message in a "warn" ACL statement was writing to the reject log as
7470 well as to the main log, which contradicts the documentation and doesn't
7471 seem right (because no rejection is happening). So I have stopped it.
7473 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7474 However, I am unable to do any testing of this.
7476 34. Fixed an infelicity in the appendfile transport. When checking directories
7477 for a mailbox, to see if any needed to be created, it was accidentally
7478 using path names with one or more superfluous leading slashes; tracing
7479 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7481 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7482 contents are added to the log line that is written for every discarded
7483 recipient. (Previously a log_message setting was ignored.)
7485 36. The ${quote: operator now quotes the string if it is empty.
7487 37. The install script runs exim in order to find its version number. If for
7488 some reason other than non-existence or emptiness, which it checks, it
7489 could not run './exim', it was installing it with an empty version number,
7490 i.e. as "exim-". This error state is now caught, and the installation is
7493 38. An argument was missing from the function that creates an error message
7494 when Exim fails to connect to the socket for saslauthd authentication.
7495 This could cause Exim to crash, or give a corrupted message.
7497 39. Added isip, isip4, and isip6 to ${if conditions.
7499 40. The ACL variables $acl_xx are now saved with the message, and can be
7500 accessed later in routers, transports, and filters.
7502 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7503 strings in the file are not string-expanded.
7505 42. If a MAIL command specified a SIZE value that was too large to fit into an
7506 int variable, the check against message_size_limit failed. Such values are
7507 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7508 day this will have to be increased, but I don't think I want to be around
7509 when emails are that large.
7516 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7517 iconv() is not standard in FreeBSD.
7519 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7520 IPv6 enabled. The observed symptom was a segmentation fault on return from
7521 the function os_common_find_running_interfaces() in src/os.c.
7523 3. In the check_special_case() function in daemon.c I had used "errno" as an
7524 argument name, which causes warnings on some systems. This was basically a
7525 typo, since it was named "eno" in the comments!
7527 4. The code that waits for the clock to tick (at a resolution of some fraction
7528 of a second) so as to ensure message-id uniqueness was always waiting for
7529 at least one whole tick, when it could have waited for less. [This is
7530 almost certainly not relevant at current processor speeds, where it is
7531 unlikely to ever wait at all. But we try to future-proof.]
7533 5. The function that sleeps for a time interval that includes fractions of a
7534 second contained a race. It did not block SIGALRM between setting the
7535 timer, and suspending (a couple of lines later). If the interval was short
7536 and the sigsuspend() was delayed until after it had expired, the suspension
7537 never ended. On busy systems this could lead to processes getting stuck for
7540 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7541 process, before it forks any delivery processes. The open lookup caching
7542 mechanism meant that the open file or database connection was passed into
7543 the delivery process. The problem was that delivery processes always tidy
7544 up cached lookup data. This could cause a problem for the next delivery
7545 process started by the queue runner, because the external queue runner
7546 process does not know about the closure. So the next delivery process
7547 still has data in the lookup cache. In the case of a file lookup, there was
7548 no problem because closing a file descriptor in a subprocess doesn't affect
7549 the parent. However, if the lookup was caching a connection to a database,
7550 the connection was closed, and the second delivery process was likely to
7551 see errors such as "PGSQL: query failed: server closed the connection
7552 unexpectedly". The problem has been fixed by closing all cached lookups
7553 in a queue runner before running a delivery process.
7555 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7556 seem to have the "const" qualifier which it has on other OS. I've
7559 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7560 *of the same type* (client or server) with the same public name that an
7561 error should be diagnosed.
7563 9. When Exim looked up a host name for an IP address, but failed to find the
7564 original IP address when looking up the host name (a safety check), it
7565 output the message "<ip address> does not match any IP for NULL", which was
7566 confusing, to say the least. The bug was that the host name should have
7567 appeared instead of "NULL".
7569 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7570 user that is built into the binary, and the -C or -D options is used, root
7571 privilege is dropped before the configuration file is read. In addition,
7572 logging is switched to stderr instead of the normal log files. If the
7573 configuration then re-defines the Exim user, the unprivileged environment
7574 is probably not what is expected, so Exim logs a panic warning message (but
7577 However, if deliver_drop_privilege is set, the unprivileged state may well
7578 be exactly what is intended, so the warning has been cut out in that case,
7579 and Exim is allowed to try to write to its normal log files.
7585 1. smtp_return_error_details was not giving details for temporary sender
7586 or receiver verification errors.
7588 2. Diagnose a configuration error if two authenticators have the same public
7591 3. Exim used not to create the message log file for a message until the first
7592 delivery attempt. This could be confusing when incoming messages were held
7593 for policy or load reasons. The message log file is now created at the time
7594 the message is received, and an initial "Received" line is written to it.
7596 4. The automatically generated man page for command line options had a minor
7597 bug that caused no ill effects; however, a more serious problem was that
7598 the procedure for building the man page automatically didn't always
7599 operate. Consequently, release 4.20 contains an out-of-date version. This
7600 shouldn't happen again.
7602 5. When building Exim with embedded Perl support, the script that builds the
7603 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7604 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7606 6. The freeze_tell option was not being used for messages that were frozen on
7607 arrival, either by an ACL or by local_scan().
7609 7. Added the smtp_incomplete_transaction log selector.
7611 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7612 was accepting AUTH without a new EHLO.
7614 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7615 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7618 10. Logging of TCP/IP connections (when configured) now happens in the main
7619 daemon process instead of the child process, so that the TCP/IP connection
7620 count is more accurate (but it can never be perfect).
7622 11. The use of "drop" in a nested ACL was not being handled correctly in the
7623 outer ACL. Now, if condition failure induced by the nested "drop" causes
7624 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7625 or "require"), the connection is dropped.
7627 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7628 that yield "discard" can now be used with an "accept" or a "discard" verb,
7629 but an error is generated for any others (because I can't see a useful way
7630 to define what should happen).
7632 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7633 are now processed in the same way as lines in the Exim configuration file.
7634 In particular, continuation lines are supported.
7636 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7638 15. Added -ti meaning -t -i.
7640 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7641 domains, and warn by logging if others are found.
7643 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7644 returned from GIFCONF when Exim is trying to find the list of interfaces on
7645 a host. The code in os.c has been modified to copy each ifreq to an aligned
7646 structure in all cases.
7648 Also, in some cases, the returned ifreq's were being copied to a 'struct
7649 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7650 means the last couple of bytes of an IPv6 address could be chopped if the
7651 ifreq contained only a normal sockaddr (14 bytes storage).
7653 18. Named domain lists were not supported in the hosts_treat_as_local option.
7654 An entry such as +xxxx was not recognized, and was treated as a literal
7657 19. Ensure that header lines added by a DATA ACL are included in the reject log
7658 if the ACL subsequently rejects the message.
7660 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7661 MD5 (which is deprecated).
7663 21. When testing a filter file using -bf, Exim was writing a message when it
7664 took the sender from a "From " line in the message, but it was not doing so
7665 when it took $return_path from a Return-Path: header line. It now does.
7667 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7668 with a valid header line field name (a series of printing characters
7669 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7671 23. Changed "disc" in the source to "disk" to conform to the documentation and
7672 the book and for uniformity.
7674 24. Ignore Sendmail's -Ooption=value command line item.
7676 25. When execve() failed while trying to run a command in a pipe transport,
7677 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7678 could be confused with a return value of 69 from the command itself. This
7679 has been changed to 127, the value the shell returns if it is asked to run
7680 a non-existent command. The wording for the related log line suggests a
7681 non-existent command as the problem.
7683 26. If received_header_text expands to an empty string, do not add a Received:
7684 header line to the message. (Well, it adds a token one on the spool, but
7685 marks it "old" so that it doesn't get used or transmitted.)
7687 27. Installed eximstats 1.28 (addition of -nt option).
7689 28. There was no check for failure on the call to getsockname() in the daemon
7690 code. This can fail if there is a shortage of resources on the system, with
7691 ENOMEM, for example. A temporary error is now given on failure.
7693 29. Contrary to the C standard, it seems that in some environments, the
7694 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7695 program. Exim now does this explicitly; it affects the formatting of
7696 timestamps using strftime().
7698 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7699 complaints. I've now initialized all the variables, to avoid this.
7701 32. Header lines added by a system filter were not being "seen" during
7702 transport-time rewrites.
7704 33. The info_callback() function passed to OpenSSL is set up with type void
7705 (*)(SSL *, int, int), as described somewhere. However, when calling the
7706 function (actually a macro) that sets it up, the type void(*)() is
7707 expected. I've put in a cast to prevent warnings from picky compilers.
7709 34. If a DNS black list lookup found a CNAME record, but there were no A
7710 records associated with the domain it pointed at, Exim crashed.
7712 35. If a DNS black list lookup returned more than one A record, Exim ignored
7713 all but the first. It now scans all returned addresses if a particular IP
7714 value is being sought. In this situation, the contents of the
7715 $dnslist_value variable are a list of all the addresses, separated by a
7718 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7719 to accept a host name and all its aliases if the forward lookup for any of
7720 them yielded the IP address of the incoming connection. Now it accepts only
7721 those names whose forward lookup yields the correct IP address. Any other
7722 names are discarded. This closes a loophole whereby a rogue DNS
7723 administrator could create reverse DNS records to break through a
7724 wildcarded host restriction in an ACL.
7726 37. If a user filter or a system filter that ran in a subprocess used any of
7727 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7728 the wrong values were passed to the pipe command ($thisaddress had the
7729 value of $0, $0 had the value of $1, etc). This bug was introduced by
7730 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7732 38. Improved the line breaking for long SMTP error messages from ACLs.
7733 Previously, if there was no break point between 40 and 75 characters, Exim
7734 left the rest of the message alone. Two changes have been made: (a) I've
7735 reduced the minimum length to 35 characters; (b) if it can't find a break
7736 point between 35 and 75 characters, it looks ahead and uses the first one
7737 that it finds. This may give the occasional overlong line, but at least the
7738 remaining text gets split now.
7740 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7741 file descriptors might be low, and that setting 1000 would always raise it.
7742 It turns out that in some environments, the limit is already over 1000 and
7743 that lowering it causes trouble. So now Exim takes care not to decrease it.
7745 40. When delivering a message, the value of $return_path is set to $sender_
7746 address at the start of routing (routers may change the value). By an
7747 oversight, this default was not being set up when an address was tested by
7748 -bt or -bv, which affected the outcome if any router or filter referred to
7751 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7752 log only when "message" or "log_message" are set. However, if one of the
7753 conditions was an address verification, or a call to a nested ACL, the
7754 messages generated by the underlying test were being passed through. This
7755 no longer happens. The underlying message is available in $acl_verify_
7756 message for both "message" and "log_message" expansions, so it can be
7757 passed through if needed.
7759 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7760 new expansion $bh_ to give the encoded byte string without charset
7761 translation. Translation happens only if iconv() is available; HAVE_ICONV
7762 indicates this at build time. HEADERS_CHARSET gives the charset to
7763 translate to; headers_charset can change it in the configuration, and
7764 "headers charset" can change it in an individual filter file.
7766 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7767 that creates RFC 2047 encoded "words" labels them as that charset instead
7768 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7769 expansion operator; (ii) when Exim creates a From: line for a local
7770 message; (iii) when a header line is rewritten to include a "phrase" part.
7772 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7773 buggy, causing it to skip the first lines of messages whose message ID
7774 ended in 'D'. This would not have bitten before Exim release 4.14, because
7775 message IDs were unlikely to end in 'D' before then. The effect was to have
7776 incorrect size information for certain domains.
7778 45. #include "config.h" was missing at the start of the crypt16.c module. This
7779 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7782 46. If there was a timeout during a "random" callout check, Exim treated it as
7783 a failure of the random address, and carried on sending RSET and the real
7784 address. If the delay was just some slowness somewhere, the response to the
7785 original RCPT would be taken as a response to RSET and so on, causing
7786 mayhem of various kinds.
7788 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7789 when I implemented it. It didn't allow for the fact that some option values
7790 may legitimately be negative (e.g. size_addition), and it didn't even do
7791 the right test for positive values.
7793 48. Domain names in DNS records are case-independent. Exim always looks them up
7794 in lower case. Some resolvers return domain names in exactly the case they
7795 appear in the zone file, that is, they may contain uppercase letters. Not
7796 all resolvers do this - some return always lower case. Exim was treating a
7797 change of case by a resolver as a change of domain, similar to a widening
7798 of a domain abbreviation. This triggered its re-routing code and so it was
7799 trying to route what was effectively the same domain again. This normally
7800 caused routing to fail (because the router wouldn't handle the domain
7801 twice). Now Exim checks for this case specially, and just changes the
7802 casing of the domain that it ultimately uses when it transmits the message
7805 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7808 50. If a filter generated a file delivery with a non-absolute name (possible if
7809 no home directory exists for the router), the forbid_file option was not
7812 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7813 the existing equality matching.
7815 52. Exim was using ints instead of ino_t variables in some places where it was
7816 dealing with inode numbers.
7818 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7819 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7820 called TMPDIR, and if it finds it is different, it changes its value.
7822 54. The smtp_printf() function is now made available to local_scan() so
7823 additional output lines can be written before returning. There is also an
7824 smtp_fflush() function to enable the detection of a dropped connection.
7825 The variables smtp_input and smtp_batched_input are exported to
7828 55. Changed the default runtime configuration: the message "Unknown user"
7829 has been removed from the ACL, and instead placed on the localuser router,
7830 using the cannot_route_message feature. This means that any verification
7831 failures that generate their own messages won't get overridden. Similarly,
7832 the "Unrouteable address" message that was in the ACL for unverifiable
7833 relay addresses has also been removed.
7835 56. Added hosts_avoid_esmtp to the smtp transport.
7837 57. The exicyclog script was not checking for the esoteric option
7838 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7839 will work only if exicyclog is run under the appropriate euid.
7841 58. Following a discussion on the list, the rules by which Exim recognises line
7842 endings on incoming messages have been changed. The -dropcr and drop_cr
7843 options are now no-ops, retained only for backwards compatibility. The
7844 following line terminators are recognized: LF CRLF CR. However, special
7845 processing applies to CR:
7847 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7848 nor a local message in the state where . is a terminator.
7850 (ii) If a bare CR is encountered in a header line, an extra space is added
7851 after the line terminator so as not to end the header. The reasoning
7852 behind this is that bare CRs in header lines are most likely either
7853 to be mistakes, or people trying to play silly games.
7855 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7856 was being incorrectly given as 18 bytes larger than it should have been.
7857 This is a VOB (very old bug).
7859 60. This may never have affected anything current, but just in case it has:
7860 When the local host is found other than at the start of a list of hosts,
7861 the local host, those with the same MX, and any that follow, are discarded.
7862 When the list in question was part of a longer list of hosts, the following
7863 hosts (not currently being processed) were also being discarded. This no
7864 longer happens. I'm not sure if this situation could ever has previously
7867 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7870 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7871 header. This is something that is recommended in a new Internet Draft, and
7872 is something that is documented as being done by Sendmail. There are two
7873 possible values. For messages generated by the autoreply transport, Exim
7876 Auto-Submitted: auto-replied
7878 whereas for all other generated messages (e.g. bounces) it adds
7880 Auto-Submitted: auto-generated
7882 63. The "personal" condition in filters now includes a test for the
7883 Auto-Submitted: header. If it contains the string "auto-" the message it
7884 not considered personal.
7886 64. Added rcpt_include_affixes as a generic transport option.
7888 65. Added queue_only_override (default true).
7890 66. Added the syslog_duplication option.
7892 67. If what should have been the first header line of a message consisted of
7893 a space followed by a colon, Exim was mis-interpreting it as a header line.
7894 It isn't of course - it is syntactically invalid and should therefore be
7895 treated as the start of the message body. The misbehaviour could have
7896 caused a number of strange effects, including loss of data in subsequent
7897 header lines, and spool format errors.
7899 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7900 client host had authenticated. This control can now be exercised by an ACL
7901 for more flexibility.
7903 69. By default, callouts do not happen when testing with -bh. There is now a
7904 variant, -bhc, which does actually run the callout code, including
7905 consulting and updating the callout cache.
7907 70. Added support for saslauthd authentication, courtesy of Alexander
7910 71. If statvfs() failed on the spool or log directories while checking their
7911 size for availability, Exim confusingly gave the error "space shortage".
7912 Furthermore, in debugging mode it crashed with a floating point exception.
7913 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7914 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7915 serious problem, Exim now writes to the main and panic logs when this
7916 happens, with details of the failure. It then refuses to accept the
7917 incoming message, giving the message "spool directory problem" or "log
7918 directory problem" with a 421 code for SMTP messages.
7920 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7921 0, 1, and 2 exist, because some OS complain for execs without them (see
7922 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7923 descriptors. However, the code omitted to check that the open succeeded,
7924 causing mysterious errors if for some reason the permissions on /dev/null
7925 got screwed. Now Exim writes a message to the main and panic logs, and
7926 bombs out if it can't open /dev/null.
7928 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7929 interact so that it is all more flexible. It is supposed to remain
7930 backwards compatible. Also added extra_local_interfaces.
7932 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7933 to bomb out with an assertion failure - to the client this appears as a
7934 connection drop. This problem occurs in the part of the code that was taken
7935 from the Samba project. Fortunately, the assertion is in a very simple
7936 function, so I have fixed this by reproducing the function inline in the
7937 one place where it is called, and arranging for authentication to fail
7938 instead of killing the process with assert().
7940 75. The SPA client code was not working when the server requested OEM rather
7941 than Unicode encoding.
7943 76. Added code to make require_files with a specific uid setting more usable in
7944 the case where statting the file as root fails - usually a non-root-mounted
7945 NFS file system. When this happens and the failure is EACCES, Exim now
7946 forks a subprocess and does the per-uid checking as the relevant uid.
7948 77. Added process_log_path.
7950 78. If log_file_path was not explicitly set, a setting of check_log_space or
7951 check_log_inodes was ignored.
7953 79. If a space check for the spool or log partitions fails, the incident is now
7954 logged. Of course, in the latter case the data may get lost...
7956 80. Added the %p formatting code to string_format() so that it can be used to
7957 print addresses in debug_print(). Adjusted all the address printing in the
7958 debugging in store.c to use %p rather than %d.
7960 81. There was a concern that a line of code in smtp_in.c could overflow a
7961 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7962 initially expressed, the concern was not well-founded, because trailing
7963 spaces are removed early. However, if the trailing spaces were followed by
7964 a NULL, they did not get removed, so the overflow was possible. Two fixes
7967 (a) I re-wrote the offending code in a cleaner fashion.
7968 (b) If an incoming SMTP command contains a NULL character, it is rejected
7971 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7972 runs initgroups(), so that if the Exim user is in any additional groups,
7973 they will be used during message reception.
7979 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob