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
137 JH/01 Avoid costly startup code when not strictly needed. This reduces time
138 for some exim process initialisations. It does mean that the logging
139 of TLS configuration problems is only done for the daemon startup.
141 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
143 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
144 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
147 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
148 DKIM signing is being done. Previously a transport_filter would always
149 disable CHUNKING, falling back to traditional DATA.
151 JH/05 Regard command-line recipients as tainted.
153 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
155 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
156 PAM library frees one of the arguments given to it, despite the
157 documentation. Therefore a plain malloc must be used.
159 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
160 on-stack buffers were used, resulting in a taint trap when DSN information
161 copied from a received message was written into the buffer.
163 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
164 the ordering of its ARC headers. This caused a crash.
166 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
167 a new record was being constructed with information from the peer, a trap
170 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
171 installation would get error messages from DMARC verify, when it hit the
172 nonexistent file indicated by the default. Distros wanting DMARC enabled
173 should both provide the file and set the option.
174 Also enforce no DMARC verification for command-line sourced messages.
176 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
177 could, depending on the platform, hang at the STARTTLS response.
179 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
180 message on a connection. Previously if one message had ARC headers and
181 the following one did not, a crash could result when adding an
182 Authentication-Results: header.
184 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
185 Exim main code and Exim-related utities. The introduction of taint
186 tracking also did many adjustments to string handling. Since then, eximon
187 frequently terminated with an assert failure.
189 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
190 check for 452 responses. This slightly helps the inefficieny of doing
191 a large alias-expansion into a recipient-limited target. The max_rcpt
192 transport option still applies (and at the current default, will override
193 the new feature). The check is done for either cause of synch, and forces
194 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
195 connection. The new facility is not tunable at this time.
197 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
198 library live data was being used, so the results became garbage. Make
199 copies while it is still usable.
201 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
202 on delivery deferred (==) and failed (**) lines (if a delivery was
203 attemtped). Previously it was only on completion (=>) lines.
205 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
206 for the expansion of the server_scram_iter and server_scram_salt options.
208 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
209 are now specifically given a NO_DATA response without hitting the system
210 resolver. The library goes on to do the now-standard TXT lookup.
211 Use of dnsdb lookups is not affected.
213 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
214 only retrieve the errormessage once. Previously two calls to dlerror()
215 were used, and the second one (for mainlog/paniclog) retrieved null
218 JH/20 Taint checking: disallow use of tainted data for
219 - the appendfile transport file and directory options
220 - the pipe transport command
221 - the autoreply transport file, log and once options
222 - file names used by the redirect router (including filter files)
224 - paths used by single-key lookups
225 Previously this was permitted.
227 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
228 adjusted the size of a major service buffer; this failed because the
229 buffer was in use at the time. Change to a compile-time increase in the
230 buffer size, when this authenticator is compiled into exim.
232 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
233 previous fast-mode was untenable in the face of glibs using mmap to
234 support larger malloc requests.
236 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
237 New values supported, if defined on system where compiled:
238 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
239 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
241 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
242 running a limited number of proceses in parallel, a benefit is gained. The
243 amount varies with the platform hardware and load. The use of the option
244 queue_run_in_order means we cannot do this, as ordering becomes
247 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
248 had introduced a string-copy (for ensuring NUL-termination) which was not
249 appropriate for that case, which can include embedded NUL bytes in the
250 block of data. Investigation showed the copy to actually be needless, the
251 data being length-specified.
253 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
254 done during a receiving connection, and both used TLS, global info was
255 used rather than per-connection info for tracking the state of data
256 queued for transmission. This could result in a connection hang.
258 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
259 Previously, when delivering serveral messages down a single connection
260 only the first would provide a SIZE. This was due to the size information
261 not being properly tracked.
263 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
264 TAI (at 37 seconds currently), pretend to be in UTC for time-related
265 expansion and logging. Previously, spurious values such as a future
266 minute could be seen.
268 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
269 it could crash from a null-deref. This could also affect the
270 ${addresses: } operator and ${readsock } item.
272 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
273 message following a mime one, the variable was not reset.
275 JH/30 When an pipelined-connect fails at the first response, assume incorrect
276 cached capability (perhaps the peer reneged?) and immediately retry in
279 JH/31 Fix spurious detection of timeout while writing to transport filter.
281 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
282 an attempt to copy the string was made before checking it.
284 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
285 taint of the lookup key was maintained; we now regard the presence in the
286 filesystem as sufficient validation.
288 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
289 argument is supplied.
291 JH/35 The dsearch lookup now requires that the directory is an absolute path.
292 Previously this was not checked, and nonempty relative paths made an
293 access under Exim's current working directory.
295 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
296 Previously no event was raised.
298 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
299 parameter supplied by the sender MAIL FROM command. Previously it was
300 ignored, and only the check_spool_space option value for the required
303 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
304 the size of the signing public-key. Previously it was instead giving
305 the size of the signature hash.
307 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
308 the default. See the (new) dkim_verify_min_keysizes option.
310 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
311 and an ACL use a lookup for checking either the local_part or domain,
312 stale data could be accessed. Ensure that variable references are
313 dropped between messages.
315 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
316 by the client was not checked as pointing within response data before
317 being used. A malicious client could thus cause an out-of-bounds read and
318 possibly gain authentication. Fix by adding the check.
320 JH/42 Internationalisation: change the default for downconversion in the smtp
321 transport to be "if needed". Previously it was "as previously set" for
322 the message, which usually meant "if needed" for message-submission but
323 "no" for everything else. However, MTAs have been seen using SMTPUTF8
324 even when the envelope addresses did not need it, resulting in forwarding
325 failures to non-supporting MTAs. A downconvert in such cases will be
326 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
327 transport. The change does mean that addresses needing conversion will
328 be converted when previously a delivery failure would occur.
330 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
331 response was received it would be used unchecked in a fail-DSN, violating
332 standards on line-length limits. Truncate if needed.
334 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
335 doesn't work. (Additionally add a new main config option to configure the
342 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
343 side, to decode a TLS 1.3 packet capture.
345 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
346 Previously the default library behaviour applied, sending two, each in
349 JH/03 Debug output for ACL now gives the config file name and line number for
352 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
354 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
356 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
357 buffer overrun for (non-chunking) other transports.
359 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
360 TLS1.3, means that a server rejecting a client certificate is not visible
361 to the client until the first read of encrypted data (typically the
362 response to EHLO). Add detection for that case and treat it as a failed
363 TLS connection attempt, so that the normal retry-in-clear can work (if
364 suitably configured).
366 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
367 and/or domain. Found and fixed by Jason Betts.
369 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
370 configuration). If a CNAME target was not a wellformed name pattern, a
373 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
374 the OS reports them interleaved with other addresses.
376 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
377 used both for input and for a verify callout, both encrypted, SMTP
378 responses being sent by the server could be lost. This resulted in
379 dropped connections and sometimes bounces generated by a peer sending
382 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
383 string. Previously it was possible to cause undefined behaviour in a
384 library routine (usually a crash). Found by "zerons".
386 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
389 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
390 API was removed, so update to use the newer ones.
392 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
393 any timeout set, is taking a long time. Previously we would hang on to a
394 rotated logfile "forever" if the input was arriving with long gaps
395 (a previous attempt to fix addressed lack, for a long time, of initial
398 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
399 shared (NFS) environment. The length of the tempfile name is now
400 4 + 16 ("hdr.$message_exim_id") which might break on file
401 systems which restrict the file name length to lower values.
404 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
405 shared (NFS) environment.
407 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
408 did for all versions <4.90). Notably -M, -m, --invert, -I may be
411 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
412 on some platforms for bit 31.
414 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
415 to changes apparently associated with TLS1.3 handling some of the APIs
416 previously used were either nonfunctional or inappropriate. Strings
417 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
418 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
419 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
420 This affects log line X= elements, the $tls_{in,out}_cipher variables,
421 and the use of specific cipher names in the encrypted= ACL condition.
423 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
425 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
426 verification result was not updated unless hosts_require_ocsp applied.
428 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
429 queue_list_requires_admin set to false, non-admin users were denied the
432 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
433 directory-of-certs mode. Previously they were advertised despite the
436 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
437 A single TCP connection by a client will now hold a TLS connection open
438 for multiple message deliveries, by default. Previously the default was to
441 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
442 default. If built with the facility, DANE will be used. The facility
443 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
445 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
446 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
447 must be defined and you must still, unless you define DISABLE_TLS, manage
448 the the include-dir and library-file requirements that go with that
449 choice. Non-TLS builds are still supported.
451 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
454 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
455 default. If the platform supports and has the facility enabled, it will
456 be requested on all coneections.
458 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
459 controlled by the build-time option SUPPORT_PIPE_CONNECT.
461 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
463 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
464 success-DSN messages. Previously the From: header was always the default
465 one for these; the option was ignored.
467 JH/28 Fix the timeout on smtp response to apply to the whole response.
468 Previously it was reset for every read, so a teergrubing peer sending
469 single bytes within the time limit could extend the connection for a
470 long time. Credit to Qualsys Security Advisory Team for the discovery.
472 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
473 delivery address, which leaked information of the results of local
474 forwarding. Change to the original envelope recipient address, per
477 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
478 requested. Previously not bounce was generated and a log entry of
479 error ignored was made.
481 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
483 JH/32 Introduce a general tainting mechanism for values read from the input
484 channel, and values derived from them. Refuse to expand any tainted
485 values, to catch one form of exploit.
487 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
488 was unused and the unexpanded text used for the test. Found and
489 fixed by Ruben Jenster.
491 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
492 an attempt to use a TLS library read routine dereffed a nul pointer,
495 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
496 them in our smtp response.
498 JH/36 Have the general router option retry_use_local_part default to true when
499 any of the restrictive preconditions are set (to anything). Previously it
500 was only for check_local user. The change removes one item of manual
501 configuration which is required for proper retries when a remote router
502 handles a subset of addresses for a domain.
504 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
505 link count into consideration.
507 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
508 caused the extension of big_buffer, the following lines were ignored.
510 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
511 accordance with RFC 2308. Previously there was no expiry, so a longlived
512 receive process (eg. due to ACL delays) versus a short SOA value could
515 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
517 JH/39 Promote DMARC support to mainline.
519 JH/40 Bug 2452: Add a References: header to DSNs.
521 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
522 parameters. The relevant library call is documented as "Deprecated: This
523 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
524 3.6.0, DH parameters are negotiated following RFC7919."
526 HS/06 Change the default of dnssec_request_domains to "*"
528 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
529 carried on and emitted a BDAT command, even when PIPELINING was not
532 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
533 buffer was used for the filename, resulting in a trap when tainted
534 arguments (eg. $domain) were used.
536 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
537 recommended to avoid a possible server-load attack. The feature can be
538 re-enabled via the openssl_options main cofiguration option.
540 JH/45 local_scan API: documented the current smtp_printf() call. This changed
541 for version 4.90 - adding a "more data" boolean to the arguments.
542 Bumped the ABI version number also, this having been missed previously;
543 release versions 4.90 to 4.92.3 inclusive were effectively broken in
544 respect of usage of smtp_printf() by either local_scan code or libraries
545 accessed via the ${dlfunc } expansion item. Both will need coding
546 adjustment for any calls to smtp_printf() to match the new function
547 signature; a FALSE value for the new argument is always safe.
549 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
550 the file-offset (which the Linux syscall does, and exim expects); this
551 resulted in an indefinite loop.
553 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
554 to do ARC verification. The Authentication-Results: header line added
555 by the configuration then had no ARC item.
561 JH/01 Remove code calling the customisable local_scan function, unless a new
562 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
564 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
565 non-signal-safe functions being used.
567 JH/03 Bug 2269: When presented with a received message having a stupidly large
568 number of DKIM-Signature headers, disable DKIM verification to avoid
569 a resource-consumption attack. The limit is set at twenty.
571 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
572 report of oldest_pass in ${authres } in consequence, and separate out
573 some descriptions of reasons for verification fail.
575 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
576 files in the spool were present and unlocked. A queue-runner could spot
577 them, resulting in a duplicate delivery. Fix that by doing the unlock
578 after the unlink. Investigation by Tim Stewart. Take the opportunity to
579 add more error-checking on spoolfile handling while that code is being
582 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
583 No known attacks, no CVE, this is defensive hardening.
585 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
586 a queue-runner could start a delivery while other operations were ongoing.
587 Cutthrough delivery was a common victim, resulting in duplicate delivery.
588 Found and investigated by Tim Stewart. Fix by using the open message data
589 file handle rather than opening another, and not locally closing it (which
590 releases a lock) for that case, while creating the temporary .eml format
591 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
593 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
594 $sender_verify_failure/$recipient_verify_failure to "random".
596 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
599 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
600 Previously this would segfault.
602 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
605 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
606 like zero, since the resolver should be doing this for us, But we need one
607 as a CNAME but no MX presence gets the CNAME returned; we need to check
608 that doesn't point to an MX to declare it "no MX returned" rather than
609 "error, loop". A new main option is added so the older capability of
610 following some limited number of chain links is maintained.
612 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
614 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
615 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
616 now-deprecated earlier definitions used only specified the range up to TLS
617 1.2 (in the older-version library docs).
619 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
621 JH/15 Rework TLS client-side context management. Stop using a global, and
622 explicitly pass a context around. This enables future use of TLS for
623 connections to service-daemons (eg. malware scanning) while a client smtp
624 connection is using TLS; with cutthrough connections this is quite likely.
626 JH/16 Fix ARC verification to do AS checks in reverse order.
628 JH/17 Support a "tls" option on the ${readsocket } expansion item.
630 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
631 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
632 Previously the "utf8" would be re-prepended for every additional message.
634 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
635 Previously thery were accepted, resulting in issues when attempting to
636 forward messages to a non-supporting MTA.
638 PP/02 Let -n work with printing macros too, not just options.
640 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
641 one parent address was copied, and bogus data was used at delivery-logging
642 time. Either a crash (after delivery) or bogus log data could result.
643 Discovery and analysis by Tim Stewart.
645 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
646 Previously if the string ended mid-character, we did not insert the
647 promised '?' replacement.
649 PP/04 Documentation: current string operators work on bytes, not codepoints.
651 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
652 should pack well giving a smaller memory footprint so better caching and
653 therefore performance. Group the declarations where this can't be done so
654 that the byte-sized flag variables are not interspersed among pointer
655 variables, giving a better chance of good packing by the compiler.
657 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
658 non-null, to avoid issues with sites running BATV. Previously reports were
659 sent with an empty envelope sender so looked like bounces.
661 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
662 The ignore_error flag wasn't being returned from the filter subprocess so
663 was not set for later routers. Investigation and fix by Matthias Kurz.
665 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
666 and a msg:complete for the whole, when a message is manually removed using
667 -Mrm. Developement by Matthias Kurz, hacked on by JH.
669 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
670 a "Gnu special" function, asprintf() in the DB utility binary builds; I
671 hope that is portable enough.
673 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
674 requiring a known-CA anchor certificate; make it now rely entirely on the
675 TLSA as an anchor. Checking the name on the leaf cert against the name
676 on the A-record for the host is still done for TA (but not for EE mode).
678 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
679 included in delivery lines for non-proxied connections, when compiled with
680 SUPPORT_SOCKS and running with proxy logging enabled.
682 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
683 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
684 move the existing event to fire before the normal logging of message
685 failure so that custom logging is bracketed by normal logging.
687 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
688 msg:fail:internal event. Developement by Matthias Kurz.
690 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
691 far too small for todays use of crypto signatures stored there. Go all
692 the way to the max DNS message size of 64kB, even though this might be
693 overmuch for IOT constrained device use.
695 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
696 copy a string over itself. The library routine is documented as not
697 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
699 JH/32 For main options check_spool_space and check_inode_space, where the
700 platform supports 64b integers, support more than the previous 2^31 kB
701 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
702 the previous G, M, k.
704 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
705 $authenticated_fail_id variable on authentication failure. Previously
708 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
709 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
710 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
711 GNUTLS_SEC_PARAM_MEDIUM.
713 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
714 side. Previously we would continue as if no SNI had been received.
716 JH/36 Harden the handling of string-lists. When a list consisted of a sole
717 "<" character, which should be a list-separator specification, we walked
718 off past the nul-terimation.
720 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
721 causes) even when the retry time is not yet met. Previously they were
722 not, meaning that when (say) an account was over-quota and temp-rejecting,
723 and multiple senders' messages were queued, only one sender would get
724 notified on each configured delay_warning cycle.
726 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
728 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
729 cipher-suites, an error can be left on the stack even for a succeeding
730 accept; this results in impossible error messages when a later operation
733 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
734 return error codes indicating retry. Under TLS1.3 this becomes required.
736 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
737 it only wrote the new authenticators, resulting in a lack of tracking of
738 peer changes of ESMTP extensions until the next cache flush.
740 JH/41 Fix the loop reading a message header line to check for integer overflow,
741 and more-often against header_maxsize. Previously a crafted message could
742 induce a crash of the recive process; now the message is cleanly rejected.
744 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
745 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
751 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
752 When redis_servers is set to a list of > 1 element, and the Redis servers
753 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
754 case of MOVED into a DEFER case instead, thus moving the query onto the
755 next server in the list. For a cluster of N elements, all N servers must
756 be defined in redis_servers.
758 GF/02 Catch and remove uninitialized value warning in exiqsumm
759 Check for existence of @ARGV before looking at $ARGV[0]
761 JH/01 Replace the store_release() internal interface with store_newblock(),
762 which internalises the check required to safely use the old one, plus
763 the allocate and data copy operations duplicated in both (!) of the
764 extant use locations.
766 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
767 modifier. This matches the restriction on the commandline.
769 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
770 Previously only the last row was returned.
772 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
773 we assumed that tags in the header were well-formed, and parsed the
774 element content after inspecting only the first char of the tag.
775 Assumptions at that stage could crash the receive process on malformed
778 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
779 While running the DKIM ACL we operate on the Permanent memory pool so that
780 variables created with "set" persist to the DATA ACL. Also (at any time)
781 DNS lookups that fail create cache records using the Permanent pool. But
782 expansions release any allocations made on the current pool - so a dnsdb
783 lookup expansion done in the DKIM ACL releases the memory used for the
784 DNS negative-cache, and bad things result. Solution is to switch to the
785 Main pool for expansions.
786 While we're in that code, add checks on the DNS cache during store_reset,
787 active in the testsuite.
788 Problem spotted, and debugging aided, by Wolfgang Breyha.
790 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
791 When none of the hosts presented to a transport match an already-open
792 connection, close it and proceed with the list. Previously we would
793 queue the message. Spotted by Lena with Yahoo, probably involving
796 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
797 Previously a spurious "250 OK id=" response was appended to the proper
800 JH/08 The "support for" informational output now, which built with Content
801 Scanning support, has a line for the malware scanner interfaces compiled
802 in. Interface can be individually included or not at build time.
804 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
805 by the template makefile "src/EDITME". The "STREAM" support for an older
806 ClamAV interface method is removed.
808 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
809 rows affected is given instead).
811 JH/11 The runtime Berkeley DB library version is now additionally output by
812 "exim -d -bV". Previously only the compile-time version was shown.
814 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
815 SMTP connection. Previously, when one had more recipients than the
816 first, an abortive onward connection was made. Move to full support for
817 multiple onward connections in sequence, handling cutthrough connection
818 for all multi-message initiating connections.
820 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
821 routers. Previously, a multi-recipient message would fail to match the
822 onward-connection opened for the first recipient, and cause its closure.
824 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
825 a timeout on read on a GnuTLS initiating connection, resulting in the
826 initiating connection being dropped. This mattered most when the callout
827 was marked defer_ok. Fix to keep the two timeout-detection methods
830 JH/15 Relax results from ACL control request to enable cutthrough, in
831 unsupported situations, from error to silently (except under debug)
832 ignoring. This covers use with PRDR, frozen messages, queue-only and
835 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
837 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
838 metadata, resulting in a crash in free().
840 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
841 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
842 Broken also in d185889f4, with init system revamp.
844 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
845 Previously we abruptly closed the connection after reading a malware-
846 found indication; now we go on to read the "scan ok" response line,
849 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
850 ACL. Previously, a crash would result.
852 JH/19 Speed up macro lookups during configuration file read, by skipping non-
853 macro text after a replacement (previously it was only once per line) and
854 by skipping builtin macros when searching for an uppercase lead character.
856 JH/20 DANE support moved from Experimental to mainline. The Makefile control
857 for the build is renamed.
859 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
860 was allocated for every new TLS startup, meaning one per message. Fix
861 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
863 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
864 reported the original. Fix to report (as far as possible) the ACL
865 result replacing the original.
867 JH/23 Fix memory leak during multi-message connections using STARTTLS under
868 OpenSSL. Certificate information is loaded for every new TLS startup,
869 and the resources needed to be freed.
871 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
873 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
876 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
877 DATA response info to the (existing) per-recipient response info for
878 the "C=" log element. It can have useful tracking info from the
879 destination system. Patch from Simon Arlott.
881 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
882 length value. Previously this would segfault.
884 HS/02 Support Avast multiline protoocol, this allows passing flags to
885 newer versions of the scanner.
887 JH/28 Ensure that variables possibly set during message acceptance are marked
888 dead before release of memory in the daemon loop. This stops complaints
889 about them when the debug_store option is enabled. Discovered specifically
890 for sender_rate_period, but applies to a whole set of variables.
891 Do the same for the queue-runner and queue-list loops, for variables set
892 from spool message files. Do the same for the SMTP per-message loop, for
893 certain variables indirectly set in ACL operations.
895 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
896 as a multi-recipient message from a mailinglist manager). The coding had
897 an arbitrary cutoff number of characters while checking for more input;
898 enforced by writing a NUL into the buffer. This corrupted long / fast
899 input. The problem was exposed more widely when more pipelineing of SMTP
900 responses was introduced, and one Exim system was feeding another.
901 The symptom is log complaints of SMTP syntax error (NUL chars) on the
902 receiving system, and refused recipients seen by the sending system
903 (propating to people being dropped from mailing lists).
904 Discovered and pinpointed by David Carter.
906 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
907 replaced by the ${authresults } expansion.
909 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
911 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
912 allows proper process termination in container environments.
914 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
915 Previously the "final dot" had a newline after it; ensure it is CR,LF.
917 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
918 and "err_perm", deprecated since 4.83 when the RFC-defined words
919 "temperror" and "permerror" were introduced.
921 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
922 transport-filters or DKIM-signing. The restriction was lost in the
923 consolidation of verify-callout and delivery SMTP handling.
924 Extend the restriction to also cover ARC-signing.
926 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
927 in defer=pass mode supply a 450 to the initiator. Previously the message
930 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
931 tls_require_ciphers is used as before.
933 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
934 "pass_unscanned". Only tmpfails from the scanner are written to
935 the paniclog, as they may require admin intervention (permission
936 denied, license issues). Other scanner errors (like decompression
937 bombs) do not cause a paniclog entry.
939 JH/36 Fix reinitialisation of DKIM logging variable between messages.
940 Previously it was possible to log spurious information in receive log
943 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
944 triggered odd behaviour from Outlook Express clients.
946 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
949 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
950 since the IETF WG has not yet settled on that versus the original
951 "bare" representation.
953 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
954 Previously the millisecond value corrupted the output.
955 Fix also for syslog_pid=no and log_selector +pid, for which the pid
956 corrupted the output.
962 JH/01 Rework error string handling in TLS interface so that the caller in
963 more cases is responsible for logging. This permits library-sourced
964 string to be attached to addresses during delivery, and collapses
965 pairs of long lines into single ones.
967 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
968 during configuration. Wildcards are allowed and expanded.
970 JH/02 Rework error string handling in DKIM to pass more info back to callers.
971 This permits better logging.
973 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
974 do not close it down and have the child transport start it up again on
975 the passed-on TCP connection. Instead, proxy the child (and any
976 subsequent ones) for TLS via a unix-domain socket channel. Logging is
977 affected: the continued delivery log lines do not have any DNSSEC, TLS
978 Certificate or OCSP information. TLS cipher information is still logged.
980 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
981 identical IP addresses on different listening ports. Will also affect
984 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
985 add noisy ifdef guards to special-case this sillyness.
986 Patch from Bernd Kuhls.
988 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
989 than 255 are no longer allowed.
991 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
992 Disable the session-cache too, which might reduce our load. Since we
993 currrectly use a new context for every connection, both as server and
994 client, there is no benefit for these.
995 GnuTLS appears to not support tickets server-side by default (we don't
996 call gnutls_session_ticket_enable_server()) but client side is enabled
997 by default on recent versions (3.1.3 +) unless the PFS priority string
1000 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1001 <https://reproducible-builds.org/specs/source-date-epoch/>.
1003 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1004 the check for any unsuccessful recipients did not notice the limit, and
1005 erroneously found still-pending ones.
1007 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1008 MSG_MORE. Only in-clear (not on TLS connections).
1010 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1011 transport-filter is involved we can buffer the headers in memory for
1012 creating the signature, and read the spool data file once for the
1013 signature and again for transmission.
1015 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1016 4.77 as the kernel support then wasn't solid, having issues in 64bit
1017 mode. Now, it's been long enough. Add support for FreeBSD also.
1019 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1020 case where the routing stage had gathered several addresses to send to
1021 a host before calling the transport for the first, we previously failed
1022 to close down TLS in the old transport process before passing the TCP
1023 connection to the new process. The new one sent a STARTTLS command
1024 which naturally failed, giving a failed delivery and bloating the retry
1025 database. Investigation and fix prototype from Wolfgang Breyha.
1027 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1028 false-negatives in the check that the sender had not preempted a response
1029 or prompt from Exim (running as a server), due to that code's lack of
1030 awareness of the SMTP input buffering.
1032 PP/04 Add commandline_checks_require_admin option.
1033 Exim drops privileges sanely, various checks such as -be aren't a
1034 security problem, as long as you trust local users with access to their
1035 own account. When invoked by services which pass untrusted data to
1036 Exim, this might be an issue. Set this option in main configuration
1037 AND make fixes to the calling application, such as using `--` to stop
1040 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1041 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1042 responses to those, into a single TLS record each way (this usually means
1043 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1046 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1047 affects you only if you're dancing at the edge of the param size limits.
1048 If you are, and this message makes sense to you, then: raise the
1049 configured limit or use OpenSSL 1.1. Nothing we can do for older
1052 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1053 cmdline element to get the documented default one. Previously it was
1056 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1057 are unsuable for non-message-specific reasons - eg. connection timeout,
1058 banner-time rejection.
1060 JH/16 Fix logging of delivery remote port, when specified by router, under
1063 PP/06 Repair manualroute's ability to take options in any order, even if one
1064 is the name of a transport.
1067 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1069 JH/17 Change the list-building routines interface to use the expanding-string
1070 triplet model, for better allocation and copying behaviour.
1072 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1073 Previously it was constructed the first time a possibly-matching string
1074 was met in the configuration file input during startup; now it is done
1077 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1078 compatible one, to avoid the (poorly documented) possibility of a config
1079 file in the working directory redirecting the DB files, possibly correpting
1080 some existing file. CVE-2017-10140 assigned for BDB.
1082 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1083 cache-hot. Previously, although the result was properly cached, the
1084 initial verify call returned a defer.
1086 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1087 the main verify for receipient in uncached-mode.
1089 JH/22 Retire historical build files to an "unsupported" subdir. These are
1090 defined as "ones for which we have no current evidence of testing".
1092 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1093 if present. Previously it was ignored.
1095 JH/24 Start using specified-initialisers in C structure init coding. This is
1096 a C99 feature (it's 2017, so now considered safe).
1098 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1099 if was a fixed-sized field and bitmask ops via macros; it is now more
1102 PP/07 GitHub PR 56: Apply MariaDB build fix.
1103 Patch provided by Jaroslav Škarvada.
1105 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1106 during Coverity cleanups [4.87 JH/47]
1107 Diagnosis and fix provided by Michael Fischer v. Mollard.
1109 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1110 the right size to place the terminating semicolon on its own folded
1111 line, the header hash was calculated to an incorrect value thanks to
1112 the (relaxed) space the fold became.
1114 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1115 and confused the parent.
1117 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1118 which could crash as a result. This could lead to undeliverable messages.
1120 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1123 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1124 matters on fast-turnover and PID-randomising systems, which were getting
1125 out-of-order delivery.
1127 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1128 a possibly-overlapping copy. The symptom was that "Remote host closed
1129 connection in response to HELO" was logged instead of the actual 4xx
1132 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1133 Previously only that bufferd was discarded, resulting in SYMTP command
1136 JH/32 DKIM: when a message has multiple signatures matching an identity given
1137 in dkim_verify_signers, run the dkim acl once for each. Previously only
1138 one run was done. Bug 2189.
1140 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1141 "panic the current process" to "deliberately defer". The panic log is
1142 still written with the problem list name; the mail and reject logs now
1143 get a temp-reject line for the message that was being handled, saying
1144 something like "domains check lookup or other defer". The SMTP 451
1145 message is still "Temporary local problem".
1147 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1148 A crafted sequence of BDAT commands could result in in-use memory beeing
1149 freed. CVE-2017-16943.
1151 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1152 from SMTP input. Previously it was always done; now only done for DATA
1153 and not BDAT commands. CVE-2017-16944.
1155 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1156 to the message (such as an overlong header line). Previously this was
1157 not done and we did not exit BDAT mode. Followon from the previous item
1158 though a different problem.
1164 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1165 than -2003 did; needs libidn2 in addition to libidn.
1167 JH/02 The path option on a pipe transport is now expanded before use.
1169 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1170 Patch provided by "Björn", documentation fix added too.
1172 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1173 missing a wire-to-host endian conversion.
1175 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1176 close after a BDAT command line could be taken as a following command,
1177 giving a synch failure. Fix by only checking for synch immediately
1178 before acknowledging the chunk.
1180 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1181 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1182 macro. Patches provided by Josh Soref.
1184 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1185 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1186 by IANA as a service extension. However, John Klensin suggests that we
1189 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1190 the dkim code may be unix-mode line endings rather than smtp wire-format
1191 CRLF, so prepend a CR to any bare LF.
1193 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1194 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1196 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1197 the first were themselves being wrongly included in the feed into dkim
1198 processing; with most chunk sizes in use this resulted in an incorrect
1199 body hash calculated value.
1201 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1202 DKIM signature block, for verification. Although advised against by
1203 standards it is specifically not ruled illegal.
1205 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1207 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1208 missing a body hash (the bh= tag).
1210 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1211 It seems that HAProxy sends the Proxy Protocol information in clear and
1212 only then does a TLS startup, so do the same.
1214 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1215 TCP connections (such as for Spamd) unless the daemon successfully set
1216 Fast Open mode on its listening sockets. This fixes breakage seen on
1217 too-old kernels or those not configured for Fast Open, at the cost of
1218 requiring both directions being enabled for TFO, and TFO never being used
1219 by non-daemon-related Exim processes.
1221 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1222 endings, at least on the first header line. Try to canonify any that get
1223 past that check, despite the cost.
1225 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1226 now limited to an arbitrary five deep, while parsing addresses with the
1227 strip_excess_angle_brackets option enabled.
1229 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1230 instead leave the unprompted TLS handshake in socket buffer for the
1231 TLS library to consume.
1233 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1235 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1237 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1238 region in one operation, for speed, and this leaves assigned pointers
1239 dangling. Add checks run only under the testsuite which checks all
1240 variables at a store-reset and panics on a dangling pointer; add code
1241 explicitly nulling out all the variables discovered. Fixes one known
1242 bug: a transport crash, where a dangling pointer for $sending_ip_address
1243 originally assigned in a verify callout, is re-used.
1245 PP/06 Drop '.' from @INC in various Perl scripts.
1247 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1249 PP/08 Reduce a number of compilation warnings under clang; building with
1250 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1251 should be warning-free.
1253 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1255 HS/01 Fix portability problems introduced by PP/08 for platforms where
1256 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1258 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1259 chunk. This allows us to accept broken chunked messages. We need a more
1260 general solution here.
1262 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1263 already-broken messages in the queue.
1265 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1267 JH/19 Fix reference counting bug in routing-generated-address tracking.
1273 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1274 supports it and a size is available (ie. the sending peer gave us one).
1276 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1277 years of being deprecated). The replacements are the ACLs
1278 acl_smtp_mime and acl_not_smtp_mime.
1280 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1281 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1282 or even an in-clear connection were permitted. Now, if the host lookup
1283 was dnssec and dane was requested then the host is only used if the
1284 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1285 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1286 if one fails this test.
1287 This means that a poorly-configured remote DNS will make it incommunicado;
1288 but it protects against a DNS-interception attack on it.
1290 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1291 non-noisy when a race steals the message being considered.
1293 JH/05 If main configuration option tls_certificate is unset, generate a
1294 self-signed certificate for inbound TLS connections.
1296 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1297 in rewrites and routers.
1299 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1300 and logged a warning sing 4.83; now they are a configuration file error.
1302 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1303 (lacking @domain). Apply the same qualification processing as RCPT.
1305 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1307 JH/10 Support ${sha256:} applied to a string (as well as the previous
1310 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1311 a cutthrough deliver is pending, as we always want to make a connection.
1312 This also avoids re-routing the message when later placing the cutthrough
1313 connection after a verify cache hit.
1314 Do not update it with the verify result either.
1316 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1317 when routing results in more than one destination address.
1319 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1320 signing (which inhibits the cutthrough capability). Previously only
1321 the presence of an option was tested; now an expansion evaluating as
1322 empty is permissible (obviously it should depend only on data available
1323 when the cutthrough connection is made).
1325 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1326 the relevant preceding SMTP command did not note the pipelining mode.
1328 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1329 Previously they were not counted.
1331 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1332 as one having no matching records. Previously we deferred the message
1333 that needed the lookup.
1335 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1336 distinguished as "(=".
1338 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1339 for missing MX records. Previously it only worked for missing A records.
1341 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1343 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1344 after the data-go-ahead and data-ack. Patch from Jason Betts.
1346 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1347 even for a "none" policy. Patch from Tony Meyer.
1349 JH/22 Fix continued use of a connection for further deliveries. If a port was
1350 specified by a router, it must also match for the delivery to be
1353 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1354 When one of the recipients of a message was unsuitable for the connection
1355 (has no matching addresses), we lost track of needing to mark it
1356 deferred. As a result mail would be lost.
1358 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1360 JH/25 Decoding ACL controls is now done using a binary search; the source code
1361 takes up less space and should be simpler to maintain. Merge the ACL
1362 condition decode tables also, with similar effect.
1364 JH/26 Fix problem with one_time used on a redirect router which returned the
1365 parent address unchanged. A retry would see the parent address marked as
1366 delivered, so not attempt the (identical) child. As a result mail would
1369 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1370 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1371 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1374 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1375 Main config options check_{log,spool}_{inodes,space} are now
1376 100 inodes, 10MB unless set otherwise in the configuration.
1378 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1379 Previously it only applied to the main-section connection policy
1382 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1384 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1385 by me. Added RFC7919 DH primes as an alternative.
1387 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1388 are not in the system include path.
1390 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1391 GnuTLS, when a session startup failed (eg because the client disconnected)
1392 Exim did stdio operations after fclose. This was exposed by a recent
1393 change which nulled out the file handle after the fclose.
1395 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1396 signed directly by the cert-signing cert, rather than an intermediate
1397 OCSP-signing cert. This is the model used by LetsEncrypt.
1399 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1401 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1402 an incoming connection.
1404 HS/02 Bug 1802: Do not half-close the connection after sending a request
1407 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1408 fallback to "prime256v1".
1410 JH/34 SECURITY: Use proper copy of DATA command in error message.
1411 Could leak key material. Remotely exploitable. CVE-2016-9963.
1417 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1418 and 3.4.4 - once the server is enabled to respond to an OCSP request
1419 it does even when not requested, resulting in a stapling non-aware
1420 client dropping the TLS connection.
1422 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1423 support variable-length bit vectors. No functional change.
1425 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1426 The I= interface field on outgoing lines is now after the H= remote
1427 host field, same as incoming lines. There is a separate
1428 outgoing_interface log selector which allows you to disable the
1431 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1432 If not running log_selector +smtp_connection the mainlog would be held
1433 open indefinitely after a "too many connections" event, including to a
1434 deleted file after a log rotate. Leave the per net connection logging
1435 leaving it open for efficiency as that will be quickly detected by the
1436 check on the next write.
1438 HS/01 Bug 1671: Fix post transport crash.
1439 Processing the wait-<transport> messages could crash the delivery
1440 process if the message IDs didn't exist for some reason. When
1441 using 'split_spool_directory=yes' the construction of the spool
1442 file name failed already, exposing the same netto behaviour.
1444 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1445 mime_regex ACL conditions.
1447 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1448 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1449 to HELO, local diagnostic string.
1451 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1452 log line to debug. Even when configured with a tls authenticator many
1453 client connections are expected to not authenticate in this way, so
1454 an authenticate fail is not an error.
1456 HS/02 Add the Exim version string to the process info. This way exiwhat
1457 gives some more detail about the running daemon.
1459 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1460 matter for fast-change records such as DNSBLs.
1462 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1463 retry record, even if constant. There may be multiple transports with
1464 different interface settings and the retry behaviour needs to be kept
1467 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1469 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1471 JH/10 Bug 840: fix log_defer_output option of pipe transport
1473 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1474 hosts_randomize. This matters a lot when combined with mua_wrapper.
1476 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1477 ${quote_pgsql:<string>} operator.
1479 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1481 JH/14 Change header file name for freeradius-client. Relevant if compiling
1482 with Radius support; from the Gentoo tree and checked under Fedora.
1484 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1486 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1487 option transport_filter, meaning no filtering.
1489 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1491 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1492 defaults to "*" (all hosts). The variable is now available when not built
1493 with TLS, default unset, mainly to enable keeping the testsuite sane.
1494 If a server certificate is not supplied (via tls_certificate) an error is
1495 logged, and clients will find TLS connections fail on startup. Presumably
1496 they will retry in-clear.
1497 Packagers of Exim are strongly encouraged to create a server certificate
1498 at installation time.
1500 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1501 with the $config_file variable.
1503 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1504 in transport context, after the attempt, and per-recipient. The latter type
1505 is per host attempted. The event data is the error message, and the errno
1506 information encodes the lookup type (A vs. MX) used for the (first) host,
1507 and the trailing two digits of the smtp 4xx response.
1509 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1510 to write to mainlog (or rejectlog, paniclog) in the window between file
1511 creation and permissions/ownership being changed. Particularly affects
1512 installations where exicyclog is run as root, rather than exim user;
1513 result is that the running daemon panics and dies.
1515 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1517 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1518 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1519 "pri" and "weight". Note that the previous implicit priority given by the
1520 list order is no longer honoured.
1522 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1523 for DKIM processing.
1525 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1526 by defining SUPPORT_SOCKS.
1528 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1529 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1530 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1531 variables are renamed to proxy_{local,external}_{address,port}.
1533 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1534 for a build by defining SUPPORT_I18N
1536 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1537 of the query string, and make ${quote_redis:} do that quoting.
1539 JH/29 Move Events support from Experimental to mainline, enabled by default
1540 and removable for a build by defining DISABLE_EVENT.
1542 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1544 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1545 cached by the daemon.
1547 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1548 by defining LOOKUP_REDIS. The libhiredis library is required.
1550 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1551 keys are given for lookup.
1553 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1554 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1555 only supported when built with TLS support. The PolarSSL SHA routines
1556 are still used when the TLS library is too old for convenient support.
1558 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1559 openssl_options), for security. OpenSSL forces this from version 1.1.0
1560 server-side so match that on older versions.
1562 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1563 allocation for $value could be released as the expansion processing
1564 concluded, but leaving the global pointer active for it.
1566 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1567 and to use the domains and local_parts ACL conditions.
1569 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1570 incorrectly not doubled on cutthrough transmission, hence seen as a
1571 body-termination at the receiving system - resulting in truncated mails.
1572 Commonly the sender saw a TCP-level error, and retransmitted the message
1573 via the normal store-and-forward channel. This could result in duplicates
1574 received - but deduplicating mailstores were liable to retain only the
1575 initial truncated version.
1577 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1579 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1581 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1582 we're in there, support oversigning also; bug 1309.
1584 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1586 HS/04 Add support for keep_environment and add_environment options.
1588 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1589 either intentional arithmetic overflow during PRNG, or testing config-
1592 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1593 delivery resulted in actual delivery. Cancel cutthrough before DATA
1596 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1597 rejects a recipient: pass the reject to the originator.
1599 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1600 Many were false-positives and ignorable, but it's worth fixing the
1603 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1604 for the new environment-manipulation done at startup. Move the routines
1605 from being local to tls.c to being global via the os.c file.
1607 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1608 an extract embedded as result-arg for a map, the first arg for extract
1609 is unavailable so we cannot tell if this is a numbered or keyed
1610 extraction. Accept either.
1616 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1619 JH/02 The smtp transport option "multi_domain" is now expanded.
1621 JH/03 The smtp transport now requests PRDR by default, if the server offers
1624 JH/04 Certificate name checking on server certificates, when exim is a client,
1625 is now done by default. The transport option tls_verify_cert_hostnames
1626 can be used to disable this per-host. The build option
1627 EXPERIMENTAL_CERTNAMES is withdrawn.
1629 JH/05 The value of the tls_verify_certificates smtp transport and main options
1630 default to the word "system" to access the system default CA bundle.
1631 For GnuTLS, only version 3.0.20 or later.
1633 JH/06 Verification of the server certificate for a TLS connection is now tried
1634 (but not required) by default. The verification status is now logged by
1635 default, for both outbound TLS and client-certificate supplying inbound
1638 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1641 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1642 Status Notification (bounce) messages are now MIME format per RFC 3464.
1643 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1644 under the control of the dsn_advertise_hosts option, and routers may
1645 have a dsn_lasthop option.
1647 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1648 default, modifiable by a malware= option. The list separator for
1649 the options can now be changed in the usual way. Bug 68.
1651 JH/10 The smtp_receive_timeout main option is now expanded before use.
1653 JH/11 The incoming_interface log option now also enables logging of the
1654 local interface on delivery outgoing connections.
1656 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1657 if the interface and destination host and port all match.
1659 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1662 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1663 Patch from Andrew Lewis.
1665 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1666 now supports optional time-restrictions, weighting, and priority
1667 modifiers per server. Patch originally by <rommer@active.by>.
1669 JH/16 The spamd_address main option now supports a mixed list of local
1670 and remote servers. Remote servers can be IPv6 addresses, and
1671 specify a port-range.
1673 JH/17 Bug 68: The spamd_address main option now supports an optional
1674 timeout value per server.
1676 JH/18 Bug 1581: Router and transport options headers_add/remove can
1677 now have the list separator specified.
1679 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1682 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1685 JH/21 Support for the A6 type of dns record is withdrawn.
1687 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1688 rather than the verbs used.
1690 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1691 from 255 to 1024 chars.
1693 JH/24 Verification callouts now attempt to use TLS by default.
1695 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1696 are generic router options now. The defaults didn't change.
1698 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1699 Original patch from Alexander Shikoff, worked over by JH.
1701 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1702 files if ZCAT_COMMAND is not executable.
1704 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1706 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1708 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1709 Normally benign, it bites when the pair was led to by a CNAME;
1710 modern usage is to not canonicalize the domain to a CNAME target
1711 (and we were inconsistent anyway for A-only vs AAAA+A).
1713 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1715 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1716 when evaluating $sender_host_dnssec.
1718 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1719 $sender_helo_dnssec variable.
1721 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1723 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1725 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1727 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1728 documented as working, but never had. Support all but $spam_report.
1730 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1731 added for tls authenticator.
1733 HS/03 Add perl_taintmode main config option
1739 TL/01 When running the test suite, the README says that variables such as
1740 no_msglog_check are global and can be placed anywhere in a specific
1741 test's script, however it was observed that placement needed to be near
1742 the beginning for it to behave that way. Changed the runtest perl
1743 script to read through the entire script once to detect and set these
1744 variables, reset to the beginning of the script, and then run through
1745 the script parsing/test process like normal.
1747 TL/02 The BSD's have an arc4random API. One of the functions to induce
1748 adding randomness was arc4random_stir(), but it has been removed in
1749 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1750 function when detected.
1752 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1753 cause callback expansion.
1755 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1756 syntax errors in an expansion can be treated as a string instead of
1757 logging or causing an error, due to the internal use of bool_lax
1758 instead of bool when processing it.
1760 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1761 server certificates when making smtp deliveries.
1763 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1765 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1767 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1769 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1770 Merged patch from Sebastian Wiedenroth.
1772 JH/05 Fix results-pipe from transport process. Several recipients, combined
1773 with certificate use, exposed issues where response data items split
1774 over buffer boundaries were not parsed properly. This eventually
1775 resulted in duplicates being sent. This issue only became common enough
1776 to notice due to the introduction of connection certificate information,
1777 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1779 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1780 size buffer was used, resulting in syntax errors when an expansion
1783 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1784 version 3.3.6 or later.
1786 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1787 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1788 both become "event_action", the variables become $event_name, $event_data
1789 and $event_defer_errno. There is a new variable $verify_mode, usable in
1790 routers, transports and related events. The tls:cert event is now also
1791 raised for inbound connections, if the main configuration event_action
1794 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1795 early OCSP support, but no stapling (appears to be less than 1.0.0).
1797 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1798 server certificate names available under the smtp transport option
1799 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1802 JH/10 Time-related extraction expansions from certificates now use the main
1803 option "timezone" setting for output formatting, and are consistent
1804 between OpenSSL and GnuTLS compilations. Bug 1541.
1806 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1807 encoded parameter in the incoming message. Bug 1558.
1809 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1810 include certificate info, eximon was claiming there were spoolfile
1813 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1815 JH/14 Log delivery-related information more consistently, using the sequence
1816 "H=<name> [<ip>]" wherever possible.
1818 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1819 are problematic for Debian distribution, omit them from the release
1822 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1824 JH/16 Fix string representation of time values on 64bit time_t architectures.
1827 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1828 output list separator was used.
1833 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1834 checkers that were complaining about end of non-void function with no
1837 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1838 This was a regression introduced in 4.83 by another bugfix.
1840 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1842 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1843 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1849 TF/01 Correctly close the server side of TLS when forking for delivery.
1851 When a message was received over SMTP with TLS, Exim failed to clear up
1852 the incoming connection properly after forking off the child process to
1853 deliver the message. In some situations the subsequent outgoing
1854 delivery connection happened to have the same fd number as the incoming
1855 connection previously had. Exim would try to use TLS and fail, logging
1856 a "Bad file descriptor" error.
1858 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1859 utilities have not been installed.
1861 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1862 temporary space as the ACL may create new global variables.
1864 TL/01 LDAP support uses per connection or global context settings, depending
1865 upon the detected version of the libraries at build time.
1867 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1868 to extract and use the src ip:port in logging and expansions as if it
1869 were a direct connection from the outside internet. PPv2 support was
1870 updated based on HAProxy spec change in May 2014.
1872 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1874 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1875 Properly escape header and check for NULL return.
1877 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1880 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1882 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1883 characters in header names, implemented as a verify condition.
1884 Contributed by Michael Fischer v. Mollard.
1886 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1887 results permerror and temperror. Previous values are deprecated but
1888 still accepted. In a future release, err_perm and err_temp will be
1889 completely removed, which will be a backward incompatibility if the
1890 ACL tests for either of these two old results. Patch contributed by
1891 user bes-internal on the mailing list.
1893 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1895 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1896 selectors, in both main and reject logs.
1898 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1901 JH/07 Add malware type "sock" for talking to simple daemon.
1903 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1905 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1906 routers/transports under cutthrough routing.
1908 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1909 numbers. Touch up "bool" conditional to keep the same definition.
1911 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1913 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1915 JH/12 Expand items in router/transport headers_add or headers_remove lists
1916 individually rather than the list as a whole. Bug 1452.
1918 Required for reasonable handling of multiple headers_ options when
1919 they may be empty; requires that headers_remove items with embedded
1920 colons must have them doubled (or the list-separator changed).
1922 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1923 view the policy declared in the DMARC record. Currently, $dmarc_status
1924 is a combined value of both the record presence and the result of the
1927 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1929 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1930 dnslookup router and the smtp transport (applying to the forward
1933 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1934 of ldap servers used for a specific lookup. Patch provided by Heiko
1937 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1938 New variable $lookup_dnssec_authenticated for observability.
1940 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1941 Patch submitted by Lars Timman.
1943 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1945 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1946 Requires trusted mode and valid format message id, aborts otherwise.
1947 Patch contributed by Heiko Schlichting.
1949 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1950 certextract with support for various fields. Bug 1358.
1952 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1953 is requested by default, modifiable by smtp transport option
1956 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1957 operate on certificate variables to give certificate fingerprints
1958 Also new ${sha256:cert_variable}.
1960 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1962 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1963 Christian Aistleitner.
1965 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1967 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1968 file. Patch from Wolfgang Breyha.
1970 JH/25 Expand the coverage of the delivery $host and $host_address to
1971 client authenticators run in verify callout. Bug 1476.
1973 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1974 align with daemon_smtp_ports. Bug 72.
1976 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1977 support and error reporting did not work properly.
1979 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1980 and is readable. Patch from Andrew Colin Kissa.
1982 TL/14 Enhance documentation of ${run expansion and how it parses the
1983 commandline after expansion, particularly in the case when an
1984 unquoted variable expansion results in an empty value.
1986 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1988 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1989 by renaming away from it. Addresses GH issue 15, reported by
1992 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1994 TL/15 SECURITY: prevent double expansion in math comparison functions
1995 (can expand unsanitized data). Not remotely exploitable.
2002 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2004 PP/02 Make -n do something, by making it not do something.
2005 When combined with -bP, the name of an option is not output.
2007 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2010 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2011 $sender_host_name and config options to manage this, and basic check
2014 PP/05 DSCP support for outbound connections and control modifier for inbound.
2016 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2017 (Only plugin which currently uses this is kerberos4, which nobody should
2018 be using, but we should make it available and other future plugins might
2019 conceivably use it, even though it would break NAT; stuff *should* be
2020 using channel bindings instead).
2022 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2023 name; added for Sendmail compatibility; requires admin caller.
2024 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2025 just ignore it); requires trusted caller.
2026 Also parse but ignore: -Ac -Am -X<logfile>
2029 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2031 TL/02 Add +smtp_confirmation as a default logging option.
2033 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2034 Patch by Magnus Holmgren from 2007-02-20.
2036 TL/04 Bugzilla 1281 - Spec typo.
2037 Bugzilla 1283 - Spec typo.
2038 Bugzilla 1290 - Spec grammar fixes.
2040 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2042 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2044 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2045 mailing list by by Dmitry Isaikin.
2047 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2049 JH/02 Support "G" suffix to numbers in ${if comparisons.
2051 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2053 NM/01 Bugzilla 1197 - Spec typo
2054 Bugzilla 1196 - Spec examples corrections
2056 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2058 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2059 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2062 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2063 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2065 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2066 "acl {{name}{arg}...}", and optional args on acl condition
2069 JH/05 Permit multiple router/transport headers_add/remove lines.
2071 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2073 JH/07 Avoid using a waiting database for a single-message-only transport.
2074 Performance patch from Paul Fisher. Bugzilla 1262.
2076 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2079 JH/09 Add $headers_added variable, with content from use of ACL modifier
2080 add_header (but not yet added to the message). Bugzilla 199.
2082 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2083 Pulled from Bugzilla 817 by Wolfgang Breyha.
2085 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2087 (nb: this is the same fix as in Exim 4.80.1)
2089 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2092 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2095 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2097 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2098 advertise SMTP AUTH mechanism to us, instead of a generic
2099 protocol violation error. Also, make Exim more robust to bad
2100 data from the Dovecot auth socket.
2102 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2104 When a queue runner is handling a message, Exim first routes the
2105 recipient addresses, during which it prunes them based on the retry
2106 hints database. After that it attempts to deliver the message to
2107 any remaining recipients. It then updates the hints database using
2110 So if a recipient address works intermittently, it can get repeatedly
2111 deferred at routing time. The retry hints record remains fresh so the
2112 address never reaches the final cutoff time.
2114 This is a fairly common occurrence when a user is bumping up against
2115 their storage quota. Exim had some logic in its local delivery code
2116 to deal with this. However it did not apply to per-recipient defers
2117 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2119 This change adds a proper retry rule check during routing so that the
2120 final cutoff time is checked against the message's age. We only do
2121 this check if there is an address retry record and there is not a
2122 domain retry record; this implies that previous attempts to handle
2123 the address had the retry_use_local_parts option turned on. We use
2124 this as an approximation for the destination being like a local
2125 delivery, as in LMTP.
2127 I suspect this new check makes the old local delivery cutoff check
2128 redundant, but I have not verified this so I left the code in place.
2130 TF/02 Correct gecos expansion when From: is a prefix of the username.
2132 Test 0254 submits a message to Exim with the header
2136 When I ran the test suite under the user fanf2, Exim expanded
2137 the header to contain my full name, whereas it should have added
2138 a Resent-Sender: header. It erroneously treats any prefix of the
2139 username as equal to the username.
2141 This change corrects that bug.
2143 GF/01 DCC debug and logging tidyup
2144 Error conditions log to paniclog rather than rejectlog.
2145 Debug lines prefixed by "DCC: " to remove any ambiguity.
2147 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2149 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2150 Bug spotted by Jeremy Harris; was flawed since initial commit.
2151 Would have resulted in OCSP responses post-SNI triggering an Exim
2152 NULL dereference and crash.
2154 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2156 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2157 Bug detection, analysis and fix by Samuel Thibault.
2158 Bugzilla 1331, Debian bug #698092.
2160 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2162 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2163 Server implementation by Todd Lyons, client by JH.
2164 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2165 config variable "prdr_enable" controls whether the server
2166 advertises the facility. If the client requests PRDR a new
2167 acl_data_smtp_prdr ACL is called once for each recipient, after
2168 the body content is received and before the acl_smtp_data ACL.
2169 The client is controlled by both of: a hosts_try_prdr option
2170 on the smtp transport, and the server advertisement.
2171 Default client logging of deliveries and rejections involving
2172 PRDR are flagged with the string "PRDR".
2174 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2175 fclose(). Diagnosis by Todd Lyons.
2177 PP/17 Update configure.default to handle IPv6 localhost better.
2178 Patch by Alain Williams (plus minor tweaks).
2181 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2182 This is now consistent with GnuTLS, and is now documented: the
2183 previous undocumented portable approach to treating the option as
2184 unset was to force an expansion failure. That still works, and
2185 an empty string is now equivalent.
2187 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2188 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2189 not performing validation itself.
2191 PP/20 Added force_command boolean option to pipe transport.
2192 Patch from Nick Koston, of cPanel Inc.
2194 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2197 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2199 PP/21 Fix eximon continuous updating with timestamped log-files.
2200 Broken in a format-string cleanup in 4.80, missed when I repaired the
2201 other false fix of the same issue.
2202 Report and fix from Heiko Schlichting.
2205 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2206 Report from Prashanth Katuri.
2208 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2209 It's SecureTransport, so affects any MacOS clients which use the
2210 system-integrated TLS libraries, including email clients.
2212 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2213 using a MIME ACL for non-SMTP local injection.
2214 Report and assistance in diagnosis by Warren Baker.
2216 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2218 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2220 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2221 last id that failed so it may be referenced in subsequent ACL's.
2223 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2226 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2227 ldap library initialization, allowing self-signed CA's to be
2228 used. Also properly sets require_cert option later in code by
2229 using NULL (global ldap config) instead of ldap handle (per
2230 session). Bug diagnosis and testing by alxgomz.
2232 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2233 the src/util/ subdirectory.
2235 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2236 renamed to Transport Post Delivery Action by Jeremy Harris, as
2239 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2240 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2241 redis_servers = needs to be configured which will be used by the redis
2242 lookup. Patch from Warren Baker, of The Packet Hub.
2244 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2246 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2247 hostname or reverse DNS when processing a host list. Used suggestions
2248 from multiple comments on this bug.
2250 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2252 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2253 Missed a few lines, added it to make the runtest require no keyboard
2256 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2257 contains upper case chars. Make router use caseful_local_part.
2259 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2260 support when GnuTLS has been built with p11-kit.
2266 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2268 This, or similar/improved, will also be change PP/11 of 4.82.
2274 PP/01 Handle short writes when writing local log-files.
2275 In practice, only affects FreeBSD (8 onwards).
2276 Bugzilla 1053, with thanks to Dmitry Isaikin.
2278 NM/01 Bugzilla 949 - Documentation tweak
2280 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2283 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2285 PP/02 Implemented gsasl authenticator.
2287 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2289 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2290 `pkg-config foo` for cflags/libs.
2292 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2293 with rest of GSASL and with heimdal_gssapi.
2295 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2296 `pkg-config foo` for cflags/libs for the TLS implementation.
2298 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2299 properties get this fed in as external SSF. A number of robustness
2300 and debugging improvements to the cyrus_sasl authenticator.
2302 PP/08 cyrus_sasl server now expands the server_realm option.
2304 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2305 Patch by Jeremy Harris.
2307 PP/10 Added dbmjz lookup type.
2309 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2311 PP/12 MAIL args handles TAB as well as SP, for better interop with
2312 non-compliant senders.
2313 Analysis and variant patch by Todd Lyons.
2315 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2316 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2317 Patch from Dirk Mueller <dmueller@suse.com>
2319 PP/13 tls_peerdn now print-escaped for spool files.
2320 Observed some $tls_peerdn in wild which contained \n, which resulted
2321 in spool file corruption.
2323 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2324 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2325 or write after TLS renegotiation, which otherwise led to messages
2328 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2329 as a tracking header (ie: a signed header comes before the signature).
2330 Patch from Wolfgang Breyha.
2332 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2333 comma-sep list; embedded commas doubled.
2335 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2337 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2339 Report and patch from Dmitry Banschikov.
2341 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2342 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2343 failures. We appear to now support TLS1.1+ with Exim.
2345 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2346 lets Exim select keys and certificates based upon TLS SNI from client.
2347 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2348 before an outbound SMTP session. New log_selector, +tls_sni.
2350 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2351 NULL dereference. Report and patch from Alun Jones.
2353 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2354 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2355 Not seeing resolver debug output on NetBSD, but suspect this is a
2356 resolver implementation change.
2358 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2359 Left warnings. Added "eximon gdb" invocation mode.
2361 PP/21 Defaulting "accept_8bitmime" to true, not false.
2363 PP/22 Added -bw for inetd wait mode support.
2365 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2366 locate the relevant includes and libraries. Made this the default.
2368 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2369 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2371 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2372 This may cause build issues on older platforms.
2374 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2375 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2376 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2377 Added SNI support via GnuTLS too.
2378 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2380 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2382 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2383 Applied second patch from Janne, implementing suggestion to default
2384 multiple-strings-in-record handling to match SPF spec.
2386 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2388 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2389 read-only, out of scope).
2390 Patch from Wolfgang Breyha, report from Stuart Northfield.
2392 PP/29 Fix three issues highlighted by clang analyser static analysis.
2393 Only crash-plausible issue would require the Cambridge-specific
2394 iplookup router and a misconfiguration.
2395 Report from Marcin Mirosław.
2397 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2399 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2400 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2401 As part of this, removing so much warning spew let me fix some minor
2402 real issues in debug logging.
2404 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2405 assignment on my part. Fixed.
2407 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2408 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2409 Janne Snabb (who went above and beyond: thank you).
2411 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2412 string otherwise requires a connection and a bunch more work and it's
2413 relatively easy to get wrong. Should also expose TLS library linkage
2416 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2417 64-bit ${eval} (JH/03).
2419 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2420 GNU libc to support some of the 64-bit stuff, should not lead to
2421 conflicts. Defined before os.h is pulled in, so if a given platform
2422 needs to override this, it can.
2424 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2425 protection layer was required, which is not implemented.
2426 Bugzilla 1254, patch from Wolfgang Breyha.
2428 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2429 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2430 tls_dhparam take prime identifiers. Also unbreak combination of
2431 OpenSSL+DH_params+TLSSNI.
2433 PP/39 Disable SSLv2 by default in OpenSSL support.
2439 PP/01 Solaris build fix for Oracle's LDAP libraries.
2440 Bugzilla 1109, patch from Stephen Usher.
2442 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2444 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2447 TF/02 Fix a couple more cases where we did not log the error message
2448 when unlink() failed. See also change 4.74-TF/03.
2450 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2451 lock up or crash if it happened to be inside a call to libc when it
2452 got a SIGUSR1 from exiwhat.
2454 The SIGUSR1 handler appends the current process status to the process
2455 log which is later printed by exiwhat. It used to use the general
2456 purpose logging code to do this, but several functions it calls are
2457 not safe for signals.
2459 The new output code in the SIGUSR1 handler is specific to the process
2460 log, and simple enough that it's easy to inspect for signal safety.
2461 Removing some special cases also simplifies the general logging code.
2462 Removing the spurious timestamps from the process log simplifies
2465 TF/04 Improved ratelimit ACL condition.
2467 The /noupdate option has been deprecated in favour of /readonly which
2468 has clearer semantics. The /leaky, /strict, and /readonly update modes
2469 are mutually exclusive. The update mode is no longer included in the
2470 database key; it just determines when the database is updated. (This
2471 means that when you upgrade Exim will forget old rate measurements.)
2473 Exim now checks that the per_* options are used with an update mode that
2474 makes sense for the current ACL. For example, when Exim is processing a
2475 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2476 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2477 must specify per_mail/readonly. If you omit the update mode it defaults to
2478 /leaky where that makes sense (as before) or /readonly where required.
2480 The /noupdate option is now undocumented but still supported for
2481 backwards compatibility. It is equivalent to /readonly except that in
2482 ACLs where /readonly is required you may specify /leaky/noupdate or
2483 /strict/noupdate which are treated the same as /readonly.
2485 A useful new feature is the /count= option. This is a generalization
2486 of the per_byte option, so that you can measure the throughput of other
2487 aggregate values. For example, the per_byte option is now equivalent
2488 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2490 The per_rcpt option has been generalized using the /count= mechanism
2491 (though it's more complicated than the per_byte equivalence). When it is
2492 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2493 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2494 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2495 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2496 Note that using per_rcpt with a non-readonly update mode in more than
2497 one ACL will cause the recipients to be double-counted. (The per_mail
2498 and per_byte options don't have this problem.)
2500 The handling of very low rates has changed slightly. If the computed rate
2501 is less than the event's count (usually one) then this event is the first
2502 after a long gap. In this case the rate is set to the same as this event's
2503 count, so that the first message of a spam run is counted properly.
2505 The major new feature is a mechanism for counting the rate of unique
2506 events. The new per_addr option counts the number of different
2507 recipients that someone has sent messages to in the last time period. It
2508 behaves like per_rcpt if all the recipient addresses are different, but
2509 duplicate recipient addresses do not increase the measured rate. Like
2510 the /count= option this is a general mechanism, so the per_addr option
2511 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2512 example, measure the rate that a client uses different sender addresses
2513 with the options per_mail/unique=$sender_address. There are further
2514 details in the main documentation.
2516 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2518 TF/06 Removed a few PCRE remnants.
2520 TF/07 Automatically extract Exim's version number from tags in the git
2521 repository when doing development or release builds.
2523 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2524 Bugzilla 879. Patch from Paul Fisher.
2526 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2527 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2530 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2532 PP/05 Variable $av_failed, true if the AV scanner deferred.
2533 Bugzilla 1078. Patch from John Horne.
2535 PP/06 Stop make process more reliably on build failure.
2536 Bugzilla 1087. Patch from Heiko Schlittermann.
2538 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2539 Bugzilla 1089. Patch from Heiko Schlittermann.
2541 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2542 Bugzilla 1131. Patch from Holger Weiß.
2544 PP/09 Handle IPv6 addresses with SPF.
2545 Bugzilla 860. Patch from Wolfgang Breyha.
2547 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2549 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2552 PP/11 match_* no longer expand right-hand-side by default.
2553 New compile-time build option, EXPAND_LISTMATCH_RHS.
2554 New expansion conditions, "inlist", "inlisti".
2556 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2558 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2560 PP/14 fix log_write() format string regression from TF/03.
2561 Bugzilla 1152. Patch from Dmitry Isaikin.
2567 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2569 PP/02 Harmonised TLS library version reporting; only show if debugging.
2570 Layout now matches that introduced for other libraries in 4.74 PP/03.
2572 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2574 PP/04 New "dns_use_edns0" global option.
2576 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2579 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2580 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2582 TK/01 Updated PolarSSL code to 0.14.2.
2583 Bugzilla 1097. Patch from Andreas Metzler.
2585 PP/07 Catch divide-by-zero in ${eval:...}.
2586 Fixes bugzilla 1102.
2588 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2591 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2592 format-string attack -- SECURITY: remote arbitrary code execution.
2594 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2595 time unintentionally subject to list matching rules, letting the header
2596 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2597 arbitrary string expansion). This allowed for information disclosure.
2599 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2600 INT_MIN/-1 -- value coerced to INT_MAX.
2606 NM/01 Workaround for PCRE version dependency in version reporting
2609 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2610 This fixes portability to compilers other than gcc, notably
2611 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2613 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2614 makefiles for portability to HP-UX and POSIX correctness.
2616 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2617 Also via indented variable definition in the Makefile.
2618 (Debugging by Oliver Heesakkers).
2620 PP/02 Restore caching of spamd results with expanded spamd_address.
2621 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2623 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2624 Improves build reliability. Fix from: Frank Elsner
2626 NM/02 Fix wide character breakage in the rfc2047 coding
2627 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2629 NM/03 Allow underscore in dnslist lookups
2630 Fixes bug 1026. Patch from Graeme Fowler
2632 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2633 Code patches from Adam Ciarcinski of NetBSD.
2635 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2638 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2639 is logged, to avoid truncation. Patch from John Horne.
2641 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2642 Patch from Jakob Hirsch.
2644 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2645 SQL string expansion failure details.
2646 Patch from Andrey Oktyabrski.
2648 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2649 Patch from Simon Arlott.
2651 PP/09 New lookups functionality failed to compile on old gcc which rejects
2652 extern declarations in function scope.
2653 Patch from Oliver Fleischmann
2655 PP/10 Use sig_atomic_t for flags set from signal handlers.
2656 Check getgroups() return and improve debugging.
2657 Fixed developed for diagnosis in bug 927 (which turned out to be
2660 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2661 Patch from Mark Zealey.
2663 PP/12 Bugzilla 1056: Improved spamd server selection.
2664 Patch from Mark Zealey.
2666 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2667 Based on patch from Heiko Schlittermann.
2669 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2670 Patch from Uwe Doering, sign-off by Michael Haardt.
2672 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2673 variable declaration deep within a block. Bug and patch from
2676 PP/15 lookups-Makefile IRIX compatibility coercion.
2678 PP/16 Make DISABLE_DKIM build knob functional.
2680 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2681 Patch by Simon Arlott
2683 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2684 variable argument macros. Our copy now differs from upstream.
2690 TF/01 Failure to get a lock on a hints database can have serious
2691 consequences so log it to the panic log.
2693 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2694 controlled using the smtp_confirmation log selector.
2696 TF/03 Include the error message when we fail to unlink a spool file.
2698 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2699 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2700 for maintaining out-of-tree patches for some time.
2702 PP/01 Bugzilla 139: Documentation and portability issues.
2703 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2704 Handle per-OS dynamic-module compilation flags.
2706 PP/02 Let /dev/null have normal permissions.
2707 The 4.73 fixes were a little too stringent and complained about the
2708 permissions on /dev/null. Exempt it from some checks.
2709 Reported by Andreas M. Kirchwitz.
2711 PP/03 Report version information for many libraries, including
2712 Exim version information for dynamically loaded libraries. Created
2713 version.h, now support a version extension string for distributors
2714 who patch heavily. Dynamic module ABI change.
2716 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2717 privilege escalation vulnerability whereby the Exim run-time user
2718 can cause root to append content of the attacker's choosing to
2721 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2724 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2725 If dropping privileges for untrusted macros, we disabled normal logging
2726 on the basis that it would fail; for the Exim run-time user, this is not
2727 the case, and it resulted in successful deliveries going unlogged.
2728 Fixed. Reported by Andreas Metzler.
2734 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2735 only prepend for the Resent-* case. Fixes regression introduced in
2736 Exim 4.70 by NM/22 for Bugzilla 607.
2738 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2739 increasing numbers of administrators be bitten by this.
2741 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2743 PP/03 Bugzilla 994: added openssl_options main configuration option.
2745 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2747 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2749 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2751 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2752 without a peer certificate, leading to a segfault because of an
2753 assumption that peers always have certificates. Be a little more
2754 paranoid. Problem reported by Martin Tscholak.
2756 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2757 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2758 NB: ClamAV planning to remove STREAM in "middle of 2010".
2759 CL also introduces -bmalware, various -d+acl logging additions and
2760 more caution in buffer sizes.
2762 PP/09 Implemented reverse_ip expansion operator.
2764 PP/10 Bugzilla 937: provide a "debug" ACL control.
2766 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2768 PP/12 Bugzilla 973: Implement --version.
2770 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2772 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2774 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2776 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2777 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2778 ignore trailing whitespace.
2780 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2782 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2785 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2786 Notification from Dr Andrew Aitchison.
2788 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2789 ExtendedDetectionInfo response format.
2790 Notification from John Horne.
2792 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2795 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2796 XSL and documented dependency on system catalogs, with examples of how
2799 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2802 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2803 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2804 configuration file which is writeable by the Exim user or group.
2806 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2807 of configuration files to cover files specified with the -C option if
2808 they are going to be used with root privileges, not just the default
2811 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2812 option (effectively making it always true).
2814 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2815 files to be used while preserving root privileges.
2817 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2818 that rogue child processes cannot use them.
2820 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2821 run-time user, instead of root.
2823 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2824 Exim run-time user without dropping privileges.
2826 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2827 result string, instead of calling string_vformat() twice with the same
2830 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2831 for other users. Others should always drop root privileges if they use
2832 -C on the command line, even for a whitelisted configure file.
2834 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2836 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2842 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2843 $data_path, and $header_path variables; fixed documentation bugs and
2846 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2847 exipick to access non-standard spools, including the "frozen" queue
2850 NM/01 Bugzilla 965: Support mysql stored procedures.
2851 Patch from Alain Williams
2853 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2855 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2856 Patch from Andreas Metzler
2858 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2859 Patch from Kirill Miazine
2861 NM/05 Bugzilla 671: Added umask to procmail example.
2863 JJ/03 installed exipick 20100323.0, fixing doc bug
2865 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2866 directory. Notification and patch from Dan Rosenberg.
2868 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2870 TK/02 Improve log output when DKIM signing operation fails.
2872 MH/01 Treat the transport option dkim_domain as a colon separated
2873 list, not as a single string, and sign the message with each element,
2874 omitting multiple occurences of the same signer.
2876 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2877 Bugzilla 985, 986. Patch by Simon Arlott
2879 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2880 Patch by Simon Arlott
2882 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2883 MBX locking. Notification from Dan Rosenberg.
2889 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2891 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2893 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2895 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2897 NM/04 Bugzilla 915: Flag broken perl installation during build.
2903 TK/01 Added patch by Johannes Berg that expands the main option
2904 "spamd_address" if it starts with a dollar sign.
2906 TK/02 Write list of recipients to X-Envelope-Sender header when building
2907 the mbox-format spool file for content scanning (suggested by Jakob
2910 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2911 (http://www.dcc-servers.net/) support via dccifd. Activated by
2912 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2914 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2915 by Mark Daniel Reidel <mr@df.eu>.
2917 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2918 When building exim an external PCRE library is now needed -
2919 PCRE is a system library on the majority of modern systems.
2920 See entry on PCRE_LIBS in EDITME file.
2922 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2923 conversation. Added nologin parameter to request.
2924 Patch contributed by Kirill Miazine.
2926 TF/01 Do not log submission mode rewrites if they do not change the address.
2928 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2930 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2931 log files in place. Contributed by Roberto Lima.
2933 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2935 TF/03 Bugzilla 615: When checking the local_parts router precondition
2936 after a local_part_suffix or local_part_prefix option, Exim now
2937 does not use the address's named list lookup cache, since this
2938 contains cached lookups for the whole local part.
2940 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2941 Robert Millan. Documentation is in experimental-spec.txt.
2943 TF/04 Bugzilla 668: Fix parallel build (make -j).
2945 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2947 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2948 Patch provided by Jan Srzednicki.
2950 TF/05 Leading white space used to be stripped from $spam_report which
2951 wrecked the formatting. Now it is preserved.
2953 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2954 that they are available at delivery time.
2956 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2958 TF/08 TLS error reporting now respects the incoming_interface and
2959 incoming_port log selectors.
2961 TF/09 Produce a more useful error message if an SMTP transport's hosts
2962 setting expands to an empty string.
2964 NM/06 Bugzilla 744: EXPN did not work under TLS.
2965 Patch provided by Phil Pennock.
2967 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2968 Patch provided by Richard Godbee.
2970 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2971 acl_smtp_notquit, added index entry.
2973 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2974 Patch provided by Eugene Bujak.
2976 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2977 accept(). Patch provided by Maxim Dounin.
2979 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2980 Patch provided by Phil Pennock.
2982 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2984 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2985 Patch provided by Brad "anomie" Jorsch.
2987 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2988 Patch provided by Dean Brooks.
2990 TK/05 Add native DKIM support (does not depend on external libraries).
2992 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2993 Patch provided by Graeme Fowler.
2995 NM/16 Bugzilla 851: Documentation example syntax fix.
2997 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2999 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3002 NM/19 Bugzilla 745: TLS version reporting.
3003 Patch provided by Phil Pennock.
3005 NM/20 Bugzilla 167: bool: condition support.
3006 Patch provided by Phil Pennock.
3008 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3009 clients. Patch provided by Phil Pennock.
3011 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3012 Patch provided by Brad "anomie" Jorsch.
3014 NM/23 Bugzilla 687: Fix misparses in eximstats.
3015 Patch provided by Heiko Schlittermann.
3017 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3018 Patch provided by Heiko Schlittermann.
3020 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3021 plus update to original patch.
3023 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3025 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3026 Patch provided by David Brownlee.
3028 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3030 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3032 NM/30 Bugzilla 888: TLS documentation bugfixes.
3034 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3036 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3037 Unlike the original bugzilla I have changed all shell scripts in src tree.
3039 NM/33 Bugzilla 898: Transport filter timeout fix.
3040 Patch by Todd Rinaldo.
3042 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3043 Patch by Serge Demonchaux.
3045 NM/35 Bugzilla 39: Base64 decode bug fixes.
3046 Patch by Jakob Hirsch.
3048 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3050 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3052 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3054 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3060 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3061 ALT-N's libdkim that I have put here:
3062 http://duncanthrax.net/exim-experimental/
3064 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3065 were called 'true' and it seems that C99 defines that as a reserved
3066 keyword to be used with 'bool' variable types. That means you could
3067 not include C99-style headers which use bools without triggering
3068 build errors in sieve.c.
3070 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3071 as mailq or other aliases. Changed the --help handling significantly
3072 to do whats expected. exim_usage() emits usage/help information.
3074 SC/01 Added the -bylocaldomain option to eximstats.
3076 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3078 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3080 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3086 PH/01 Another patch from the Sieve maintainer.
3088 PH/02 When an IPv6 address is converted to a string for single-key lookup
3089 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3090 dots are used instead of colons so that keys in lsearch files need not
3091 contain colons. This was done some time before quoting was made available
3092 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3093 (notated using the quote facility) so as to distinguish them from IPv4
3094 keys. This meant that lookups for IP addresses in host lists did not work
3095 for iplsearch lookups.
3097 This has been fixed by arranging for IPv6 addresses to be expressed with
3098 colons if the lookup type is iplsearch. This is not incompatible, because
3099 previously such lookups could never work.
3101 The situation is now rather anomalous, since one *can* have colons in
3102 ordinary lsearch keys. However, making the change in all cases is
3103 incompatible and would probably break a number of configurations.
3105 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3108 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3109 conversion specification without a maximum field width, thereby enabling
3110 a rogue spamd server to cause a buffer overflow. While nobody in their
3111 right mind would setup Exim to query an untrusted spamd server, an
3112 attacker that gains access to a server running spamd could potentially
3113 exploit this vulnerability to run arbitrary code as the Exim user.
3115 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3116 $primary_hostname instead of what libspf2 thinks the hosts name is.
3118 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3119 a directory entry by the name of the lookup key. Previously, if a
3120 symlink pointed to a non-existing file or a file in a directory that
3121 Exim lacked permissions to read, a lookup for a key matching that
3122 symlink would fail. Now it is enough that a matching directory entry
3123 exists, symlink or not. (Bugzilla 503.)
3125 PH/03 The body_linecount and body_zerocount variables are now exported in the
3128 PH/04 Added the $dnslist_matched variable.
3130 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3131 This means they are set thereafter only if the connection becomes
3134 PH/06 Added the client_condition to authenticators so that some can be skipped
3135 by clients under certain conditions.
3137 PH/07 The error message for a badly-placed control=no_multiline_responses left
3138 "_responses" off the end of the name.
3140 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3142 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3143 (without spaces) instead of just copying the configuration text.
3145 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3147 PH/11 Added $max_received_linelength.
3149 PH/12 Added +ignore_defer and +include_defer to host lists.
3151 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3152 way in which PCRE > 7.0 is built.
3154 PH/14 Implemented queue_only_load_latch.
3156 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3157 MAIL command. The effect was to mangle the value on 64-bit systems.
3159 PH/16 Another patch from the Sieve maintainer.
3161 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3163 PH/18 If a system quota error occurred while trying to create the file for
3164 a maildir delivery, the message "Mailbox is full" was not appended to the
3165 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3166 only to a quota excession during the actual writing of the file.
3168 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3169 characters?) which causes problems in log lines. The DN values are now
3170 passed through string_printing() before being added to log lines.
3172 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3173 and InterBase are left for another time.)
3175 PH/21 Added message_body_newlines option.
3177 PH/22 Guard against possible overflow in moan_check_errorcopy().
3179 PH/23 POSIX allows open() to be a macro; guard against that.
3181 PH/24 If the recipient of an error message contained an @ in the local part
3182 (suitably quoted, of course), incorrect values were put in $domain and
3183 $local_part during the evaluation of errors_copy.
3189 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3190 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3193 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3194 issue a MAIL command.
3196 PH/02 In an ACL statement such as
3198 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3200 if a client was not listed at all, or was listed with a value other than
3201 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3202 the condition was not true (as it should be), so access was not denied.
3203 The bug was that the ! inversion was incorrectly passed on to the second
3204 item. This has been fixed.
3206 PH/03 Added additional dnslists conditions == and =& which are different from
3207 = and & when the dns lookup returns more than one IP address.
3209 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3210 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3212 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3213 FSYNC, which compiles an option called disable_fsync that allows for
3214 bypassing fsync(). The documentation is heavily laced with warnings.
3216 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3218 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3219 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3220 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3221 including adding "make clean"; (3) Added -fPIC when compiling the test
3222 dynamically loaded module, to get rid of a warning.
3224 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3225 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3226 = 0s. The bug is otherwise harmless.
3228 PH/07 There was a bug in the dovecot authenticator such that the value of
3229 $auth1 could be overwritten, and so not correctly preserved, after a
3230 successful authentication. This usually meant that the value preserved by
3231 the server_setid option was incorrect.
3233 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3235 PH/09 Installed PCRE release 7.0.
3237 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3238 run for batched SMTP input. It is now run at the start of every message
3239 in the batch. While fixing this I discovered that the process information
3240 (output by running exiwhat) was not always getting set for -bs and -bS
3241 input. This is fixed, and it now also says "batched" for BSMTP.
3243 PH/11 Added control=no_pipelining.
3245 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3246 patch, slightly modified), and move the expansion of helo_data till after
3247 the connection is made in the smtp transport (so it can use these
3250 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3252 PH/14 Added log_selector = +pid.
3254 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3256 PH/16 Add ${if forany and ${if forall.
3258 PH/17 Added dsn_from option to vary the From: line in DSNs.
3260 PH/18 Flush SMTP output before performing a callout, unless control =
3261 no_callout_flush is set.
3263 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3264 was true (the default) a successful delivery failed to delete the retry
3265 item, thus causing premature timeout of the address. The bug is now
3268 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3270 PH/21 Long custom messages for fakedefer and fakereject are now split up
3271 into multiline responses in the same way that messages for "deny" and
3272 other ACL rejections are.
3274 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3275 with slight modification.
3277 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3278 draft, changing the syntax and factoring some duplicate code".
3280 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3281 for deliveries of the second and subsequent messages over the same SMTP
3284 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3285 ${reduce, with only minor "tidies".
3287 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3289 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3290 expansion side effects.
3292 PH/27 When a message times out after an over-quota error from an Exim-imposed
3293 quota, the bounce message says "mailbox is full". This message was not
3294 being given when it was a system quota that was exceeded. It now should
3297 MH/03 Made $recipients available in local_scan(). local_scan() already has
3298 better access to the recipient list through recipients_list[], but
3299 $recipients can be useful in postmaster-provided expansion strings.
3301 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3302 in the case of a MAIL command with additional options following the
3303 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3304 were accidentally chopped off.
3306 PH/29 SMTP synchronization checks are implemented when a command is read -
3307 there is a check that no more input is waiting when there shouldn't be
3308 any. However, for some commands, a delay in an ACL can mean that it is
3309 some time before the response is written. In this time, more input might
3310 arrive, invalidly. So now there are extra checks after an ACL has run for
3311 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3312 pipelining has not been advertised.
3314 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3316 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3317 returned in the SMTP response when smtp_return_error_details was false.
3318 This has been fixed.
3320 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3321 instead of the C I/O that was originally supplied, because problems were
3322 reported on Solaris.
3324 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3325 Exim which did not show up earlier: it was assuming that a call to
3326 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3327 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3328 was a macro that became an assignment, so it seemed to work. This has
3329 changed to a proper function call with a void return, hence the compile
3330 error. Exim's code has been fixed.
3332 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3335 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3337 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3339 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3340 because it turns out to be too big or there is a timeout) there is no
3341 "Completed" line in the log. When some messages of this type were
3342 selected by exigrep, they were listed as "not completed". Others were
3343 picked up by some special patterns. I have improved the selection
3344 criteria to be more general.
3346 PH/38 The host_find_failed option in the manualroute router can now be set
3347 to "ignore", to completely ignore a host whose IP address cannot be
3348 found. If all hosts are ignored, the behaviour is controlled by the new
3349 host_all_ignored option.
3351 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3352 homing or because of multiple MX records with /mx) generated more than
3353 one IP address, and the following item turned out to be the local host,
3354 all the secondary addresses of the first item were incorrectly removed
3355 from the list, along with the local host and any following hosts (which
3356 is what is supposed to happen).
3358 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3359 whoever called Exim into the -H file. In the case of the daemon it was
3360 behaving confusingly. When first started, it used values for whoever
3361 started the daemon, but after a SIGHUP it used the Exim user (because it
3362 calls itself on a restart). I have changed the code so that it now always
3365 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3366 message are rejected with the same error (e.g. no authentication or bad
3367 sender address), and a DATA command is nevertheless sent (as can happen
3368 with PIPELINING or a stupid MUA), the error message that was given to the
3369 RCPT commands is included in the rejection of the DATA command. This is
3370 intended to be helpful for MUAs that show only the final error to their
3373 PH/42 Another patch from the Sieve maintainer.
3375 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3376 Eximstats - Fixed some broken HTML links and added missing column headers
3378 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3379 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3381 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3383 PH/43 Yet another patch from the Sieve maintainer.
3385 PH/44 I found a way to check for a TCP/IP connection going away before sending
3386 the response to the final '.' that terminates a message, but only in the
3387 case where the client has not sent further data following the '.'
3388 (unfortunately, this is allowed). However, in many cases there won't be
3389 any further data because there won't be any more messages to send. A call
3390 to select() can be used: if it shows that the input is "ready", there is
3391 either input waiting, or the socket has been closed. An attempt to read
3392 the next input character can distinguish the two cases. Previously, Exim
3393 would have sent an OK response which the client would never have see.
3394 This could lead to message repetition. This fix should cure that, at
3395 least in a lot of common cases.
3397 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3398 advertised in response to EHLO.
3404 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3405 fixed by 4.65/MH/01 (is this a record?) are fixed:
3407 (i) An empty string was always treated as zero by the numeric comparison
3408 operators. This behaviour has been restored.
3410 (ii) It is documented that the numeric comparison operators always treat
3411 their arguments as decimal numbers. This was broken in that numbers
3412 starting with 0 were being interpreted as octal.
3414 While fixing these problems I realized that there was another issue that
3415 hadn't been noticed. Values of message_size_limit (both the global option
3416 and the transport option) were treated as octal if they started with 0.
3417 The documentation was vague. These values are now always treated as
3418 decimal, and I will make that clear in the documentation.
3424 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3425 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3428 MH/01 Don't check that the operands of numeric comparison operators are
3429 integers when their expansion is in "skipping" mode (fixes bug
3430 introduced by 4.64-PH/07).
3432 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3433 child addresses, Exim now panics and dies. Previously, because the count
3434 is held in a short int, deliveries were likely to be lost. As such a
3435 large number of recipients for a single message is ridiculous
3436 (performance will be very, very poor), I have chosen to impose a limit
3437 rather than extend the field.
3443 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3444 leftover -K file (the existence of which was triggered by #402).
3445 While we were at it, introduced process PID as part of the -K
3446 filename. This should rule out race conditions when creating
3449 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3450 processing considerably. Previous code took too long for large mails,
3451 triggering a timeout which in turn triggers #401.
3453 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3454 in the DK code in transports.c. sendfile() is not really portable,
3455 hence the _LINUX specificness.
3457 TF/01 In the add_headers option to the mail command in an Exim filter,
3458 there was a bug that Exim would claim a syntax error in any
3459 header after the first one which had an odd number of characters
3462 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3463 callout verification, Exim cached a "reject" for the entire domain. This
3464 is correct for most verifications, but it is not correct for a recipient
3465 verification with use_sender or use_postmaster set, because in that case
3466 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3467 case of MAIL FROM:<> rejection from other early rejections (e.g.
3468 rejection of HELO). When verifying a recipient using a non-null MAIL
3469 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3470 Whatever the result of the callout, the value of the domain cache is
3471 left unchanged (for any other kind of callout, getting as far as trying
3472 RCPT means that the domain itself is ok).
3474 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3477 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3478 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3479 session, the daemon ignores ECONNECT errors and logs others; it now
3480 ignores EPIPE as well.
3482 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3483 (quoted-printable decoding).
3485 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3486 later the small subsequent patch to fix an introduced bug.
3488 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3490 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3492 PH/08 An error is now given if message_size_limit is specified negative.
3494 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3495 to be given (somewhat) arbitrary names.
3497 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3500 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3501 miscellaneous code fixes
3503 PH/10 Added the log_reject_target ACL modifier to specify where to log
3506 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3507 hostname. This is wrong, because it relates to the incoming message (and
3508 probably the interface on which it is arriving) and not to the outgoing
3509 callout (which could be using a different interface). This has been
3510 changed to use the value of the helo_data option from the smtp transport
3511 instead - this is what is used when a message is actually being sent. If
3512 there is no remote transport (possible with a router that sets up host
3513 addresses), $smtp_active_hostname is used.
3515 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3516 tweaks were necessary in order to get it to work (see also 21 below):
3517 (a) The code assumed that strncpy() returns a negative number on buffer
3518 overflow, which isn't the case. Replaced with Exim's string_format()
3520 (b) There were several signed/unsigned issues. I just did the minimum
3521 hacking in of casts. There is scope for a larger refactoring.
3522 (c) The code used strcasecmp() which is not a standard C function.
3523 Replaced with Exim's strcmpic() function.
3524 (d) The code set only $1; it now sets $auth1 as well.
3525 (e) A simple test gave the error "authentication client didn't specify
3526 service in request". It would seem that Dovecot has changed its
3527 interface. Fortunately there's a specification; I followed it and
3528 changed what the client sends and it appears to be working now.
3530 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3533 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3534 address A is aliased to B and C, where B exists and C does not. Without
3535 -v the output is "A verified" because verification stops after a
3536 successful redirection if more than one address is generated. However,
3537 with -v the child addresses are also verified. Exim was outputting "A
3538 failed to verify" and then showing the successful verification for C,
3539 with its parentage. It now outputs "B failed to verify", showing B's
3540 parentage before showing the successful verification of C.
3542 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3543 look up a TXT record in a specific list after matching in a combined
3546 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3547 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3548 they consult the DNS. I had assumed they would set it the way they
3549 wanted; and indeed my experiments on Linux seem to show that in some
3550 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3551 To be on the safe side, however, I have now made the interface to
3552 host_find_byname() similar to host_find_bydns(), with an argument
3553 containing the DNS resolver options. The host_find_byname() function now
3554 sets these options at its start, just as host_find_bydns() does. The smtp
3555 transport options dns_qualify_single and dns_search_parents are passed to
3556 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3557 of host_find_byname() use the default settings of RES_DEFNAMES
3558 (qualify_single) but not RES_DNSRCH (search_parents).
3560 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3561 spool_read_header() do less string testing, by means of a preliminary
3562 switch on the second character of optional "-foo" lines. (This is
3563 overdue, caused by the large number of possibilities that now exist.
3564 Originally there were few.) While I was there, I also converted the
3565 str(n)cmp tests so they don't re-test the leading "-" and the first
3566 character, in the hope this might squeeze out yet more improvement.
3568 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3569 flag allowing group syntax was set by the header_syntax check but not
3570 turned off, possible causing trouble later; (2) The flag was not being
3571 set at all for the header_verify test, causing "group"-style headers to
3572 be rejected. I have now set it in this case, and also caused header_
3573 verify to ignore an empty address taken from a group. While doing this, I
3574 came across some other cases where the code for allowing group syntax
3575 while scanning a header line wasn't quite right (mostly, not resetting
3576 the flag correctly in the right place). These bugs could have caused
3577 trouble for malformed header lines. I hope it is now all correct.
3579 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3580 with the "reply" argument non-NULL. The code, however (which originally
3581 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3582 but it didn't always do it. This confused somebody who was copying the
3583 code for some other use. I have removed all the tests.
3585 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3586 feature that was used to support insecure browsers during the U.S. crypto
3587 embargo. It requires special client support, and Exim is probably the
3588 only MTA that supported it -- and would never use it because real RSA is
3589 always available. This code has been removed, because it had the bad
3590 effect of slowing Exim down by computing (never used) parameters for the
3591 RSA_EXPORT functionality.
3593 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3594 authenticator to fail if there's a tab character in the incoming data
3595 (there should never be unless someone is messing about, as it's supposed
3596 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3597 if the connection is using TLS or if the remote IP is the same as the
3598 local IP, and the "valid-client-cert option" if a client certificate has
3601 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3602 authenticators. This can be used for authorization after authentication
3603 succeeds. (In the case of plaintext, it servers for both authentication
3606 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3607 if any retry times were supplied.
3609 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3610 connection, where there is no client IP address to check. In this
3611 situation, the verify now always succeeds.
3613 PH/25 Applied John Jetmore's -Mset patch.
3615 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3617 PH/27 In a string expansion for a processed (not raw) header when multiple
3618 headers of the same name were present, leading whitespace was being
3619 removed from all of them, but trailing whitespace was being removed only
3620 from the last one. Now trailing whitespace is removed from each header
3621 before concatenation. Completely empty headers in a concatenation (as
3622 before) are ignored.
3624 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3625 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3627 PH/29 [Removed. This was a change that I later backed out, and forgot to
3628 correct the ChangeLog entry (that I had efficiently created) before
3629 committing the later change.]
3631 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3632 address errors (4xx response to RCPT) over the same connection as other
3633 messages routed to the same hosts. Such deliveries are always "forced",
3634 so retry times are not inspected. This resulted in far too many retries
3635 for the affected addresses. The effect occurred only when there were more
3636 hosts than the hosts_max_try setting in the smtp transport when it had
3637 the 4xx errors. Those hosts that it had tried were not added to the list
3638 of hosts for which the message was waiting, so if all were tried, there
3639 was no problem. Two fixes have been applied:
3641 (i) If there are any address or message errors in an SMTP delivery, none
3642 of the hosts (tried or untried) are now added to the list of hosts
3643 for which the message is waiting, so the message should not be a
3644 candidate for sending over the same connection that was used for a
3645 successful delivery of some other message. This seems entirely
3646 reasonable: after all the message is NOT "waiting for some host".
3647 This is so "obvious" that I'm not sure why it wasn't done
3648 previously. Hope I haven't missed anything, but it can't do any
3649 harm, as the worst effect is to miss an optimization.
3651 (ii) If, despite (i), such a delivery is accidentally attempted, the
3652 routing retry time is respected, so at least it doesn't keep
3653 hammering the server.
3655 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3656 in ${readsocket because some servers need this prod.
3658 PH/32 Added some extra debug output when updating a wait-xxx database.
3660 PH/33 The hint "could be header name not terminated by colon", which has been
3661 given for certain expansion errors for a long time, was not being given
3662 for the ${if def:h_colon_omitted{... case.
3664 PH/34 The spec says: "With one important exception, whenever a domain list is
3665 being scanned, $domain contains the subject domain." There was at least
3666 one case where this was not true.
3668 PH/35 The error "getsockname() failed: connection reset by peer" was being
3669 written to the panic log as well as the main log, but it isn't really
3670 panic-worthy as it just means the connection died rather early on. I have
3671 removed the panic log writing for the ECONNRESET error when getsockname()
3674 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3675 runs only) independently of the message's sender address. This meant
3676 that, if the 4xx error was in fact related to the sender, a different
3677 message to the same recipient with a different sender could confuse
3678 things. In particular, this can happen when sending to a greylisting
3679 server, but other circumstances could also provoke similar problems.
3680 I have changed the default so that the retry time for these errors is now
3681 based a combination of the sender and recipient addresses. This change
3682 can be overridden by setting address_retry_include_sender=false in the
3685 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3686 remote server are returned as part of bounce messages. This was not
3687 happening for LMTP over a pipe (the lmtp transport), but now it is the
3688 same for both kinds of LMTP.
3690 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3691 in header lines that were in fact CNAMEs. This is no longer the case.
3693 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3694 and queue runs started by the daemon processed all messages. This has
3695 been fixed so that -R and -S can now usefully be given with -q<time>.
3697 PH/40 Import PCRE release 6.7 (fixes some bugs).
3699 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3701 PH/42 Give an error if -q is specified more than once.
3703 PH/43 Renamed the variables $interface_address and $interface_port as
3704 $received_ip_address and $received_port, to make it clear that these
3705 values apply to message reception, and not to the outgoing interface when
3706 a message is delivered. (The old names remain recognized, of course.)
3708 PH/44 There was no timeout on the connect() call when using a Unix domain
3709 socket in the ${readsocket expansion. There now is.
3711 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3712 be meaningful with "accept".
3714 SC/01 Eximstats V1.43
3715 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3717 SC/02 Eximstats V1.44
3718 Use a glob alias rather than an array ref in the generated
3719 parser. This improves both readability and performance.
3721 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3722 Collect SpamAssassin and rejection statistics.
3723 Don't display local sender or destination tables unless
3724 there is data to show.
3725 Added average volumes into the top table text output.
3727 SC/04 Eximstats V1.46
3728 Collect data on the number of addresses (recipients)
3729 as well as the number of messages.
3731 SC/05 Eximstats V1.47
3732 Added 'Message too big' to the list of mail rejection
3733 reasons (thanks to Marco Gaiarin).
3735 SC/06 Eximstats V1.48
3736 Mainlog lines which have GMT offsets and are too short to
3737 have a flag are now skipped.
3739 SC/07 Eximstats V1.49 (Alain Williams)
3740 Added the -emptyok flag.
3742 SC/08 Eximstats V1.50
3743 Fixes for obtaining the IP address from reject messages.
3745 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3746 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3747 whitespace changes from 4.64-PH/27
3749 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3752 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3753 are found, allow negative numbers in numeric criteria)
3755 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3757 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3758 to match changes made in 4.64-PH/43
3760 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3762 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3763 "domain" is actually an IP address, and force a failure. This locks out
3764 those revolvers/nameservers that support "A-for-A" lookups, in
3765 contravention of the specifications.
3767 PH/48 When a host name was looked up from an IP address, and the subsequent
3768 forward lookup of the name timed out, the host name was left in
3769 $sender_host_name, contrary to the specification.
3771 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3772 restricted to single-key lookups, Exim was not diagnosing an error if
3773 * or *@ was used with a query-style lookup.
3775 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3777 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3778 long ago, but noone interested enough thought of it. Let's just say that
3779 the "1.1" means that there are some new functions that weren't there at
3780 some point in the past.
3782 PH/51 Error processing for expansion failure of helo_data from an smtp
3783 transport during callout processing was broken.
3785 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3786 tested/used via the -bh/-bhc/-bs options.
3788 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3789 bug, fixed in subsequent PCRE releases).
3791 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3792 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3794 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3800 SC/01 Use a glob alias rather than an array ref in eximstats generated
3801 parser. This improves both readability and performance.
3803 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3804 Don't display local sender or destination tables in eximstats unless
3805 there is data to show.
3806 Added average volumes into the eximstats top table text output.
3808 SC/03 Collect data on the number of addresses (recipients) as well
3809 as the number of messages in eximstats.
3811 TF/01 Correct an error in the documentation for the redirect router. Exim
3812 does (usually) call initgroups() when daemonizing.
3814 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3815 with consistent privilege compared to when running as a daemon.
3817 TF/03 Note in the spec that $authenticated_id is not set for local
3818 submissions from trusted users.
3820 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3821 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3823 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3824 by adding some example configuration directives to the default
3825 configuration file. A little bit of work is required to uncomment the
3826 directives and define how usernames and passwords are checked, but
3827 there is now a framework to start from.
3829 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3830 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3831 without this. I don't know how relevant this is to other LDAP libraries.
3833 PH/02 Add the verb name to the "unknown ACL verb" error.
3835 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3837 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3839 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3840 directory not expanded when it should be if an expanded home directory
3841 was set for the address (which is overridden by the transport).
3843 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3846 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3847 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3848 because it is too late at that time, and has no effect.
3850 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3851 security issue with \' (bugzilla #107). I could not use the
3852 PQescapeStringConn() function, because it needs a PGconn value as one of
3855 PH/08 When testing addresses using -bt, indicate those final addresses that
3856 are duplicates that would not cause an additional delivery. At least one
3857 person was confused, thinking that -bt output corresponded to deliveries.
3858 (Suppressing duplicates isn't a good idea as you lose the information
3859 about possibly different redirections that led to the duplicates.)
3861 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3862 systems where poll() doesn't work, in particular OS X.
3864 PH/10 Added more information to debugging output for retry time not reached.
3866 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3867 operations in malware.c.
3869 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3872 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3873 syslog_duplication set false, log lines that would normally be written
3874 both the the main log and to the reject log were not written to syslog at
3877 PH/14 In the default configuration, change the use of "message" in ACL warn
3878 statements to "add_header".
3880 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3881 not followed by a command (e.g. "seen endif").
3883 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3884 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3887 PH/17 Added extra conditions to the default value of delay_warning_condition
3891 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3892 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3893 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3896 The Auto-Submitted: and various List- headers are standardised, whereas I
3897 don't think Precedence: ever was.
3899 PH/18 Refactored debugging code in route_finduser() to show more information,
3900 in particular, the error code if getpwnam() issues one.
3902 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3903 This is apparently needed in addition to the PH/07 change above to avoid
3904 any possible encoding problems.
3906 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3907 but not after initializing Perl.
3909 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3910 output them only if debugging. By default they are written stderr,
3911 apparently, which is not desirable.
3913 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3916 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3919 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3921 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3922 authenticated or an ident call has been made. Suppress the default
3923 values for $authenticated_id and $authenticated_sender (but permit -oMai
3924 and -oMas) when testing with -bh.
3926 PH/24 Re-jigged the order of the tests in the default configuration so that the
3927 tests for valid domains and recipients precede the DNS black list and CSA
3928 tests, on the grounds that those ones are more expensive.
3930 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3931 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3932 This bug exists in every version of Exim that I still have, right back to
3935 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3936 However, an attempt to turn on case-sensitivity in a regex key by
3937 including (?-i) didn't work because the subject string was already
3938 lowercased, and the effects were non-intuitive. It turns out that a
3939 one-line patch can be used to allow (?-i) to work as expected.
3945 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3946 other effects) broke the use of negated acl sub-conditions.
3948 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3951 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3952 "Deny" causes Exim to reject the incoming connection with a 554 error.
3953 Unfortunately, if there is a major crisis, such as a disk failure,
3954 tcp-wrappers gives "deny", whereas what one would like would be some
3955 kind of temporary error. A kludge has been added to help with this.
3956 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3957 554 error is used if errno is still zero or contains ENOENT (which occurs
3958 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3961 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3963 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3964 errors. Otherwise a message that provokes a temporary error (when other
3965 messages do not) can cause a whole host to time out.
3967 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3968 addresses were routed directly to files or pipes from a redirect router.
3969 File deliveries just didn't batch; pipe deliveries might have suffered
3972 PH/06 A failure to get a lock for a hints database would erroneously always say
3973 "Failed to get write lock", even when it was really a read lock.
3975 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3976 of 0600. This has been changed to use the value of the lockfile_mode
3977 option (which defaults to 0600).
3979 PH/08 Applied small patch from the Sieve maintainer.
3981 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3982 folder from quota calculations, a direct delivery into this folder messed
3983 up the contents of the maildirsize file. This was because the regex was
3984 used only to exclude .Trash (or whatever) when the size of the mailbox
3985 was calculated. There was no check that a delivery was happening into an
3986 excluded directory. This bug has been fixed by ignoring all quota
3987 processing for deliveries into excluded directories.
3989 PH/10 Added the maildirfolder_create_regex option to appendfile.
3995 PH/01 The code for finding all the local interface addresses on a FreeBSD
3996 system running IPv6 was broken. This may well have applied to all BSD
3997 systems, as well as to others that have similar system calls. The broken
3998 code found IPv4 interfaces correctly, but gave incorrect values for the
3999 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4000 that it would not match correctly against @[] and not recognize the IPv6
4003 PH/02 The ipliteral router was not recognizing addresses of the form user@
4004 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4006 PH/03 Added disable_ipv6.
4008 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4009 decoded form, because it is most often used to construct To: headers
4010 lines in autoreplies, and the decoded form may well be syntactically
4011 invalid. However, $reply_address has leading white space removed, and all
4012 newlines turned into spaces so that the autoreply transport does not
4015 PH/05 If group was specified without a user on a router, and no group or user
4016 was specified on a transport, the group from the router was ignored.
4018 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4019 for visible compile-time settings that can be used to change these
4020 numbers, for those that want even more. Backwards compatibility with old
4021 spool files has been maintained. However, going back to a previous Exim
4022 release will lost any variables that are in spool files.
4024 PH/07 Two small changes when running in the test harness: increase delay when
4025 passing a TCP/IP connection to a new process, in case the original
4026 process has to generate a bounce, and remove special handling of
4027 127.0.0.2 (sic), which is no longer necessary.
4029 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4030 be the same on different OS.
4032 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4035 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4036 whether --show-vars was specified or not
4038 JJ/02 exipick: Added support for new ACL variable spool format introduced
4041 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4042 syntactically invalid From: or Reply-to: line, and a filter used this to
4043 generate an autoreply, and therefore failed to obtain an address for the
4044 autoreply, Exim could try to deliver to a non-existent relative file
4045 name, causing unrelated and misleading errors. What now happens is that
4046 it logs this as a hard delivery error, but does not attempt to create a
4049 PH/11 The exinext utility has a -C option for testing purposes, but although
4050 the given file was scanned by exinext itself; it wasn't being passed on
4051 when Exim was called.
4053 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4054 an end-of-file indication when reading a command response.
4056 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4057 compiled. In many other places in Exim, IPv6 addresses are always
4058 recognized, so I have changed this. It also means that IPv4 domain
4059 literals of the form [IPV4:n.n.n.n] are now always recognized.
4061 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4062 used if the router is not running as root, for example, when verifying at
4063 ACL time, or when using -bh. The debugging output from this situation was
4064 non-existent - all you got was a failure to exec. I have made two
4067 (a) Failures to set uid/gid, the current directory, or a process leader
4068 in a subprocess such as that created by queryprogram now generate
4069 suitable debugging output when -d is set.
4071 (b) The queryprogram router detects when it is not running as root,
4072 outputs suitable debugging information if -d is set, and then runs
4073 the subprocess without attempting to change uid/gid.
4075 PH/15 Minor change to Makefile for building test_host (undocumented testing
4078 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4079 additional section of a DNS packet that returns MX or SRV records.
4080 Instead, it always explicitly searches for A/AAAA records. This avoids
4081 major problems that occur when a DNS server includes only records of one
4082 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4083 fixed another bug: if SRV records were looked up and the corresponding
4084 address records were *not* found in the additional section, the port
4085 values from the SRV records were lost.
4087 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4088 using the correct key (the original address) when searching the retry
4089 rules in order to find which one to use for generating the retry hint.
4091 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4092 adding the default one. Similarly, if it contains a Reply-To: header, the
4093 errors_reply_to option, if set, is not used.
4095 PH/19 When calculating a retry time, Exim used to measure the "time since
4096 failure" by looking at the "first failed" field in the retry record. Now
4097 it does not use this if it is later than than the arrival time of the
4098 message. Instead it uses the arrival time. This makes for better
4099 behaviour in cases where some deliveries succeed, thus re-setting the
4100 "first failed" field. An example is a quota failure for a huge message
4101 when small messages continue to be delivered. Without this change, the
4102 "time since failure" will always be short, possible causing more frequent
4103 delivery attempts for the huge message than are intended.
4104 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4106 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4107 $1, $2, $3) because the numerical variables can be reset during some
4108 expansion items (e.g. "match"), thereby losing the authentication data.
4110 PH/21 Make -bV show the size of off_t variables so that the test suite can
4111 decide whether to run tests for quotas > 2G.
4113 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4114 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4115 filecount value is greater than 2G or if a quota value is greater than 2G
4116 on a system where the size of off_t is not greater than 4, a panic error
4119 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4120 never match. The debug and -bh output now contains an explicit error
4121 message indicating a malformed IPv4 address or mask.
4123 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4124 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4125 PH/23 above applies.
4127 PH/25 Do not write to syslog when running in the test harness. The only
4128 occasion when this arises is a failure to open the main or panic logs
4129 (for which there is an explicit test).
4131 PH/26 Added the /no_tell option to "control=freeze".
4133 PH/27 If a host name lookup failed very early in a connection, for example, if
4134 the IP address matched host_lookup and the reverse lookup yielded a name
4135 that did not have a forward lookup, an error message of the form "no IP
4136 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4137 could be logged. Now it outputs the IP address instead of "NULL".
4139 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4140 allows the sender and the authenticated sender to be set when
4141 submitting a message from within Exim. Since child_open_exim() is
4142 documented for local_scan(), the new function should be too.
4144 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4145 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4146 results in an empty string is now treated as unset.
4148 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4150 PH/31 Added sender_verify_fail logging option.
4152 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4153 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4154 tidied the source and removed it altogether.
4156 PH/33 When a queue run was abandoned because the load average was too high, a
4157 log line was always written; now it is written only if the queue_run log
4158 selector is set. In addition, the log line for abandonment now contains
4159 information about the queue run such as the pid. This is always present
4160 in "start" and "stop" lines but was omitted from the "abandon" line.
4162 PH/34 Omit spaces between a header name and the colon in the error message that
4163 is given when verify = headers_syntax fails (if there are lots of them,
4164 the message gets confusing).
4166 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4167 names, as there are now some PTR records that contain slashes. This check
4168 is only to protect against broken name servers that fall over on strange
4169 characters, so the fact that it applies to all lookups doesn't matter.
4171 PH/36 Now that the new test suite is complete, we can remove some of the
4172 special code in Exim that was needed for the old test suite. For example,
4173 sorting DNS records because real resolvers return them in an arbitrary
4174 order. The new test suite's fake resolver always returns records in the
4177 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4178 bounces) except when queue_only is set, to avoid logging races between
4179 the different processes.
4181 PH/38 Panic-die if .include specifies a non-absolute path.
4183 PH/39 A tweak to the "H" retry rule from its user.
4185 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4186 a label. They prevented compilation on older perls.
4188 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4189 a warning to be raised on newish perls.
4191 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4192 on queue. Changes to match documented behaviour of showing count of
4193 messages matching specified criteria.
4195 PH/40 Changed the default ident timeout from 30s to 5s.
4197 PH/41 Added support for the use of login_cap features, on those BSD systems
4198 that have them, for controlling the resources used by pipe deliveries.
4200 PH/42 The content-scanning code uses fopen() to create files in which to put
4201 message data. Previously it was not paying any attention to the mode of
4202 the files. Exim runs with umask(0) because the rest of the code creates
4203 files with open(), and sets the required mode explicitly. Thus, these
4204 files were ending up world-writeable. This was not a big issue, because,
4205 being within the spool directory, they were not world-accessible. I have
4206 created a function called modefopen, which takes an additional mode
4207 argument. It sets umask(777), creates the file, chmods it to the required
4208 mode, then resets the umask. All the relevant calls to fopen() in the
4209 content scanning code have been changed to use this function.
4211 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4212 to 24 hours. This avoids potential overflow problems when processing G
4213 and H retry rules. I suspect nobody ever tinkers with this value.
4215 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4217 PH/45 When the plaintext authenticator is running as a client, the server's
4218 challenges are checked to ensure they are valid base64 strings. By
4219 default, the authentication attempt is cancelled if an invalid string is
4220 received. Setting client_ignore_invalid_base64 true ignores these errors.
4221 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4222 they are received. Thus, the responses can be made to depend on the
4223 challenges. If an invalid string is ignored, an empty string is placed in
4226 PH/46 Messages that are created by the autoreply transport now contains a
4227 References: header, in accordance with RFCs 2822 and 3834.
4229 PH/47 Added authenticated_sender_force to the smtp transport.
4231 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4233 PH/49 Installed latest patch from the Sieve maintainer.
4235 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4236 was also set, the appendfile transport was unnecessarily scanning a
4237 directory of message files (e.g. for maildir delivery) to find the count
4238 of files (along with the size), even though it did not need this
4239 information. It now does the scan only if it needs to find either the
4240 size of the count of files.
4242 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4244 PH/52 Two bugs concerned with error handling when the smtp transport is
4247 (i) Exim was not creating retry information for temporary errors given
4248 for individual recipients after the DATA command when the smtp transport
4249 was used in LMTP mode. This meant that they could be retried too
4250 frequently, and not timed out correctly.
4252 (ii) Exim was setting the flag that allows error details to be returned
4253 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4254 recipients that were returned after the DATA command.
4256 PH/53 This is related to PH/52, but is more general: for any failing address,
4257 when detailed error information was permitted to be returned to the
4258 sender, but the error was temporary, then after the final timeout, only
4259 "retry timeout exceeded" was returned. Now it returns the full error as
4260 well as "retry timeout exceeded".
4262 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4263 do this, and (what is worse) MTAs that accept it.
4265 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4266 will now be deprecated.
4268 PH/56 New os.c-cygwin from the Cygwin maintainer.
4270 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4271 formats (previously only available in exim formats via -bpr, -bpru,
4272 and -bpra. Now also available in native and exiqgrep formats)
4274 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4275 with very large, slow to parse queues
4277 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4279 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4281 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4282 responses to authentication challenges, though it was showing the
4283 challenges; (ii) I've removed the CR characters from the debug output for
4286 PH/58 Allow for the insertion of a newline as well as a space when a string
4287 is turned into more than one encoded-word during RFC 2047 encoding. The
4288 Sieve code now uses this.
4290 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4291 data_4xx, lost_connection, tls_required.
4293 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4294 message was being sent as an SMTP response.
4296 PH/61 Add -l and -k options to exicyclog.
4298 PH/62 When verifying, if an address was redirected to one new address, so that
4299 verification continued, and the new address failed or deferred after
4300 having set something in $address_data, the value of $address_data was not
4301 passed back to the ACL. This was different to the case when no
4302 redirection occurred. The value is now passed back in both cases.
4304 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4305 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4306 use login_cap.h, so on its own it isn't the distinguishing feature. The
4307 new name refers directly to the setclassresources() function.
4309 PH/65 Added configuration files for NetBSD3.
4311 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4313 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4314 is preferred over IPv4.
4316 PH/68 The bounce_return_message and bounce_return_body options were not being
4317 honoured for bounces generated during the reception of non-SMTP messages.
4318 In particular, this applied to messages rejected by the ACL. This bug has
4319 been fixed. However, if bounce_return_message is true and bounce_return_
4320 body is false, the headers that are returned for a non-SMTP message
4321 include only those that have been read before the error was detected.
4322 (In the case of an ACL rejection, they have all been read.)
4324 PH/69 The HTML version of the specification is now built in a directory called
4325 spec_html instead of spec.html, because the latter looks like a path with
4326 a MIME-type, and this confuses some software.
4328 PH/70 Catch two compiler warnings in sieve.c.
4330 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4331 function verify_get_ident() calls ip_connect() to connect a socket, but
4332 if the "connect()" function timed out, ip_connect() used to close the
4333 socket. However, verify_get_ident() also closes the socket later, and in
4334 between Exim writes to the log, which may get opened at this point. When
4335 the socket was closed in ip_connect(), the log could get the same file
4336 descriptor number as the socket. This naturally causes chaos. The fix is
4337 not to close the socket in ip_connect(); the socket should be closed by
4338 the function that creates it. There was only one place in the code where
4339 this was missing, in the iplookup router, which I don't think anybody now
4340 uses, but I've fixed it anyway.
4342 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4343 well as to direct DNS lookups. Otherwise the handling of names in host
4344 lists is inconsistent and therefore confusing.
4350 PH/01 Two changes to the default runtime configuration:
4352 (1) Move the checks for relay_from_hosts and authenticated clients from
4353 after to before the (commented out) DNS black list checks.
4355 (2) Add control=submission to the relay_from_hosts and authenticated
4356 clients checks, on the grounds that messages accepted by these
4357 statements are most likely to be submissions.
4359 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4361 (1) Generate an error if the third argument for the ${prvs expansion is
4364 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4367 (3) Reset the variables that are obtained from the first argument of
4368 ${prvscheck and used in the second argument before leaving the code,
4369 because their memory is reclaimed, so using them afterwards may do
4372 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4373 one (it's much easier than Tom thought :-).
4375 (5) Because of (4), we can now allow for the use of $prvscheck_result
4376 inside the third argument.
4378 PH/03 For some reason, the default setting of PATH when running a command from
4379 a pipe transport was just "/usr/bin". I have changed it to
4382 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4383 anything to be listed in the output from -bV.
4385 PH/05 When a filter generated an autoreply, the entire To: header line was
4386 quoted in the delivery log line, like this:
4388 => >A.N.Other <ano@some.domain> <original@ddress> ...
4390 This has been changed so that it extracts the operative address. There
4391 may be more than one such address. If so, they are comma-separated, like
4394 => >ano@some.domain,ona@other.domain <original@ddress> ...
4396 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4397 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4398 not being looked up in the rDNS to get a host name, Exim was showing the
4399 IP address twice in Received: lines, even though the IP addresses were
4400 identical. For example:
4402 Received: from [1.2.3.4] (helo=[1.2.3.4])
4404 However, if the real host name was known, it was omitting the HELO data
4405 if it matched the actual IP address. This has been tidied up so that it
4406 doesn't show the same IP address twice.
4408 PH/07 When both +timestamp and +memory debugging was on, the value given by
4409 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4410 called by the debug printing, thereby overwriting the timestamp buffer.
4411 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4413 PH/08 When the original message was included in an autoreply transport, it
4414 always said "this is a copy of the message, including all the headers",
4415 even if body_only or headers_only was set. It now gives an appropriate
4418 PH/09 Applied a patch from the Sieve maintainer which:
4420 o fixes some comments
4421 o adds the (disabled) notify extension core
4422 o adds some debug output for the result of if/elsif tests
4423 o points to the current vacation draft in the documentation
4424 and documents the missing references header update
4428 o fixes a bug in processing the envelope test (when testing
4429 multiple envelope elements, the last element determined the
4432 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4433 Electronic Mail") by including:
4435 Auto-submitted: auto-generated
4437 in the messages that it generates (bounce messages and others, such as
4438 warnings). In the case of bounce messages for non-SMTP messages, there was
4439 also a typo: it was using "Auto_submitted" (underscore instead of
4440 hyphen). Since every message generated by Exim is necessarily in response
4441 to another message, thes have all been changed to:
4443 Auto-Submitted: auto-replied
4445 in accordance with these statements in the RFC:
4447 The auto-replied keyword:
4449 - SHOULD be used on messages sent in direct response to another
4450 message by an automatic process,
4452 - MUST NOT be used on manually-generated messages,
4454 - MAY be used on Delivery Status Notifications (DSNs) and Message
4455 Disposition Notifications (MDNs),
4457 - MUST NOT be used on messages generated by automatic or periodic
4458 processes, except for messages which are automatic responses to
4461 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4462 to the default Received: header definition.
4464 PH/12 Added log selector acl_warn_skipped (default on).
4466 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4467 variables because (a) they are in the wrong storage pool and (b) even if
4468 they were copied, it wouldn't work properly because of the caching.
4470 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4471 checking when decoding. Apparently there are clients that generate
4472 overlong encoded strings. Why am I not surprised?
4474 PH/15 If the first argument of "${if match_address" was not empty, but did not
4475 contain an "@" character, Exim crashed. Now it writes a panic log message
4476 and treats the condition as false.
4478 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4480 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4481 extension "envelope-auth". The code is finished and in agreement with
4482 other implementations, but there is no documentation so far and in fact,
4483 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4484 not changing the active code.
4486 Print executed "if" and "elsif" statements when debugging is used. This
4487 helps a great deal to understand what a filter does.
4489 Document more things not specified clearly in RFC3028. I had all this
4490 sorted out, when out of a sudden new issues came to my mind. Oops."
4492 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4495 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4496 canonical form (as documented). However, after a host name lookup from
4497 the IP address, check_host() was doing a simple string comparison with
4498 addresses acquired from the DNS when checking that the found name did
4499 have the original IP as one of its addresses. Since any found IPv6
4500 addresses are likely to be in abbreviated form, the comparison could
4501 fail. Luckily, there already exists a function for doing the comparison
4502 by converting both addresses to binary, so now that is used instead of
4503 the text comparison.
4505 PH/20 There was another similar case to PH/19, when a complete host name was
4506 given in a host list; looking up its IP address could give an abbreviated
4507 form, whereas the current host's name might or might not be abbreviated.
4508 The same fix has been applied.
4514 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4515 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4518 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4519 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4521 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4523 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4524 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4525 Also, for "Auto-Submitted", treat anything other than "no" as
4526 non-personal, in accordance with RFC 3834. (Previously it treated
4527 anything starting "auto-" as non-personal.)
4529 TF/01 The control=submission/name=... option had a problem with syntax
4530 errors if the name included a slash character. The /name= option
4531 now slurps the rest of the string, so it can include any characters
4532 but it must come last in the list of options (after /sender_retain
4535 PH/05 Some modifications to the interface to the fake nameserver for the new
4543 TK/01 Added the "success_on_redirect" address verification option. See
4544 NewStuff for rationale and an example.
4546 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4548 PH/02 Patch to exigrep to allow it to work on syslog lines.
4550 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4551 fread() to skip over the body file's header line, because in Cygwin the
4552 header line is locked and is inaccessible.
4554 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4555 co-exist for some time) to make it clear that it is the Exim ID that is
4556 referenced, not the Message-ID: header line.
4558 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4559 string_format() function, because snprintf() does not exist on all
4562 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4563 ${stat: expansion item.
4565 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4566 protocol synchronization error", to keep the pedants happy.
4568 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4569 well as for IRIX systems, when gcc is being used. See the host.c source
4572 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4574 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4577 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4578 transport and to the smtp transport in LMTP mode.
4580 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4582 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4583 run a filter in a subprocess. This could lead to confusion in subsequent
4584 lookups in the parent process. There should also be a search_tidyup() at
4585 the end of the subprocess.
4587 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4588 only if the host matched helo_try_verify_hosts, which caused the
4589 verification to occur when the EHLO/HELO command was issued. The ACL just
4590 tested the remembered result. Now, if a previous verification attempt has
4591 not happened, "verify = helo" does it there and then.
4593 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4595 TK/03 Fix log output including CR from clamd.
4597 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4598 exist provoked a memory error which could cause a segfault.
4600 PH/15 Installed PCRE 6.2
4602 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4604 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4605 of the problem. Specifically, suggested +O2 rather than +O1 for the
4608 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4610 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4611 record, the port was not correctly logged when the outgoing_port log
4612 selector was set (it logged the transort's default port).
4614 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4615 fallback_hosts, and "hosts" in the smtp transport.
4617 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4618 host errors such as "Connection refused".
4620 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4621 authentication with radiusclient 0.4.9:
4623 - Error returned from rc_read_config was caught wrongly
4624 - Username/password not passed on to radius server due to wrong length.
4626 The presumption is that some radiusclient API changes for 4.51/PH/17
4627 were not taken care of correctly. The code is still untested by me (my
4628 Linux distribution still has 0.3.2 of radiusclient), but it was
4629 contributed by a Radius user.
4631 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4632 expanding the "port" option of the smtp transport.
4634 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4635 while reading a MIME header. Thanks to Tom Hughes for a patch.
4637 PH/24 Include config.h inside local_scan.h so that configuration settings are
4640 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4641 that in an ACL for RCPT (for example), you can examine exactly what was
4644 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4645 commands, but it was not correctly comparing the address with the actual
4646 client host address. Thus, it would show the EHLO address in Received:
4647 header lines when this was not necessary.
4649 PH/27 Added the % operator to ${eval:}.
4651 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4652 it should be ignoring failures (because with -C, for example, it has lost
4653 privilege). It wasn't ignoring creation failures other than "already
4656 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4659 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4660 because an input line was too long, either on its own, or by virtue of
4661 too many continuations, the temporary file was not being removed, and the
4662 return code was incorrect.
4664 PH/31 Missing "BOOL" in function definition in filtertest.c.
4666 PH/32 Applied Sieve patches from the maintainer.
4668 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4670 PH/33 Added "verify = not_blind".
4672 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4673 Local/Makefile (with some defaults set). These are used in built scripts
4674 such as exicyclog, but they have never been used in the exim_install
4675 script (though there are many overriding facilities there). I have
4676 arranged that the exim_install script now takes note of these two
4679 PH/35 Installed configuration files for Dragonfly.
4681 PH/36 When a locally submitted message by a trusted user did not contain a
4682 From: header, and the sender address was obtained from -f or from an SMTP
4683 MAIL command, and the trusted user did not use -F to supply a sender
4684 name, $originator_name was incorrectly used when constructing a From:
4685 header. Furthermore, $originator_name was used for submission mode
4686 messages from external hosts without From: headers in a similar way,
4687 which is clearly wrong.
4689 PH/37 Added control=suppress_local_fixups.
4691 PH/38 When log_selector = +received_sender was set, and the addition of the
4692 sender made the log line's construction buffer exactly full, or one byte
4693 less than full, an overflow happened when the terminating "\n" was
4696 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4697 when the result of a list match is failure because a DNS lookup failed.
4699 PH/40 RM_COMMAND is now used in the building process.
4701 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4702 the "build-* directories that it finds.
4704 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4705 address in a domain literal was a prefix of an interface address.
4707 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4708 when verifying a sender address, unless rewrite_headers is false.
4710 PH/44 Wrote a long comment about why errors_to addresses are verified as
4711 recipients, not senders.
4713 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4714 the ratelimit ACL was added.
4716 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4718 PH/46 Added extra information about PostgreSQL errors to the error string.
4720 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4721 avoiding the need to install special zones in a real server. This is
4722 backwards compatible; if it can't find the fake resolver, it drops back.
4723 Thus, both old and new test suites can be run.
4725 TF/02 Added util/ratelimit.pl
4727 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4728 clock is set back in time.
4730 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4731 Candler <B.Candler@pobox.com>.
4733 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4734 OK for addresses that are the result of redirections.
4736 PH/48 A number of further additions for the benefit of the new test suite,
4737 including a fake gethostbyname() that interfaces to the fake DNS resolver
4740 TF/06 The fix for widen_domains has also been applied to qualify_single and
4741 search_parents which are the other dnslookup options that can cause
4744 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4747 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4749 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4750 service" if it encountered an MX record with an empty target hostname.
4751 The message is now "an MX or SRV record indicated no SMTP service".
4753 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4754 if the DNS of the sending site is misconfigured. This is quite a
4755 common situation. This change restores the behaviour of treating a
4756 helo verification defer as a failure.
4758 PH/51 If self=fail was set on a router, the bounce message did not include the
4759 actual error message.
4765 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4767 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4768 command itself ended in error, the underlying message about the transport
4769 filter timeout was being overwritten with the pipe command error. Now the
4770 underlying error message should be appended to the second error message.
4772 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4774 PH/02 Reduce the amount of output that "make" produces by default. Full output
4775 can still be requested.
4777 PH/03 The warning log line about a condition test deferring for a "warn" verb
4778 was being output only once per connection, rather than after each
4779 occurrence (because it was using the same function as for successful
4780 "warn" verbs). This seems wrong, so I have changed it.
4782 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4783 it should not have, which might have caused a crash in the right
4784 circumstances, but probably never did.
4786 PH/04 Installed a modified version of Tony Finch's patch to make submission
4787 mode fix the return path as well as the Sender: header line, and to
4788 add a /name= option so that you can make the user's friendly name appear
4791 TF/03 Added the control = fakedefer ACL modifier.
4793 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4794 Mark Lowes for thorough testing.
4796 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4798 TK/03 Merged latest SRS patch from Miles Wilton.
4800 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4801 with the definition in sysexits.h (which is #included earlier).
4802 Fortunately, Exim does not actually use EX_OK. The code used to try to
4803 preserve the sysexits.h value, by assuming that macro definitions were
4804 scanned for macro replacements. I have been disabused of this notion,
4805 so now the code just undefines EX_OK before #including unistd.h.
4807 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4808 in the smtp transport. When a block could not be written in a single
4809 write() function, the timeout was being re-applied to each part-write.
4810 This seems wrong - if the receiver was accepting one byte at a time it
4811 would take for ever. The timeout is now adjusted when this happens. It
4812 doesn't have to be particularly precise.
4814 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4815 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4817 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4818 without a domain if the check to <postmaster@domain> fails.
4820 SC/01 Eximstats: added -xls and the ability to specify output files
4821 (patch written by Frank Heydlauf).
4823 SC/02 Eximstats: use FileHandles for outputting results.
4825 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4827 SC/04 Eximstats: fixed display of large numbers with -nvr option
4829 SC/05 Eximstats: fixed merging of reports with empty tables.
4831 SC/06 Eximstats: added the -include_original_destination flag
4833 SC/07 Eximstats: removed tabs and trailing whitespace.
4835 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4837 TK/06 MBOX spool code: Add real "From " MBOX separator line
4838 so the .eml file is really in mbox format (even though
4839 most programs do not really care). Patch from Alex Miller.
4841 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4842 The latter is generated from $received_to and is only set if the
4843 message has one envelope recipient. SA can use these headers,
4844 obviously out-of-the-box. Patch from Alex Miller.
4846 PH/08 The ${def test on a variable was returning false if the variable's
4847 value was "0", contrary to what the specification has always said!
4848 The result should be true unless the variable is empty.
4850 PH/09 The syntax error of a character other than { following "${if
4851 def:variable_name" (after optional whitespace) was not being diagnosed.
4852 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4853 accidental colon was present, for example, could give incorrect results.
4855 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4856 result is used (not including appendfile, where other changes are about
4859 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4860 This involved changing a lot of size variables from int to off_t. It
4861 should work with maildirs and everything.
4863 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4864 spamd dying while we are connected to it.
4866 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4869 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4870 function for BDB 4.3.
4872 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4874 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4875 the sources for PCRE have been re-arranged and more files are now
4878 PH/15 The code I had for printing potentially long long variables in PH/11
4879 above was not the best (it lost precision). The length of off_t variables
4880 is now inspected at build time, and an appropriate printing format (%ld
4881 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4882 to be "long long int" or "long int". This is needed for the internal
4883 formatting function string_vformat().
4885 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4886 the configuration file to be ":syslog", then the script "guesses" where
4887 the logs files are, rather than using the compiled in default. In our
4888 case the guess is not the same as the compiled default, so the script
4889 suddenly stopped working when I started to use syslog. The patch checks
4890 to see if log_file_path is "". If so, it attempts to read it from exim
4891 with no configuration file to get the compiled in version, before it
4892 falls back to the previous guessing code."
4894 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4895 implementing BATV in an Exim configuration. See NewStuff for the gory
4898 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4899 Makefile that are specific to HP-UX.
4901 PH/18 If the "use_postmaster" option was set for a recipient callout together
4902 with the "random" option, the postmaster address was used as the MAIL
4903 FROM address for the random test, but not for the subsequent recipient
4904 test. It is now used for both.
4906 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4907 patch removes a few documentation additions to RFC 3028, because the
4908 latest draft now contains them. It adds the new en;ascii-case comparator
4909 and a new error check for 8bit text in MIME parts. Comparator and
4910 require names are now matched exactly. I enabled the subaddress
4911 extension, but it is not well tested yet (read: it works for me)."
4913 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4914 rework some of the code of TK/09 above to avoid the hardwired use of
4915 "%lld" and "long long". Replaced the call to snprintf() with a call to
4918 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4919 records point to non-existent hosts", "retry timeout exceeded", and
4920 "retry time not reached for any host after a long failure period".
4922 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4923 experimental DomainKeys support:
4925 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4926 (2) On an error such as an illegally used "control", the wrong name for
4927 the control was given.
4929 These problems did NOT occur unless DomainKeys support was compiled.
4931 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4933 PH/24 Added ${if match_ip condition.
4935 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4936 databases so that it will be absolutely obvious if a crash occurs in the
4937 DB library. This is a regular occurrence (often caused by mis-matched
4940 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4941 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4942 picked up on a user's system that detects such things. There doesn't seem
4943 to be a gcc warning option for this - only an attribute that has to be
4944 put on the function's prototype. It seems that in Fedora Core 4 they have
4945 set this on a number of new functions. No doubt there will be more in due
4948 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4949 specify a transport. However, if an address that was verified by such a
4950 router was the subject of a callout, Exim crashed because it tried to
4951 read the rcpt_include_affixes from the non-existent transport. Now it
4952 just assumes that the setting of that option is false. This bug was
4953 introduced by 4.51/PH/31.
4955 PH/28 Changed -d+all to exclude +memory, because that information is very
4956 rarely of interest, but it makes the output a lot bigger. People tend to
4957 do -d+all out of habit.
4959 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4960 code in os-type was giving problems when libc.so lives in lib64, like on
4963 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4964 aren't the modern standard, and it seems that some systems' include files
4965 don't always have them. Exim was already checking for some of the newer
4966 ones like T_AAAA, and defining it itself. I've added checks for all the
4967 record types that Exim uses.
4969 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4970 not automatically generating a new one, as it is supposed to. This
4971 prevented TLS from working. If the file did exist, but contained invalid
4972 data, a new version was generated, as expected. It was only the case of a
4973 non-existent file that was broken.
4975 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4976 with a change in libdomainkeys > 0.64.
4978 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4979 from DNS. If the selector record carries the flag, it now has
4980 precedence over the domain-wide flag.
4982 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4984 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4985 the use of a port name that isn't defined in /etc/services) occurred, the
4986 message was deferred as in a normal delivery, and thus remained on the
4987 spool, instead of being failed because of the mua_wrapper setting. This
4988 is now fixed, and I tidied up some of the mua_wrapper messages at the
4991 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4992 the messages in a hash of arrays rather than using individual hashes.
4993 This is a bit cleaner and results in dramatic memory savings, albeit
4994 at a slight CPU cost.
4996 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4997 as requested by Marc Sherman.
4999 SC/10 Eximstats: added histograms for user specified patterns as requested
5002 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5004 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5005 fopen() in the content-scanning modules that did not already have it.
5011 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5012 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5014 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5016 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5018 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5019 to test Sieve filters that use "vacation".
5021 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5022 that changes the way the GnuTLS parameters are stored in the cache file.
5023 The new format can be generated externally. For backward compatibility,
5024 if the data in the cache doesn't make sense, Exim assumes it has read an
5025 old-format file, and it generates new data and writes a new file. This
5026 means that you can't go back to an older release without removing the
5029 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5030 work if there are any delivery delays because "one_time" forces the
5031 parent to be marked "delivered", so its unseen clone is never tried
5032 again. For this reason, Exim now forbids the simultaneous setting of
5035 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5036 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5037 entry if you want to know the details. The fix, however, neglected to
5038 consider the case when local delivery batching is involved. The test for
5039 "previously delivered" was not happening when checking to see if an
5040 address could be batched with a previous (undelivered) one; under
5041 certain circumstances this could lead to multiple deliveries to the same
5044 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5045 in its include files, and this causes problems building Exim.
5047 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5048 header_syntax) but Exim was just ignoring anything given after a slash.
5049 In particular, this caused confusion with an attempt to use "verify =
5050 reverse_host_lookup/defer_ok". An error is now given when options are
5051 supplied for verify items that do not have them. (Maybe reverse_host_
5052 lookup should have a defer_ok option, but that's a different point.)
5054 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5055 defined by RFC 821) to 2048, because there were problems with some AUTH
5056 commands, and RFC 1869 says the size should be increased for extended
5057 SMTP commands that take arguments.
5059 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5062 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5063 "unknown" error; now it says that the functionality isn't in the binary.
5065 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5066 an address' error message when a string expansion fails (syntax or
5067 whatever). Otherwise the password may appear in the log. Following change
5068 PH/42 below, there is no longer a chance of it appearing in a bounce
5071 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5073 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5074 of its addresses was ever tried. (Bugzilla bug #2.)
5076 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5077 the result incorrectly in the debug output. (It correctly added a newline
5078 to what was transported.)
5080 TF/01 Added $received_time.
5082 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5083 commented out examples of how to interface to a virus scanner and to
5084 SpamAssassin. Also added commented examples of av_scanner and
5085 spamd_address settings.
5087 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5088 and controls are allowed in which ACLs. There were a couple of minor
5089 errors. Some of the entries in the conditions table (which is a table of
5090 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5091 negation of where the condition IS allowed.
5093 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5095 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5096 header file does not have a version number, so I've had to invent a new
5097 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5098 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5099 radiusclient), but it was contributed by a Radius user.
5101 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5102 files or directories, for OpenSSL.
5104 PH/19 When an Exim process that is running as root has to create an Exim log
5105 file, it does so in a subprocess that runs as exim:exim so as to get the
5106 ownership right at creation (otherwise, other Exim processes might see
5107 the file with the wrong ownership). There was no test for failure of this
5108 fork() call, which would lead to the process getting stuck as it waited
5109 for a non-existent subprocess. Forks do occasionally fail when resources
5110 run out. I reviewed all the other calls to fork(); they all seem to check
5113 PH/20 When checking for unexpected SMTP input at connect time (before writing
5114 the banner), Exim was not dealing correctly with a non-positive return
5115 from the read() function. If the client had disconnected by this time,
5116 the result was a log entry for a synchronization error with an empty
5117 string after "input=" when read() returned zero. If read() returned -1
5118 (an event I could not check), uninitialized data bytes were printed.
5119 There were reports of junk text (parts of files, etc) appearing after
5122 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5124 PH/22 Added support for macro redefinition, and (re)definition in between
5125 driver and ACL definitions.
5127 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5128 forgetting to use the resulting value; it was using the unexpanded value.
5130 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5131 hadn't been configured. The fix is from Juergen Kreileder, who
5132 understands it better than I do:
5134 "Here's what I see happening with three configured cyrus_sasl
5135 authenticators configured (plain, login, cram-md5):
5137 On startup auth_cyrus_sasl_init() gets called for each of these.
5138 This means three calls to sasl_listmech() without a specified mech_list.
5139 => SASL tests which mechs of all available mechs actually work
5140 => three warnings about OTP not working
5141 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5143 With the patch, sasl_listmech() also gets called three times. But now
5144 SASL's mech_list option is set to the server_mech specified in the the
5145 authenticator. Or in other words, the answer from sasl_listmech()
5146 gets limited to just the mech you're testing for (which is different
5148 => the return list contains just 'plain' or 'login', 'cram-md5' or
5149 nothing depending on the value of ob->server_mech.
5151 I've just tested the patch: Authentication still works fine,
5152 unavailable mechs specified in the exim configuration are still
5153 caught, and the auth.log warnings about OTP are gone."
5155 PH/25 When debugging is enabled, the contents of the command line are added
5156 to the debugging output, even when log_selector=+arguments is not
5159 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5160 answer is "GNU", and only if the return is "GNU/something" is the answer
5163 PH/27 $acl_verify_message is now set immediately after the failure of a
5164 verification in an ACL, and so is available in subsequent modifiers. In
5165 particular, the message can be preserved by coding like this:
5167 warn !verify = sender
5168 set acl_m0 = $acl_verify_message
5170 Previously, $acl_verify_message was set only while expanding "message"
5171 and "log_message" when a very denied access.
5173 PH/28 Modified OS/os.c-Linux with
5175 -#ifndef OS_LOAD_AVERAGE
5176 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5178 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5179 nomenclature these days.)
5181 PH/29 Installed patch from the Sieve maintainer that adds the options
5182 sieve_useraddress and sieve_subaddress to the redirect router.
5184 PH/30 In these circumstances:
5185 . Two addresses routed to the same list of hosts;
5186 . First host does not offer TLS;
5187 . First host accepts first address;
5188 . First host gives temporary error to second address;
5189 . Second host offers TLS and a TLS session is established;
5190 . Second host accepts second address.
5191 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5192 and peerdn, if requested) that were in fact used only for the second
5195 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5196 attention to any local part prefix or suffix that was matched by the
5197 router that accepted the address. It now behaves in the same way as it
5198 does for delivery: the affixes are removed from the local part unless
5199 rcpt_include_affixes is set on the transport.
5201 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5202 timeout during the DATA phase of an incoming message.
5204 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5205 applied a patch sanctioned by the Sieve maintainer.
5207 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5208 the uid or gid is negative. A case of a negative gid caused this to be
5209 noticed. The fix allows for either to be negative.
5211 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5212 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5215 PH/36 The change PH/12 above was broken. Fixed it.
5217 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5218 the grounds that routing the same address twice would always produce the
5219 same answer. This might have been true once, but it is certainly no
5220 longer true now. Routing a child address may depend on the previous
5221 routing that produced that child. Some complicated redirection strategies
5222 went wrong when messages had multiple recipients, and made Exim's
5223 behaviour dependent on the order in which the addresses were given.
5225 I have moved the duplicate checking until after the routing is complete.
5226 Exim scans the addresses that are assigned to local and remote
5227 transports, and removes any duplicates. This means that more work will be
5228 done, as duplicates will always all be routed, but duplicates are
5229 presumably rare, so I don't expect this is of any significance.
5231 For deliveries to pipes, files, and autoreplies, the duplicate checking
5232 still happens during the routing process, since they are not going to be
5235 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5236 It corrects a timeout issue with spamd. This is Ian's comment: "The
5237 background is that sometimes spamd either never reads data from a
5238 connection it has accepted, or it never writes response data. The exiscan
5239 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5240 blindly assumes that writes won't block so it may never time out."
5242 PH/39 Allow G after quota size as well as K and M.
5244 PH/40 The value set for $authenticated_id in an authenticator may not contain
5245 binary zeroes or newlines because the value is written to log lines and
5246 to spool files. There was no check on this. Now the value is run through
5247 the string_printing() function so that such characters are converted to
5248 printable escape sequences.
5250 PH/41 $message_linecount is a new variable that contains the total number of
5251 lines in the message. Compare $body_linecount, which is the count for the
5254 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5255 bounce and delay warning messages, except in certain special cases, which
5258 (a) An SMTP error message from a remote host;
5259 (b) A message specified in a :fail: redirection;
5260 (c) A message specified in a "fail" command in a system filter;
5261 (d) A message specified in a FAIL return from the queryprogram router;
5262 (e) A message specified by the cannot_route_message router option.
5264 In these cases only, Exim does include the error details in bounce and
5265 warning messages. There are also a few cases where bland messages such
5266 as "unrouteable address" or "local delivery error" are given.
5268 PH/43 $value is now also set for the "else" part of a ${run expansion.
5270 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5271 being worked on, but at least Exim now implements the latest version to
5274 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5275 process to complete was treated as a delivery failure, a timeout while
5276 writing the message to the pipe was logged, but erroneously treated as a
5277 successful delivery. Such timeouts include transport filter timeouts. For
5278 consistency with the overall process timeout, these timeouts are now
5279 treated as errors, giving rise to delivery failures by default. However,
5280 there is now a new Boolean option for the pipe transport called
5281 timeout_defer, which, if set TRUE, converts the failures into defers for
5282 both kinds of timeout. A transport filter timeout is now identified in
5285 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5286 systems where "make" and "gmake" are different, calling "gmake" at top
5287 level broke things. I've arranged for the value of $(MAKE) to be passed
5288 from the Makefile to this script so that it can call the same version of
5292 A note about Exim versions 4.44 and 4.50
5293 ----------------------------------------
5295 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5296 changes of various kinds. As a consequence, a big documentation update was
5297 needed. This delayed the release for rather longer than seemed good, especially
5298 in the light of a couple of (minor) security issues. Therefore, the changes
5299 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5300 release. So 4.44 and 4.50 are in effect two different branches that both start
5303 I have left the 4.50 change log unchanged; it contains all the changes since
5304 4.43. The change log for 4.44 is below; many of its items are identical to
5305 those for 4.50. This seems to be the most sensible way to preserve the
5306 historical information.
5312 1. Minor wording change to the doc/README.SIEVE file.
5314 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5315 computation of the current number of files was incorrect.
5317 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5318 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5321 4. Give more explanation in the error message when the command for a transport
5322 filter fails to execute.
5324 5. There are several places where Exim runs a non-Exim command in a
5325 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5326 was being done only for the command run by the queryprogram router. It is
5327 now done for all such subprocesses. The other cases are: ${run, transport
5328 filters, and the commands run by the lmtp and pipe transports.
5330 6. Added CONFIGURE_GROUP build-time option.
5332 7. Some older OS have a limit of 256 on the maximum number of file
5333 descriptors. Exim was using setrlimit() to set 1000 as a large value
5334 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5335 systems. I've change it so that if it can't get 1000, it tries for 256.
5337 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5338 was an oversight, and furthermore, ever since the addition of extra
5339 controls (e.g. 4.43/32), the checks on when to allow different forms of
5340 "control" were broken. There should now be diagnostics for all cases when a
5341 control that does not make sense is encountered.
5343 9. Added the /retain_sender option to "control=submission".
5345 10. $recipients is now available in the predata ACL (oversight).
5347 11. Tidy the search cache before the fork to do a delivery from a message
5348 received from the command line. Otherwise the child will trigger a lookup
5349 failure and thereby defer the delivery if it tries to use (for example) a
5350 cached ldap connection that the parent has called unbind on.
5352 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5353 of $address_data from the recipient verification was clobbered by the
5354 sender verification.
5356 13. The value of address_data from a sender verification is now available in
5357 $sender_address_data in subsequent conditions in the ACL statement.
5359 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5361 15. Added a new option "connect=<time>" to callout options, to set a different
5364 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5365 was its contents. (It was OK if the option was not defined at all.)
5367 17. A "Completed" log line is now written for messages that are removed from
5368 the spool by the -Mrm option.
5370 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5371 information about exactly what failed.
5373 19. Added -dd to debug only the daemon process.
5375 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5376 handles timeouts, both on the server side and network timeouts. Renamed the
5377 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5379 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5380 It is now set to "smtps".
5382 22. $host_address is now set to the target address during the checking of
5383 ignore_target_hosts.
5385 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5386 being passed; this would have caused $sender_host_name to have been used if
5387 matching the list had actually called for a host name (not very likely,
5388 since this list is usually IP addresses). A host name is now passed as
5391 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5392 code that specifies a non-restarting handler (typically sigaction() in
5393 modern systems) in an attempt to fix a rare and obscure crash bug.
5395 25. Narrowed the window for a race in the daemon that could cause it to ignore
5396 SIGCHLD signals. This is not a major problem, because they are used only to
5397 wake it up if nothing else does.
5399 26. A malformed maildirsize file could cause Exim to calculate negative values
5400 for the mailbox size or file count. Odd effects could occur as a result.
5401 The maildirsize information is now recalculated if the size or filecount
5404 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5405 support for a long time. Removed HAVE_SYS_VFS_H.
5407 28. Installed the latest version of exipick from John Jetmore.
5409 29. In an address list, if the pattern was not a regular expression, an empty
5410 subject address (from a bounce message) matched only if the pattern was an
5411 empty string. Non-empty patterns were not even tested. This was the wrong
5412 because it is perfectly reasonable to use an empty address as part of a
5413 database query. An empty address is now tested by patterns that are
5414 lookups. However, all the other forms of pattern expect the subject to
5415 contain a local part and a domain, and therefore, for them, an empty
5416 address still always fails if the pattern is not itself empty.
5418 30. Exim went into a mad DNS loop when attempting to do a callout where the
5419 host was specified on an smtp transport, and looking it up yielded more
5420 than one IP address.
5422 31. Re-factored the code for checking spool and log partition space into a
5423 function that finds that data and another that does the check. The former
5424 is then used to implement four new variables: $spool_space, $log_space,
5425 $spool_inodes, and $log_inodes.
5427 32. The RFC2047 encoding function was originally intended for short strings
5428 such as real names; it was not keeping to the 75-character limit for
5429 encoded words that the RFC imposes. It now respects the limit, and
5430 generates multiple encoded words if necessary. To be on the safe side, I
5431 have increased the buffer size for the ${rfc2047: expansion operator from
5434 33. It is now permitted to omit both strings after an "if" condition; if the
5435 condition is true, the result is "true". As before, when the second string
5436 is omitted, a false condition yields an empty string. This makes it less
5437 cumbersome to write custom ACL and router conditions.
5439 34. Failure to deliver a bounce message always caused it to be frozen, even if
5440 there was an errors_to setting on the router. The errors_to setting is now
5443 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5444 canonical form (fully expanded) before being placed in
5445 $sender_host_address.
5447 36. The table in the code that translates DNS record types into text (T_A to
5448 "A" for instance) was missing entries for NS and CNAME. It is just possible
5449 that this could have caused confusion if both these types were looked up
5450 for the same domain, because the text type is used as part of Exim's
5451 per-process caching. But the chance of anyone hitting this buglet seems
5454 37. The dnsdb lookup has been extended in a number of ways.
5456 (1) There is a new type, "zns", which walks up the domain tree until it
5457 finds some nameserver records. It should be used with care.
5459 (2) There is a new type, "mxh", which is like "mx" except that it returns
5460 just the host names, not the priorities.
5462 (3) It is now possible to give a list of domains (or IP addresses) to be
5463 looked up. The behaviour when one of the lookups defers can be
5464 controlled by a keyword.
5466 (4) It is now possible to specify the separator character for use when
5467 multiple records are returned.
5469 38. The dnslists ACL condition has been extended: it is now possible to supply
5470 a list of IP addresses and/or domains to be looked up in a particular DNS
5473 39. Added log_selector=+queue_time_overall.
5475 40. When running the queue in the test harness, wait just a tad after forking a
5476 delivery process, to get repeatability of debugging output.
5478 41. Include certificate and key file names in error message when GnuTLS fails
5479 to set them up, because the GnuTLS error message doesn't include the name
5480 of the failing file when there is a problem reading it.
5482 42. Allow both -bf and -bF in the same test run.
5484 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5486 44. The "Exiscan patch" is now merged into the mainline Exim source.
5488 45. Sometimes the final signoff response after QUIT could fail to get
5489 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5490 < 0 before doing a fflush(). This bug looks as though it goes back to the
5491 introduction of TLS in release 3.20, but "sometimes" must have been rare
5492 because the tests only now provoked it.
5494 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5495 (this can affect the format of dates).
5497 47. exim_tidydb, when checking for the continued existence of a message for
5498 which it has found a message-specific retry record, was not finding
5499 messages that were in split spool directories. Consequently, it was
5500 deleting retry records that should have stayed in existence.
5502 48. Steve fixed some bugs in eximstats.
5504 49. The SPA authentication driver was not abandoning authentication and moving
5505 on to the next authenticator when an expansion was forced to fail,
5506 contradicting the general specification for all authenticators. Instead it
5507 was generating a temporary error. It now behaves as specified.
5509 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5510 (the order specifies the preference for clients). The order is now AES256,
5511 AES128, 3DES, ARCFOUR128.
5513 51. Small patch to Sieve code - explicitly set From: when generating an
5516 52. Exim crashed if a remote delivery caused a very long error message to be
5517 recorded - for instance if somebody sent an entire SpamAssassin report back
5518 as a large number of 550 error lines. This bug was coincidentally fixed by
5519 increasing the size of one of Exim's internal buffers (big_buffer) that
5520 happened as part of the Exiscan merge. However, to be on the safe side, I
5521 have made the code more robust (and fixed the comments that describe what
5524 53. Now that there can be additional text after "Completed" in log lines (if
5525 the queue_time_overall log selector is set), a one-byte patch to exigrep
5526 was needed to allow it to recognize "Completed" as not the last thing in
5529 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5530 patch that reportedly fixes this has been added. I am not expert enough to
5531 create a test for it. This is what the patch creator wrote:
5533 "I found a little strange behaviour of ldap code when working with
5534 Windows 2003 AD Domain, where users was placed in more than one
5535 Organization Units. When I tried to give exim partial DN, the exit code
5536 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5537 But simultaneously result of request was absolutely normal ldap result,
5538 so I produce this patch..."
5540 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5541 REFERENCE, so I have modified the code to exclude the patch when that macro
5544 55. Some experimental protocols are using DNS PTR records for new purposes. The
5545 keys for these records are domain names, not reversed IP addresses. The
5546 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5547 leaves it alone. Component reversal etc. now happens only for IP addresses.
5550 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5552 57. Double the size of the debug message buffer (to 2048) so that more of very
5553 long debug lines gets shown.
5555 58. The exicyclog utility now does better if the number of log files to keep
5556 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5558 59. Two changes related to the smtp_active_hostname option:
5560 (1) $smtp_active_hostname is now available as a variable.
5561 (2) The default for smtp_banner uses $smtp_active_hostname instead
5562 of $primary_hostname.
5564 60. The host_aton() function is supposed to be passed a string that is known
5565 to be a valid IP address. However, in the case of IPv6 addresses, it was
5566 not checking this. This is a hostage to fortune. Exim now panics and dies
5567 if the condition is not met. A case was found where this could be provoked
5568 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5569 components; fortuitously, this particular loophole had already been fixed
5570 by change 4.50/55 above.
5572 If there are any other similar loopholes, the new check in host_aton()
5573 itself should stop them being exploited. The report I received stated that
5574 data on the command line could provoke the exploit when Exim was running as
5575 exim, but did not say which command line option was involved. All I could
5576 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5577 running as the user.
5580 61. There was a buffer overflow vulnerability in the SPA authentication code
5581 (which came originally from the Samba project). I have added a test to the
5582 spa_base64_to_bits() function which I hope fixes it.
5585 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5586 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5588 63. The daemon start-up calls getloadavg() while still root for those OS that
5589 need the first call to be done as root, but it missed one case: when
5590 deliver_queue_load_max is set with deliver_drop_privilege. This is
5591 necessary for the benefit of the queue runner, because there is no re-exec
5592 when deliver_drop_privilege is set.
5594 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5595 This has been fixed.
5597 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5598 was in use, was not putting the data itself into the right store pool;
5599 consequently, it could be overwritten for a subsequent message in the same
5600 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5603 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5605 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5606 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5607 the function were treating the return as a boolean value, which happened to
5608 work because 0=false and not-0=true, but is not correct code.
5610 68. The host_aton() function was not handling scoped IPv6 addresses (those
5611 with, for example, "%eth0" on the end) correctly.
5613 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5614 negated items (that is, ~something) in unsigned ints. Some compilers
5615 apparently mutter when there is no cast.
5617 70. If an address verification called from an ACL failed, and did not produce a
5618 user-specific message (i.e. there was only a "system" message), nothing was
5619 put in $acl_verify_message. In this situation, it now puts the system
5622 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5623 session; change 4.31/43 added the unwanted input to the log line - except
5624 that it did not do this in the start of session case. It now does.
5626 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5627 This is wrong and can cause the other end to generate a synchronization
5628 error if it is another Exim or anything else that does the synchronization
5629 check. A QUIT command is no longer sent after a timeout.
5631 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5632 during host lookups.
5634 74. The defer_ok option of callout verification was not working if it was used
5635 when verifying addresses in header lines, that is, for this case:
5637 verify = header_sender/callout=defer_ok
5639 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5640 those file descriptors could be used for SMTP connections. If anything
5641 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5642 could be sent to the SMTP client, causing chaos. The daemon now opens
5643 stdin, stdout, and stderr to /dev/null when it puts itself into the
5646 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5647 log by default. The user can override this with suitable Perl magic.
5649 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5650 the log message when discard triggers, was not working for the DATA ACL or
5651 for the non-SMTP ACL.
5653 78. Error message wording change in sieve.c.
5655 79. If smtp_accept_max_per_host was set, the number of connections could be
5656 restricted to fewer than expected, because the daemon was trying to set up
5657 a new connection before checking whether the processes handling previous
5658 connections had finished. The check for completed processes is now done
5659 earlier. On busy systems, this bug wouldn't be noticed because something
5660 else would have woken the daemon, and it would have reaped the completed
5663 80. If a message was submitted locally by a user whose login name contained one
5664 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5665 It caused a spool format error. I have fixed the spool reading code. A
5666 related problem was that the "from" clause in the Received: line became
5667 illegal because of the space(s). It is now covered by ${quote_local_part.
5669 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5672 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5674 83. Patch from Sieve maintainer to fix unterminated string problem in
5675 "vacation" handling.
5677 84. Some minor changes to the Linux configuration files to help with other
5678 OS variants using glibc.
5680 85. One more patch for Sieve to update vacation handling to latest spec.
5683 ----------------------------------------------------
5684 See the note above about the 4.44 and 4.50 releases.
5685 ----------------------------------------------------
5691 1. Change 4.43/35 introduced a bug that caused file counts to be
5692 incorrectly computed when quota_filecount was set in an appendfile
5695 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5696 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5699 3. Give more explanation in the error message when the command for a transport
5700 filter fails to execute.
5702 4. There are several places where Exim runs a non-Exim command in a
5703 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5704 was being done only for the command run by the queryprogram router. It is
5705 now done for all such subprocesses. The other cases are: ${run, transport
5706 filters, and the commands run by the lmtp and pipe transports.
5708 5. Some older OS have a limit of 256 on the maximum number of file
5709 descriptors. Exim was using setrlimit() to set 1000 as a large value
5710 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5711 systems. I've change it so that if it can't get 1000, it tries for 256.
5713 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5714 was an oversight, and furthermore, ever since the addition of extra
5715 controls (e.g. 4.43/32), the checks on when to allow different forms of
5716 "control" were broken. There should now be diagnostics for all cases when a
5717 control that does not make sense is encountered.
5719 7. $recipients is now available in the predata ACL (oversight).
5721 8. Tidy the search cache before the fork to do a delivery from a message
5722 received from the command line. Otherwise the child will trigger a lookup
5723 failure and thereby defer the delivery if it tries to use (for example) a
5724 cached ldap connection that the parent has called unbind on.
5726 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5727 of $address_data from the recipient verification was clobbered by the
5728 sender verification.
5730 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5731 was its contents. (It was OK if the option was not defined at all.)
5733 11. A "Completed" log line is now written for messages that are removed from
5734 the spool by the -Mrm option.
5736 12. $host_address is now set to the target address during the checking of
5737 ignore_target_hosts.
5739 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5740 being passed; this would have caused $sender_host_name to have been used if
5741 matching the list had actually called for a host name (not very likely,
5742 since this list is usually IP addresses). A host name is now passed as
5745 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5746 code that specifies a non-restarting handler (typically sigaction() in
5747 modern systems) in an attempt to fix a rare and obscure crash bug.
5749 15. Narrowed the window for a race in the daemon that could cause it to ignore
5750 SIGCHLD signals. This is not a major problem, because they are used only to
5751 wake it up if nothing else does.
5753 16. A malformed maildirsize file could cause Exim to calculate negative values
5754 for the mailbox size or file count. Odd effects could occur as a result.
5755 The maildirsize information is now recalculated if the size or filecount
5758 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5759 support for a long time. Removed HAVE_SYS_VFS_H.
5761 18. Updated exipick to current release from John Jetmore.
5763 19. Allow an empty sender to be matched against a lookup in an address list.
5764 Previously the only cases considered were a regular expression, or an
5767 20. Exim went into a mad DNS lookup loop when doing a callout where the
5768 host was specified on the transport, if the DNS lookup yielded more than
5771 21. The RFC2047 encoding function was originally intended for short strings
5772 such as real names; it was not keeping to the 75-character limit for
5773 encoded words that the RFC imposes. It now respects the limit, and
5774 generates multiple encoded words if necessary. To be on the safe side, I
5775 have increased the buffer size for the ${rfc2047: expansion operator from
5778 22. Failure to deliver a bounce message always caused it to be frozen, even if
5779 there was an errors_to setting on the router. The errors_to setting is now
5782 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5783 canonical form (fully expanded) before being placed in
5784 $sender_host_address.
5786 24. Updated eximstats to version 1.33
5788 25. Include certificate and key file names in error message when GnuTLS fails
5789 to set them up, because the GnuTLS error message doesn't include the name
5790 of the failing file when there is a problem reading it.
5792 26. Expand error message when OpenSSL has problems setting up cert/key files.
5795 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5796 (this can affect the format of dates).
5798 28. exim_tidydb, when checking for the continued existence of a message for
5799 which it has found a message-specific retry record, was not finding
5800 messages that were in split spool directories. Consequently, it was
5801 deleting retry records that should have stayed in existence.
5803 29. eximstats updated to version 1.35
5804 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5805 1.35 - bugfix such that pie charts by volume are generated correctly
5807 30. The SPA authentication driver was not abandoning authentication and moving
5808 on to the next authenticator when an expansion was forced to fail,
5809 contradicting the general specification for all authenticators. Instead it
5810 was generating a temporary error. It now behaves as specified.
5812 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5813 (the order specifies the preference for clients). The order is now AES256,
5814 AES128, 3DES, ARCFOUR128.
5816 31. Small patch to Sieve code - explicitly set From: when generating an
5819 32. Exim crashed if a remote delivery caused a very long error message to be
5820 recorded - for instance if somebody sent an entire SpamAssassin report back
5821 as a large number of 550 error lines. This bug was coincidentally fixed by
5822 increasing the size of one of Exim's internal buffers (big_buffer) that
5823 happened as part of the Exiscan merge. However, to be on the safe side, I
5824 have made the code more robust (and fixed the comments that describe what
5827 33. Some experimental protocols are using DNS PTR records for new purposes. The
5828 keys for these records are domain names, not reversed IP addresses. The
5829 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5830 leaves it alone. Component reversal etc. now happens only for IP addresses.
5833 34. The host_aton() function is supposed to be passed a string that is known
5834 to be a valid IP address. However, in the case of IPv6 addresses, it was
5835 not checking this. This is a hostage to fortune. Exim now panics and dies
5836 if the condition is not met. A case was found where this could be provoked
5837 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5838 components; fortuitously, this particular loophole had already been fixed
5839 by change 4.50/55 or 4.44/33 above.
5841 If there are any other similar loopholes, the new check in host_aton()
5842 itself should stop them being exploited. The report I received stated that
5843 data on the command line could provoke the exploit when Exim was running as
5844 exim, but did not say which command line option was involved. All I could
5845 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5846 running as the user.
5849 35. There was a buffer overflow vulnerability in the SPA authentication code
5850 (which came originally from the Samba project). I have added a test to the
5851 spa_base64_to_bits() function which I hope fixes it.
5854 36. The daemon start-up calls getloadavg() while still root for those OS that
5855 need the first call to be done as root, but it missed one case: when
5856 deliver_queue_load_max is set with deliver_drop_privilege. This is
5857 necessary for the benefit of the queue runner, because there is no re-exec
5858 when deliver_drop_privilege is set.
5860 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5861 was in use, was not putting the data itself into the right store pool;
5862 consequently, it could be overwritten for a subsequent message in the same
5863 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5866 38. Sometimes the final signoff response after QUIT could fail to get
5867 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5868 < 0 before doing a fflush(). This bug looks as though it goes back to the
5869 introduction of TLS in release 3.20, but "sometimes" must have been rare
5870 because the tests only now provoked it.
5876 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5877 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5878 Now it returns an integer. A number of calls were still expecting a T/F
5879 return. Fortuitously, in all cases, the tests worked in OK situations,
5880 which is the norm. However, things would have gone wrong on any write
5881 failures on the smtp file descriptor. This function is used when sending
5882 messages over SMTP and also when doing verify callouts.
5884 2. When Exim is called to do synchronous delivery of a locally submitted
5885 message (the -odf or -odi options), it no longer closes stderr before doing
5888 3. Implemented the mua_wrapper option.
5890 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5892 5. Implemented the functions header_remove(), header_testname(),
5893 header_add_at_position(), and receive_remove_recipient(), and exported them
5896 6. If an ACL "warn" statement specified the addition of headers, Exim already
5897 inserted X-ACL-Warn: at the start if there was no header name. However, it
5898 was not making this test for the second and subsequent header lines if
5899 there were newlines in the string. This meant that an invalid header could
5900 be inserted if Exim was badly configured.
5902 7. Allow an ACL "warn" statement to add header lines at the start or after all
5903 the Received: headers, as well as at the end.
5905 8. Added the rcpt_4xx retry error code.
5907 9. Added postmaster_mailfrom=xxx to callout verification option.
5909 10. Added mailfrom=xxxx to the callout verification option, for verify=
5912 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5913 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5915 12. Inserted some casts to stop certain compilers warning when using pointer
5916 differences as field lengths or precisions in printf-type calls (mostly
5917 affecting debugging statements).
5919 13. Added optional readline() support for -be (dynamically loaded).
5921 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5922 same clock tick as a message's arrival, so that its received time was the
5923 same as the "first fail" time on the retry record, and that message
5924 remained on the queue past the ultimate address timeout, every queue runner
5925 would try a delivery (because it was past the ultimate address timeout) but
5926 after another failure, the ultimate address timeout, which should have then
5927 bounced the address, did not kick in. This was a "< instead of <=" error;
5928 in most cases the first failure would have been in the next clock tick
5929 after the received time, and all would be well.
5931 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5932 being recognized when the domain list was tested by the match_domain
5933 condition in an expansion string.
5935 16. Added the ${str2b64: operator.
5937 17. Exim was always calling setrlimit() to set a large limit for the number of
5938 processes, without checking whether the existing limit was already
5939 adequate. (It did check for the limit on file descriptors.) Furthermore,
5940 errors from getrlimit() and setrlimit() were being ignored. Now they are
5941 logged to the main and panic logs, but Exim does carry on, to try to do its
5942 job under whatever limits there are.
5944 18. Imported PCRE 5.0.
5946 19. Trivial typo in log message " temporarily refused connection" (the leading
5949 20. If the log selector return_path_on_delivery was set and an address was
5950 redirected to /dev/null, the delivery process crashed because it assumed
5951 that a return path would always be set for a "successful" delivery. In this
5952 case, the whole delivery is bypassed as an optimization, and therefore no
5955 21. Internal re-arrangement: the function for sending a challenge and reading
5956 a response while authentication was assuming a zero-terminated challenge
5957 string. It's now changed to take a pointer and a length, to allow for
5958 binary data in such strings.
5960 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5962 23. Exim was not respecting finduser_retries when seeking the login of the
5963 uid under which it was called; it was always trying 10 times. (The default
5964 setting of finduser_retries is zero.) Also, it was sleeping after the final
5965 failure, which is pointless.
5967 24. Implemented tls_on_connect_ports.
5969 25. Implemented acl_smtp_predata.
5971 26. If the domain in control=submission is set empty, Exim assumes that the
5972 authenticated id is a complete email address when it generates From: or
5973 Sender: header lines.
5975 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5976 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5977 chown and chgrp in /bin and hostname in /usr/bin.
5979 28. Exim was keeping the "process log" file open after each use, just as it
5980 does for the main log. This opens the possibility of it remaining open for
5981 long periods when the USR1 signal hits a daemon. Occasional processlog
5982 errors were reported, that could have been caused by this. Anyway, it seems
5983 much more sensible not to leave this file open at all, so that is what now
5986 29. The long-running daemon process does not normally write to the log once it
5987 has entered its main loop, and it closes the log before doing so. This is
5988 so that log files can straightforwardly be renamed and moved. However,
5989 there are a couple of unusual error situations where the daemon does write
5990 log entries, and I had neglected to close the log afterwards.
5992 30. The text of an SMTP error response that was received during a remote
5993 delivery was being truncated at 512 bytes. This is too short for some of
5994 the long messages that one sometimes sees. I've increased the limit to
5997 31. It is now possible to make retry rules that apply only when a message has a
5998 specific sender, in particular, an empty sender.
6000 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6001 it possible to be selective about when SMTP synchronization is enforced.
6003 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6005 32. Implemented hosts_connection_nolog.
6007 33. Added an ACL for QUIT.
6009 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6012 35. Added mailbox_size and mailbox_filecount to appendfile.
6014 36. Added control = no_multiline_responses to ACLs.
6016 37. There was a bug in the logic of the code that waits for the clock to tick
6017 in the case where the clock went backwards by a substantial amount such
6018 that the microsecond fraction of "now" was more than the microsecond
6019 fraction of "then" (but the whole seconds number was less).
6021 38. Added support for the libradius Radius client library this is found on
6022 FreeBSD (previously only the radiusclient library was supported).
6028 1. When certain lookups returned multiple values in the form name=value, the
6029 quoting of the values was not always being done properly. Specifically:
6030 (a) If the value started with a double quote, but contained no whitespace,
6032 (b) If the value contained whitespace other than a space character (i.e.
6033 tabs or newlines or carriage returns) it was not quoted.
6034 This fix has been applied to the mysql and pgsql lookups by writing a
6035 separate quoting function and calling it from the lookup code. The fix
6036 should probably also be applied to nisplus, ibase and oracle lookups, but
6037 since I cannot test any of those, I have not disturbed their existing code.
6039 2. A hit in the callout cache for a specific address caused a log line with no
6040 reason for rejecting RCPT. Now it says "Previous (cached) callout
6041 verification failure".
6043 3. There was an off-by-one bug in the queryprogram router. An over-long
6044 return line was truncated at 256 instead of 255 characters, thereby
6045 overflowing its buffer with the terminating zero. As well as fixing this, I
6046 have increased the buffer size to 1024 (and made a note to document this).
6048 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6049 when Exim is waiting for an SMTP response from a remote server, Exim
6050 restarts its select() call on the socket, thereby resetting its timeout.
6051 This is not a problem when such interrupts are rare. Somebody set up a cron
6052 job to run exiwhat every 2 minutes, which is less than the normal select()
6053 timeout (5 or 10 minutes). This meant that the select() timeout never
6054 kicked in because it was always reset. I have fixed this by comparing the
6055 time when an interrupt arrives with the time at the start of the first call
6056 to select(). If more time than the timeout has elapsed, the interrupt is
6057 treated as a timeout.
6059 5. Some internal re-factoring in preparation for the addition of Sieve
6060 extensions (by MH). In particular, the "personal" test is moved to a
6061 separate function, and given an option for scanning Cc: and Bcc: (which is
6062 not set for Exim filters).
6064 6. When Exim created an email address using the login of the caller as the
6065 local part (e.g. when creating a From: or Sender: header line), it was not
6066 quoting the local part when it contained special characters such as @.
6068 7. Installed new OpenBSD configuration files.
6070 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6071 try to make them clearer.
6073 9. Callout options, other than the timeout value, were being ignored when
6074 verifying sender addresses in header lines. For example, when using
6076 verify = header_sender/callout=no_cache
6078 the cache was (incorrectly) being used.
6080 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6081 only the Cygwin environment.
6083 11. When return_path_on_delivery was set as a log selector, if different remote
6084 addresses in the same message used different return paths and parallel
6085 remote delivery occurred, the wrong values would sometimes be logged.
6086 (Whenever a remote delivery process finished, the return path value from
6087 the most recently started remote delivery process was logged.)
6089 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6090 lines when AUTH and/or TLS are in use. This is the "received protocol"
6091 field. Exim used to use "asmtp" for authenticated SMTP, without any
6092 indication (in the protocol name) for TLS use. Now it follows the RFC and
6093 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6094 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6095 names appear in log lines as well as in Received: header lines.
6097 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6098 extensions, and comparison tests, and to fix some bugs.
6100 14. Changes to the "personal" filter test:
6102 (1) The test was buggy in that it was just doing the equivalent of
6103 "contains" tests on header lines. For example, if a user's address was
6104 anne@some.where, the "personal" test would incorrectly be true for
6106 To: susanne@some.where
6108 This test is now done by extracting each address from the header in turn,
6109 and checking the entire address. Other tests that are part of "personal"
6110 are now done using regular expressions (for example, to check local parts
6111 of addresses in From: header lines).
6113 (2) The list of non-personal local parts in From: addresses has been
6114 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6115 taken from the Sieve specification recommendations.
6117 (3) If the message contains any header line starting with "List-" it is
6118 treated as non-personal.
6120 (4) The test for "circular" in the Subject: header line has been removed
6121 because it now seems ill-conceived.
6123 15. Minor typos in src/EDITME comments corrected.
6125 16. Installed latest exipick from John Jetmore.
6127 17. If headers_add on a router specified a text string that was too long for
6128 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6129 of string_sprintf() is now avoided.
6131 18. $message_body_size was not set (it was always zero) when running the DATA
6132 ACL and the local_scan() function.
6134 19. For the "mail" command in an Exim filter, no default was being set for
6135 the once_repeat time, causing a random time value to be used if "once" was
6136 specified. (If the value happened to be <= 0, no repeat happened.) The
6137 default is now 0s, meaning "never repeat". The "vacation" command was OK
6138 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6139 (I found it when inspecting the code).
6141 20. There is now an overall timeout for performing a callout verification. It
6142 defaults to 4 times the callout timeout, which applies to individual SMTP
6143 commands during the callout. The overall timeout applies when there is more
6144 than one host that can be tried. The timeout is checked before trying the
6145 next host. This prevents very long delays if there are a large number of
6146 hosts and all are timing out (e.g. when the network connections are timing
6147 out). The value of the overall timeout can be changed by specifying an
6148 additional sub-option for "callout", called "maxwait". For example:
6150 verify = sender/callout=5s,maxwait=20s
6152 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6153 to the end before writing, but this should make it even safer).
6155 22. Exim was forgetting that it had advertised PIPELINING for the second and
6156 subsequent messages on an SMTP connection. It was also not resetting its
6157 memory on STARTTLS and an internal HELO.
6159 23. When Exim logs an SMTP synchronization error within a session, it now
6160 records whether PIPELINING has been advertised or not.
6162 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6163 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6164 rather than long int.
6166 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6168 26. Added the never_mail option to autoreply.
6174 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6175 crash if the getsockname() call failed; this can happen if a connection is
6176 closed very soon after it is established. The problem was simply in the
6177 order in which certain operations were done, causing Exim to try to write
6178 to the SMTP stream before it had set up the file descriptor. The bug has
6179 been fixed by making things happen in the correct order.
6185 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6186 before the connection was closed, thus losing the rejection response.
6188 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6189 some early Solaris releases, but causes trouble in current releases where
6190 socklen_t is defined.
6192 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6195 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6198 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6199 that releasing the top of it at the end releases what was used for sub-
6200 expansions (unless the block got too big). However, discard this block if
6201 the first thing is a variable or header, so that we can use its block when
6202 it is dynamic (useful for very large $message_headers, for example).
6204 6. Lookups now cache *every* query, not just the most recent. A new, separate
6205 store pool is used for this. It can be recovered when all lookup caches are
6206 flushed. Lookups now release memory at the end of their result strings.
6207 This has involved some general refactoring of the lookup sources.
6209 7. Some code has been added to the store_xxx() functions to reduce the amount
6210 of flapping under certain conditions.
6212 8. log_incoming_interface used to affect only the <= reception log lines. Now
6213 it causes the local interface and port to be added to several more SMTP log
6214 lines, for example "SMTP connection from", and rejection lines.
6216 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6218 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6220 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6221 could be overwritten at the end of the current message (or the start of a
6222 new message if it was set in a HELO ACL). The value is now preserved for
6223 the duration of the SMTP connection.
6225 12. If a transport had a headers_rewrite setting, and a matching header line
6226 contained an unqualified address, that address was qualified, even if it
6227 did not match any rewriting rules. The underlying bug was that the values
6228 of the flags that permit the existence of unqualified sender and recipient
6229 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6230 non-local messages, and by -bnq for local messages) were not being
6231 preserved with the message after it was received.
6233 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6234 "next input=" as part of the text comprising the host identity instead of
6235 the correct text. The code was using the same buffer for two different
6236 strings. However, depending on which order the printing function evaluated
6237 its arguments, the bug did not always show up. Under Linux, for example, my
6238 test suite worked just fine.
6240 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6241 doesn't work with some older versions of Perl. It has been changed to "my",
6242 which in any case is probably the better facility to use.
6244 15. A really picky compiler found some instances of statements for creating
6245 error messages that either had too many or two few arguments for the format
6248 16. The size of the buffer for calls to the DNS resolver has been increased
6249 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6250 for addresses that have a lot of PTR records. This alleviates a problem; it
6251 does not fully solve it.
6253 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6254 buffer now truncates the list and logs the incident, which is the same
6255 action as happens when Exim is looking up a host name and its aliases.
6256 Previously in this situation something unpredictable would happen;
6257 sometimes it was "internal error: store_reset failed".
6259 18. If a server dropped the connection unexpectedly when an Exim client was
6260 using GnuTLS and trying to read a response, the client delivery process
6261 crashed while trying to generate an error log message.
6263 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6264 string, for example:
6266 warn message = H1: something\nH2: something
6268 the text was added as a single header line from Exim's point of view
6269 though it ended up OK in the delivered message. However, searching for the
6270 second and subsequent header lines using $h_h2: did not work. This has been
6271 fixed. Similarly, if a system filter added multiple headers in this way,
6272 the routers could not see them.
6274 20. Expanded the error message when iplsearch is called with an invalid key to
6275 suggest using net-iplsearch in a host list.
6277 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6278 ACLs are no longer actually imposed (and a message to that effect is
6281 22. If a "gecos" field in a passwd entry contained escaped characters, in
6282 particular, if it contained a \" sequence, Exim got it wrong when building
6283 a From: or a Sender: header from that name. A second bug also caused
6284 incorrect handling when an unquoted " was present following a character
6285 that needed quoting.
6287 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6288 was not being matched caselessly.
6290 24. Arranged for all hyphens in the exim.8 source to be escaped with
6293 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6294 properly. Recipient callout cache records were still being keyed to include
6295 the sender, even when use_sender was set false. This led to far more
6296 callouts that were necessary. The sender is no longer included in the key
6297 when use_sender is false.
6299 26. Added "control = submission" modifier to ACLs.
6301 27. Added the ${base62d: operator to decode base 62 numbers.
6303 28. dnsdb lookups can now access SRV records.
6305 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6306 the configuration file.
6308 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6309 (-v) mode. This makes the output for a verbose queue run more intelligible.
6311 31. Added a use_postmaster feature to recipient callouts.
6313 32. Added the $body_zerocount variable, containing the number of binary zero
6314 bytes in the message body.
6316 33. The time of last modification of the "new" subdirectory is now used as the
6317 "mailbox time last read" when there is a quota error for a maildir
6320 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6322 35. Added +ignore_unknown as a special item in host lists.
6324 36. Code for decoding IPv6 addresses in host lists is now included, even if
6325 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6326 address was recognized as an IP address, but was then not correctly decoded
6327 into binary, causing unexpected and incorrect effects when compared with
6334 1. Very minor rewording of debugging text in manualroute to say "list of
6335 hosts" instead of "hostlist".
6337 2. If verify=header_syntax was set, and a header line with an unqualified
6338 address (no domain) and a large number of spaces between the end of the
6339 name and the colon was received, the reception process suffered a buffer
6340 overflow, and (when I tested it) crashed. This was caused by some obsolete
6341 code that should have been removed. The fix is to remove it!
6343 3. When running in the test harness, delay a bit after writing a bounce
6344 message to get a bit more predictability in the log output.
6346 4. Added a call to search_tidyup() just before forking a reception process. In
6347 theory, someone could use a lookup in the expansion of smtp_accept_max_
6348 per_host which, without the tidyup, could leave open a database connection.
6350 5. Added the variables $recipient_data and $sender_data which get set from a
6351 lookup success in an ACL "recipients" or "senders" condition, or a router
6352 "senders" option, similar to $domain_data and $local_part_data.
6354 6. Moved the writing of debug_print from before to after the "senders" test
6357 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6358 problems for message scanning, either using a data ACL, or using
6359 local_scan() because the Received: header was not generated till after they
6360 were called (in order to set the time as the time of reception completion).
6361 I have revised the way this works. The header is now generated after the
6362 body is received, but before the ACL or local_scan() are called. After they
6363 are run, the timestamp in the header is updated.
6369 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6370 before starting a queue runner without re-exec. This happened only when
6371 deliver_drop_privilege was set or when the Exim user was set to root. The
6372 effect of the bug was that timeouts during subsequent deliveries caused
6373 crashes instead of being properly handled. The handler is now left at its
6374 default (and expected) setting.
6376 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6377 message, again when deliver_drop_privilege is set or Exim is run as root.
6378 The bug described in (1) was not present in this case, but the tidying up
6379 of the other signals was missing. I have made the two cases consistent.
6381 3. The ignore_target_hosts setting on a manualroute router was being ignored
6382 for hosts that were looked up using the /MX notation.
6384 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6387 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6388 operated on the sender address. After changing the $sender_address to <>
6389 for the sender address verify, Exim was re-instated it as the original
6390 (before rewriting) address, but remembering that it had rewritten it, so it
6391 wasn't rewriting it again. This bug also had the effect of breaking the
6392 sender address verification caching when the sender address was rewritten.
6394 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6395 This has been changed so that if the ip literal address matches
6396 ignore_target_hosts, the router declines.
6398 7. Added expansion conditions match_domain, match_address, and match_local_
6399 part (NOT match_host).
6401 8. The placeholder for the Received: header didn't have a length field set.
6403 9. Added code to Exim itself and to exim_lock to test for a specific race
6404 condition that could lead to file corruption when using MBX delivery. The
6405 issue is with the lockfile that is created in /tmp. If this file is removed
6406 after a process has opened it but before that process has acquired a lock,
6407 there is the potential for a second process to recreate the file and also
6408 acquire a lock. This could lead to two Exim processes writing to the file
6409 at the same time. The added code performs the same test as UW imapd; it
6410 checks after acquiring the lock that its file descriptor still refers to
6411 the same named file.
6413 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6414 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6417 11. Added the smtp_active_hostname option. If used, this will typically be made
6418 to depend on the incoming interface address. Because $interface_address is
6419 not set up until the daemon has forked a reception process, error responses
6420 that can happen earlier (such as "too many connections") no longer contain
6423 12. If an expansion in a condition on a "warn" statement fails because a lookup
6424 defers, the "warn" statement is abandoned, and the next ACL statement is
6425 processed. Previously this caused the whole ACL to be aborted.
6427 13. Added the iplsearch lookup type.
6429 14. Added ident_timeout as a log selector.
6431 15. Added tls_certificate_verified as a log selector.
6433 16. Added a global option tls_require_ciphers (compare the smtp transport
6434 option of the same name). This controls incoming TLS connections.
6436 17. I finally figured out how to make tls_require_ciphers do a similar thing
6437 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6438 before starting the TLS session.
6440 18. Tabs are now shown as \t in -bP output.
6442 19. If the log selector return_path_on_delivery was set, Exim crashed when
6443 bouncing a message because it had too many Received: header lines.
6445 20. If two routers both had headers_remove settings, and the first one included
6446 a superfluous trailing colon, the final name in the first list and the
6447 first name in the second list were incorrectly joined into one item (with a
6448 colon in the middle).
6454 1. Added -C and -D options to the exinext utility, mainly to make it easier
6455 to include in the automated testing, but these could be helpful when
6456 multiple configurations are in use.
6458 2. The exinext utility was not formatting the output nicely when there was
6459 an alternate port involved in the retry record key, nor when there was a
6460 message id as well (for retries that were specific to a specific message
6461 and a specific host). It was also confused by IPv6 addresses, because of
6462 the additional colons they contain. I have fixed the IPv4 problem, and
6463 patched it up to do a reasonable job for IPv6.
6465 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6466 delivery, the log line now contains "pipelined" if PIPELINING was used.
6468 4. An SMTP transport process used to panic and die if the bind() call to set
6469 an explicit outgoing interface failed. This has been changed; it is now
6470 treated in the same way as a connect() failure.
6472 5. A reference to $sender_host_name in the part of a conditional expansion
6473 that was being skipped was still causing a DNS lookup. This no longer
6476 6. The def: expansion condition was not recognizing references to header lines
6477 that used bh_ and bheader_.
6479 7. Added the _cache feature to named lists.
6481 8. The code for checking quota_filecount in the appendfile transport was
6482 allowing one more file than it should have been.
6484 9. For compatibility with Sendmail, the command line option
6492 and sets the incoming protocol and host name (for trusted callers). The
6493 host name and its colon can be omitted when only the protocol is to be set.
6494 Note the Exim already has two private options, -pd and -ps, that refer to
6495 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6496 "s", but I don't think that's a major issue.
6498 10. A number of refactoring changes to the code, none of which should affect
6501 (a) The number of logging options was getting close to filling up the
6502 32-bit word that was used as a bit map. I have split them into two classes:
6503 those that are passed in the argument to log_write(), and those that are
6504 only ever tested independently outside of that function. These are now in
6505 separate 32-bit words, so there is plenty of room for expansion again.
6506 There is no change in the user interface or the logging behaviour.
6508 (b) When building, for example, log lines, the code previously used a
6509 macro that called string_cat() twice, in order to add two strings. This is
6510 not really sufficiently general. Furthermore, there was one instance where
6511 it was actually wrong because one of the argument was used twice, and in
6512 one call a function was used. (As it happened, calling the function twice
6513 did not affect the overall behaviour.) The macro has been replaced by a
6514 function that can join an arbitrary number of extra strings onto a growing
6517 (c) The code for expansion conditions now uses a table and a binary chop
6518 instead of a serial search (which was left over from when there were very
6519 few conditions). Also, it now recognizes conditions like "pam" even when
6520 the relevant support is not compiled in: a suitably worded error message is
6521 given if an attempt is made to use such a condition.
6523 11. Added ${time_interval:xxxxx}.
6525 12. A bug was causing one of the ddress fields not to be passed back correctly
6526 from remote delivery subprocesses. The field in question was not being
6527 subsequently used, so this caused to problems in practice.
6529 13. Added new log selectors queue_time and deliver_time.
6531 14. Might have fixed a bug in maildirsizefile handling that threw up
6532 "unexpected character" debug warnings, and recalculated the data
6533 unnecessarily. In any case, I expanded the warning message to give more
6536 15. Added the message "Restricted characters in address" to the statements in
6537 the default ACL that block characters like @ and % in local parts.
6539 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6540 Three changes have been made:
6542 (a) There was a serious bug; a negative response to MAIL caused the whole
6543 recipient domain to be cached as invalid, thereby blocking all messages
6544 to all local parts at the same domain, from all senders. This bug has
6545 been fixed. The domain is no longer cached after a negative response to
6546 MAIL if the sender used is not empty.
6548 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6551 (c) A new callout option, "use_sender" has been added for people who want
6552 the modified behaviour.
6558 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6561 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6562 indeed breaks things for older releases.
6564 3. Added additional logging to the case where there is a problem reading data
6565 from a filter that is running in a subprocess using a pipe, in order to
6566 try to track down a specific problem.
6568 4. Testing facility fudge: when running in the test harness and attempting
6569 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6570 getting "No route to host". Convert this to a timeout.
6572 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6575 6. Some OS don't have socklen_t but use size_t instead. This affects the
6576 fifth argument of getsockopt() amongst other things. This is now
6577 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6578 can be set for individual OS. I have set it for SunOS5, OSF1, and
6579 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6580 some earlier ones do not.
6582 7. Change 4.30/15 was not doing the test caselessly.
6584 8. The standard form for an IPv6 address literal was being rejected by address
6585 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6586 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6587 this, as well as the form without the "IPv6" on the front (but only when
6588 address literals are enabled, of course).
6590 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6592 10. Exim crashed if a message with an empty sender address specified by -f
6593 encountered a router with an errors_to setting. This could be provoked only
6594 by a command such as
6598 where an empty string was supplied; "<>" did not hit this bug.
6600 11. Installed PCRE release 4.5.
6602 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6603 remained set. It is now erased.
6605 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6606 times from message ids (which are base 36 rather than the normal 62).
6608 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6609 were being counted as actual protocol errors, and logged if the log
6610 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6611 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6612 and DATA following a set of rejected RCPTs do not count as protocol errors.
6613 In other words, Exim assumes they were pipelined, though this may not
6614 actually be the case. Of course, in all cases the client gets an
6615 appropriate error code.
6617 15. If a lookup fails in an ACL condition, a message about the failure may
6618 be available; it is used if testing the ACL cannot continue, because most
6619 such messages specify what the cause of the deferral is. However, some
6620 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6621 that caused an old message to be retained and used if a later statement
6622 caused a defer, replacing the real cause of the deferral.
6624 16. If an IP address had so many PTR records that the DNS lookup buffer
6625 was not large enough to hold them, Exim could crash while trying to process
6626 the truncated data. It now detects and logs this case.
6628 17. Further to 4.21/58, another change has been made: if (and only if) the
6629 first line of a message (the first header line) ends with CRLF, a bare LF
6630 in a subsequent header line has a space inserted after it, so as not to
6631 terminate the header.
6633 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6634 unnecessarily, used atoi() instead of strtol(), and didn't check the
6635 termination when getting file sizes from file names by regex.
6637 19. Completely re-implemented the support for maildirsize files, in the light
6638 of a number of problems with the previous contributed implementation
6639 (4.30/29). In particular:
6641 . If the quota is zero, the maildirsize file is maintained, but no quota is
6644 . If the maildir directory does not exist, it is created before any attempt
6645 to write a maildirsize file.
6647 . The quota value in the file is just a cache; if the quota is changed in
6648 the transport, the new value overrides.
6650 . A regular expression is available for excluding directories from the
6653 20. The autoreply transport checks the characters in options that define the
6654 message's headers; it allows continued headers, but it was checking with
6655 isspace() after an embedded newline instead of explicitly looking for a
6658 21. If all the "regular" hosts to which an address was routed had passed their
6659 expiry times, and had not reached their retry times, the address was
6660 bounced, even if fallback hosts were defined. Now Exim should go on to try
6663 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6664 equivalent code in the SMTP transport. Some hosts send humungous responses
6665 to HELO/EHLO, more than 1024 it seems.
6667 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6668 gives compiler warnings in some environments. I've now done it "properly",
6671 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6672 (because of problems with the built-in one) was declared to return uschar *
6673 instead of char *, causing compiler failure.
6675 25. Fixed a file descriptor leak when processing alias/forward files.
6677 26. Fixed a minor format string issue in dbfn.c.
6679 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6681 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6682 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6685 29. When a sender address is verified, it is cached, to save repeating the test
6686 when there is more than one recipient in a message. However, when the
6687 verification involves a callout, it is possible for different callout
6688 options to be set for different recipients. It is too complicated to keep
6689 track of this in the cache, so now Exim always runs a verification when a
6690 callout is required, relying on the callout cache for the optimization.
6691 The overhead is duplication of the address routing, but this should not be
6694 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6695 to be verified with callout=postmaster, and the main callout worked but the
6696 postmaster check failed, the verification correctly failed. However, if a
6697 subsequent RCPT command asked for sender verification *without* the
6698 postmaster check, incorrect caching caused this verification also to fail,
6701 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6702 it was not caching the DNS options (qualify_single, search_parents) that
6703 were used when the lookup failed. A subsequent lookup with different
6704 options therefore always gave the same answer, though there were cases
6705 where it should not have. (Example: a "domains = !$mx_any" option on a
6706 dnslookup router: the "domains" option is always processed without any
6707 widening, but the router might have qualify_single set.) Now Exim uses the
6708 cached value only when the same options are set.
6710 32. Added John Jetmore's "exipick" utility to the distribution.
6712 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6713 than a timeout (e.g. a certificate is required, and is not provided), an
6714 Exim server now closes the connection immediately. Previously it waited for
6715 the client to close - but if the client is SSL, it seems that they each
6716 wait for each other, leading to a delay before one of them times out.
6718 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6719 maintained 0.8.x compatibility because I don't think many are using it, and
6720 it is clearly obsolete.
6722 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6725 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6726 client certificate was expired. A simple patch fixes this, though I don't
6727 understand the full logic of why the verify callback is called multiple
6730 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6731 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6732 which causes problems with some clients (such as the Certicom SSL Plus
6733 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6734 disables the coutermeasure allowing Eudora to connect."
6736 38. Exim was not checking that a write() to a log file succeeded. This could
6737 lead to Bad Things if a log got too big, in particular if it hit a file
6738 size limit. Exim now panics and dies if it cannot write to a log file, just
6739 as it does if it cannot open a log file.
6741 39. Modified OS/Makefile-Linux so that it now contains
6743 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6745 The two -D definitions ensure that Exim is compiled with large file
6746 support, which makes it possible to handle log files that are bigger than
6749 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6750 instance) a domain was checked against a named list that involved a lookup,
6751 causing $domain_data to be set, then another domain was checked against the
6752 same list, then the first domain was re-checked, the value of $domain_data
6753 after the final check could be wrong. In particular, if the second check
6754 failed, it could be set empty. This bug probably also applied to
6757 41. The strip_trailing_dot option was not being applied to the address given
6758 with the -f command-line option.
6760 42. The code for reading a message's header from the spool was incrementing
6761 $received_count, but never initializing it. This meant that the value was
6762 incorrect (doubled) while delivering a message in the same process in which
6763 it was received. In the most common configuration of Exim, this never
6764 happens - a fresh exec is done - but it can happen when
6765 deliver_drop_privilege is set.
6767 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6768 it now includes up to 150 characters of the unexpected data in the log
6771 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6772 and building data strings. The size of both of these buffers was 10 000
6773 bytes - far larger than anybody would *ever* want, thought I. Needless to
6774 say, somebody hit the limit. I have increased the maximum line length to
6775 20 000 and the maximum data length of concatenated lines to 100 000. I have
6776 also fixed two bugs, because there was no checking on these buffers. Tsk,
6777 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6778 buffer is too small.
6780 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6781 lsearch lookups. Now it does.
6783 46. When parsing a route_list item in a manualroute router, a fixed-length
6784 buffer was used for the list of hosts. I made this 1024 bytes long,
6785 thinking that nobody would ever have a list of hosts that long. Wrong.
6786 Somebody had a whole pile of complicated expansion conditions, and the
6787 string was silently truncated, leading to an expansion error. It turns out
6788 that it is easier to change to an unlimited length (owing to other changes
6789 that have happened since this code was originally written) than to build
6790 structure for giving a limitation error. The length of the item that
6791 expands into the list of hosts is now unlimited.
6793 47. The lsearch lookup could not handle data where the length of text line was
6794 more than 4095 characters. Such lines were truncated, leading to shortened
6795 data being returned. It should now handle lines of any length.
6797 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6798 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6801 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6802 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6803 no longer happens in comment lines. A list of replacements is now placed
6804 at the head of all of the source files, except those whose only change is
6805 to replace PERL_COMMAND in the very first #! line.
6807 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6808 messages on the queue, with a bottom-up merge sort, using code contributed
6809 by Michael Haardt. This should make operations like -bp somewhat faster on
6810 large queues. It won't affect queue runners, except when queue_run_in_order
6813 51. Installed eximstats 1.31 in the distribution.
6815 52. Added support for SRV lookups to the dnslookup router.
6817 53. If an ACL referred to $message_body or $message_body_end, the value was not
6818 reset for any messages that followed in the same SMTP session.
6820 54. The store-handling optimization for building very long strings was not
6821 differentiating between the different store pools. I don't think this
6822 actually made any difference in practice, but I've tidied it.
6824 55. While running the routers to verify a sender address, $sender_address
6825 was still set to the sender address. This is wrong, because when routing to
6826 send a bounce to the sender, it would be empty. Therefore, I have changed
6827 it so that, while verifying a sender address, $sender_address is set to <>.
6828 (There is no change to what happens when verifying a recipient address.)
6830 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6831 target A or AAAA records (if not already returned) without resetting the
6832 qualify_single or search_parents options of the DNS resolver. These are
6833 inappropriate in this case because the targets of MX and SRV records must
6834 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6835 target that, when qualified, matched something in the local domain. These
6836 two options are now turned off when doing these lookups.
6838 57. It seems that at least some releases of Reiserfs (which does not have the
6839 concept of a fixed number of inodes) returns zero and not -1 for the
6840 number of available inodes. This interacted badly with check_spool_inodes,
6841 which assumed that -1 was the "no such thing" setting. What I have done is
6842 to check that the total number of inodes is greater than zero before doing
6843 the test of how many are available.
6845 58. When a "warn" ACL statement has a log_message modifier, the message is
6846 remembered, and not repeated. This is to avoid a lot of repetition when a
6847 message has many recipients that cause the same warning to be written.
6848 However, Exim was preserving the list of already written lines for an
6849 entire SMTP session, which doesn't seem right. The memory is now reset if a
6850 new message is started.
6852 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6853 debugging output unless log_rewrite was also set.
6855 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6856 of (int)(handle) when we know that handle contains (void *)(-1).
6858 61. The Exim daemon panic-logs an error return when it closes the incoming
6859 connection. However "connection reset by peer" seems to be common, and
6860 isn't really an error worthy of noting specially, so that particular error
6863 62. When Exim is trying to find all the local interfaces, it used to panic and
6864 die if the ioctl to get the interface flags failed. However, it seems that
6865 on at least one OS (Solaris 9) it is possible to have an interface that is
6866 included in the list of interfaces, but for which you get a failure error
6867 for this call. This happens when the interface is not "plumbed" into a
6868 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6869 failure of the "get flags" call assumes that the interface is down.
6871 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6872 makes life easier for people who are doing arithmetic on fields extracted
6873 from dates, where you often get leading zeros that should not be
6874 interpreted as octal.
6876 64. Added qualify_domain to the redirect router, to override the global
6879 65. If a pathologically long header line contained very many addresses (the
6880 report of this problem mentioned 10 000) and each of them was rewritten,
6881 Exim could use up a very large amount of memory. (It kept on making new
6882 copies of the header line as it rewrote, and never released the old ones.)
6883 At the expense of a bit more processing, the header rewriting function has
6884 been changed so that it no longer eats memory in this way.
6886 66. The generation of the Received: header has been moved from the time that a
6887 message starts to be received, to the time that it finishes. The timestamp
6888 in the Received: header should now be very close to that of the <= log
6889 line. There are two side-effects of this change:
6891 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6892 logged header lines no longer include the local Received: line, because
6893 it has not yet been created. The same applies to a copy of the message
6894 that is returned to a non-SMTP sender when a message is rejected.
6896 (b) When a filter file is tested using -bf, no additional Received: header
6897 is added to the test message. After some thought, I decided that this
6900 This change does not affect the value of $received_for. It is still set
6901 after address rewriting, but before local_scan() is called.
6903 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6905 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6906 gave an unhelpful panic error message, and a defer error. I have managed to
6907 change this behaviour so that it now rejects any supplied certificate,
6908 which seems right, as the list of acceptable certificates is empty.
6910 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6911 gave an unhelpful defer error. I have not managed to make this reject any
6912 supplied certificates, but the error message it gives is "no certificate
6913 supplied", which is not helpful.
6915 70. exigrep's output now also includes lines that are not associated with any
6916 message, but which match the given pattern. Implemented by a patch from
6917 Martin Sluka, which also tidied up the Perl a bit.
6919 71. Recipient callout verification, like sender verification, was using <> in
6920 the MAIL FROM command. This isn't really the right thing, since the actual
6921 sender may affect whether the remote host accepts the recipient or not. I
6922 have changed it to use the actual sender in the callout; this means that
6923 the cache record is now keyed on a recipient/sender pair, not just the
6924 recipient address. There doesn't seem to be a real danger of callout loops,
6925 since a callout by the remote host to check the sender would use <>.
6926 [SEE ABOVE: changed after hitting problems.]
6928 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6929 temporary errors. However, in the case of such a code being given after
6930 the end of a data transmission (i.e. after ".") Exim was failing to write
6931 a retry record for the message. (Yes, there was some broken host that was
6932 actually sending 8xx at this point.)
6934 73. An unknown lookup type in a host list could cause Exim to panic-die when
6935 the list was checked. (An example that provoked this was putting <; in the
6936 middle of a list instead of at the start.) If this happened during a DATA
6937 ACL check, a -D file could be left lying around. This kind of configuration
6938 error no longer causes Exim to die; instead it causes a defer error. The
6939 incident is still logged to the main and panic logs.
6941 74. Buglet left over from Exim 3 conversion. The message "too many messages
6942 in one connection" was written to the rejectlog but not the mainlog, except
6943 when address rewriting (yes!) was being logged.
6945 75. Added write_rejectlog option.
6947 76. When a system filter was run not as root (that is, when system_filter_user
6948 was set), the values of the $n variables were not being returned to the
6949 main process; thus, they were not subsequently available in the $sn
6952 77. Added +return_path_on_delivery log selector.
6954 78. A connection timeout was being treated differently from recipients deferred
6955 when testing hosts_max_try with a message that was older than the host's
6956 retry timeout. (The host should not be counted, thus allowing all hosts to
6957 be tried at least once before bouncing.) This may have been the cause of an
6958 occasionally reported bug whereby a message would remain on the queue
6959 longer than the retry timeout, but would be bounced if a delivery was
6960 forced. I say "may" because I never totally pinned down the problem;
6961 setting up timeout/retry tests is difficult. See also the next item.
6963 79. The ultimate address timeout was not being applied to errors that involved
6964 a combination of host plus message (for example, a timeout on a MAIL
6965 command). When an address resolved to a number of possible hosts, and they
6966 were not all tried for each delivery (e.g. because of hosts_max_try), a
6967 message could remain on the queue longer than the retry timeout.
6969 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6972 81. Fixed an obscure SMTP outgoing bug which required at least the following
6973 conditions: (a) there was another message waiting for the same server;
6974 (b) the server returned 5xx to all RCPT commands in the first message so
6975 that the message was not completed; (c) the server dropped the connection
6976 or gave a negative response to the RSET that Exim sends to abort the
6977 transaction. The observed case was a dropped connection after DATA that had
6978 been sent in pipelining mode. That is, the server had advertised PIPELINING
6979 but was not implementing it correctly. The effect of the bug was incorrect
6980 behaviour, such as trying another host, and this could lead to a crash.
6986 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6987 and daemon.c were passed as pointers to ints; they should have been
6988 pointers to socklen_t variables (which are typically unsigned ints).
6990 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6993 3. Fixed a really odd bug that affected only the testing scheme; patching a
6994 certain fixed string in the binary changed the value of another string that
6995 happened to be identical to the end of the original first string.
6997 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6998 name", it returns that address as the IP address. On some operating
6999 systems (e.g. Solaris), it also passes back the IP address string as the
7000 "host name". However, on others (e.g. Linux), it passes back an empty
7001 string. Exim wasn't checking for this, and was changing the host name to an
7002 empty string, assuming it had been canonicalized.
7004 5. Although rare, it is permitted to have more than one PTR record for a given
7005 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7006 all the names associated with an address, because they do in Solaris.
7007 However, it seems that they do not in Linux for data that comes from the
7008 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7009 I found this out when I moved to a new Linux workstation and tried to run
7010 the Exim test suite.
7012 To get round this problem I have changed the code so that it now does its
7013 own call to the DNS to look up PTR records when searching for a host name.
7014 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7015 addresses that are only in /etc/hosts are still found.
7017 This behaviour is, however, controlled by an option called host_lookup_
7018 order, which defaults to "bydns:byaddr". If people want to use the other
7019 order, or indeed, just use one or the other means of lookup, they can
7020 specify it in this variable.
7022 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7023 some operating systems, this comes back from gethostbyaddr() as an empty
7024 string, and this is what Exim used to test for. However, it seems that in
7025 other systems, "." is yielded. Exim now tests for this case too.
7027 7. The values of check_spool_space and check_log_space are now held internally
7028 as a number of kilobytes instead of an absolute number of bytes. If a
7029 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7030 kilobyte. This means that much larger values can be stored.
7032 8. Exim monitor: an attempt to get the action menu when not actually pointing
7033 at a message produces an empty menu entitled "No message selected". This
7034 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7035 no entries in it ("Shell widget menu has zero width and/or height"). So I
7036 have added a single, blank menu entry in this case.
7038 9. Added ${quote_local_part.
7040 10. MIME decoding is now applied to the contents of Subject: header lines when
7043 11. Now that a reference to $sender_host_address automatically causes a reverse
7044 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7045 host lookup before query-style lookups in lists that might use this
7046 variable. This has therefore been abolished, and the "net-" prefix is no
7047 longer necessary for query-style lookups.
7049 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7050 have been a typo for LFLAGS, so it has been changed.
7052 13. The install script calls Exim with "-C /dev/null" in order to find the
7053 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7054 to be output. However, since Exim outputs its version number before the
7055 error, it didn't break the script. It just looked ugly. I fixed this by
7056 always allowing "-C /dev/null" if the caller is root.
7058 14. Ignore overlarge ACL variable number when reading spool file - insurance
7059 against a later release with more variables having written the file.
7061 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7062 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7063 this, as well as the form without the "IPv6" on the front.
7065 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7066 OS/Makefile-Darwin file.
7068 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7069 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7071 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7072 variable. However, some OS use ints for inodes. I've added cast to long int
7073 to get rid of the compiler warning.
7075 19. I had forgotten to lock out "/../" in configuration file names when
7076 ALT_CONFIG_PREFIX was set.
7078 20. Routers used for verification do not need to specify transports. However,
7079 if such a router generated a host list, and callout was configured, Exim
7080 crashed, because it could not find a port number from the (non-existent)
7081 transport. It now assumes port 25 in this circumstance.
7083 21. Added the -t option to exigrep.
7085 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7086 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7087 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7088 them is set, LOOKUP_LSEARCH is forced.
7090 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7092 24. Added sender and host information to the "rejected by local_scan()" log
7093 line; previously there was no indication of these.
7095 25. Added .include_if_exists.
7097 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7098 Linux. It turns out that not all file systems support this. Apparently some
7099 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7100 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7101 not supported on the file descriptor, is now ignored when Exim is trying to
7102 sync a directory. This applies only to Linux.
7104 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7106 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7107 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7110 29. Added support for maildirsize files from supplied patch (modified a bit).
7112 30. The use of :fail: followed by an empty string could lead Exim to respond to
7113 sender verification failures with (e.g.):
7115 550 Verification failed for <xxx>
7116 550 Sender verify failed
7118 where the first response line was missing the '-' that indicates it is not
7119 the final line of the response.
7121 31. The loop for finding the name of the user that called Exim had a hardwired
7122 limit of 10; it now uses the value of finduser_retries, which is used for
7123 all other user lookups.
7125 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7128 33. Exim was neglecting to zero errno before one call of strtol() when
7129 expanding a string and expecting an integer value. On some systems this
7130 resulted in spurious "integer overflow" errors. Also, it was casting the
7131 result into an int without checking.
7133 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7134 did not work. The code looks as if it has *never* worked, though it appears
7135 to have been documented since at least release 1.62. I have made it work.
7137 35. The "timeout_DNS" error in retry rules, also documented since at least
7138 1.62, also never worked. As it isn't clear exactly what this means, and
7139 clearly it isn't a major issue, I have abolished the feature by treating it
7140 as "timeout", and writing a warning to the main and panic logs.
7142 36. The display of retry rules for -brt wasn't always showing the error code
7145 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7146 timeout_connect_A, timeout_connect_MX.
7148 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7149 to the empty sender.
7151 39. The daemon was not analysing the content of -oX till after it had closed
7152 stderr and disconnected from the controlling terminal. This meant that any
7153 syntax errors were only noted on the panic log, and the return code from
7154 the command was 0. By re-arranging the code a little, I've made the
7155 decoding happen first, so such errors now appear on stderr, and the return
7156 code is 1. However, the actual setting up of the sockets still happens in
7157 the disconnected process, so errors there are still only recorded on the
7160 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7161 connections (as happens on some IP stacks) was logged at start up time as
7162 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7163 it from "IPv6 and IPv4", which means that two separate sockets are being
7166 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7167 says whether AF_INET or AF_INET6 was passed as an argument.
7169 42. Exiwhat output was messed up when time zones were included in log
7172 43. Exiwhat now gives more information about the daemon's listening ports,
7173 and whether -tls-on-connect was used.
7175 44. The "port" option of the smtp transport is now expanded.
7177 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7178 silently ignored. Now an error message is written to the main and panic
7181 46. There's a new ACL modifier called "logwrite" which writes to a log file
7182 as soon as it is encountered.
7184 47. Added $local_user_uid and $local_user_gid at routing time.
7186 48. Exim crashed when trying to verify a sender address that was being
7189 49. Exim was recognizing only a space character after ".include". It now also
7190 recognizes a tab character.
7192 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7193 extracting the relevant information from the specification. The man page no
7194 longer contains scrambled data for the -d option, and I've added a section
7195 at the front about calling Exim under different names.
7197 51. Added "extra_headers" argument to the "mail" command in filter files.
7199 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7202 53. Installed eximstats 1.29.
7204 54. Added transport_filter_timeout as a generic transport option.
7206 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7207 Cc: header lines. This was required by RFC 822, but it not required by RFC
7210 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7211 incoming messages that did not have them. Now it does so only if the
7212 message originates locally, that is, if there is no associated remote host
7213 address. When Resent- header lines are present, this applies to the Resent-
7214 lines rather than the non-Resent- lines.
7216 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7217 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7219 58. Messages for configuration errors now include the name of the main
7220 configuration file - useful now that there may be more than one file in a
7221 list (.included file names were always shown).
7223 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7224 for those rare installations that do not start the daemon as root or run it
7225 setuid root. I've cut out the call to initgroups() if the daemon is not
7228 60. The Exim user and group can now be bound into the binary as text strings
7229 that are looked up at the start of Exim's processing.
7231 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7233 62. Added $mailstore_basename variable.
7235 63. Installed patch to sieve.c from Michael Haardt.
7237 64. When Exim failed to open the panic log after failing to open the main log,
7238 the original message it was trying to log was written to stderr and debug
7239 output, but if they were not available (the usual case in production), it
7240 was lost. Now it is written to syslog before the two lines that record the
7241 failures to open the logs.
7243 65. Users' Exim filters run in subprocesses under the user's uid. It is
7244 possible for a "deliver" command or an alias in a "personal" command to
7245 provoke an address rewrite. If logging of address rewriting is configured,
7246 this fails because the process is not running as root or exim. There may be
7247 a better way of dealing with this, but for the moment (because 4.30 needs
7248 to be released), I have disabled address rewrite logging when running a
7249 filter in a non-root, non-exim process.
7255 1. The buildconfig auxiliary program wasn't quoting the value set for
7256 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7257 not defined. This bug was masked in 4.22 by the effect that was fixed in
7260 2. Some messages that were rejected after a message id was allocated were
7261 shown as "incomplete" by exigrep. It no longer does this for messages that
7262 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7264 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7265 have allow_domain_literals set, the ID did not get logged in the <= line.
7266 Domain literals are now always recognized in Message-ID: header lines.
7268 4. The first argument for a ${extract expansion item is the key name or field
7269 number. Leading and trailing spaces in this item were not being ignored,
7270 causing some misleading effects.
7272 5. When deliver_drop_privilege was set, single queue runner processes started
7273 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7274 same command in the process it spins off) were not dropping privilege.
7276 6. When the daemon running as "exim" started a queue runner, it always
7277 re-executed Exim in the spun-off process. This is a waste of effort when
7278 deliver_drop_privilege is set. The new process now just calls the
7279 queue-runner function directly.
7285 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7288 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7289 ignored. Though the use of -oP was forcing the writing of a pid file, it
7290 was always written to the default place.
7292 3. If the message "no IP address found for host xxxx" is generated during
7293 incoming verification, it is now followed by identification of the incoming
7294 connection (so you can more easily find what provoked it).
7296 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7298 5. Added some features to "harden" Exim a bit more against certain attacks:
7300 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7301 be put in Local/Makefile. This is like the never_users runtime option,
7302 but it cannot be overridden. The default setting is "root".
7304 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7305 prefix string with which any file named in a -C command line option
7308 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7309 is retained for -C and -D only if the caller of Exim is root. Without
7310 it, the exim user may also use -C and -D and retain privilege.
7312 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7313 command line option is disabled.
7315 6. Macro names set by the -D option must start with an upper case letter, just
7316 like macro names defined in the configuration file.
7318 7. Added "dereference=" facility to LDAP.
7320 8. Two instances of the typo "uknown" in the source files are fixed.
7322 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7323 the Configure-Makefile script screwed up while processing it.
7325 10. Incorporated PCRE 4.4.
7327 11. The SMTP synchronization check was not operating right at the start of an
7328 SMTP session. For example, it could not catch a HELO sent before the client
7329 waited for the greeting. There is now a check for outstanding input at the
7330 point when the greeting is written. Because of the duplex, asynchronous
7331 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7332 way, but not yet received, when the check is performed.
7334 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7335 on TCP/IP sockets, because this apparently causes some broken clients to
7338 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7339 unchanged) from the Cygwin maintainer.
7341 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7342 was supported instead of testing for mbx. Effectively a typo.
7344 15. The spa authenticator server code was not checking that the input it
7345 received was valid base64.
7347 16. The debug output line for the "set" modifier in ACLs was not showing the
7348 name of the variable that was being set.
7350 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7352 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7353 lookup on its own. Something else was needed to trigger the lookup. For
7354 example, a match in host_lookup or the need for a host name in a host list.
7355 Now, if $sender_host_name is referenced and the host name has not yet been
7356 looked up, a lookup is performed. If the lookup fails, the variable remains
7357 empty, and $host_lookup_failed is set to "1".
7359 19. Added "eqi" as a case-independent comparison operator.
7361 20. The saslauthd authentication condition could segfault if neither service
7362 nor realm was specified.
7364 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7365 error message that was logged was misleading, and incoming SMTP
7366 connections were dropped. The message is now more accurate, and temporary
7367 errors are given to SMTP connections.
7369 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7370 (except RSET) with a 421 error, until QUIT is received. However, it was
7371 failing to send a response to QUIT.
7373 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7374 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7375 ACL is now run after the helo_try_verify_hosts code.
7377 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7378 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7379 case-independent, so other case variants are also recognized). Apparently
7380 some systems use these upper case variants.
7382 25. If more than two messages were waiting for the same host, and a transport
7383 filter was specified for the transport, Exim sent two messages over the
7384 same TCP/IP connection, and then failed with "socket operation on non-
7385 socket" when it tried to send the third.
7387 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7389 27. The extern definition of crypt16() in expand.c was not being excluded when
7390 the OS had its own crypt16() function.
7392 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7393 as a preferred synonym for return_size_limit, both as an option and as an
7396 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7398 30. Changed the default configuration ACL to relax the local part checking rule
7399 for addresses that are not in any local domains. For these addresses,
7400 slashes and pipe symbols are allowed within local parts, but the sequence
7401 /../ is explicitly forbidden.
7403 31. SPA server authentication was not clearing the challenge buffer before
7406 32. log_message in a "warn" ACL statement was writing to the reject log as
7407 well as to the main log, which contradicts the documentation and doesn't
7408 seem right (because no rejection is happening). So I have stopped it.
7410 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7411 However, I am unable to do any testing of this.
7413 34. Fixed an infelicity in the appendfile transport. When checking directories
7414 for a mailbox, to see if any needed to be created, it was accidentally
7415 using path names with one or more superfluous leading slashes; tracing
7416 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7418 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7419 contents are added to the log line that is written for every discarded
7420 recipient. (Previously a log_message setting was ignored.)
7422 36. The ${quote: operator now quotes the string if it is empty.
7424 37. The install script runs exim in order to find its version number. If for
7425 some reason other than non-existence or emptiness, which it checks, it
7426 could not run './exim', it was installing it with an empty version number,
7427 i.e. as "exim-". This error state is now caught, and the installation is
7430 38. An argument was missing from the function that creates an error message
7431 when Exim fails to connect to the socket for saslauthd authentication.
7432 This could cause Exim to crash, or give a corrupted message.
7434 39. Added isip, isip4, and isip6 to ${if conditions.
7436 40. The ACL variables $acl_xx are now saved with the message, and can be
7437 accessed later in routers, transports, and filters.
7439 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7440 strings in the file are not string-expanded.
7442 42. If a MAIL command specified a SIZE value that was too large to fit into an
7443 int variable, the check against message_size_limit failed. Such values are
7444 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7445 day this will have to be increased, but I don't think I want to be around
7446 when emails are that large.
7453 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7454 iconv() is not standard in FreeBSD.
7456 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7457 IPv6 enabled. The observed symptom was a segmentation fault on return from
7458 the function os_common_find_running_interfaces() in src/os.c.
7460 3. In the check_special_case() function in daemon.c I had used "errno" as an
7461 argument name, which causes warnings on some systems. This was basically a
7462 typo, since it was named "eno" in the comments!
7464 4. The code that waits for the clock to tick (at a resolution of some fraction
7465 of a second) so as to ensure message-id uniqueness was always waiting for
7466 at least one whole tick, when it could have waited for less. [This is
7467 almost certainly not relevant at current processor speeds, where it is
7468 unlikely to ever wait at all. But we try to future-proof.]
7470 5. The function that sleeps for a time interval that includes fractions of a
7471 second contained a race. It did not block SIGALRM between setting the
7472 timer, and suspending (a couple of lines later). If the interval was short
7473 and the sigsuspend() was delayed until after it had expired, the suspension
7474 never ended. On busy systems this could lead to processes getting stuck for
7477 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7478 process, before it forks any delivery processes. The open lookup caching
7479 mechanism meant that the open file or database connection was passed into
7480 the delivery process. The problem was that delivery processes always tidy
7481 up cached lookup data. This could cause a problem for the next delivery
7482 process started by the queue runner, because the external queue runner
7483 process does not know about the closure. So the next delivery process
7484 still has data in the lookup cache. In the case of a file lookup, there was
7485 no problem because closing a file descriptor in a subprocess doesn't affect
7486 the parent. However, if the lookup was caching a connection to a database,
7487 the connection was closed, and the second delivery process was likely to
7488 see errors such as "PGSQL: query failed: server closed the connection
7489 unexpectedly". The problem has been fixed by closing all cached lookups
7490 in a queue runner before running a delivery process.
7492 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7493 seem to have the "const" qualifier which it has on other OS. I've
7496 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7497 *of the same type* (client or server) with the same public name that an
7498 error should be diagnosed.
7500 9. When Exim looked up a host name for an IP address, but failed to find the
7501 original IP address when looking up the host name (a safety check), it
7502 output the message "<ip address> does not match any IP for NULL", which was
7503 confusing, to say the least. The bug was that the host name should have
7504 appeared instead of "NULL".
7506 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7507 user that is built into the binary, and the -C or -D options is used, root
7508 privilege is dropped before the configuration file is read. In addition,
7509 logging is switched to stderr instead of the normal log files. If the
7510 configuration then re-defines the Exim user, the unprivileged environment
7511 is probably not what is expected, so Exim logs a panic warning message (but
7514 However, if deliver_drop_privilege is set, the unprivileged state may well
7515 be exactly what is intended, so the warning has been cut out in that case,
7516 and Exim is allowed to try to write to its normal log files.
7522 1. smtp_return_error_details was not giving details for temporary sender
7523 or receiver verification errors.
7525 2. Diagnose a configuration error if two authenticators have the same public
7528 3. Exim used not to create the message log file for a message until the first
7529 delivery attempt. This could be confusing when incoming messages were held
7530 for policy or load reasons. The message log file is now created at the time
7531 the message is received, and an initial "Received" line is written to it.
7533 4. The automatically generated man page for command line options had a minor
7534 bug that caused no ill effects; however, a more serious problem was that
7535 the procedure for building the man page automatically didn't always
7536 operate. Consequently, release 4.20 contains an out-of-date version. This
7537 shouldn't happen again.
7539 5. When building Exim with embedded Perl support, the script that builds the
7540 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7541 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7543 6. The freeze_tell option was not being used for messages that were frozen on
7544 arrival, either by an ACL or by local_scan().
7546 7. Added the smtp_incomplete_transaction log selector.
7548 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7549 was accepting AUTH without a new EHLO.
7551 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7552 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7555 10. Logging of TCP/IP connections (when configured) now happens in the main
7556 daemon process instead of the child process, so that the TCP/IP connection
7557 count is more accurate (but it can never be perfect).
7559 11. The use of "drop" in a nested ACL was not being handled correctly in the
7560 outer ACL. Now, if condition failure induced by the nested "drop" causes
7561 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7562 or "require"), the connection is dropped.
7564 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7565 that yield "discard" can now be used with an "accept" or a "discard" verb,
7566 but an error is generated for any others (because I can't see a useful way
7567 to define what should happen).
7569 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7570 are now processed in the same way as lines in the Exim configuration file.
7571 In particular, continuation lines are supported.
7573 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7575 15. Added -ti meaning -t -i.
7577 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7578 domains, and warn by logging if others are found.
7580 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7581 returned from GIFCONF when Exim is trying to find the list of interfaces on
7582 a host. The code in os.c has been modified to copy each ifreq to an aligned
7583 structure in all cases.
7585 Also, in some cases, the returned ifreq's were being copied to a 'struct
7586 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7587 means the last couple of bytes of an IPv6 address could be chopped if the
7588 ifreq contained only a normal sockaddr (14 bytes storage).
7590 18. Named domain lists were not supported in the hosts_treat_as_local option.
7591 An entry such as +xxxx was not recognized, and was treated as a literal
7594 19. Ensure that header lines added by a DATA ACL are included in the reject log
7595 if the ACL subsequently rejects the message.
7597 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7598 MD5 (which is deprecated).
7600 21. When testing a filter file using -bf, Exim was writing a message when it
7601 took the sender from a "From " line in the message, but it was not doing so
7602 when it took $return_path from a Return-Path: header line. It now does.
7604 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7605 with a valid header line field name (a series of printing characters
7606 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7608 23. Changed "disc" in the source to "disk" to conform to the documentation and
7609 the book and for uniformity.
7611 24. Ignore Sendmail's -Ooption=value command line item.
7613 25. When execve() failed while trying to run a command in a pipe transport,
7614 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7615 could be confused with a return value of 69 from the command itself. This
7616 has been changed to 127, the value the shell returns if it is asked to run
7617 a non-existent command. The wording for the related log line suggests a
7618 non-existent command as the problem.
7620 26. If received_header_text expands to an empty string, do not add a Received:
7621 header line to the message. (Well, it adds a token one on the spool, but
7622 marks it "old" so that it doesn't get used or transmitted.)
7624 27. Installed eximstats 1.28 (addition of -nt option).
7626 28. There was no check for failure on the call to getsockname() in the daemon
7627 code. This can fail if there is a shortage of resources on the system, with
7628 ENOMEM, for example. A temporary error is now given on failure.
7630 29. Contrary to the C standard, it seems that in some environments, the
7631 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7632 program. Exim now does this explicitly; it affects the formatting of
7633 timestamps using strftime().
7635 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7636 complaints. I've now initialized all the variables, to avoid this.
7638 32. Header lines added by a system filter were not being "seen" during
7639 transport-time rewrites.
7641 33. The info_callback() function passed to OpenSSL is set up with type void
7642 (*)(SSL *, int, int), as described somewhere. However, when calling the
7643 function (actually a macro) that sets it up, the type void(*)() is
7644 expected. I've put in a cast to prevent warnings from picky compilers.
7646 34. If a DNS black list lookup found a CNAME record, but there were no A
7647 records associated with the domain it pointed at, Exim crashed.
7649 35. If a DNS black list lookup returned more than one A record, Exim ignored
7650 all but the first. It now scans all returned addresses if a particular IP
7651 value is being sought. In this situation, the contents of the
7652 $dnslist_value variable are a list of all the addresses, separated by a
7655 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7656 to accept a host name and all its aliases if the forward lookup for any of
7657 them yielded the IP address of the incoming connection. Now it accepts only
7658 those names whose forward lookup yields the correct IP address. Any other
7659 names are discarded. This closes a loophole whereby a rogue DNS
7660 administrator could create reverse DNS records to break through a
7661 wildcarded host restriction in an ACL.
7663 37. If a user filter or a system filter that ran in a subprocess used any of
7664 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7665 the wrong values were passed to the pipe command ($thisaddress had the
7666 value of $0, $0 had the value of $1, etc). This bug was introduced by
7667 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7669 38. Improved the line breaking for long SMTP error messages from ACLs.
7670 Previously, if there was no break point between 40 and 75 characters, Exim
7671 left the rest of the message alone. Two changes have been made: (a) I've
7672 reduced the minimum length to 35 characters; (b) if it can't find a break
7673 point between 35 and 75 characters, it looks ahead and uses the first one
7674 that it finds. This may give the occasional overlong line, but at least the
7675 remaining text gets split now.
7677 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7678 file descriptors might be low, and that setting 1000 would always raise it.
7679 It turns out that in some environments, the limit is already over 1000 and
7680 that lowering it causes trouble. So now Exim takes care not to decrease it.
7682 40. When delivering a message, the value of $return_path is set to $sender_
7683 address at the start of routing (routers may change the value). By an
7684 oversight, this default was not being set up when an address was tested by
7685 -bt or -bv, which affected the outcome if any router or filter referred to
7688 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7689 log only when "message" or "log_message" are set. However, if one of the
7690 conditions was an address verification, or a call to a nested ACL, the
7691 messages generated by the underlying test were being passed through. This
7692 no longer happens. The underlying message is available in $acl_verify_
7693 message for both "message" and "log_message" expansions, so it can be
7694 passed through if needed.
7696 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7697 new expansion $bh_ to give the encoded byte string without charset
7698 translation. Translation happens only if iconv() is available; HAVE_ICONV
7699 indicates this at build time. HEADERS_CHARSET gives the charset to
7700 translate to; headers_charset can change it in the configuration, and
7701 "headers charset" can change it in an individual filter file.
7703 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7704 that creates RFC 2047 encoded "words" labels them as that charset instead
7705 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7706 expansion operator; (ii) when Exim creates a From: line for a local
7707 message; (iii) when a header line is rewritten to include a "phrase" part.
7709 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7710 buggy, causing it to skip the first lines of messages whose message ID
7711 ended in 'D'. This would not have bitten before Exim release 4.14, because
7712 message IDs were unlikely to end in 'D' before then. The effect was to have
7713 incorrect size information for certain domains.
7715 45. #include "config.h" was missing at the start of the crypt16.c module. This
7716 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7719 46. If there was a timeout during a "random" callout check, Exim treated it as
7720 a failure of the random address, and carried on sending RSET and the real
7721 address. If the delay was just some slowness somewhere, the response to the
7722 original RCPT would be taken as a response to RSET and so on, causing
7723 mayhem of various kinds.
7725 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7726 when I implemented it. It didn't allow for the fact that some option values
7727 may legitimately be negative (e.g. size_addition), and it didn't even do
7728 the right test for positive values.
7730 48. Domain names in DNS records are case-independent. Exim always looks them up
7731 in lower case. Some resolvers return domain names in exactly the case they
7732 appear in the zone file, that is, they may contain uppercase letters. Not
7733 all resolvers do this - some return always lower case. Exim was treating a
7734 change of case by a resolver as a change of domain, similar to a widening
7735 of a domain abbreviation. This triggered its re-routing code and so it was
7736 trying to route what was effectively the same domain again. This normally
7737 caused routing to fail (because the router wouldn't handle the domain
7738 twice). Now Exim checks for this case specially, and just changes the
7739 casing of the domain that it ultimately uses when it transmits the message
7742 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7745 50. If a filter generated a file delivery with a non-absolute name (possible if
7746 no home directory exists for the router), the forbid_file option was not
7749 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7750 the existing equality matching.
7752 52. Exim was using ints instead of ino_t variables in some places where it was
7753 dealing with inode numbers.
7755 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7756 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7757 called TMPDIR, and if it finds it is different, it changes its value.
7759 54. The smtp_printf() function is now made available to local_scan() so
7760 additional output lines can be written before returning. There is also an
7761 smtp_fflush() function to enable the detection of a dropped connection.
7762 The variables smtp_input and smtp_batched_input are exported to
7765 55. Changed the default runtime configuration: the message "Unknown user"
7766 has been removed from the ACL, and instead placed on the localuser router,
7767 using the cannot_route_message feature. This means that any verification
7768 failures that generate their own messages won't get overridden. Similarly,
7769 the "Unrouteable address" message that was in the ACL for unverifiable
7770 relay addresses has also been removed.
7772 56. Added hosts_avoid_esmtp to the smtp transport.
7774 57. The exicyclog script was not checking for the esoteric option
7775 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7776 will work only if exicyclog is run under the appropriate euid.
7778 58. Following a discussion on the list, the rules by which Exim recognises line
7779 endings on incoming messages have been changed. The -dropcr and drop_cr
7780 options are now no-ops, retained only for backwards compatibility. The
7781 following line terminators are recognized: LF CRLF CR. However, special
7782 processing applies to CR:
7784 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7785 nor a local message in the state where . is a terminator.
7787 (ii) If a bare CR is encountered in a header line, an extra space is added
7788 after the line terminator so as not to end the header. The reasoning
7789 behind this is that bare CRs in header lines are most likely either
7790 to be mistakes, or people trying to play silly games.
7792 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7793 was being incorrectly given as 18 bytes larger than it should have been.
7794 This is a VOB (very old bug).
7796 60. This may never have affected anything current, but just in case it has:
7797 When the local host is found other than at the start of a list of hosts,
7798 the local host, those with the same MX, and any that follow, are discarded.
7799 When the list in question was part of a longer list of hosts, the following
7800 hosts (not currently being processed) were also being discarded. This no
7801 longer happens. I'm not sure if this situation could ever has previously
7804 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7807 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7808 header. This is something that is recommended in a new Internet Draft, and
7809 is something that is documented as being done by Sendmail. There are two
7810 possible values. For messages generated by the autoreply transport, Exim
7813 Auto-Submitted: auto-replied
7815 whereas for all other generated messages (e.g. bounces) it adds
7817 Auto-Submitted: auto-generated
7819 63. The "personal" condition in filters now includes a test for the
7820 Auto-Submitted: header. If it contains the string "auto-" the message it
7821 not considered personal.
7823 64. Added rcpt_include_affixes as a generic transport option.
7825 65. Added queue_only_override (default true).
7827 66. Added the syslog_duplication option.
7829 67. If what should have been the first header line of a message consisted of
7830 a space followed by a colon, Exim was mis-interpreting it as a header line.
7831 It isn't of course - it is syntactically invalid and should therefore be
7832 treated as the start of the message body. The misbehaviour could have
7833 caused a number of strange effects, including loss of data in subsequent
7834 header lines, and spool format errors.
7836 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7837 client host had authenticated. This control can now be exercised by an ACL
7838 for more flexibility.
7840 69. By default, callouts do not happen when testing with -bh. There is now a
7841 variant, -bhc, which does actually run the callout code, including
7842 consulting and updating the callout cache.
7844 70. Added support for saslauthd authentication, courtesy of Alexander
7847 71. If statvfs() failed on the spool or log directories while checking their
7848 size for availability, Exim confusingly gave the error "space shortage".
7849 Furthermore, in debugging mode it crashed with a floating point exception.
7850 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7851 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7852 serious problem, Exim now writes to the main and panic logs when this
7853 happens, with details of the failure. It then refuses to accept the
7854 incoming message, giving the message "spool directory problem" or "log
7855 directory problem" with a 421 code for SMTP messages.
7857 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7858 0, 1, and 2 exist, because some OS complain for execs without them (see
7859 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7860 descriptors. However, the code omitted to check that the open succeeded,
7861 causing mysterious errors if for some reason the permissions on /dev/null
7862 got screwed. Now Exim writes a message to the main and panic logs, and
7863 bombs out if it can't open /dev/null.
7865 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7866 interact so that it is all more flexible. It is supposed to remain
7867 backwards compatible. Also added extra_local_interfaces.
7869 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7870 to bomb out with an assertion failure - to the client this appears as a
7871 connection drop. This problem occurs in the part of the code that was taken
7872 from the Samba project. Fortunately, the assertion is in a very simple
7873 function, so I have fixed this by reproducing the function inline in the
7874 one place where it is called, and arranging for authentication to fail
7875 instead of killing the process with assert().
7877 75. The SPA client code was not working when the server requested OEM rather
7878 than Unicode encoding.
7880 76. Added code to make require_files with a specific uid setting more usable in
7881 the case where statting the file as root fails - usually a non-root-mounted
7882 NFS file system. When this happens and the failure is EACCES, Exim now
7883 forks a subprocess and does the per-uid checking as the relevant uid.
7885 77. Added process_log_path.
7887 78. If log_file_path was not explicitly set, a setting of check_log_space or
7888 check_log_inodes was ignored.
7890 79. If a space check for the spool or log partitions fails, the incident is now
7891 logged. Of course, in the latter case the data may get lost...
7893 80. Added the %p formatting code to string_format() so that it can be used to
7894 print addresses in debug_print(). Adjusted all the address printing in the
7895 debugging in store.c to use %p rather than %d.
7897 81. There was a concern that a line of code in smtp_in.c could overflow a
7898 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7899 initially expressed, the concern was not well-founded, because trailing
7900 spaces are removed early. However, if the trailing spaces were followed by
7901 a NULL, they did not get removed, so the overflow was possible. Two fixes
7904 (a) I re-wrote the offending code in a cleaner fashion.
7905 (b) If an incoming SMTP command contains a NULL character, it is rejected
7908 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7909 runs initgroups(), so that if the Exim user is in any additional groups,
7910 they will be used during message reception.
7916 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob