1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
9 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
10 related applications. Previously an "H" was used where available info
11 says that "M" should be, so change to match.
13 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
14 as arguments, so an implementation trying to copy these into a local
15 buffer was taking a taint-enforcement trap. Fix by using dynamically
16 created buffers. Similar fix for radius expansion condition.
18 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
19 reasonable, eg. to count headers. Fix by using dynamically created
20 buffers rather than a local. Do similar fixes for ACL actions "dcc",
21 "log_reject_target", "malware" and "spam"; the arguments are expanded
22 so could be handling tainted values.
24 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
25 broken the (no-op) support for this sendmail command. Restore it
26 to doing nothing, silently, and returning good status.
28 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
29 record path was given (or the default used) without a leading directory
30 path, an error occurred on trying to open it. Use the transport's working
33 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
34 transport. Previously it was the name on the DNS A-record; use instead
35 the head of the CNAME chain leading there (if there is one). This seems
36 to align better with RFC 6125.
38 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
39 smtp_accept_max_per_host allocated resources which were not released
40 when the limit was exceeded. This eventually crashed the daemon. Fix
41 by adding a relase action in that path.
43 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
44 expanded; previously using tainted values was rejected. Fix by using
45 dynamically-created buffers.
47 JH/09 Relax restrictions on ACL verify condition needing access to message
48 headers. Previously they were only permitted in data and non-smtp ACLs;
49 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
50 not_blind, header_sender and header_names_ascii verification.
52 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
53 Previously a macro used one argument twice; when called with the
54 argument as an expression having side-effects, incorrect operation
55 resulted. Use an inlineable function.
57 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
58 held open for a verify callout. Previously this wan not accounted for
59 and a corrupt onward SMTP conversation resulted.
61 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
62 Previously they were embedded naively in the constructed address; when
63 needed, strip the quoting and quote the entire local_part.
64 Also make the inbound_srs expansion condition handle quoting.
66 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
67 excluded, not matching the documentation.
69 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
70 was given for the sqlite_dbfile a trap resulted.
72 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
73 "name" argument resulted in a trap. There is no reason to disallow such;
74 this was a coding error.
76 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
77 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
78 spent suspended, ignoring the Posix definition. Previously we assumed
79 it did and a constant offset from real time could be used as a correction.
80 Change to using the same clock source for the start-of-message and the
81 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
82 exists, just to get a clock slightly more aligned to reality.
84 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
85 RFC says it is optional some validators care. The missing char was not
86 intended but triggered by a line-wrap alignement. Discovery and fix by
87 Guillaume Outters, hacked on by JH.
89 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
90 name being quoted was tainted a trap would be taken. Fix by using
91 dynamicaly created buffers. The routine could have been called by a
92 rewrite with the "h" flag, by using the "-F" command-line option, or
93 by using a "name=" option on a control=submission ACL modifier.
95 JH/19 SPF: change the Authentication-Results expansion component to give
96 smtp.helo when the sender domain is empty. Previously it gave
99 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
100 not in 127.0.0.0/8 to help in spotting list domains taken over by a
101 domain-parking registrar.
103 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
104 Previously when a whitespace character was specified it was not inserted
105 after removing the newline.
107 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
108 the domain part of the recipient address. This overrides any tls_sni
109 option set, which was previously used.
111 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
114 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
115 is_tainted() had an off-by-one error in the overenthusiastic direction.
116 Find and fix by Gavan. Although NetBSD is not a supported platform for
117 4.94 this bug could affect other platforms.
119 PP/01 Fix default prime selection to be consistent.
120 One path used ike23 still, instead of exim.dev.20160529.3; now both
121 execution flows will use the same DH primes (currently
122 exim.dev.20160529.3).
124 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
125 option in smtp transport, to match the documentation. Previously
126 verification was not being done in some cases where it should have been.
128 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
129 than one server was defined and depending on the platform memory layout
130 details, an internal consistency trap could be hit while walking the list
133 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
134 files. The value is used by the authresults expansion item. Previously
135 if this was used in a router or transport, a crash could result.
137 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
138 have a sufficient incidence of EINTR returns from select that an
139 interaction with other operations done by the main daemon loop exposed
140 a bug in the error-handling. This was benign apart from the log
147 JH/01 Avoid costly startup code when not strictly needed. This reduces time
148 for some exim process initialisations. It does mean that the logging
149 of TLS configuration problems is only done for the daemon startup.
151 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
153 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
154 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
157 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
158 DKIM signing is being done. Previously a transport_filter would always
159 disable CHUNKING, falling back to traditional DATA.
161 JH/05 Regard command-line recipients as tainted.
163 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
165 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
166 PAM library frees one of the arguments given to it, despite the
167 documentation. Therefore a plain malloc must be used.
169 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
170 on-stack buffers were used, resulting in a taint trap when DSN information
171 copied from a received message was written into the buffer.
173 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
174 the ordering of its ARC headers. This caused a crash.
176 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
177 a new record was being constructed with information from the peer, a trap
180 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
181 installation would get error messages from DMARC verify, when it hit the
182 nonexistent file indicated by the default. Distros wanting DMARC enabled
183 should both provide the file and set the option.
184 Also enforce no DMARC verification for command-line sourced messages.
186 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
187 could, depending on the platform, hang at the STARTTLS response.
189 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
190 message on a connection. Previously if one message had ARC headers and
191 the following one did not, a crash could result when adding an
192 Authentication-Results: header.
194 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
195 Exim main code and Exim-related utities. The introduction of taint
196 tracking also did many adjustments to string handling. Since then, eximon
197 frequently terminated with an assert failure.
199 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
200 check for 452 responses. This slightly helps the inefficieny of doing
201 a large alias-expansion into a recipient-limited target. The max_rcpt
202 transport option still applies (and at the current default, will override
203 the new feature). The check is done for either cause of synch, and forces
204 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
205 connection. The new facility is not tunable at this time.
207 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
208 library live data was being used, so the results became garbage. Make
209 copies while it is still usable.
211 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
212 on delivery deferred (==) and failed (**) lines (if a delivery was
213 attemtped). Previously it was only on completion (=>) lines.
215 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
216 for the expansion of the server_scram_iter and server_scram_salt options.
218 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
219 are now specifically given a NO_DATA response without hitting the system
220 resolver. The library goes on to do the now-standard TXT lookup.
221 Use of dnsdb lookups is not affected.
223 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
224 only retrieve the errormessage once. Previously two calls to dlerror()
225 were used, and the second one (for mainlog/paniclog) retrieved null
228 JH/20 Taint checking: disallow use of tainted data for
229 - the appendfile transport file and directory options
230 - the pipe transport command
231 - the autoreply transport file, log and once options
232 - file names used by the redirect router (including filter files)
234 - paths used by single-key lookups
235 Previously this was permitted.
237 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
238 adjusted the size of a major service buffer; this failed because the
239 buffer was in use at the time. Change to a compile-time increase in the
240 buffer size, when this authenticator is compiled into exim.
242 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
243 previous fast-mode was untenable in the face of glibs using mmap to
244 support larger malloc requests.
246 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
247 New values supported, if defined on system where compiled:
248 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
249 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
251 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
252 running a limited number of proceses in parallel, a benefit is gained. The
253 amount varies with the platform hardware and load. The use of the option
254 queue_run_in_order means we cannot do this, as ordering becomes
257 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
258 had introduced a string-copy (for ensuring NUL-termination) which was not
259 appropriate for that case, which can include embedded NUL bytes in the
260 block of data. Investigation showed the copy to actually be needless, the
261 data being length-specified.
263 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
264 done during a receiving connection, and both used TLS, global info was
265 used rather than per-connection info for tracking the state of data
266 queued for transmission. This could result in a connection hang.
268 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
269 Previously, when delivering serveral messages down a single connection
270 only the first would provide a SIZE. This was due to the size information
271 not being properly tracked.
273 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
274 TAI (at 37 seconds currently), pretend to be in UTC for time-related
275 expansion and logging. Previously, spurious values such as a future
276 minute could be seen.
278 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
279 it could crash from a null-deref. This could also affect the
280 ${addresses: } operator and ${readsock } item.
282 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
283 message following a mime one, the variable was not reset.
285 JH/30 When an pipelined-connect fails at the first response, assume incorrect
286 cached capability (perhaps the peer reneged?) and immediately retry in
289 JH/31 Fix spurious detection of timeout while writing to transport filter.
291 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
292 an attempt to copy the string was made before checking it.
294 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
295 taint of the lookup key was maintained; we now regard the presence in the
296 filesystem as sufficient validation.
298 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
299 argument is supplied.
301 JH/35 The dsearch lookup now requires that the directory is an absolute path.
302 Previously this was not checked, and nonempty relative paths made an
303 access under Exim's current working directory.
305 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
306 Previously no event was raised.
308 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
309 parameter supplied by the sender MAIL FROM command. Previously it was
310 ignored, and only the check_spool_space option value for the required
313 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
314 the size of the signing public-key. Previously it was instead giving
315 the size of the signature hash.
317 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
318 the default. See the (new) dkim_verify_min_keysizes option.
320 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
321 and an ACL use a lookup for checking either the local_part or domain,
322 stale data could be accessed. Ensure that variable references are
323 dropped between messages.
325 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
326 by the client was not checked as pointing within response data before
327 being used. A malicious client could thus cause an out-of-bounds read and
328 possibly gain authentication. Fix by adding the check.
330 JH/42 Internationalisation: change the default for downconversion in the smtp
331 transport to be "if needed". Previously it was "as previously set" for
332 the message, which usually meant "if needed" for message-submission but
333 "no" for everything else. However, MTAs have been seen using SMTPUTF8
334 even when the envelope addresses did not need it, resulting in forwarding
335 failures to non-supporting MTAs. A downconvert in such cases will be
336 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
337 transport. The change does mean that addresses needing conversion will
338 be converted when previously a delivery failure would occur.
340 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
341 response was received it would be used unchecked in a fail-DSN, violating
342 standards on line-length limits. Truncate if needed.
344 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
345 doesn't work. (Additionally add a new main config option to configure the
352 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
353 side, to decode a TLS 1.3 packet capture.
355 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
356 Previously the default library behaviour applied, sending two, each in
359 JH/03 Debug output for ACL now gives the config file name and line number for
362 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
364 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
366 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
367 buffer overrun for (non-chunking) other transports.
369 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
370 TLS1.3, means that a server rejecting a client certificate is not visible
371 to the client until the first read of encrypted data (typically the
372 response to EHLO). Add detection for that case and treat it as a failed
373 TLS connection attempt, so that the normal retry-in-clear can work (if
374 suitably configured).
376 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
377 and/or domain. Found and fixed by Jason Betts.
379 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
380 configuration). If a CNAME target was not a wellformed name pattern, a
383 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
384 the OS reports them interleaved with other addresses.
386 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
387 used both for input and for a verify callout, both encrypted, SMTP
388 responses being sent by the server could be lost. This resulted in
389 dropped connections and sometimes bounces generated by a peer sending
392 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
393 string. Previously it was possible to cause undefined behaviour in a
394 library routine (usually a crash). Found by "zerons".
396 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
399 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
400 API was removed, so update to use the newer ones.
402 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
403 any timeout set, is taking a long time. Previously we would hang on to a
404 rotated logfile "forever" if the input was arriving with long gaps
405 (a previous attempt to fix addressed lack, for a long time, of initial
408 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
409 shared (NFS) environment. The length of the tempfile name is now
410 4 + 16 ("hdr.$message_exim_id") which might break on file
411 systems which restrict the file name length to lower values.
414 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
415 shared (NFS) environment.
417 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
418 did for all versions <4.90). Notably -M, -m, --invert, -I may be
421 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
422 on some platforms for bit 31.
424 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
425 to changes apparently associated with TLS1.3 handling some of the APIs
426 previously used were either nonfunctional or inappropriate. Strings
427 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
428 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
429 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
430 This affects log line X= elements, the $tls_{in,out}_cipher variables,
431 and the use of specific cipher names in the encrypted= ACL condition.
433 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
435 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
436 verification result was not updated unless hosts_require_ocsp applied.
438 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
439 queue_list_requires_admin set to false, non-admin users were denied the
442 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
443 directory-of-certs mode. Previously they were advertised despite the
446 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
447 A single TCP connection by a client will now hold a TLS connection open
448 for multiple message deliveries, by default. Previously the default was to
451 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
452 default. If built with the facility, DANE will be used. The facility
453 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
455 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
456 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
457 must be defined and you must still, unless you define DISABLE_TLS, manage
458 the the include-dir and library-file requirements that go with that
459 choice. Non-TLS builds are still supported.
461 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
464 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
465 default. If the platform supports and has the facility enabled, it will
466 be requested on all coneections.
468 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
469 controlled by the build-time option SUPPORT_PIPE_CONNECT.
471 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
473 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
474 success-DSN messages. Previously the From: header was always the default
475 one for these; the option was ignored.
477 JH/28 Fix the timeout on smtp response to apply to the whole response.
478 Previously it was reset for every read, so a teergrubing peer sending
479 single bytes within the time limit could extend the connection for a
480 long time. Credit to Qualsys Security Advisory Team for the discovery.
482 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
483 delivery address, which leaked information of the results of local
484 forwarding. Change to the original envelope recipient address, per
487 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
488 requested. Previously not bounce was generated and a log entry of
489 error ignored was made.
491 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
493 JH/32 Introduce a general tainting mechanism for values read from the input
494 channel, and values derived from them. Refuse to expand any tainted
495 values, to catch one form of exploit.
497 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
498 was unused and the unexpanded text used for the test. Found and
499 fixed by Ruben Jenster.
501 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
502 an attempt to use a TLS library read routine dereffed a nul pointer,
505 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
506 them in our smtp response.
508 JH/36 Have the general router option retry_use_local_part default to true when
509 any of the restrictive preconditions are set (to anything). Previously it
510 was only for check_local user. The change removes one item of manual
511 configuration which is required for proper retries when a remote router
512 handles a subset of addresses for a domain.
514 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
515 link count into consideration.
517 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
518 caused the extension of big_buffer, the following lines were ignored.
520 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
521 accordance with RFC 2308. Previously there was no expiry, so a longlived
522 receive process (eg. due to ACL delays) versus a short SOA value could
525 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
527 JH/39 Promote DMARC support to mainline.
529 JH/40 Bug 2452: Add a References: header to DSNs.
531 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
532 parameters. The relevant library call is documented as "Deprecated: This
533 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
534 3.6.0, DH parameters are negotiated following RFC7919."
536 HS/06 Change the default of dnssec_request_domains to "*"
538 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
539 carried on and emitted a BDAT command, even when PIPELINING was not
542 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
543 buffer was used for the filename, resulting in a trap when tainted
544 arguments (eg. $domain) were used.
546 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
547 recommended to avoid a possible server-load attack. The feature can be
548 re-enabled via the openssl_options main cofiguration option.
550 JH/45 local_scan API: documented the current smtp_printf() call. This changed
551 for version 4.90 - adding a "more data" boolean to the arguments.
552 Bumped the ABI version number also, this having been missed previously;
553 release versions 4.90 to 4.92.3 inclusive were effectively broken in
554 respect of usage of smtp_printf() by either local_scan code or libraries
555 accessed via the ${dlfunc } expansion item. Both will need coding
556 adjustment for any calls to smtp_printf() to match the new function
557 signature; a FALSE value for the new argument is always safe.
559 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
560 the file-offset (which the Linux syscall does, and exim expects); this
561 resulted in an indefinite loop.
563 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
564 to do ARC verification. The Authentication-Results: header line added
565 by the configuration then had no ARC item.
571 JH/01 Remove code calling the customisable local_scan function, unless a new
572 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
574 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
575 non-signal-safe functions being used.
577 JH/03 Bug 2269: When presented with a received message having a stupidly large
578 number of DKIM-Signature headers, disable DKIM verification to avoid
579 a resource-consumption attack. The limit is set at twenty.
581 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
582 report of oldest_pass in ${authres } in consequence, and separate out
583 some descriptions of reasons for verification fail.
585 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
586 files in the spool were present and unlocked. A queue-runner could spot
587 them, resulting in a duplicate delivery. Fix that by doing the unlock
588 after the unlink. Investigation by Tim Stewart. Take the opportunity to
589 add more error-checking on spoolfile handling while that code is being
592 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
593 No known attacks, no CVE, this is defensive hardening.
595 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
596 a queue-runner could start a delivery while other operations were ongoing.
597 Cutthrough delivery was a common victim, resulting in duplicate delivery.
598 Found and investigated by Tim Stewart. Fix by using the open message data
599 file handle rather than opening another, and not locally closing it (which
600 releases a lock) for that case, while creating the temporary .eml format
601 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
603 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
604 $sender_verify_failure/$recipient_verify_failure to "random".
606 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
609 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
610 Previously this would segfault.
612 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
615 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
616 like zero, since the resolver should be doing this for us, But we need one
617 as a CNAME but no MX presence gets the CNAME returned; we need to check
618 that doesn't point to an MX to declare it "no MX returned" rather than
619 "error, loop". A new main option is added so the older capability of
620 following some limited number of chain links is maintained.
622 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
624 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
625 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
626 now-deprecated earlier definitions used only specified the range up to TLS
627 1.2 (in the older-version library docs).
629 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
631 JH/15 Rework TLS client-side context management. Stop using a global, and
632 explicitly pass a context around. This enables future use of TLS for
633 connections to service-daemons (eg. malware scanning) while a client smtp
634 connection is using TLS; with cutthrough connections this is quite likely.
636 JH/16 Fix ARC verification to do AS checks in reverse order.
638 JH/17 Support a "tls" option on the ${readsocket } expansion item.
640 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
641 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
642 Previously the "utf8" would be re-prepended for every additional message.
644 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
645 Previously thery were accepted, resulting in issues when attempting to
646 forward messages to a non-supporting MTA.
648 PP/02 Let -n work with printing macros too, not just options.
650 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
651 one parent address was copied, and bogus data was used at delivery-logging
652 time. Either a crash (after delivery) or bogus log data could result.
653 Discovery and analysis by Tim Stewart.
655 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
656 Previously if the string ended mid-character, we did not insert the
657 promised '?' replacement.
659 PP/04 Documentation: current string operators work on bytes, not codepoints.
661 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
662 should pack well giving a smaller memory footprint so better caching and
663 therefore performance. Group the declarations where this can't be done so
664 that the byte-sized flag variables are not interspersed among pointer
665 variables, giving a better chance of good packing by the compiler.
667 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
668 non-null, to avoid issues with sites running BATV. Previously reports were
669 sent with an empty envelope sender so looked like bounces.
671 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
672 The ignore_error flag wasn't being returned from the filter subprocess so
673 was not set for later routers. Investigation and fix by Matthias Kurz.
675 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
676 and a msg:complete for the whole, when a message is manually removed using
677 -Mrm. Developement by Matthias Kurz, hacked on by JH.
679 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
680 a "Gnu special" function, asprintf() in the DB utility binary builds; I
681 hope that is portable enough.
683 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
684 requiring a known-CA anchor certificate; make it now rely entirely on the
685 TLSA as an anchor. Checking the name on the leaf cert against the name
686 on the A-record for the host is still done for TA (but not for EE mode).
688 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
689 included in delivery lines for non-proxied connections, when compiled with
690 SUPPORT_SOCKS and running with proxy logging enabled.
692 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
693 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
694 move the existing event to fire before the normal logging of message
695 failure so that custom logging is bracketed by normal logging.
697 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
698 msg:fail:internal event. Developement by Matthias Kurz.
700 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
701 far too small for todays use of crypto signatures stored there. Go all
702 the way to the max DNS message size of 64kB, even though this might be
703 overmuch for IOT constrained device use.
705 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
706 copy a string over itself. The library routine is documented as not
707 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
709 JH/32 For main options check_spool_space and check_inode_space, where the
710 platform supports 64b integers, support more than the previous 2^31 kB
711 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
712 the previous G, M, k.
714 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
715 $authenticated_fail_id variable on authentication failure. Previously
718 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
719 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
720 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
721 GNUTLS_SEC_PARAM_MEDIUM.
723 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
724 side. Previously we would continue as if no SNI had been received.
726 JH/36 Harden the handling of string-lists. When a list consisted of a sole
727 "<" character, which should be a list-separator specification, we walked
728 off past the nul-terimation.
730 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
731 causes) even when the retry time is not yet met. Previously they were
732 not, meaning that when (say) an account was over-quota and temp-rejecting,
733 and multiple senders' messages were queued, only one sender would get
734 notified on each configured delay_warning cycle.
736 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
738 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
739 cipher-suites, an error can be left on the stack even for a succeeding
740 accept; this results in impossible error messages when a later operation
743 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
744 return error codes indicating retry. Under TLS1.3 this becomes required.
746 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
747 it only wrote the new authenticators, resulting in a lack of tracking of
748 peer changes of ESMTP extensions until the next cache flush.
750 JH/41 Fix the loop reading a message header line to check for integer overflow,
751 and more-often against header_maxsize. Previously a crafted message could
752 induce a crash of the recive process; now the message is cleanly rejected.
754 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
755 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
761 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
762 When redis_servers is set to a list of > 1 element, and the Redis servers
763 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
764 case of MOVED into a DEFER case instead, thus moving the query onto the
765 next server in the list. For a cluster of N elements, all N servers must
766 be defined in redis_servers.
768 GF/02 Catch and remove uninitialized value warning in exiqsumm
769 Check for existence of @ARGV before looking at $ARGV[0]
771 JH/01 Replace the store_release() internal interface with store_newblock(),
772 which internalises the check required to safely use the old one, plus
773 the allocate and data copy operations duplicated in both (!) of the
774 extant use locations.
776 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
777 modifier. This matches the restriction on the commandline.
779 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
780 Previously only the last row was returned.
782 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
783 we assumed that tags in the header were well-formed, and parsed the
784 element content after inspecting only the first char of the tag.
785 Assumptions at that stage could crash the receive process on malformed
788 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
789 While running the DKIM ACL we operate on the Permanent memory pool so that
790 variables created with "set" persist to the DATA ACL. Also (at any time)
791 DNS lookups that fail create cache records using the Permanent pool. But
792 expansions release any allocations made on the current pool - so a dnsdb
793 lookup expansion done in the DKIM ACL releases the memory used for the
794 DNS negative-cache, and bad things result. Solution is to switch to the
795 Main pool for expansions.
796 While we're in that code, add checks on the DNS cache during store_reset,
797 active in the testsuite.
798 Problem spotted, and debugging aided, by Wolfgang Breyha.
800 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
801 When none of the hosts presented to a transport match an already-open
802 connection, close it and proceed with the list. Previously we would
803 queue the message. Spotted by Lena with Yahoo, probably involving
806 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
807 Previously a spurious "250 OK id=" response was appended to the proper
810 JH/08 The "support for" informational output now, which built with Content
811 Scanning support, has a line for the malware scanner interfaces compiled
812 in. Interface can be individually included or not at build time.
814 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
815 by the template makefile "src/EDITME". The "STREAM" support for an older
816 ClamAV interface method is removed.
818 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
819 rows affected is given instead).
821 JH/11 The runtime Berkeley DB library version is now additionally output by
822 "exim -d -bV". Previously only the compile-time version was shown.
824 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
825 SMTP connection. Previously, when one had more recipients than the
826 first, an abortive onward connection was made. Move to full support for
827 multiple onward connections in sequence, handling cutthrough connection
828 for all multi-message initiating connections.
830 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
831 routers. Previously, a multi-recipient message would fail to match the
832 onward-connection opened for the first recipient, and cause its closure.
834 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
835 a timeout on read on a GnuTLS initiating connection, resulting in the
836 initiating connection being dropped. This mattered most when the callout
837 was marked defer_ok. Fix to keep the two timeout-detection methods
840 JH/15 Relax results from ACL control request to enable cutthrough, in
841 unsupported situations, from error to silently (except under debug)
842 ignoring. This covers use with PRDR, frozen messages, queue-only and
845 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
847 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
848 metadata, resulting in a crash in free().
850 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
851 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
852 Broken also in d185889f4, with init system revamp.
854 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
855 Previously we abruptly closed the connection after reading a malware-
856 found indication; now we go on to read the "scan ok" response line,
859 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
860 ACL. Previously, a crash would result.
862 JH/19 Speed up macro lookups during configuration file read, by skipping non-
863 macro text after a replacement (previously it was only once per line) and
864 by skipping builtin macros when searching for an uppercase lead character.
866 JH/20 DANE support moved from Experimental to mainline. The Makefile control
867 for the build is renamed.
869 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
870 was allocated for every new TLS startup, meaning one per message. Fix
871 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
873 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
874 reported the original. Fix to report (as far as possible) the ACL
875 result replacing the original.
877 JH/23 Fix memory leak during multi-message connections using STARTTLS under
878 OpenSSL. Certificate information is loaded for every new TLS startup,
879 and the resources needed to be freed.
881 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
883 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
886 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
887 DATA response info to the (existing) per-recipient response info for
888 the "C=" log element. It can have useful tracking info from the
889 destination system. Patch from Simon Arlott.
891 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
892 length value. Previously this would segfault.
894 HS/02 Support Avast multiline protoocol, this allows passing flags to
895 newer versions of the scanner.
897 JH/28 Ensure that variables possibly set during message acceptance are marked
898 dead before release of memory in the daemon loop. This stops complaints
899 about them when the debug_store option is enabled. Discovered specifically
900 for sender_rate_period, but applies to a whole set of variables.
901 Do the same for the queue-runner and queue-list loops, for variables set
902 from spool message files. Do the same for the SMTP per-message loop, for
903 certain variables indirectly set in ACL operations.
905 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
906 as a multi-recipient message from a mailinglist manager). The coding had
907 an arbitrary cutoff number of characters while checking for more input;
908 enforced by writing a NUL into the buffer. This corrupted long / fast
909 input. The problem was exposed more widely when more pipelineing of SMTP
910 responses was introduced, and one Exim system was feeding another.
911 The symptom is log complaints of SMTP syntax error (NUL chars) on the
912 receiving system, and refused recipients seen by the sending system
913 (propating to people being dropped from mailing lists).
914 Discovered and pinpointed by David Carter.
916 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
917 replaced by the ${authresults } expansion.
919 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
921 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
922 allows proper process termination in container environments.
924 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
925 Previously the "final dot" had a newline after it; ensure it is CR,LF.
927 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
928 and "err_perm", deprecated since 4.83 when the RFC-defined words
929 "temperror" and "permerror" were introduced.
931 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
932 transport-filters or DKIM-signing. The restriction was lost in the
933 consolidation of verify-callout and delivery SMTP handling.
934 Extend the restriction to also cover ARC-signing.
936 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
937 in defer=pass mode supply a 450 to the initiator. Previously the message
940 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
941 tls_require_ciphers is used as before.
943 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
944 "pass_unscanned". Only tmpfails from the scanner are written to
945 the paniclog, as they may require admin intervention (permission
946 denied, license issues). Other scanner errors (like decompression
947 bombs) do not cause a paniclog entry.
949 JH/36 Fix reinitialisation of DKIM logging variable between messages.
950 Previously it was possible to log spurious information in receive log
953 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
954 triggered odd behaviour from Outlook Express clients.
956 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
959 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
960 since the IETF WG has not yet settled on that versus the original
961 "bare" representation.
963 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
964 Previously the millisecond value corrupted the output.
965 Fix also for syslog_pid=no and log_selector +pid, for which the pid
966 corrupted the output.
972 JH/01 Rework error string handling in TLS interface so that the caller in
973 more cases is responsible for logging. This permits library-sourced
974 string to be attached to addresses during delivery, and collapses
975 pairs of long lines into single ones.
977 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
978 during configuration. Wildcards are allowed and expanded.
980 JH/02 Rework error string handling in DKIM to pass more info back to callers.
981 This permits better logging.
983 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
984 do not close it down and have the child transport start it up again on
985 the passed-on TCP connection. Instead, proxy the child (and any
986 subsequent ones) for TLS via a unix-domain socket channel. Logging is
987 affected: the continued delivery log lines do not have any DNSSEC, TLS
988 Certificate or OCSP information. TLS cipher information is still logged.
990 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
991 identical IP addresses on different listening ports. Will also affect
994 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
995 add noisy ifdef guards to special-case this sillyness.
996 Patch from Bernd Kuhls.
998 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
999 than 255 are no longer allowed.
1001 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1002 Disable the session-cache too, which might reduce our load. Since we
1003 currrectly use a new context for every connection, both as server and
1004 client, there is no benefit for these.
1005 GnuTLS appears to not support tickets server-side by default (we don't
1006 call gnutls_session_ticket_enable_server()) but client side is enabled
1007 by default on recent versions (3.1.3 +) unless the PFS priority string
1010 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1011 <https://reproducible-builds.org/specs/source-date-epoch/>.
1013 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1014 the check for any unsuccessful recipients did not notice the limit, and
1015 erroneously found still-pending ones.
1017 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1018 MSG_MORE. Only in-clear (not on TLS connections).
1020 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1021 transport-filter is involved we can buffer the headers in memory for
1022 creating the signature, and read the spool data file once for the
1023 signature and again for transmission.
1025 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1026 4.77 as the kernel support then wasn't solid, having issues in 64bit
1027 mode. Now, it's been long enough. Add support for FreeBSD also.
1029 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1030 case where the routing stage had gathered several addresses to send to
1031 a host before calling the transport for the first, we previously failed
1032 to close down TLS in the old transport process before passing the TCP
1033 connection to the new process. The new one sent a STARTTLS command
1034 which naturally failed, giving a failed delivery and bloating the retry
1035 database. Investigation and fix prototype from Wolfgang Breyha.
1037 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1038 false-negatives in the check that the sender had not preempted a response
1039 or prompt from Exim (running as a server), due to that code's lack of
1040 awareness of the SMTP input buffering.
1042 PP/04 Add commandline_checks_require_admin option.
1043 Exim drops privileges sanely, various checks such as -be aren't a
1044 security problem, as long as you trust local users with access to their
1045 own account. When invoked by services which pass untrusted data to
1046 Exim, this might be an issue. Set this option in main configuration
1047 AND make fixes to the calling application, such as using `--` to stop
1050 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1051 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1052 responses to those, into a single TLS record each way (this usually means
1053 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1056 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1057 affects you only if you're dancing at the edge of the param size limits.
1058 If you are, and this message makes sense to you, then: raise the
1059 configured limit or use OpenSSL 1.1. Nothing we can do for older
1062 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1063 cmdline element to get the documented default one. Previously it was
1066 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1067 are unsuable for non-message-specific reasons - eg. connection timeout,
1068 banner-time rejection.
1070 JH/16 Fix logging of delivery remote port, when specified by router, under
1073 PP/06 Repair manualroute's ability to take options in any order, even if one
1074 is the name of a transport.
1077 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1079 JH/17 Change the list-building routines interface to use the expanding-string
1080 triplet model, for better allocation and copying behaviour.
1082 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1083 Previously it was constructed the first time a possibly-matching string
1084 was met in the configuration file input during startup; now it is done
1087 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1088 compatible one, to avoid the (poorly documented) possibility of a config
1089 file in the working directory redirecting the DB files, possibly correpting
1090 some existing file. CVE-2017-10140 assigned for BDB.
1092 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1093 cache-hot. Previously, although the result was properly cached, the
1094 initial verify call returned a defer.
1096 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1097 the main verify for receipient in uncached-mode.
1099 JH/22 Retire historical build files to an "unsupported" subdir. These are
1100 defined as "ones for which we have no current evidence of testing".
1102 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1103 if present. Previously it was ignored.
1105 JH/24 Start using specified-initialisers in C structure init coding. This is
1106 a C99 feature (it's 2017, so now considered safe).
1108 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1109 if was a fixed-sized field and bitmask ops via macros; it is now more
1112 PP/07 GitHub PR 56: Apply MariaDB build fix.
1113 Patch provided by Jaroslav Škarvada.
1115 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1116 during Coverity cleanups [4.87 JH/47]
1117 Diagnosis and fix provided by Michael Fischer v. Mollard.
1119 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1120 the right size to place the terminating semicolon on its own folded
1121 line, the header hash was calculated to an incorrect value thanks to
1122 the (relaxed) space the fold became.
1124 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1125 and confused the parent.
1127 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1128 which could crash as a result. This could lead to undeliverable messages.
1130 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1133 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1134 matters on fast-turnover and PID-randomising systems, which were getting
1135 out-of-order delivery.
1137 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1138 a possibly-overlapping copy. The symptom was that "Remote host closed
1139 connection in response to HELO" was logged instead of the actual 4xx
1142 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1143 Previously only that bufferd was discarded, resulting in SYMTP command
1146 JH/32 DKIM: when a message has multiple signatures matching an identity given
1147 in dkim_verify_signers, run the dkim acl once for each. Previously only
1148 one run was done. Bug 2189.
1150 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1151 "panic the current process" to "deliberately defer". The panic log is
1152 still written with the problem list name; the mail and reject logs now
1153 get a temp-reject line for the message that was being handled, saying
1154 something like "domains check lookup or other defer". The SMTP 451
1155 message is still "Temporary local problem".
1157 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1158 A crafted sequence of BDAT commands could result in in-use memory beeing
1159 freed. CVE-2017-16943.
1161 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1162 from SMTP input. Previously it was always done; now only done for DATA
1163 and not BDAT commands. CVE-2017-16944.
1165 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1166 to the message (such as an overlong header line). Previously this was
1167 not done and we did not exit BDAT mode. Followon from the previous item
1168 though a different problem.
1174 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1175 than -2003 did; needs libidn2 in addition to libidn.
1177 JH/02 The path option on a pipe transport is now expanded before use.
1179 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1180 Patch provided by "Björn", documentation fix added too.
1182 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1183 missing a wire-to-host endian conversion.
1185 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1186 close after a BDAT command line could be taken as a following command,
1187 giving a synch failure. Fix by only checking for synch immediately
1188 before acknowledging the chunk.
1190 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1191 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1192 macro. Patches provided by Josh Soref.
1194 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1195 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1196 by IANA as a service extension. However, John Klensin suggests that we
1199 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1200 the dkim code may be unix-mode line endings rather than smtp wire-format
1201 CRLF, so prepend a CR to any bare LF.
1203 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1204 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1206 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1207 the first were themselves being wrongly included in the feed into dkim
1208 processing; with most chunk sizes in use this resulted in an incorrect
1209 body hash calculated value.
1211 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1212 DKIM signature block, for verification. Although advised against by
1213 standards it is specifically not ruled illegal.
1215 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1217 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1218 missing a body hash (the bh= tag).
1220 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1221 It seems that HAProxy sends the Proxy Protocol information in clear and
1222 only then does a TLS startup, so do the same.
1224 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1225 TCP connections (such as for Spamd) unless the daemon successfully set
1226 Fast Open mode on its listening sockets. This fixes breakage seen on
1227 too-old kernels or those not configured for Fast Open, at the cost of
1228 requiring both directions being enabled for TFO, and TFO never being used
1229 by non-daemon-related Exim processes.
1231 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1232 endings, at least on the first header line. Try to canonify any that get
1233 past that check, despite the cost.
1235 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1236 now limited to an arbitrary five deep, while parsing addresses with the
1237 strip_excess_angle_brackets option enabled.
1239 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1240 instead leave the unprompted TLS handshake in socket buffer for the
1241 TLS library to consume.
1243 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1245 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1247 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1248 region in one operation, for speed, and this leaves assigned pointers
1249 dangling. Add checks run only under the testsuite which checks all
1250 variables at a store-reset and panics on a dangling pointer; add code
1251 explicitly nulling out all the variables discovered. Fixes one known
1252 bug: a transport crash, where a dangling pointer for $sending_ip_address
1253 originally assigned in a verify callout, is re-used.
1255 PP/06 Drop '.' from @INC in various Perl scripts.
1257 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1259 PP/08 Reduce a number of compilation warnings under clang; building with
1260 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1261 should be warning-free.
1263 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1265 HS/01 Fix portability problems introduced by PP/08 for platforms where
1266 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1268 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1269 chunk. This allows us to accept broken chunked messages. We need a more
1270 general solution here.
1272 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1273 already-broken messages in the queue.
1275 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1277 JH/19 Fix reference counting bug in routing-generated-address tracking.
1283 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1284 supports it and a size is available (ie. the sending peer gave us one).
1286 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1287 years of being deprecated). The replacements are the ACLs
1288 acl_smtp_mime and acl_not_smtp_mime.
1290 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1291 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1292 or even an in-clear connection were permitted. Now, if the host lookup
1293 was dnssec and dane was requested then the host is only used if the
1294 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1295 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1296 if one fails this test.
1297 This means that a poorly-configured remote DNS will make it incommunicado;
1298 but it protects against a DNS-interception attack on it.
1300 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1301 non-noisy when a race steals the message being considered.
1303 JH/05 If main configuration option tls_certificate is unset, generate a
1304 self-signed certificate for inbound TLS connections.
1306 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1307 in rewrites and routers.
1309 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1310 and logged a warning sing 4.83; now they are a configuration file error.
1312 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1313 (lacking @domain). Apply the same qualification processing as RCPT.
1315 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1317 JH/10 Support ${sha256:} applied to a string (as well as the previous
1320 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1321 a cutthrough deliver is pending, as we always want to make a connection.
1322 This also avoids re-routing the message when later placing the cutthrough
1323 connection after a verify cache hit.
1324 Do not update it with the verify result either.
1326 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1327 when routing results in more than one destination address.
1329 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1330 signing (which inhibits the cutthrough capability). Previously only
1331 the presence of an option was tested; now an expansion evaluating as
1332 empty is permissible (obviously it should depend only on data available
1333 when the cutthrough connection is made).
1335 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1336 the relevant preceding SMTP command did not note the pipelining mode.
1338 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1339 Previously they were not counted.
1341 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1342 as one having no matching records. Previously we deferred the message
1343 that needed the lookup.
1345 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1346 distinguished as "(=".
1348 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1349 for missing MX records. Previously it only worked for missing A records.
1351 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1353 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1354 after the data-go-ahead and data-ack. Patch from Jason Betts.
1356 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1357 even for a "none" policy. Patch from Tony Meyer.
1359 JH/22 Fix continued use of a connection for further deliveries. If a port was
1360 specified by a router, it must also match for the delivery to be
1363 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1364 When one of the recipients of a message was unsuitable for the connection
1365 (has no matching addresses), we lost track of needing to mark it
1366 deferred. As a result mail would be lost.
1368 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1370 JH/25 Decoding ACL controls is now done using a binary search; the source code
1371 takes up less space and should be simpler to maintain. Merge the ACL
1372 condition decode tables also, with similar effect.
1374 JH/26 Fix problem with one_time used on a redirect router which returned the
1375 parent address unchanged. A retry would see the parent address marked as
1376 delivered, so not attempt the (identical) child. As a result mail would
1379 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1380 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1381 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1384 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1385 Main config options check_{log,spool}_{inodes,space} are now
1386 100 inodes, 10MB unless set otherwise in the configuration.
1388 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1389 Previously it only applied to the main-section connection policy
1392 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1394 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1395 by me. Added RFC7919 DH primes as an alternative.
1397 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1398 are not in the system include path.
1400 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1401 GnuTLS, when a session startup failed (eg because the client disconnected)
1402 Exim did stdio operations after fclose. This was exposed by a recent
1403 change which nulled out the file handle after the fclose.
1405 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1406 signed directly by the cert-signing cert, rather than an intermediate
1407 OCSP-signing cert. This is the model used by LetsEncrypt.
1409 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1411 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1412 an incoming connection.
1414 HS/02 Bug 1802: Do not half-close the connection after sending a request
1417 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1418 fallback to "prime256v1".
1420 JH/34 SECURITY: Use proper copy of DATA command in error message.
1421 Could leak key material. Remotely exploitable. CVE-2016-9963.
1427 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1428 and 3.4.4 - once the server is enabled to respond to an OCSP request
1429 it does even when not requested, resulting in a stapling non-aware
1430 client dropping the TLS connection.
1432 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1433 support variable-length bit vectors. No functional change.
1435 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1436 The I= interface field on outgoing lines is now after the H= remote
1437 host field, same as incoming lines. There is a separate
1438 outgoing_interface log selector which allows you to disable the
1441 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1442 If not running log_selector +smtp_connection the mainlog would be held
1443 open indefinitely after a "too many connections" event, including to a
1444 deleted file after a log rotate. Leave the per net connection logging
1445 leaving it open for efficiency as that will be quickly detected by the
1446 check on the next write.
1448 HS/01 Bug 1671: Fix post transport crash.
1449 Processing the wait-<transport> messages could crash the delivery
1450 process if the message IDs didn't exist for some reason. When
1451 using 'split_spool_directory=yes' the construction of the spool
1452 file name failed already, exposing the same netto behaviour.
1454 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1455 mime_regex ACL conditions.
1457 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1458 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1459 to HELO, local diagnostic string.
1461 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1462 log line to debug. Even when configured with a tls authenticator many
1463 client connections are expected to not authenticate in this way, so
1464 an authenticate fail is not an error.
1466 HS/02 Add the Exim version string to the process info. This way exiwhat
1467 gives some more detail about the running daemon.
1469 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1470 matter for fast-change records such as DNSBLs.
1472 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1473 retry record, even if constant. There may be multiple transports with
1474 different interface settings and the retry behaviour needs to be kept
1477 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1479 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1481 JH/10 Bug 840: fix log_defer_output option of pipe transport
1483 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1484 hosts_randomize. This matters a lot when combined with mua_wrapper.
1486 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1487 ${quote_pgsql:<string>} operator.
1489 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1491 JH/14 Change header file name for freeradius-client. Relevant if compiling
1492 with Radius support; from the Gentoo tree and checked under Fedora.
1494 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1496 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1497 option transport_filter, meaning no filtering.
1499 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1501 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1502 defaults to "*" (all hosts). The variable is now available when not built
1503 with TLS, default unset, mainly to enable keeping the testsuite sane.
1504 If a server certificate is not supplied (via tls_certificate) an error is
1505 logged, and clients will find TLS connections fail on startup. Presumably
1506 they will retry in-clear.
1507 Packagers of Exim are strongly encouraged to create a server certificate
1508 at installation time.
1510 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1511 with the $config_file variable.
1513 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1514 in transport context, after the attempt, and per-recipient. The latter type
1515 is per host attempted. The event data is the error message, and the errno
1516 information encodes the lookup type (A vs. MX) used for the (first) host,
1517 and the trailing two digits of the smtp 4xx response.
1519 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1520 to write to mainlog (or rejectlog, paniclog) in the window between file
1521 creation and permissions/ownership being changed. Particularly affects
1522 installations where exicyclog is run as root, rather than exim user;
1523 result is that the running daemon panics and dies.
1525 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1527 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1528 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1529 "pri" and "weight". Note that the previous implicit priority given by the
1530 list order is no longer honoured.
1532 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1533 for DKIM processing.
1535 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1536 by defining SUPPORT_SOCKS.
1538 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1539 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1540 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1541 variables are renamed to proxy_{local,external}_{address,port}.
1543 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1544 for a build by defining SUPPORT_I18N
1546 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1547 of the query string, and make ${quote_redis:} do that quoting.
1549 JH/29 Move Events support from Experimental to mainline, enabled by default
1550 and removable for a build by defining DISABLE_EVENT.
1552 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1554 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1555 cached by the daemon.
1557 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1558 by defining LOOKUP_REDIS. The libhiredis library is required.
1560 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1561 keys are given for lookup.
1563 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1564 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1565 only supported when built with TLS support. The PolarSSL SHA routines
1566 are still used when the TLS library is too old for convenient support.
1568 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1569 openssl_options), for security. OpenSSL forces this from version 1.1.0
1570 server-side so match that on older versions.
1572 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1573 allocation for $value could be released as the expansion processing
1574 concluded, but leaving the global pointer active for it.
1576 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1577 and to use the domains and local_parts ACL conditions.
1579 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1580 incorrectly not doubled on cutthrough transmission, hence seen as a
1581 body-termination at the receiving system - resulting in truncated mails.
1582 Commonly the sender saw a TCP-level error, and retransmitted the message
1583 via the normal store-and-forward channel. This could result in duplicates
1584 received - but deduplicating mailstores were liable to retain only the
1585 initial truncated version.
1587 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1589 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1591 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1592 we're in there, support oversigning also; bug 1309.
1594 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1596 HS/04 Add support for keep_environment and add_environment options.
1598 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1599 either intentional arithmetic overflow during PRNG, or testing config-
1602 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1603 delivery resulted in actual delivery. Cancel cutthrough before DATA
1606 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1607 rejects a recipient: pass the reject to the originator.
1609 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1610 Many were false-positives and ignorable, but it's worth fixing the
1613 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1614 for the new environment-manipulation done at startup. Move the routines
1615 from being local to tls.c to being global via the os.c file.
1617 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1618 an extract embedded as result-arg for a map, the first arg for extract
1619 is unavailable so we cannot tell if this is a numbered or keyed
1620 extraction. Accept either.
1626 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1629 JH/02 The smtp transport option "multi_domain" is now expanded.
1631 JH/03 The smtp transport now requests PRDR by default, if the server offers
1634 JH/04 Certificate name checking on server certificates, when exim is a client,
1635 is now done by default. The transport option tls_verify_cert_hostnames
1636 can be used to disable this per-host. The build option
1637 EXPERIMENTAL_CERTNAMES is withdrawn.
1639 JH/05 The value of the tls_verify_certificates smtp transport and main options
1640 default to the word "system" to access the system default CA bundle.
1641 For GnuTLS, only version 3.0.20 or later.
1643 JH/06 Verification of the server certificate for a TLS connection is now tried
1644 (but not required) by default. The verification status is now logged by
1645 default, for both outbound TLS and client-certificate supplying inbound
1648 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1651 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1652 Status Notification (bounce) messages are now MIME format per RFC 3464.
1653 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1654 under the control of the dsn_advertise_hosts option, and routers may
1655 have a dsn_lasthop option.
1657 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1658 default, modifiable by a malware= option. The list separator for
1659 the options can now be changed in the usual way. Bug 68.
1661 JH/10 The smtp_receive_timeout main option is now expanded before use.
1663 JH/11 The incoming_interface log option now also enables logging of the
1664 local interface on delivery outgoing connections.
1666 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1667 if the interface and destination host and port all match.
1669 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1672 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1673 Patch from Andrew Lewis.
1675 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1676 now supports optional time-restrictions, weighting, and priority
1677 modifiers per server. Patch originally by <rommer@active.by>.
1679 JH/16 The spamd_address main option now supports a mixed list of local
1680 and remote servers. Remote servers can be IPv6 addresses, and
1681 specify a port-range.
1683 JH/17 Bug 68: The spamd_address main option now supports an optional
1684 timeout value per server.
1686 JH/18 Bug 1581: Router and transport options headers_add/remove can
1687 now have the list separator specified.
1689 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1692 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1695 JH/21 Support for the A6 type of dns record is withdrawn.
1697 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1698 rather than the verbs used.
1700 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1701 from 255 to 1024 chars.
1703 JH/24 Verification callouts now attempt to use TLS by default.
1705 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1706 are generic router options now. The defaults didn't change.
1708 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1709 Original patch from Alexander Shikoff, worked over by JH.
1711 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1712 files if ZCAT_COMMAND is not executable.
1714 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1716 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1718 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1719 Normally benign, it bites when the pair was led to by a CNAME;
1720 modern usage is to not canonicalize the domain to a CNAME target
1721 (and we were inconsistent anyway for A-only vs AAAA+A).
1723 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1725 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1726 when evaluating $sender_host_dnssec.
1728 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1729 $sender_helo_dnssec variable.
1731 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1733 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1735 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1737 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1738 documented as working, but never had. Support all but $spam_report.
1740 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1741 added for tls authenticator.
1743 HS/03 Add perl_taintmode main config option
1749 TL/01 When running the test suite, the README says that variables such as
1750 no_msglog_check are global and can be placed anywhere in a specific
1751 test's script, however it was observed that placement needed to be near
1752 the beginning for it to behave that way. Changed the runtest perl
1753 script to read through the entire script once to detect and set these
1754 variables, reset to the beginning of the script, and then run through
1755 the script parsing/test process like normal.
1757 TL/02 The BSD's have an arc4random API. One of the functions to induce
1758 adding randomness was arc4random_stir(), but it has been removed in
1759 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1760 function when detected.
1762 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1763 cause callback expansion.
1765 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1766 syntax errors in an expansion can be treated as a string instead of
1767 logging or causing an error, due to the internal use of bool_lax
1768 instead of bool when processing it.
1770 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1771 server certificates when making smtp deliveries.
1773 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1775 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1777 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1779 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1780 Merged patch from Sebastian Wiedenroth.
1782 JH/05 Fix results-pipe from transport process. Several recipients, combined
1783 with certificate use, exposed issues where response data items split
1784 over buffer boundaries were not parsed properly. This eventually
1785 resulted in duplicates being sent. This issue only became common enough
1786 to notice due to the introduction of connection certificate information,
1787 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1789 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1790 size buffer was used, resulting in syntax errors when an expansion
1793 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1794 version 3.3.6 or later.
1796 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1797 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1798 both become "event_action", the variables become $event_name, $event_data
1799 and $event_defer_errno. There is a new variable $verify_mode, usable in
1800 routers, transports and related events. The tls:cert event is now also
1801 raised for inbound connections, if the main configuration event_action
1804 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1805 early OCSP support, but no stapling (appears to be less than 1.0.0).
1807 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1808 server certificate names available under the smtp transport option
1809 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1812 JH/10 Time-related extraction expansions from certificates now use the main
1813 option "timezone" setting for output formatting, and are consistent
1814 between OpenSSL and GnuTLS compilations. Bug 1541.
1816 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1817 encoded parameter in the incoming message. Bug 1558.
1819 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1820 include certificate info, eximon was claiming there were spoolfile
1823 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1825 JH/14 Log delivery-related information more consistently, using the sequence
1826 "H=<name> [<ip>]" wherever possible.
1828 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1829 are problematic for Debian distribution, omit them from the release
1832 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1834 JH/16 Fix string representation of time values on 64bit time_t architectures.
1837 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1838 output list separator was used.
1843 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1844 checkers that were complaining about end of non-void function with no
1847 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1848 This was a regression introduced in 4.83 by another bugfix.
1850 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1852 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1853 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1859 TF/01 Correctly close the server side of TLS when forking for delivery.
1861 When a message was received over SMTP with TLS, Exim failed to clear up
1862 the incoming connection properly after forking off the child process to
1863 deliver the message. In some situations the subsequent outgoing
1864 delivery connection happened to have the same fd number as the incoming
1865 connection previously had. Exim would try to use TLS and fail, logging
1866 a "Bad file descriptor" error.
1868 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1869 utilities have not been installed.
1871 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1872 temporary space as the ACL may create new global variables.
1874 TL/01 LDAP support uses per connection or global context settings, depending
1875 upon the detected version of the libraries at build time.
1877 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1878 to extract and use the src ip:port in logging and expansions as if it
1879 were a direct connection from the outside internet. PPv2 support was
1880 updated based on HAProxy spec change in May 2014.
1882 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1884 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1885 Properly escape header and check for NULL return.
1887 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1890 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1892 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1893 characters in header names, implemented as a verify condition.
1894 Contributed by Michael Fischer v. Mollard.
1896 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1897 results permerror and temperror. Previous values are deprecated but
1898 still accepted. In a future release, err_perm and err_temp will be
1899 completely removed, which will be a backward incompatibility if the
1900 ACL tests for either of these two old results. Patch contributed by
1901 user bes-internal on the mailing list.
1903 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1905 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1906 selectors, in both main and reject logs.
1908 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1911 JH/07 Add malware type "sock" for talking to simple daemon.
1913 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1915 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1916 routers/transports under cutthrough routing.
1918 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1919 numbers. Touch up "bool" conditional to keep the same definition.
1921 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1923 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1925 JH/12 Expand items in router/transport headers_add or headers_remove lists
1926 individually rather than the list as a whole. Bug 1452.
1928 Required for reasonable handling of multiple headers_ options when
1929 they may be empty; requires that headers_remove items with embedded
1930 colons must have them doubled (or the list-separator changed).
1932 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1933 view the policy declared in the DMARC record. Currently, $dmarc_status
1934 is a combined value of both the record presence and the result of the
1937 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1939 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1940 dnslookup router and the smtp transport (applying to the forward
1943 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1944 of ldap servers used for a specific lookup. Patch provided by Heiko
1947 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1948 New variable $lookup_dnssec_authenticated for observability.
1950 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1951 Patch submitted by Lars Timman.
1953 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1955 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1956 Requires trusted mode and valid format message id, aborts otherwise.
1957 Patch contributed by Heiko Schlichting.
1959 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1960 certextract with support for various fields. Bug 1358.
1962 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1963 is requested by default, modifiable by smtp transport option
1966 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1967 operate on certificate variables to give certificate fingerprints
1968 Also new ${sha256:cert_variable}.
1970 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1972 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1973 Christian Aistleitner.
1975 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1977 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1978 file. Patch from Wolfgang Breyha.
1980 JH/25 Expand the coverage of the delivery $host and $host_address to
1981 client authenticators run in verify callout. Bug 1476.
1983 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1984 align with daemon_smtp_ports. Bug 72.
1986 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1987 support and error reporting did not work properly.
1989 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1990 and is readable. Patch from Andrew Colin Kissa.
1992 TL/14 Enhance documentation of ${run expansion and how it parses the
1993 commandline after expansion, particularly in the case when an
1994 unquoted variable expansion results in an empty value.
1996 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1998 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1999 by renaming away from it. Addresses GH issue 15, reported by
2002 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2004 TL/15 SECURITY: prevent double expansion in math comparison functions
2005 (can expand unsanitized data). Not remotely exploitable.
2012 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2014 PP/02 Make -n do something, by making it not do something.
2015 When combined with -bP, the name of an option is not output.
2017 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2020 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2021 $sender_host_name and config options to manage this, and basic check
2024 PP/05 DSCP support for outbound connections and control modifier for inbound.
2026 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2027 (Only plugin which currently uses this is kerberos4, which nobody should
2028 be using, but we should make it available and other future plugins might
2029 conceivably use it, even though it would break NAT; stuff *should* be
2030 using channel bindings instead).
2032 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2033 name; added for Sendmail compatibility; requires admin caller.
2034 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2035 just ignore it); requires trusted caller.
2036 Also parse but ignore: -Ac -Am -X<logfile>
2039 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2041 TL/02 Add +smtp_confirmation as a default logging option.
2043 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2044 Patch by Magnus Holmgren from 2007-02-20.
2046 TL/04 Bugzilla 1281 - Spec typo.
2047 Bugzilla 1283 - Spec typo.
2048 Bugzilla 1290 - Spec grammar fixes.
2050 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2052 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2054 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2055 mailing list by by Dmitry Isaikin.
2057 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2059 JH/02 Support "G" suffix to numbers in ${if comparisons.
2061 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2063 NM/01 Bugzilla 1197 - Spec typo
2064 Bugzilla 1196 - Spec examples corrections
2066 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2068 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2069 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2072 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2073 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2075 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2076 "acl {{name}{arg}...}", and optional args on acl condition
2079 JH/05 Permit multiple router/transport headers_add/remove lines.
2081 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2083 JH/07 Avoid using a waiting database for a single-message-only transport.
2084 Performance patch from Paul Fisher. Bugzilla 1262.
2086 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2089 JH/09 Add $headers_added variable, with content from use of ACL modifier
2090 add_header (but not yet added to the message). Bugzilla 199.
2092 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2093 Pulled from Bugzilla 817 by Wolfgang Breyha.
2095 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2097 (nb: this is the same fix as in Exim 4.80.1)
2099 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2102 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2105 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2107 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2108 advertise SMTP AUTH mechanism to us, instead of a generic
2109 protocol violation error. Also, make Exim more robust to bad
2110 data from the Dovecot auth socket.
2112 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2114 When a queue runner is handling a message, Exim first routes the
2115 recipient addresses, during which it prunes them based on the retry
2116 hints database. After that it attempts to deliver the message to
2117 any remaining recipients. It then updates the hints database using
2120 So if a recipient address works intermittently, it can get repeatedly
2121 deferred at routing time. The retry hints record remains fresh so the
2122 address never reaches the final cutoff time.
2124 This is a fairly common occurrence when a user is bumping up against
2125 their storage quota. Exim had some logic in its local delivery code
2126 to deal with this. However it did not apply to per-recipient defers
2127 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2129 This change adds a proper retry rule check during routing so that the
2130 final cutoff time is checked against the message's age. We only do
2131 this check if there is an address retry record and there is not a
2132 domain retry record; this implies that previous attempts to handle
2133 the address had the retry_use_local_parts option turned on. We use
2134 this as an approximation for the destination being like a local
2135 delivery, as in LMTP.
2137 I suspect this new check makes the old local delivery cutoff check
2138 redundant, but I have not verified this so I left the code in place.
2140 TF/02 Correct gecos expansion when From: is a prefix of the username.
2142 Test 0254 submits a message to Exim with the header
2146 When I ran the test suite under the user fanf2, Exim expanded
2147 the header to contain my full name, whereas it should have added
2148 a Resent-Sender: header. It erroneously treats any prefix of the
2149 username as equal to the username.
2151 This change corrects that bug.
2153 GF/01 DCC debug and logging tidyup
2154 Error conditions log to paniclog rather than rejectlog.
2155 Debug lines prefixed by "DCC: " to remove any ambiguity.
2157 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2159 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2160 Bug spotted by Jeremy Harris; was flawed since initial commit.
2161 Would have resulted in OCSP responses post-SNI triggering an Exim
2162 NULL dereference and crash.
2164 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2166 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2167 Bug detection, analysis and fix by Samuel Thibault.
2168 Bugzilla 1331, Debian bug #698092.
2170 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2172 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2173 Server implementation by Todd Lyons, client by JH.
2174 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2175 config variable "prdr_enable" controls whether the server
2176 advertises the facility. If the client requests PRDR a new
2177 acl_data_smtp_prdr ACL is called once for each recipient, after
2178 the body content is received and before the acl_smtp_data ACL.
2179 The client is controlled by both of: a hosts_try_prdr option
2180 on the smtp transport, and the server advertisement.
2181 Default client logging of deliveries and rejections involving
2182 PRDR are flagged with the string "PRDR".
2184 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2185 fclose(). Diagnosis by Todd Lyons.
2187 PP/17 Update configure.default to handle IPv6 localhost better.
2188 Patch by Alain Williams (plus minor tweaks).
2191 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2192 This is now consistent with GnuTLS, and is now documented: the
2193 previous undocumented portable approach to treating the option as
2194 unset was to force an expansion failure. That still works, and
2195 an empty string is now equivalent.
2197 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2198 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2199 not performing validation itself.
2201 PP/20 Added force_command boolean option to pipe transport.
2202 Patch from Nick Koston, of cPanel Inc.
2204 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2207 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2209 PP/21 Fix eximon continuous updating with timestamped log-files.
2210 Broken in a format-string cleanup in 4.80, missed when I repaired the
2211 other false fix of the same issue.
2212 Report and fix from Heiko Schlichting.
2215 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2216 Report from Prashanth Katuri.
2218 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2219 It's SecureTransport, so affects any MacOS clients which use the
2220 system-integrated TLS libraries, including email clients.
2222 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2223 using a MIME ACL for non-SMTP local injection.
2224 Report and assistance in diagnosis by Warren Baker.
2226 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2228 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2230 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2231 last id that failed so it may be referenced in subsequent ACL's.
2233 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2236 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2237 ldap library initialization, allowing self-signed CA's to be
2238 used. Also properly sets require_cert option later in code by
2239 using NULL (global ldap config) instead of ldap handle (per
2240 session). Bug diagnosis and testing by alxgomz.
2242 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2243 the src/util/ subdirectory.
2245 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2246 renamed to Transport Post Delivery Action by Jeremy Harris, as
2249 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2250 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2251 redis_servers = needs to be configured which will be used by the redis
2252 lookup. Patch from Warren Baker, of The Packet Hub.
2254 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2256 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2257 hostname or reverse DNS when processing a host list. Used suggestions
2258 from multiple comments on this bug.
2260 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2262 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2263 Missed a few lines, added it to make the runtest require no keyboard
2266 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2267 contains upper case chars. Make router use caseful_local_part.
2269 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2270 support when GnuTLS has been built with p11-kit.
2276 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2278 This, or similar/improved, will also be change PP/11 of 4.82.
2284 PP/01 Handle short writes when writing local log-files.
2285 In practice, only affects FreeBSD (8 onwards).
2286 Bugzilla 1053, with thanks to Dmitry Isaikin.
2288 NM/01 Bugzilla 949 - Documentation tweak
2290 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2293 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2295 PP/02 Implemented gsasl authenticator.
2297 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2299 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2300 `pkg-config foo` for cflags/libs.
2302 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2303 with rest of GSASL and with heimdal_gssapi.
2305 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2306 `pkg-config foo` for cflags/libs for the TLS implementation.
2308 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2309 properties get this fed in as external SSF. A number of robustness
2310 and debugging improvements to the cyrus_sasl authenticator.
2312 PP/08 cyrus_sasl server now expands the server_realm option.
2314 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2315 Patch by Jeremy Harris.
2317 PP/10 Added dbmjz lookup type.
2319 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2321 PP/12 MAIL args handles TAB as well as SP, for better interop with
2322 non-compliant senders.
2323 Analysis and variant patch by Todd Lyons.
2325 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2326 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2327 Patch from Dirk Mueller <dmueller@suse.com>
2329 PP/13 tls_peerdn now print-escaped for spool files.
2330 Observed some $tls_peerdn in wild which contained \n, which resulted
2331 in spool file corruption.
2333 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2334 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2335 or write after TLS renegotiation, which otherwise led to messages
2338 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2339 as a tracking header (ie: a signed header comes before the signature).
2340 Patch from Wolfgang Breyha.
2342 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2343 comma-sep list; embedded commas doubled.
2345 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2347 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2349 Report and patch from Dmitry Banschikov.
2351 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2352 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2353 failures. We appear to now support TLS1.1+ with Exim.
2355 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2356 lets Exim select keys and certificates based upon TLS SNI from client.
2357 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2358 before an outbound SMTP session. New log_selector, +tls_sni.
2360 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2361 NULL dereference. Report and patch from Alun Jones.
2363 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2364 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2365 Not seeing resolver debug output on NetBSD, but suspect this is a
2366 resolver implementation change.
2368 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2369 Left warnings. Added "eximon gdb" invocation mode.
2371 PP/21 Defaulting "accept_8bitmime" to true, not false.
2373 PP/22 Added -bw for inetd wait mode support.
2375 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2376 locate the relevant includes and libraries. Made this the default.
2378 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2379 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2381 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2382 This may cause build issues on older platforms.
2384 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2385 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2386 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2387 Added SNI support via GnuTLS too.
2388 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2390 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2392 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2393 Applied second patch from Janne, implementing suggestion to default
2394 multiple-strings-in-record handling to match SPF spec.
2396 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2398 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2399 read-only, out of scope).
2400 Patch from Wolfgang Breyha, report from Stuart Northfield.
2402 PP/29 Fix three issues highlighted by clang analyser static analysis.
2403 Only crash-plausible issue would require the Cambridge-specific
2404 iplookup router and a misconfiguration.
2405 Report from Marcin Mirosław.
2407 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2409 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2410 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2411 As part of this, removing so much warning spew let me fix some minor
2412 real issues in debug logging.
2414 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2415 assignment on my part. Fixed.
2417 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2418 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2419 Janne Snabb (who went above and beyond: thank you).
2421 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2422 string otherwise requires a connection and a bunch more work and it's
2423 relatively easy to get wrong. Should also expose TLS library linkage
2426 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2427 64-bit ${eval} (JH/03).
2429 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2430 GNU libc to support some of the 64-bit stuff, should not lead to
2431 conflicts. Defined before os.h is pulled in, so if a given platform
2432 needs to override this, it can.
2434 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2435 protection layer was required, which is not implemented.
2436 Bugzilla 1254, patch from Wolfgang Breyha.
2438 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2439 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2440 tls_dhparam take prime identifiers. Also unbreak combination of
2441 OpenSSL+DH_params+TLSSNI.
2443 PP/39 Disable SSLv2 by default in OpenSSL support.
2449 PP/01 Solaris build fix for Oracle's LDAP libraries.
2450 Bugzilla 1109, patch from Stephen Usher.
2452 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2454 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2457 TF/02 Fix a couple more cases where we did not log the error message
2458 when unlink() failed. See also change 4.74-TF/03.
2460 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2461 lock up or crash if it happened to be inside a call to libc when it
2462 got a SIGUSR1 from exiwhat.
2464 The SIGUSR1 handler appends the current process status to the process
2465 log which is later printed by exiwhat. It used to use the general
2466 purpose logging code to do this, but several functions it calls are
2467 not safe for signals.
2469 The new output code in the SIGUSR1 handler is specific to the process
2470 log, and simple enough that it's easy to inspect for signal safety.
2471 Removing some special cases also simplifies the general logging code.
2472 Removing the spurious timestamps from the process log simplifies
2475 TF/04 Improved ratelimit ACL condition.
2477 The /noupdate option has been deprecated in favour of /readonly which
2478 has clearer semantics. The /leaky, /strict, and /readonly update modes
2479 are mutually exclusive. The update mode is no longer included in the
2480 database key; it just determines when the database is updated. (This
2481 means that when you upgrade Exim will forget old rate measurements.)
2483 Exim now checks that the per_* options are used with an update mode that
2484 makes sense for the current ACL. For example, when Exim is processing a
2485 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2486 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2487 must specify per_mail/readonly. If you omit the update mode it defaults to
2488 /leaky where that makes sense (as before) or /readonly where required.
2490 The /noupdate option is now undocumented but still supported for
2491 backwards compatibility. It is equivalent to /readonly except that in
2492 ACLs where /readonly is required you may specify /leaky/noupdate or
2493 /strict/noupdate which are treated the same as /readonly.
2495 A useful new feature is the /count= option. This is a generalization
2496 of the per_byte option, so that you can measure the throughput of other
2497 aggregate values. For example, the per_byte option is now equivalent
2498 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2500 The per_rcpt option has been generalized using the /count= mechanism
2501 (though it's more complicated than the per_byte equivalence). When it is
2502 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2503 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2504 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2505 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2506 Note that using per_rcpt with a non-readonly update mode in more than
2507 one ACL will cause the recipients to be double-counted. (The per_mail
2508 and per_byte options don't have this problem.)
2510 The handling of very low rates has changed slightly. If the computed rate
2511 is less than the event's count (usually one) then this event is the first
2512 after a long gap. In this case the rate is set to the same as this event's
2513 count, so that the first message of a spam run is counted properly.
2515 The major new feature is a mechanism for counting the rate of unique
2516 events. The new per_addr option counts the number of different
2517 recipients that someone has sent messages to in the last time period. It
2518 behaves like per_rcpt if all the recipient addresses are different, but
2519 duplicate recipient addresses do not increase the measured rate. Like
2520 the /count= option this is a general mechanism, so the per_addr option
2521 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2522 example, measure the rate that a client uses different sender addresses
2523 with the options per_mail/unique=$sender_address. There are further
2524 details in the main documentation.
2526 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2528 TF/06 Removed a few PCRE remnants.
2530 TF/07 Automatically extract Exim's version number from tags in the git
2531 repository when doing development or release builds.
2533 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2534 Bugzilla 879. Patch from Paul Fisher.
2536 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2537 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2540 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2542 PP/05 Variable $av_failed, true if the AV scanner deferred.
2543 Bugzilla 1078. Patch from John Horne.
2545 PP/06 Stop make process more reliably on build failure.
2546 Bugzilla 1087. Patch from Heiko Schlittermann.
2548 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2549 Bugzilla 1089. Patch from Heiko Schlittermann.
2551 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2552 Bugzilla 1131. Patch from Holger Weiß.
2554 PP/09 Handle IPv6 addresses with SPF.
2555 Bugzilla 860. Patch from Wolfgang Breyha.
2557 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2559 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2562 PP/11 match_* no longer expand right-hand-side by default.
2563 New compile-time build option, EXPAND_LISTMATCH_RHS.
2564 New expansion conditions, "inlist", "inlisti".
2566 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2568 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2570 PP/14 fix log_write() format string regression from TF/03.
2571 Bugzilla 1152. Patch from Dmitry Isaikin.
2577 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2579 PP/02 Harmonised TLS library version reporting; only show if debugging.
2580 Layout now matches that introduced for other libraries in 4.74 PP/03.
2582 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2584 PP/04 New "dns_use_edns0" global option.
2586 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2589 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2590 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2592 TK/01 Updated PolarSSL code to 0.14.2.
2593 Bugzilla 1097. Patch from Andreas Metzler.
2595 PP/07 Catch divide-by-zero in ${eval:...}.
2596 Fixes bugzilla 1102.
2598 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2601 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2602 format-string attack -- SECURITY: remote arbitrary code execution.
2604 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2605 time unintentionally subject to list matching rules, letting the header
2606 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2607 arbitrary string expansion). This allowed for information disclosure.
2609 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2610 INT_MIN/-1 -- value coerced to INT_MAX.
2616 NM/01 Workaround for PCRE version dependency in version reporting
2619 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2620 This fixes portability to compilers other than gcc, notably
2621 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2623 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2624 makefiles for portability to HP-UX and POSIX correctness.
2626 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2627 Also via indented variable definition in the Makefile.
2628 (Debugging by Oliver Heesakkers).
2630 PP/02 Restore caching of spamd results with expanded spamd_address.
2631 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2633 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2634 Improves build reliability. Fix from: Frank Elsner
2636 NM/02 Fix wide character breakage in the rfc2047 coding
2637 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2639 NM/03 Allow underscore in dnslist lookups
2640 Fixes bug 1026. Patch from Graeme Fowler
2642 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2643 Code patches from Adam Ciarcinski of NetBSD.
2645 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2648 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2649 is logged, to avoid truncation. Patch from John Horne.
2651 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2652 Patch from Jakob Hirsch.
2654 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2655 SQL string expansion failure details.
2656 Patch from Andrey Oktyabrski.
2658 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2659 Patch from Simon Arlott.
2661 PP/09 New lookups functionality failed to compile on old gcc which rejects
2662 extern declarations in function scope.
2663 Patch from Oliver Fleischmann
2665 PP/10 Use sig_atomic_t for flags set from signal handlers.
2666 Check getgroups() return and improve debugging.
2667 Fixed developed for diagnosis in bug 927 (which turned out to be
2670 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2671 Patch from Mark Zealey.
2673 PP/12 Bugzilla 1056: Improved spamd server selection.
2674 Patch from Mark Zealey.
2676 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2677 Based on patch from Heiko Schlittermann.
2679 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2680 Patch from Uwe Doering, sign-off by Michael Haardt.
2682 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2683 variable declaration deep within a block. Bug and patch from
2686 PP/15 lookups-Makefile IRIX compatibility coercion.
2688 PP/16 Make DISABLE_DKIM build knob functional.
2690 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2691 Patch by Simon Arlott
2693 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2694 variable argument macros. Our copy now differs from upstream.
2700 TF/01 Failure to get a lock on a hints database can have serious
2701 consequences so log it to the panic log.
2703 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2704 controlled using the smtp_confirmation log selector.
2706 TF/03 Include the error message when we fail to unlink a spool file.
2708 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2709 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2710 for maintaining out-of-tree patches for some time.
2712 PP/01 Bugzilla 139: Documentation and portability issues.
2713 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2714 Handle per-OS dynamic-module compilation flags.
2716 PP/02 Let /dev/null have normal permissions.
2717 The 4.73 fixes were a little too stringent and complained about the
2718 permissions on /dev/null. Exempt it from some checks.
2719 Reported by Andreas M. Kirchwitz.
2721 PP/03 Report version information for many libraries, including
2722 Exim version information for dynamically loaded libraries. Created
2723 version.h, now support a version extension string for distributors
2724 who patch heavily. Dynamic module ABI change.
2726 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2727 privilege escalation vulnerability whereby the Exim run-time user
2728 can cause root to append content of the attacker's choosing to
2731 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2734 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2735 If dropping privileges for untrusted macros, we disabled normal logging
2736 on the basis that it would fail; for the Exim run-time user, this is not
2737 the case, and it resulted in successful deliveries going unlogged.
2738 Fixed. Reported by Andreas Metzler.
2744 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2745 only prepend for the Resent-* case. Fixes regression introduced in
2746 Exim 4.70 by NM/22 for Bugzilla 607.
2748 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2749 increasing numbers of administrators be bitten by this.
2751 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2753 PP/03 Bugzilla 994: added openssl_options main configuration option.
2755 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2757 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2759 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2761 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2762 without a peer certificate, leading to a segfault because of an
2763 assumption that peers always have certificates. Be a little more
2764 paranoid. Problem reported by Martin Tscholak.
2766 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2767 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2768 NB: ClamAV planning to remove STREAM in "middle of 2010".
2769 CL also introduces -bmalware, various -d+acl logging additions and
2770 more caution in buffer sizes.
2772 PP/09 Implemented reverse_ip expansion operator.
2774 PP/10 Bugzilla 937: provide a "debug" ACL control.
2776 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2778 PP/12 Bugzilla 973: Implement --version.
2780 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2782 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2784 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2786 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2787 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2788 ignore trailing whitespace.
2790 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2792 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2795 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2796 Notification from Dr Andrew Aitchison.
2798 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2799 ExtendedDetectionInfo response format.
2800 Notification from John Horne.
2802 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2805 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2806 XSL and documented dependency on system catalogs, with examples of how
2809 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2812 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2813 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2814 configuration file which is writeable by the Exim user or group.
2816 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2817 of configuration files to cover files specified with the -C option if
2818 they are going to be used with root privileges, not just the default
2821 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2822 option (effectively making it always true).
2824 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2825 files to be used while preserving root privileges.
2827 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2828 that rogue child processes cannot use them.
2830 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2831 run-time user, instead of root.
2833 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2834 Exim run-time user without dropping privileges.
2836 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2837 result string, instead of calling string_vformat() twice with the same
2840 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2841 for other users. Others should always drop root privileges if they use
2842 -C on the command line, even for a whitelisted configure file.
2844 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2846 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2852 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2853 $data_path, and $header_path variables; fixed documentation bugs and
2856 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2857 exipick to access non-standard spools, including the "frozen" queue
2860 NM/01 Bugzilla 965: Support mysql stored procedures.
2861 Patch from Alain Williams
2863 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2865 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2866 Patch from Andreas Metzler
2868 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2869 Patch from Kirill Miazine
2871 NM/05 Bugzilla 671: Added umask to procmail example.
2873 JJ/03 installed exipick 20100323.0, fixing doc bug
2875 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2876 directory. Notification and patch from Dan Rosenberg.
2878 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2880 TK/02 Improve log output when DKIM signing operation fails.
2882 MH/01 Treat the transport option dkim_domain as a colon separated
2883 list, not as a single string, and sign the message with each element,
2884 omitting multiple occurences of the same signer.
2886 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2887 Bugzilla 985, 986. Patch by Simon Arlott
2889 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2890 Patch by Simon Arlott
2892 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2893 MBX locking. Notification from Dan Rosenberg.
2899 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2901 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2903 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2905 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2907 NM/04 Bugzilla 915: Flag broken perl installation during build.
2913 TK/01 Added patch by Johannes Berg that expands the main option
2914 "spamd_address" if it starts with a dollar sign.
2916 TK/02 Write list of recipients to X-Envelope-Sender header when building
2917 the mbox-format spool file for content scanning (suggested by Jakob
2920 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2921 (http://www.dcc-servers.net/) support via dccifd. Activated by
2922 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2924 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2925 by Mark Daniel Reidel <mr@df.eu>.
2927 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2928 When building exim an external PCRE library is now needed -
2929 PCRE is a system library on the majority of modern systems.
2930 See entry on PCRE_LIBS in EDITME file.
2932 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2933 conversation. Added nologin parameter to request.
2934 Patch contributed by Kirill Miazine.
2936 TF/01 Do not log submission mode rewrites if they do not change the address.
2938 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2940 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2941 log files in place. Contributed by Roberto Lima.
2943 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2945 TF/03 Bugzilla 615: When checking the local_parts router precondition
2946 after a local_part_suffix or local_part_prefix option, Exim now
2947 does not use the address's named list lookup cache, since this
2948 contains cached lookups for the whole local part.
2950 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2951 Robert Millan. Documentation is in experimental-spec.txt.
2953 TF/04 Bugzilla 668: Fix parallel build (make -j).
2955 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2957 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2958 Patch provided by Jan Srzednicki.
2960 TF/05 Leading white space used to be stripped from $spam_report which
2961 wrecked the formatting. Now it is preserved.
2963 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2964 that they are available at delivery time.
2966 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2968 TF/08 TLS error reporting now respects the incoming_interface and
2969 incoming_port log selectors.
2971 TF/09 Produce a more useful error message if an SMTP transport's hosts
2972 setting expands to an empty string.
2974 NM/06 Bugzilla 744: EXPN did not work under TLS.
2975 Patch provided by Phil Pennock.
2977 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2978 Patch provided by Richard Godbee.
2980 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2981 acl_smtp_notquit, added index entry.
2983 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2984 Patch provided by Eugene Bujak.
2986 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2987 accept(). Patch provided by Maxim Dounin.
2989 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2990 Patch provided by Phil Pennock.
2992 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2994 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2995 Patch provided by Brad "anomie" Jorsch.
2997 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2998 Patch provided by Dean Brooks.
3000 TK/05 Add native DKIM support (does not depend on external libraries).
3002 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3003 Patch provided by Graeme Fowler.
3005 NM/16 Bugzilla 851: Documentation example syntax fix.
3007 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3009 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3012 NM/19 Bugzilla 745: TLS version reporting.
3013 Patch provided by Phil Pennock.
3015 NM/20 Bugzilla 167: bool: condition support.
3016 Patch provided by Phil Pennock.
3018 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3019 clients. Patch provided by Phil Pennock.
3021 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3022 Patch provided by Brad "anomie" Jorsch.
3024 NM/23 Bugzilla 687: Fix misparses in eximstats.
3025 Patch provided by Heiko Schlittermann.
3027 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3028 Patch provided by Heiko Schlittermann.
3030 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3031 plus update to original patch.
3033 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3035 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3036 Patch provided by David Brownlee.
3038 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3040 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3042 NM/30 Bugzilla 888: TLS documentation bugfixes.
3044 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3046 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3047 Unlike the original bugzilla I have changed all shell scripts in src tree.
3049 NM/33 Bugzilla 898: Transport filter timeout fix.
3050 Patch by Todd Rinaldo.
3052 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3053 Patch by Serge Demonchaux.
3055 NM/35 Bugzilla 39: Base64 decode bug fixes.
3056 Patch by Jakob Hirsch.
3058 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3060 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3062 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3064 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3070 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3071 ALT-N's libdkim that I have put here:
3072 http://duncanthrax.net/exim-experimental/
3074 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3075 were called 'true' and it seems that C99 defines that as a reserved
3076 keyword to be used with 'bool' variable types. That means you could
3077 not include C99-style headers which use bools without triggering
3078 build errors in sieve.c.
3080 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3081 as mailq or other aliases. Changed the --help handling significantly
3082 to do whats expected. exim_usage() emits usage/help information.
3084 SC/01 Added the -bylocaldomain option to eximstats.
3086 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3088 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3090 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3096 PH/01 Another patch from the Sieve maintainer.
3098 PH/02 When an IPv6 address is converted to a string for single-key lookup
3099 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3100 dots are used instead of colons so that keys in lsearch files need not
3101 contain colons. This was done some time before quoting was made available
3102 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3103 (notated using the quote facility) so as to distinguish them from IPv4
3104 keys. This meant that lookups for IP addresses in host lists did not work
3105 for iplsearch lookups.
3107 This has been fixed by arranging for IPv6 addresses to be expressed with
3108 colons if the lookup type is iplsearch. This is not incompatible, because
3109 previously such lookups could never work.
3111 The situation is now rather anomalous, since one *can* have colons in
3112 ordinary lsearch keys. However, making the change in all cases is
3113 incompatible and would probably break a number of configurations.
3115 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3118 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3119 conversion specification without a maximum field width, thereby enabling
3120 a rogue spamd server to cause a buffer overflow. While nobody in their
3121 right mind would setup Exim to query an untrusted spamd server, an
3122 attacker that gains access to a server running spamd could potentially
3123 exploit this vulnerability to run arbitrary code as the Exim user.
3125 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3126 $primary_hostname instead of what libspf2 thinks the hosts name is.
3128 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3129 a directory entry by the name of the lookup key. Previously, if a
3130 symlink pointed to a non-existing file or a file in a directory that
3131 Exim lacked permissions to read, a lookup for a key matching that
3132 symlink would fail. Now it is enough that a matching directory entry
3133 exists, symlink or not. (Bugzilla 503.)
3135 PH/03 The body_linecount and body_zerocount variables are now exported in the
3138 PH/04 Added the $dnslist_matched variable.
3140 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3141 This means they are set thereafter only if the connection becomes
3144 PH/06 Added the client_condition to authenticators so that some can be skipped
3145 by clients under certain conditions.
3147 PH/07 The error message for a badly-placed control=no_multiline_responses left
3148 "_responses" off the end of the name.
3150 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3152 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3153 (without spaces) instead of just copying the configuration text.
3155 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3157 PH/11 Added $max_received_linelength.
3159 PH/12 Added +ignore_defer and +include_defer to host lists.
3161 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3162 way in which PCRE > 7.0 is built.
3164 PH/14 Implemented queue_only_load_latch.
3166 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3167 MAIL command. The effect was to mangle the value on 64-bit systems.
3169 PH/16 Another patch from the Sieve maintainer.
3171 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3173 PH/18 If a system quota error occurred while trying to create the file for
3174 a maildir delivery, the message "Mailbox is full" was not appended to the
3175 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3176 only to a quota excession during the actual writing of the file.
3178 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3179 characters?) which causes problems in log lines. The DN values are now
3180 passed through string_printing() before being added to log lines.
3182 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3183 and InterBase are left for another time.)
3185 PH/21 Added message_body_newlines option.
3187 PH/22 Guard against possible overflow in moan_check_errorcopy().
3189 PH/23 POSIX allows open() to be a macro; guard against that.
3191 PH/24 If the recipient of an error message contained an @ in the local part
3192 (suitably quoted, of course), incorrect values were put in $domain and
3193 $local_part during the evaluation of errors_copy.
3199 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3200 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3203 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3204 issue a MAIL command.
3206 PH/02 In an ACL statement such as
3208 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3210 if a client was not listed at all, or was listed with a value other than
3211 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3212 the condition was not true (as it should be), so access was not denied.
3213 The bug was that the ! inversion was incorrectly passed on to the second
3214 item. This has been fixed.
3216 PH/03 Added additional dnslists conditions == and =& which are different from
3217 = and & when the dns lookup returns more than one IP address.
3219 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3220 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3222 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3223 FSYNC, which compiles an option called disable_fsync that allows for
3224 bypassing fsync(). The documentation is heavily laced with warnings.
3226 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3228 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3229 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3230 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3231 including adding "make clean"; (3) Added -fPIC when compiling the test
3232 dynamically loaded module, to get rid of a warning.
3234 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3235 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3236 = 0s. The bug is otherwise harmless.
3238 PH/07 There was a bug in the dovecot authenticator such that the value of
3239 $auth1 could be overwritten, and so not correctly preserved, after a
3240 successful authentication. This usually meant that the value preserved by
3241 the server_setid option was incorrect.
3243 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3245 PH/09 Installed PCRE release 7.0.
3247 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3248 run for batched SMTP input. It is now run at the start of every message
3249 in the batch. While fixing this I discovered that the process information
3250 (output by running exiwhat) was not always getting set for -bs and -bS
3251 input. This is fixed, and it now also says "batched" for BSMTP.
3253 PH/11 Added control=no_pipelining.
3255 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3256 patch, slightly modified), and move the expansion of helo_data till after
3257 the connection is made in the smtp transport (so it can use these
3260 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3262 PH/14 Added log_selector = +pid.
3264 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3266 PH/16 Add ${if forany and ${if forall.
3268 PH/17 Added dsn_from option to vary the From: line in DSNs.
3270 PH/18 Flush SMTP output before performing a callout, unless control =
3271 no_callout_flush is set.
3273 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3274 was true (the default) a successful delivery failed to delete the retry
3275 item, thus causing premature timeout of the address. The bug is now
3278 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3280 PH/21 Long custom messages for fakedefer and fakereject are now split up
3281 into multiline responses in the same way that messages for "deny" and
3282 other ACL rejections are.
3284 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3285 with slight modification.
3287 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3288 draft, changing the syntax and factoring some duplicate code".
3290 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3291 for deliveries of the second and subsequent messages over the same SMTP
3294 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3295 ${reduce, with only minor "tidies".
3297 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3299 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3300 expansion side effects.
3302 PH/27 When a message times out after an over-quota error from an Exim-imposed
3303 quota, the bounce message says "mailbox is full". This message was not
3304 being given when it was a system quota that was exceeded. It now should
3307 MH/03 Made $recipients available in local_scan(). local_scan() already has
3308 better access to the recipient list through recipients_list[], but
3309 $recipients can be useful in postmaster-provided expansion strings.
3311 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3312 in the case of a MAIL command with additional options following the
3313 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3314 were accidentally chopped off.
3316 PH/29 SMTP synchronization checks are implemented when a command is read -
3317 there is a check that no more input is waiting when there shouldn't be
3318 any. However, for some commands, a delay in an ACL can mean that it is
3319 some time before the response is written. In this time, more input might
3320 arrive, invalidly. So now there are extra checks after an ACL has run for
3321 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3322 pipelining has not been advertised.
3324 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3326 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3327 returned in the SMTP response when smtp_return_error_details was false.
3328 This has been fixed.
3330 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3331 instead of the C I/O that was originally supplied, because problems were
3332 reported on Solaris.
3334 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3335 Exim which did not show up earlier: it was assuming that a call to
3336 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3337 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3338 was a macro that became an assignment, so it seemed to work. This has
3339 changed to a proper function call with a void return, hence the compile
3340 error. Exim's code has been fixed.
3342 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3345 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3347 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3349 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3350 because it turns out to be too big or there is a timeout) there is no
3351 "Completed" line in the log. When some messages of this type were
3352 selected by exigrep, they were listed as "not completed". Others were
3353 picked up by some special patterns. I have improved the selection
3354 criteria to be more general.
3356 PH/38 The host_find_failed option in the manualroute router can now be set
3357 to "ignore", to completely ignore a host whose IP address cannot be
3358 found. If all hosts are ignored, the behaviour is controlled by the new
3359 host_all_ignored option.
3361 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3362 homing or because of multiple MX records with /mx) generated more than
3363 one IP address, and the following item turned out to be the local host,
3364 all the secondary addresses of the first item were incorrectly removed
3365 from the list, along with the local host and any following hosts (which
3366 is what is supposed to happen).
3368 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3369 whoever called Exim into the -H file. In the case of the daemon it was
3370 behaving confusingly. When first started, it used values for whoever
3371 started the daemon, but after a SIGHUP it used the Exim user (because it
3372 calls itself on a restart). I have changed the code so that it now always
3375 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3376 message are rejected with the same error (e.g. no authentication or bad
3377 sender address), and a DATA command is nevertheless sent (as can happen
3378 with PIPELINING or a stupid MUA), the error message that was given to the
3379 RCPT commands is included in the rejection of the DATA command. This is
3380 intended to be helpful for MUAs that show only the final error to their
3383 PH/42 Another patch from the Sieve maintainer.
3385 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3386 Eximstats - Fixed some broken HTML links and added missing column headers
3388 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3389 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3391 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3393 PH/43 Yet another patch from the Sieve maintainer.
3395 PH/44 I found a way to check for a TCP/IP connection going away before sending
3396 the response to the final '.' that terminates a message, but only in the
3397 case where the client has not sent further data following the '.'
3398 (unfortunately, this is allowed). However, in many cases there won't be
3399 any further data because there won't be any more messages to send. A call
3400 to select() can be used: if it shows that the input is "ready", there is
3401 either input waiting, or the socket has been closed. An attempt to read
3402 the next input character can distinguish the two cases. Previously, Exim
3403 would have sent an OK response which the client would never have see.
3404 This could lead to message repetition. This fix should cure that, at
3405 least in a lot of common cases.
3407 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3408 advertised in response to EHLO.
3414 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3415 fixed by 4.65/MH/01 (is this a record?) are fixed:
3417 (i) An empty string was always treated as zero by the numeric comparison
3418 operators. This behaviour has been restored.
3420 (ii) It is documented that the numeric comparison operators always treat
3421 their arguments as decimal numbers. This was broken in that numbers
3422 starting with 0 were being interpreted as octal.
3424 While fixing these problems I realized that there was another issue that
3425 hadn't been noticed. Values of message_size_limit (both the global option
3426 and the transport option) were treated as octal if they started with 0.
3427 The documentation was vague. These values are now always treated as
3428 decimal, and I will make that clear in the documentation.
3434 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3435 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3438 MH/01 Don't check that the operands of numeric comparison operators are
3439 integers when their expansion is in "skipping" mode (fixes bug
3440 introduced by 4.64-PH/07).
3442 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3443 child addresses, Exim now panics and dies. Previously, because the count
3444 is held in a short int, deliveries were likely to be lost. As such a
3445 large number of recipients for a single message is ridiculous
3446 (performance will be very, very poor), I have chosen to impose a limit
3447 rather than extend the field.
3453 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3454 leftover -K file (the existence of which was triggered by #402).
3455 While we were at it, introduced process PID as part of the -K
3456 filename. This should rule out race conditions when creating
3459 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3460 processing considerably. Previous code took too long for large mails,
3461 triggering a timeout which in turn triggers #401.
3463 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3464 in the DK code in transports.c. sendfile() is not really portable,
3465 hence the _LINUX specificness.
3467 TF/01 In the add_headers option to the mail command in an Exim filter,
3468 there was a bug that Exim would claim a syntax error in any
3469 header after the first one which had an odd number of characters
3472 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3473 callout verification, Exim cached a "reject" for the entire domain. This
3474 is correct for most verifications, but it is not correct for a recipient
3475 verification with use_sender or use_postmaster set, because in that case
3476 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3477 case of MAIL FROM:<> rejection from other early rejections (e.g.
3478 rejection of HELO). When verifying a recipient using a non-null MAIL
3479 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3480 Whatever the result of the callout, the value of the domain cache is
3481 left unchanged (for any other kind of callout, getting as far as trying
3482 RCPT means that the domain itself is ok).
3484 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3487 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3488 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3489 session, the daemon ignores ECONNECT errors and logs others; it now
3490 ignores EPIPE as well.
3492 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3493 (quoted-printable decoding).
3495 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3496 later the small subsequent patch to fix an introduced bug.
3498 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3500 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3502 PH/08 An error is now given if message_size_limit is specified negative.
3504 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3505 to be given (somewhat) arbitrary names.
3507 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3510 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3511 miscellaneous code fixes
3513 PH/10 Added the log_reject_target ACL modifier to specify where to log
3516 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3517 hostname. This is wrong, because it relates to the incoming message (and
3518 probably the interface on which it is arriving) and not to the outgoing
3519 callout (which could be using a different interface). This has been
3520 changed to use the value of the helo_data option from the smtp transport
3521 instead - this is what is used when a message is actually being sent. If
3522 there is no remote transport (possible with a router that sets up host
3523 addresses), $smtp_active_hostname is used.
3525 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3526 tweaks were necessary in order to get it to work (see also 21 below):
3527 (a) The code assumed that strncpy() returns a negative number on buffer
3528 overflow, which isn't the case. Replaced with Exim's string_format()
3530 (b) There were several signed/unsigned issues. I just did the minimum
3531 hacking in of casts. There is scope for a larger refactoring.
3532 (c) The code used strcasecmp() which is not a standard C function.
3533 Replaced with Exim's strcmpic() function.
3534 (d) The code set only $1; it now sets $auth1 as well.
3535 (e) A simple test gave the error "authentication client didn't specify
3536 service in request". It would seem that Dovecot has changed its
3537 interface. Fortunately there's a specification; I followed it and
3538 changed what the client sends and it appears to be working now.
3540 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3543 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3544 address A is aliased to B and C, where B exists and C does not. Without
3545 -v the output is "A verified" because verification stops after a
3546 successful redirection if more than one address is generated. However,
3547 with -v the child addresses are also verified. Exim was outputting "A
3548 failed to verify" and then showing the successful verification for C,
3549 with its parentage. It now outputs "B failed to verify", showing B's
3550 parentage before showing the successful verification of C.
3552 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3553 look up a TXT record in a specific list after matching in a combined
3556 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3557 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3558 they consult the DNS. I had assumed they would set it the way they
3559 wanted; and indeed my experiments on Linux seem to show that in some
3560 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3561 To be on the safe side, however, I have now made the interface to
3562 host_find_byname() similar to host_find_bydns(), with an argument
3563 containing the DNS resolver options. The host_find_byname() function now
3564 sets these options at its start, just as host_find_bydns() does. The smtp
3565 transport options dns_qualify_single and dns_search_parents are passed to
3566 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3567 of host_find_byname() use the default settings of RES_DEFNAMES
3568 (qualify_single) but not RES_DNSRCH (search_parents).
3570 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3571 spool_read_header() do less string testing, by means of a preliminary
3572 switch on the second character of optional "-foo" lines. (This is
3573 overdue, caused by the large number of possibilities that now exist.
3574 Originally there were few.) While I was there, I also converted the
3575 str(n)cmp tests so they don't re-test the leading "-" and the first
3576 character, in the hope this might squeeze out yet more improvement.
3578 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3579 flag allowing group syntax was set by the header_syntax check but not
3580 turned off, possible causing trouble later; (2) The flag was not being
3581 set at all for the header_verify test, causing "group"-style headers to
3582 be rejected. I have now set it in this case, and also caused header_
3583 verify to ignore an empty address taken from a group. While doing this, I
3584 came across some other cases where the code for allowing group syntax
3585 while scanning a header line wasn't quite right (mostly, not resetting
3586 the flag correctly in the right place). These bugs could have caused
3587 trouble for malformed header lines. I hope it is now all correct.
3589 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3590 with the "reply" argument non-NULL. The code, however (which originally
3591 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3592 but it didn't always do it. This confused somebody who was copying the
3593 code for some other use. I have removed all the tests.
3595 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3596 feature that was used to support insecure browsers during the U.S. crypto
3597 embargo. It requires special client support, and Exim is probably the
3598 only MTA that supported it -- and would never use it because real RSA is
3599 always available. This code has been removed, because it had the bad
3600 effect of slowing Exim down by computing (never used) parameters for the
3601 RSA_EXPORT functionality.
3603 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3604 authenticator to fail if there's a tab character in the incoming data
3605 (there should never be unless someone is messing about, as it's supposed
3606 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3607 if the connection is using TLS or if the remote IP is the same as the
3608 local IP, and the "valid-client-cert option" if a client certificate has
3611 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3612 authenticators. This can be used for authorization after authentication
3613 succeeds. (In the case of plaintext, it servers for both authentication
3616 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3617 if any retry times were supplied.
3619 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3620 connection, where there is no client IP address to check. In this
3621 situation, the verify now always succeeds.
3623 PH/25 Applied John Jetmore's -Mset patch.
3625 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3627 PH/27 In a string expansion for a processed (not raw) header when multiple
3628 headers of the same name were present, leading whitespace was being
3629 removed from all of them, but trailing whitespace was being removed only
3630 from the last one. Now trailing whitespace is removed from each header
3631 before concatenation. Completely empty headers in a concatenation (as
3632 before) are ignored.
3634 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3635 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3637 PH/29 [Removed. This was a change that I later backed out, and forgot to
3638 correct the ChangeLog entry (that I had efficiently created) before
3639 committing the later change.]
3641 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3642 address errors (4xx response to RCPT) over the same connection as other
3643 messages routed to the same hosts. Such deliveries are always "forced",
3644 so retry times are not inspected. This resulted in far too many retries
3645 for the affected addresses. The effect occurred only when there were more
3646 hosts than the hosts_max_try setting in the smtp transport when it had
3647 the 4xx errors. Those hosts that it had tried were not added to the list
3648 of hosts for which the message was waiting, so if all were tried, there
3649 was no problem. Two fixes have been applied:
3651 (i) If there are any address or message errors in an SMTP delivery, none
3652 of the hosts (tried or untried) are now added to the list of hosts
3653 for which the message is waiting, so the message should not be a
3654 candidate for sending over the same connection that was used for a
3655 successful delivery of some other message. This seems entirely
3656 reasonable: after all the message is NOT "waiting for some host".
3657 This is so "obvious" that I'm not sure why it wasn't done
3658 previously. Hope I haven't missed anything, but it can't do any
3659 harm, as the worst effect is to miss an optimization.
3661 (ii) If, despite (i), such a delivery is accidentally attempted, the
3662 routing retry time is respected, so at least it doesn't keep
3663 hammering the server.
3665 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3666 in ${readsocket because some servers need this prod.
3668 PH/32 Added some extra debug output when updating a wait-xxx database.
3670 PH/33 The hint "could be header name not terminated by colon", which has been
3671 given for certain expansion errors for a long time, was not being given
3672 for the ${if def:h_colon_omitted{... case.
3674 PH/34 The spec says: "With one important exception, whenever a domain list is
3675 being scanned, $domain contains the subject domain." There was at least
3676 one case where this was not true.
3678 PH/35 The error "getsockname() failed: connection reset by peer" was being
3679 written to the panic log as well as the main log, but it isn't really
3680 panic-worthy as it just means the connection died rather early on. I have
3681 removed the panic log writing for the ECONNRESET error when getsockname()
3684 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3685 runs only) independently of the message's sender address. This meant
3686 that, if the 4xx error was in fact related to the sender, a different
3687 message to the same recipient with a different sender could confuse
3688 things. In particular, this can happen when sending to a greylisting
3689 server, but other circumstances could also provoke similar problems.
3690 I have changed the default so that the retry time for these errors is now
3691 based a combination of the sender and recipient addresses. This change
3692 can be overridden by setting address_retry_include_sender=false in the
3695 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3696 remote server are returned as part of bounce messages. This was not
3697 happening for LMTP over a pipe (the lmtp transport), but now it is the
3698 same for both kinds of LMTP.
3700 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3701 in header lines that were in fact CNAMEs. This is no longer the case.
3703 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3704 and queue runs started by the daemon processed all messages. This has
3705 been fixed so that -R and -S can now usefully be given with -q<time>.
3707 PH/40 Import PCRE release 6.7 (fixes some bugs).
3709 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3711 PH/42 Give an error if -q is specified more than once.
3713 PH/43 Renamed the variables $interface_address and $interface_port as
3714 $received_ip_address and $received_port, to make it clear that these
3715 values apply to message reception, and not to the outgoing interface when
3716 a message is delivered. (The old names remain recognized, of course.)
3718 PH/44 There was no timeout on the connect() call when using a Unix domain
3719 socket in the ${readsocket expansion. There now is.
3721 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3722 be meaningful with "accept".
3724 SC/01 Eximstats V1.43
3725 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3727 SC/02 Eximstats V1.44
3728 Use a glob alias rather than an array ref in the generated
3729 parser. This improves both readability and performance.
3731 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3732 Collect SpamAssassin and rejection statistics.
3733 Don't display local sender or destination tables unless
3734 there is data to show.
3735 Added average volumes into the top table text output.
3737 SC/04 Eximstats V1.46
3738 Collect data on the number of addresses (recipients)
3739 as well as the number of messages.
3741 SC/05 Eximstats V1.47
3742 Added 'Message too big' to the list of mail rejection
3743 reasons (thanks to Marco Gaiarin).
3745 SC/06 Eximstats V1.48
3746 Mainlog lines which have GMT offsets and are too short to
3747 have a flag are now skipped.
3749 SC/07 Eximstats V1.49 (Alain Williams)
3750 Added the -emptyok flag.
3752 SC/08 Eximstats V1.50
3753 Fixes for obtaining the IP address from reject messages.
3755 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3756 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3757 whitespace changes from 4.64-PH/27
3759 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3762 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3763 are found, allow negative numbers in numeric criteria)
3765 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3767 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3768 to match changes made in 4.64-PH/43
3770 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3772 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3773 "domain" is actually an IP address, and force a failure. This locks out
3774 those revolvers/nameservers that support "A-for-A" lookups, in
3775 contravention of the specifications.
3777 PH/48 When a host name was looked up from an IP address, and the subsequent
3778 forward lookup of the name timed out, the host name was left in
3779 $sender_host_name, contrary to the specification.
3781 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3782 restricted to single-key lookups, Exim was not diagnosing an error if
3783 * or *@ was used with a query-style lookup.
3785 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3787 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3788 long ago, but noone interested enough thought of it. Let's just say that
3789 the "1.1" means that there are some new functions that weren't there at
3790 some point in the past.
3792 PH/51 Error processing for expansion failure of helo_data from an smtp
3793 transport during callout processing was broken.
3795 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3796 tested/used via the -bh/-bhc/-bs options.
3798 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3799 bug, fixed in subsequent PCRE releases).
3801 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3802 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3804 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3810 SC/01 Use a glob alias rather than an array ref in eximstats generated
3811 parser. This improves both readability and performance.
3813 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3814 Don't display local sender or destination tables in eximstats unless
3815 there is data to show.
3816 Added average volumes into the eximstats top table text output.
3818 SC/03 Collect data on the number of addresses (recipients) as well
3819 as the number of messages in eximstats.
3821 TF/01 Correct an error in the documentation for the redirect router. Exim
3822 does (usually) call initgroups() when daemonizing.
3824 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3825 with consistent privilege compared to when running as a daemon.
3827 TF/03 Note in the spec that $authenticated_id is not set for local
3828 submissions from trusted users.
3830 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3831 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3833 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3834 by adding some example configuration directives to the default
3835 configuration file. A little bit of work is required to uncomment the
3836 directives and define how usernames and passwords are checked, but
3837 there is now a framework to start from.
3839 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3840 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3841 without this. I don't know how relevant this is to other LDAP libraries.
3843 PH/02 Add the verb name to the "unknown ACL verb" error.
3845 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3847 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3849 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3850 directory not expanded when it should be if an expanded home directory
3851 was set for the address (which is overridden by the transport).
3853 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3856 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3857 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3858 because it is too late at that time, and has no effect.
3860 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3861 security issue with \' (bugzilla #107). I could not use the
3862 PQescapeStringConn() function, because it needs a PGconn value as one of
3865 PH/08 When testing addresses using -bt, indicate those final addresses that
3866 are duplicates that would not cause an additional delivery. At least one
3867 person was confused, thinking that -bt output corresponded to deliveries.
3868 (Suppressing duplicates isn't a good idea as you lose the information
3869 about possibly different redirections that led to the duplicates.)
3871 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3872 systems where poll() doesn't work, in particular OS X.
3874 PH/10 Added more information to debugging output for retry time not reached.
3876 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3877 operations in malware.c.
3879 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3882 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3883 syslog_duplication set false, log lines that would normally be written
3884 both the the main log and to the reject log were not written to syslog at
3887 PH/14 In the default configuration, change the use of "message" in ACL warn
3888 statements to "add_header".
3890 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3891 not followed by a command (e.g. "seen endif").
3893 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3894 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3897 PH/17 Added extra conditions to the default value of delay_warning_condition
3901 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3902 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3903 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3906 The Auto-Submitted: and various List- headers are standardised, whereas I
3907 don't think Precedence: ever was.
3909 PH/18 Refactored debugging code in route_finduser() to show more information,
3910 in particular, the error code if getpwnam() issues one.
3912 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3913 This is apparently needed in addition to the PH/07 change above to avoid
3914 any possible encoding problems.
3916 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3917 but not after initializing Perl.
3919 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3920 output them only if debugging. By default they are written stderr,
3921 apparently, which is not desirable.
3923 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3926 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3929 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3931 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3932 authenticated or an ident call has been made. Suppress the default
3933 values for $authenticated_id and $authenticated_sender (but permit -oMai
3934 and -oMas) when testing with -bh.
3936 PH/24 Re-jigged the order of the tests in the default configuration so that the
3937 tests for valid domains and recipients precede the DNS black list and CSA
3938 tests, on the grounds that those ones are more expensive.
3940 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3941 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3942 This bug exists in every version of Exim that I still have, right back to
3945 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3946 However, an attempt to turn on case-sensitivity in a regex key by
3947 including (?-i) didn't work because the subject string was already
3948 lowercased, and the effects were non-intuitive. It turns out that a
3949 one-line patch can be used to allow (?-i) to work as expected.
3955 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3956 other effects) broke the use of negated acl sub-conditions.
3958 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3961 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3962 "Deny" causes Exim to reject the incoming connection with a 554 error.
3963 Unfortunately, if there is a major crisis, such as a disk failure,
3964 tcp-wrappers gives "deny", whereas what one would like would be some
3965 kind of temporary error. A kludge has been added to help with this.
3966 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3967 554 error is used if errno is still zero or contains ENOENT (which occurs
3968 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3971 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3973 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3974 errors. Otherwise a message that provokes a temporary error (when other
3975 messages do not) can cause a whole host to time out.
3977 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3978 addresses were routed directly to files or pipes from a redirect router.
3979 File deliveries just didn't batch; pipe deliveries might have suffered
3982 PH/06 A failure to get a lock for a hints database would erroneously always say
3983 "Failed to get write lock", even when it was really a read lock.
3985 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3986 of 0600. This has been changed to use the value of the lockfile_mode
3987 option (which defaults to 0600).
3989 PH/08 Applied small patch from the Sieve maintainer.
3991 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3992 folder from quota calculations, a direct delivery into this folder messed
3993 up the contents of the maildirsize file. This was because the regex was
3994 used only to exclude .Trash (or whatever) when the size of the mailbox
3995 was calculated. There was no check that a delivery was happening into an
3996 excluded directory. This bug has been fixed by ignoring all quota
3997 processing for deliveries into excluded directories.
3999 PH/10 Added the maildirfolder_create_regex option to appendfile.
4005 PH/01 The code for finding all the local interface addresses on a FreeBSD
4006 system running IPv6 was broken. This may well have applied to all BSD
4007 systems, as well as to others that have similar system calls. The broken
4008 code found IPv4 interfaces correctly, but gave incorrect values for the
4009 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4010 that it would not match correctly against @[] and not recognize the IPv6
4013 PH/02 The ipliteral router was not recognizing addresses of the form user@
4014 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4016 PH/03 Added disable_ipv6.
4018 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4019 decoded form, because it is most often used to construct To: headers
4020 lines in autoreplies, and the decoded form may well be syntactically
4021 invalid. However, $reply_address has leading white space removed, and all
4022 newlines turned into spaces so that the autoreply transport does not
4025 PH/05 If group was specified without a user on a router, and no group or user
4026 was specified on a transport, the group from the router was ignored.
4028 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4029 for visible compile-time settings that can be used to change these
4030 numbers, for those that want even more. Backwards compatibility with old
4031 spool files has been maintained. However, going back to a previous Exim
4032 release will lost any variables that are in spool files.
4034 PH/07 Two small changes when running in the test harness: increase delay when
4035 passing a TCP/IP connection to a new process, in case the original
4036 process has to generate a bounce, and remove special handling of
4037 127.0.0.2 (sic), which is no longer necessary.
4039 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4040 be the same on different OS.
4042 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4045 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4046 whether --show-vars was specified or not
4048 JJ/02 exipick: Added support for new ACL variable spool format introduced
4051 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4052 syntactically invalid From: or Reply-to: line, and a filter used this to
4053 generate an autoreply, and therefore failed to obtain an address for the
4054 autoreply, Exim could try to deliver to a non-existent relative file
4055 name, causing unrelated and misleading errors. What now happens is that
4056 it logs this as a hard delivery error, but does not attempt to create a
4059 PH/11 The exinext utility has a -C option for testing purposes, but although
4060 the given file was scanned by exinext itself; it wasn't being passed on
4061 when Exim was called.
4063 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4064 an end-of-file indication when reading a command response.
4066 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4067 compiled. In many other places in Exim, IPv6 addresses are always
4068 recognized, so I have changed this. It also means that IPv4 domain
4069 literals of the form [IPV4:n.n.n.n] are now always recognized.
4071 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4072 used if the router is not running as root, for example, when verifying at
4073 ACL time, or when using -bh. The debugging output from this situation was
4074 non-existent - all you got was a failure to exec. I have made two
4077 (a) Failures to set uid/gid, the current directory, or a process leader
4078 in a subprocess such as that created by queryprogram now generate
4079 suitable debugging output when -d is set.
4081 (b) The queryprogram router detects when it is not running as root,
4082 outputs suitable debugging information if -d is set, and then runs
4083 the subprocess without attempting to change uid/gid.
4085 PH/15 Minor change to Makefile for building test_host (undocumented testing
4088 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4089 additional section of a DNS packet that returns MX or SRV records.
4090 Instead, it always explicitly searches for A/AAAA records. This avoids
4091 major problems that occur when a DNS server includes only records of one
4092 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4093 fixed another bug: if SRV records were looked up and the corresponding
4094 address records were *not* found in the additional section, the port
4095 values from the SRV records were lost.
4097 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4098 using the correct key (the original address) when searching the retry
4099 rules in order to find which one to use for generating the retry hint.
4101 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4102 adding the default one. Similarly, if it contains a Reply-To: header, the
4103 errors_reply_to option, if set, is not used.
4105 PH/19 When calculating a retry time, Exim used to measure the "time since
4106 failure" by looking at the "first failed" field in the retry record. Now
4107 it does not use this if it is later than than the arrival time of the
4108 message. Instead it uses the arrival time. This makes for better
4109 behaviour in cases where some deliveries succeed, thus re-setting the
4110 "first failed" field. An example is a quota failure for a huge message
4111 when small messages continue to be delivered. Without this change, the
4112 "time since failure" will always be short, possible causing more frequent
4113 delivery attempts for the huge message than are intended.
4114 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4116 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4117 $1, $2, $3) because the numerical variables can be reset during some
4118 expansion items (e.g. "match"), thereby losing the authentication data.
4120 PH/21 Make -bV show the size of off_t variables so that the test suite can
4121 decide whether to run tests for quotas > 2G.
4123 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4124 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4125 filecount value is greater than 2G or if a quota value is greater than 2G
4126 on a system where the size of off_t is not greater than 4, a panic error
4129 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4130 never match. The debug and -bh output now contains an explicit error
4131 message indicating a malformed IPv4 address or mask.
4133 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4134 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4135 PH/23 above applies.
4137 PH/25 Do not write to syslog when running in the test harness. The only
4138 occasion when this arises is a failure to open the main or panic logs
4139 (for which there is an explicit test).
4141 PH/26 Added the /no_tell option to "control=freeze".
4143 PH/27 If a host name lookup failed very early in a connection, for example, if
4144 the IP address matched host_lookup and the reverse lookup yielded a name
4145 that did not have a forward lookup, an error message of the form "no IP
4146 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4147 could be logged. Now it outputs the IP address instead of "NULL".
4149 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4150 allows the sender and the authenticated sender to be set when
4151 submitting a message from within Exim. Since child_open_exim() is
4152 documented for local_scan(), the new function should be too.
4154 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4155 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4156 results in an empty string is now treated as unset.
4158 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4160 PH/31 Added sender_verify_fail logging option.
4162 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4163 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4164 tidied the source and removed it altogether.
4166 PH/33 When a queue run was abandoned because the load average was too high, a
4167 log line was always written; now it is written only if the queue_run log
4168 selector is set. In addition, the log line for abandonment now contains
4169 information about the queue run such as the pid. This is always present
4170 in "start" and "stop" lines but was omitted from the "abandon" line.
4172 PH/34 Omit spaces between a header name and the colon in the error message that
4173 is given when verify = headers_syntax fails (if there are lots of them,
4174 the message gets confusing).
4176 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4177 names, as there are now some PTR records that contain slashes. This check
4178 is only to protect against broken name servers that fall over on strange
4179 characters, so the fact that it applies to all lookups doesn't matter.
4181 PH/36 Now that the new test suite is complete, we can remove some of the
4182 special code in Exim that was needed for the old test suite. For example,
4183 sorting DNS records because real resolvers return them in an arbitrary
4184 order. The new test suite's fake resolver always returns records in the
4187 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4188 bounces) except when queue_only is set, to avoid logging races between
4189 the different processes.
4191 PH/38 Panic-die if .include specifies a non-absolute path.
4193 PH/39 A tweak to the "H" retry rule from its user.
4195 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4196 a label. They prevented compilation on older perls.
4198 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4199 a warning to be raised on newish perls.
4201 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4202 on queue. Changes to match documented behaviour of showing count of
4203 messages matching specified criteria.
4205 PH/40 Changed the default ident timeout from 30s to 5s.
4207 PH/41 Added support for the use of login_cap features, on those BSD systems
4208 that have them, for controlling the resources used by pipe deliveries.
4210 PH/42 The content-scanning code uses fopen() to create files in which to put
4211 message data. Previously it was not paying any attention to the mode of
4212 the files. Exim runs with umask(0) because the rest of the code creates
4213 files with open(), and sets the required mode explicitly. Thus, these
4214 files were ending up world-writeable. This was not a big issue, because,
4215 being within the spool directory, they were not world-accessible. I have
4216 created a function called modefopen, which takes an additional mode
4217 argument. It sets umask(777), creates the file, chmods it to the required
4218 mode, then resets the umask. All the relevant calls to fopen() in the
4219 content scanning code have been changed to use this function.
4221 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4222 to 24 hours. This avoids potential overflow problems when processing G
4223 and H retry rules. I suspect nobody ever tinkers with this value.
4225 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4227 PH/45 When the plaintext authenticator is running as a client, the server's
4228 challenges are checked to ensure they are valid base64 strings. By
4229 default, the authentication attempt is cancelled if an invalid string is
4230 received. Setting client_ignore_invalid_base64 true ignores these errors.
4231 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4232 they are received. Thus, the responses can be made to depend on the
4233 challenges. If an invalid string is ignored, an empty string is placed in
4236 PH/46 Messages that are created by the autoreply transport now contains a
4237 References: header, in accordance with RFCs 2822 and 3834.
4239 PH/47 Added authenticated_sender_force to the smtp transport.
4241 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4243 PH/49 Installed latest patch from the Sieve maintainer.
4245 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4246 was also set, the appendfile transport was unnecessarily scanning a
4247 directory of message files (e.g. for maildir delivery) to find the count
4248 of files (along with the size), even though it did not need this
4249 information. It now does the scan only if it needs to find either the
4250 size of the count of files.
4252 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4254 PH/52 Two bugs concerned with error handling when the smtp transport is
4257 (i) Exim was not creating retry information for temporary errors given
4258 for individual recipients after the DATA command when the smtp transport
4259 was used in LMTP mode. This meant that they could be retried too
4260 frequently, and not timed out correctly.
4262 (ii) Exim was setting the flag that allows error details to be returned
4263 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4264 recipients that were returned after the DATA command.
4266 PH/53 This is related to PH/52, but is more general: for any failing address,
4267 when detailed error information was permitted to be returned to the
4268 sender, but the error was temporary, then after the final timeout, only
4269 "retry timeout exceeded" was returned. Now it returns the full error as
4270 well as "retry timeout exceeded".
4272 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4273 do this, and (what is worse) MTAs that accept it.
4275 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4276 will now be deprecated.
4278 PH/56 New os.c-cygwin from the Cygwin maintainer.
4280 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4281 formats (previously only available in exim formats via -bpr, -bpru,
4282 and -bpra. Now also available in native and exiqgrep formats)
4284 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4285 with very large, slow to parse queues
4287 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4289 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4291 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4292 responses to authentication challenges, though it was showing the
4293 challenges; (ii) I've removed the CR characters from the debug output for
4296 PH/58 Allow for the insertion of a newline as well as a space when a string
4297 is turned into more than one encoded-word during RFC 2047 encoding. The
4298 Sieve code now uses this.
4300 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4301 data_4xx, lost_connection, tls_required.
4303 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4304 message was being sent as an SMTP response.
4306 PH/61 Add -l and -k options to exicyclog.
4308 PH/62 When verifying, if an address was redirected to one new address, so that
4309 verification continued, and the new address failed or deferred after
4310 having set something in $address_data, the value of $address_data was not
4311 passed back to the ACL. This was different to the case when no
4312 redirection occurred. The value is now passed back in both cases.
4314 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4315 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4316 use login_cap.h, so on its own it isn't the distinguishing feature. The
4317 new name refers directly to the setclassresources() function.
4319 PH/65 Added configuration files for NetBSD3.
4321 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4323 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4324 is preferred over IPv4.
4326 PH/68 The bounce_return_message and bounce_return_body options were not being
4327 honoured for bounces generated during the reception of non-SMTP messages.
4328 In particular, this applied to messages rejected by the ACL. This bug has
4329 been fixed. However, if bounce_return_message is true and bounce_return_
4330 body is false, the headers that are returned for a non-SMTP message
4331 include only those that have been read before the error was detected.
4332 (In the case of an ACL rejection, they have all been read.)
4334 PH/69 The HTML version of the specification is now built in a directory called
4335 spec_html instead of spec.html, because the latter looks like a path with
4336 a MIME-type, and this confuses some software.
4338 PH/70 Catch two compiler warnings in sieve.c.
4340 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4341 function verify_get_ident() calls ip_connect() to connect a socket, but
4342 if the "connect()" function timed out, ip_connect() used to close the
4343 socket. However, verify_get_ident() also closes the socket later, and in
4344 between Exim writes to the log, which may get opened at this point. When
4345 the socket was closed in ip_connect(), the log could get the same file
4346 descriptor number as the socket. This naturally causes chaos. The fix is
4347 not to close the socket in ip_connect(); the socket should be closed by
4348 the function that creates it. There was only one place in the code where
4349 this was missing, in the iplookup router, which I don't think anybody now
4350 uses, but I've fixed it anyway.
4352 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4353 well as to direct DNS lookups. Otherwise the handling of names in host
4354 lists is inconsistent and therefore confusing.
4360 PH/01 Two changes to the default runtime configuration:
4362 (1) Move the checks for relay_from_hosts and authenticated clients from
4363 after to before the (commented out) DNS black list checks.
4365 (2) Add control=submission to the relay_from_hosts and authenticated
4366 clients checks, on the grounds that messages accepted by these
4367 statements are most likely to be submissions.
4369 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4371 (1) Generate an error if the third argument for the ${prvs expansion is
4374 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4377 (3) Reset the variables that are obtained from the first argument of
4378 ${prvscheck and used in the second argument before leaving the code,
4379 because their memory is reclaimed, so using them afterwards may do
4382 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4383 one (it's much easier than Tom thought :-).
4385 (5) Because of (4), we can now allow for the use of $prvscheck_result
4386 inside the third argument.
4388 PH/03 For some reason, the default setting of PATH when running a command from
4389 a pipe transport was just "/usr/bin". I have changed it to
4392 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4393 anything to be listed in the output from -bV.
4395 PH/05 When a filter generated an autoreply, the entire To: header line was
4396 quoted in the delivery log line, like this:
4398 => >A.N.Other <ano@some.domain> <original@ddress> ...
4400 This has been changed so that it extracts the operative address. There
4401 may be more than one such address. If so, they are comma-separated, like
4404 => >ano@some.domain,ona@other.domain <original@ddress> ...
4406 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4407 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4408 not being looked up in the rDNS to get a host name, Exim was showing the
4409 IP address twice in Received: lines, even though the IP addresses were
4410 identical. For example:
4412 Received: from [1.2.3.4] (helo=[1.2.3.4])
4414 However, if the real host name was known, it was omitting the HELO data
4415 if it matched the actual IP address. This has been tidied up so that it
4416 doesn't show the same IP address twice.
4418 PH/07 When both +timestamp and +memory debugging was on, the value given by
4419 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4420 called by the debug printing, thereby overwriting the timestamp buffer.
4421 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4423 PH/08 When the original message was included in an autoreply transport, it
4424 always said "this is a copy of the message, including all the headers",
4425 even if body_only or headers_only was set. It now gives an appropriate
4428 PH/09 Applied a patch from the Sieve maintainer which:
4430 o fixes some comments
4431 o adds the (disabled) notify extension core
4432 o adds some debug output for the result of if/elsif tests
4433 o points to the current vacation draft in the documentation
4434 and documents the missing references header update
4438 o fixes a bug in processing the envelope test (when testing
4439 multiple envelope elements, the last element determined the
4442 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4443 Electronic Mail") by including:
4445 Auto-submitted: auto-generated
4447 in the messages that it generates (bounce messages and others, such as
4448 warnings). In the case of bounce messages for non-SMTP messages, there was
4449 also a typo: it was using "Auto_submitted" (underscore instead of
4450 hyphen). Since every message generated by Exim is necessarily in response
4451 to another message, thes have all been changed to:
4453 Auto-Submitted: auto-replied
4455 in accordance with these statements in the RFC:
4457 The auto-replied keyword:
4459 - SHOULD be used on messages sent in direct response to another
4460 message by an automatic process,
4462 - MUST NOT be used on manually-generated messages,
4464 - MAY be used on Delivery Status Notifications (DSNs) and Message
4465 Disposition Notifications (MDNs),
4467 - MUST NOT be used on messages generated by automatic or periodic
4468 processes, except for messages which are automatic responses to
4471 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4472 to the default Received: header definition.
4474 PH/12 Added log selector acl_warn_skipped (default on).
4476 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4477 variables because (a) they are in the wrong storage pool and (b) even if
4478 they were copied, it wouldn't work properly because of the caching.
4480 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4481 checking when decoding. Apparently there are clients that generate
4482 overlong encoded strings. Why am I not surprised?
4484 PH/15 If the first argument of "${if match_address" was not empty, but did not
4485 contain an "@" character, Exim crashed. Now it writes a panic log message
4486 and treats the condition as false.
4488 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4490 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4491 extension "envelope-auth". The code is finished and in agreement with
4492 other implementations, but there is no documentation so far and in fact,
4493 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4494 not changing the active code.
4496 Print executed "if" and "elsif" statements when debugging is used. This
4497 helps a great deal to understand what a filter does.
4499 Document more things not specified clearly in RFC3028. I had all this
4500 sorted out, when out of a sudden new issues came to my mind. Oops."
4502 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4505 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4506 canonical form (as documented). However, after a host name lookup from
4507 the IP address, check_host() was doing a simple string comparison with
4508 addresses acquired from the DNS when checking that the found name did
4509 have the original IP as one of its addresses. Since any found IPv6
4510 addresses are likely to be in abbreviated form, the comparison could
4511 fail. Luckily, there already exists a function for doing the comparison
4512 by converting both addresses to binary, so now that is used instead of
4513 the text comparison.
4515 PH/20 There was another similar case to PH/19, when a complete host name was
4516 given in a host list; looking up its IP address could give an abbreviated
4517 form, whereas the current host's name might or might not be abbreviated.
4518 The same fix has been applied.
4524 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4525 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4528 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4529 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4531 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4533 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4534 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4535 Also, for "Auto-Submitted", treat anything other than "no" as
4536 non-personal, in accordance with RFC 3834. (Previously it treated
4537 anything starting "auto-" as non-personal.)
4539 TF/01 The control=submission/name=... option had a problem with syntax
4540 errors if the name included a slash character. The /name= option
4541 now slurps the rest of the string, so it can include any characters
4542 but it must come last in the list of options (after /sender_retain
4545 PH/05 Some modifications to the interface to the fake nameserver for the new
4553 TK/01 Added the "success_on_redirect" address verification option. See
4554 NewStuff for rationale and an example.
4556 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4558 PH/02 Patch to exigrep to allow it to work on syslog lines.
4560 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4561 fread() to skip over the body file's header line, because in Cygwin the
4562 header line is locked and is inaccessible.
4564 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4565 co-exist for some time) to make it clear that it is the Exim ID that is
4566 referenced, not the Message-ID: header line.
4568 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4569 string_format() function, because snprintf() does not exist on all
4572 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4573 ${stat: expansion item.
4575 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4576 protocol synchronization error", to keep the pedants happy.
4578 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4579 well as for IRIX systems, when gcc is being used. See the host.c source
4582 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4584 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4587 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4588 transport and to the smtp transport in LMTP mode.
4590 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4592 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4593 run a filter in a subprocess. This could lead to confusion in subsequent
4594 lookups in the parent process. There should also be a search_tidyup() at
4595 the end of the subprocess.
4597 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4598 only if the host matched helo_try_verify_hosts, which caused the
4599 verification to occur when the EHLO/HELO command was issued. The ACL just
4600 tested the remembered result. Now, if a previous verification attempt has
4601 not happened, "verify = helo" does it there and then.
4603 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4605 TK/03 Fix log output including CR from clamd.
4607 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4608 exist provoked a memory error which could cause a segfault.
4610 PH/15 Installed PCRE 6.2
4612 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4614 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4615 of the problem. Specifically, suggested +O2 rather than +O1 for the
4618 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4620 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4621 record, the port was not correctly logged when the outgoing_port log
4622 selector was set (it logged the transort's default port).
4624 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4625 fallback_hosts, and "hosts" in the smtp transport.
4627 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4628 host errors such as "Connection refused".
4630 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4631 authentication with radiusclient 0.4.9:
4633 - Error returned from rc_read_config was caught wrongly
4634 - Username/password not passed on to radius server due to wrong length.
4636 The presumption is that some radiusclient API changes for 4.51/PH/17
4637 were not taken care of correctly. The code is still untested by me (my
4638 Linux distribution still has 0.3.2 of radiusclient), but it was
4639 contributed by a Radius user.
4641 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4642 expanding the "port" option of the smtp transport.
4644 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4645 while reading a MIME header. Thanks to Tom Hughes for a patch.
4647 PH/24 Include config.h inside local_scan.h so that configuration settings are
4650 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4651 that in an ACL for RCPT (for example), you can examine exactly what was
4654 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4655 commands, but it was not correctly comparing the address with the actual
4656 client host address. Thus, it would show the EHLO address in Received:
4657 header lines when this was not necessary.
4659 PH/27 Added the % operator to ${eval:}.
4661 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4662 it should be ignoring failures (because with -C, for example, it has lost
4663 privilege). It wasn't ignoring creation failures other than "already
4666 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4669 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4670 because an input line was too long, either on its own, or by virtue of
4671 too many continuations, the temporary file was not being removed, and the
4672 return code was incorrect.
4674 PH/31 Missing "BOOL" in function definition in filtertest.c.
4676 PH/32 Applied Sieve patches from the maintainer.
4678 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4680 PH/33 Added "verify = not_blind".
4682 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4683 Local/Makefile (with some defaults set). These are used in built scripts
4684 such as exicyclog, but they have never been used in the exim_install
4685 script (though there are many overriding facilities there). I have
4686 arranged that the exim_install script now takes note of these two
4689 PH/35 Installed configuration files for Dragonfly.
4691 PH/36 When a locally submitted message by a trusted user did not contain a
4692 From: header, and the sender address was obtained from -f or from an SMTP
4693 MAIL command, and the trusted user did not use -F to supply a sender
4694 name, $originator_name was incorrectly used when constructing a From:
4695 header. Furthermore, $originator_name was used for submission mode
4696 messages from external hosts without From: headers in a similar way,
4697 which is clearly wrong.
4699 PH/37 Added control=suppress_local_fixups.
4701 PH/38 When log_selector = +received_sender was set, and the addition of the
4702 sender made the log line's construction buffer exactly full, or one byte
4703 less than full, an overflow happened when the terminating "\n" was
4706 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4707 when the result of a list match is failure because a DNS lookup failed.
4709 PH/40 RM_COMMAND is now used in the building process.
4711 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4712 the "build-* directories that it finds.
4714 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4715 address in a domain literal was a prefix of an interface address.
4717 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4718 when verifying a sender address, unless rewrite_headers is false.
4720 PH/44 Wrote a long comment about why errors_to addresses are verified as
4721 recipients, not senders.
4723 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4724 the ratelimit ACL was added.
4726 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4728 PH/46 Added extra information about PostgreSQL errors to the error string.
4730 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4731 avoiding the need to install special zones in a real server. This is
4732 backwards compatible; if it can't find the fake resolver, it drops back.
4733 Thus, both old and new test suites can be run.
4735 TF/02 Added util/ratelimit.pl
4737 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4738 clock is set back in time.
4740 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4741 Candler <B.Candler@pobox.com>.
4743 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4744 OK for addresses that are the result of redirections.
4746 PH/48 A number of further additions for the benefit of the new test suite,
4747 including a fake gethostbyname() that interfaces to the fake DNS resolver
4750 TF/06 The fix for widen_domains has also been applied to qualify_single and
4751 search_parents which are the other dnslookup options that can cause
4754 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4757 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4759 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4760 service" if it encountered an MX record with an empty target hostname.
4761 The message is now "an MX or SRV record indicated no SMTP service".
4763 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4764 if the DNS of the sending site is misconfigured. This is quite a
4765 common situation. This change restores the behaviour of treating a
4766 helo verification defer as a failure.
4768 PH/51 If self=fail was set on a router, the bounce message did not include the
4769 actual error message.
4775 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4777 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4778 command itself ended in error, the underlying message about the transport
4779 filter timeout was being overwritten with the pipe command error. Now the
4780 underlying error message should be appended to the second error message.
4782 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4784 PH/02 Reduce the amount of output that "make" produces by default. Full output
4785 can still be requested.
4787 PH/03 The warning log line about a condition test deferring for a "warn" verb
4788 was being output only once per connection, rather than after each
4789 occurrence (because it was using the same function as for successful
4790 "warn" verbs). This seems wrong, so I have changed it.
4792 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4793 it should not have, which might have caused a crash in the right
4794 circumstances, but probably never did.
4796 PH/04 Installed a modified version of Tony Finch's patch to make submission
4797 mode fix the return path as well as the Sender: header line, and to
4798 add a /name= option so that you can make the user's friendly name appear
4801 TF/03 Added the control = fakedefer ACL modifier.
4803 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4804 Mark Lowes for thorough testing.
4806 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4808 TK/03 Merged latest SRS patch from Miles Wilton.
4810 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4811 with the definition in sysexits.h (which is #included earlier).
4812 Fortunately, Exim does not actually use EX_OK. The code used to try to
4813 preserve the sysexits.h value, by assuming that macro definitions were
4814 scanned for macro replacements. I have been disabused of this notion,
4815 so now the code just undefines EX_OK before #including unistd.h.
4817 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4818 in the smtp transport. When a block could not be written in a single
4819 write() function, the timeout was being re-applied to each part-write.
4820 This seems wrong - if the receiver was accepting one byte at a time it
4821 would take for ever. The timeout is now adjusted when this happens. It
4822 doesn't have to be particularly precise.
4824 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4825 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4827 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4828 without a domain if the check to <postmaster@domain> fails.
4830 SC/01 Eximstats: added -xls and the ability to specify output files
4831 (patch written by Frank Heydlauf).
4833 SC/02 Eximstats: use FileHandles for outputting results.
4835 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4837 SC/04 Eximstats: fixed display of large numbers with -nvr option
4839 SC/05 Eximstats: fixed merging of reports with empty tables.
4841 SC/06 Eximstats: added the -include_original_destination flag
4843 SC/07 Eximstats: removed tabs and trailing whitespace.
4845 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4847 TK/06 MBOX spool code: Add real "From " MBOX separator line
4848 so the .eml file is really in mbox format (even though
4849 most programs do not really care). Patch from Alex Miller.
4851 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4852 The latter is generated from $received_to and is only set if the
4853 message has one envelope recipient. SA can use these headers,
4854 obviously out-of-the-box. Patch from Alex Miller.
4856 PH/08 The ${def test on a variable was returning false if the variable's
4857 value was "0", contrary to what the specification has always said!
4858 The result should be true unless the variable is empty.
4860 PH/09 The syntax error of a character other than { following "${if
4861 def:variable_name" (after optional whitespace) was not being diagnosed.
4862 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4863 accidental colon was present, for example, could give incorrect results.
4865 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4866 result is used (not including appendfile, where other changes are about
4869 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4870 This involved changing a lot of size variables from int to off_t. It
4871 should work with maildirs and everything.
4873 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4874 spamd dying while we are connected to it.
4876 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4879 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4880 function for BDB 4.3.
4882 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4884 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4885 the sources for PCRE have been re-arranged and more files are now
4888 PH/15 The code I had for printing potentially long long variables in PH/11
4889 above was not the best (it lost precision). The length of off_t variables
4890 is now inspected at build time, and an appropriate printing format (%ld
4891 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4892 to be "long long int" or "long int". This is needed for the internal
4893 formatting function string_vformat().
4895 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4896 the configuration file to be ":syslog", then the script "guesses" where
4897 the logs files are, rather than using the compiled in default. In our
4898 case the guess is not the same as the compiled default, so the script
4899 suddenly stopped working when I started to use syslog. The patch checks
4900 to see if log_file_path is "". If so, it attempts to read it from exim
4901 with no configuration file to get the compiled in version, before it
4902 falls back to the previous guessing code."
4904 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4905 implementing BATV in an Exim configuration. See NewStuff for the gory
4908 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4909 Makefile that are specific to HP-UX.
4911 PH/18 If the "use_postmaster" option was set for a recipient callout together
4912 with the "random" option, the postmaster address was used as the MAIL
4913 FROM address for the random test, but not for the subsequent recipient
4914 test. It is now used for both.
4916 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4917 patch removes a few documentation additions to RFC 3028, because the
4918 latest draft now contains them. It adds the new en;ascii-case comparator
4919 and a new error check for 8bit text in MIME parts. Comparator and
4920 require names are now matched exactly. I enabled the subaddress
4921 extension, but it is not well tested yet (read: it works for me)."
4923 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4924 rework some of the code of TK/09 above to avoid the hardwired use of
4925 "%lld" and "long long". Replaced the call to snprintf() with a call to
4928 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4929 records point to non-existent hosts", "retry timeout exceeded", and
4930 "retry time not reached for any host after a long failure period".
4932 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4933 experimental DomainKeys support:
4935 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4936 (2) On an error such as an illegally used "control", the wrong name for
4937 the control was given.
4939 These problems did NOT occur unless DomainKeys support was compiled.
4941 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4943 PH/24 Added ${if match_ip condition.
4945 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4946 databases so that it will be absolutely obvious if a crash occurs in the
4947 DB library. This is a regular occurrence (often caused by mis-matched
4950 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4951 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4952 picked up on a user's system that detects such things. There doesn't seem
4953 to be a gcc warning option for this - only an attribute that has to be
4954 put on the function's prototype. It seems that in Fedora Core 4 they have
4955 set this on a number of new functions. No doubt there will be more in due
4958 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4959 specify a transport. However, if an address that was verified by such a
4960 router was the subject of a callout, Exim crashed because it tried to
4961 read the rcpt_include_affixes from the non-existent transport. Now it
4962 just assumes that the setting of that option is false. This bug was
4963 introduced by 4.51/PH/31.
4965 PH/28 Changed -d+all to exclude +memory, because that information is very
4966 rarely of interest, but it makes the output a lot bigger. People tend to
4967 do -d+all out of habit.
4969 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4970 code in os-type was giving problems when libc.so lives in lib64, like on
4973 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4974 aren't the modern standard, and it seems that some systems' include files
4975 don't always have them. Exim was already checking for some of the newer
4976 ones like T_AAAA, and defining it itself. I've added checks for all the
4977 record types that Exim uses.
4979 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4980 not automatically generating a new one, as it is supposed to. This
4981 prevented TLS from working. If the file did exist, but contained invalid
4982 data, a new version was generated, as expected. It was only the case of a
4983 non-existent file that was broken.
4985 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4986 with a change in libdomainkeys > 0.64.
4988 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4989 from DNS. If the selector record carries the flag, it now has
4990 precedence over the domain-wide flag.
4992 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4994 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4995 the use of a port name that isn't defined in /etc/services) occurred, the
4996 message was deferred as in a normal delivery, and thus remained on the
4997 spool, instead of being failed because of the mua_wrapper setting. This
4998 is now fixed, and I tidied up some of the mua_wrapper messages at the
5001 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5002 the messages in a hash of arrays rather than using individual hashes.
5003 This is a bit cleaner and results in dramatic memory savings, albeit
5004 at a slight CPU cost.
5006 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5007 as requested by Marc Sherman.
5009 SC/10 Eximstats: added histograms for user specified patterns as requested
5012 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5014 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5015 fopen() in the content-scanning modules that did not already have it.
5021 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5022 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5024 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5026 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5028 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5029 to test Sieve filters that use "vacation".
5031 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5032 that changes the way the GnuTLS parameters are stored in the cache file.
5033 The new format can be generated externally. For backward compatibility,
5034 if the data in the cache doesn't make sense, Exim assumes it has read an
5035 old-format file, and it generates new data and writes a new file. This
5036 means that you can't go back to an older release without removing the
5039 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5040 work if there are any delivery delays because "one_time" forces the
5041 parent to be marked "delivered", so its unseen clone is never tried
5042 again. For this reason, Exim now forbids the simultaneous setting of
5045 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5046 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5047 entry if you want to know the details. The fix, however, neglected to
5048 consider the case when local delivery batching is involved. The test for
5049 "previously delivered" was not happening when checking to see if an
5050 address could be batched with a previous (undelivered) one; under
5051 certain circumstances this could lead to multiple deliveries to the same
5054 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5055 in its include files, and this causes problems building Exim.
5057 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5058 header_syntax) but Exim was just ignoring anything given after a slash.
5059 In particular, this caused confusion with an attempt to use "verify =
5060 reverse_host_lookup/defer_ok". An error is now given when options are
5061 supplied for verify items that do not have them. (Maybe reverse_host_
5062 lookup should have a defer_ok option, but that's a different point.)
5064 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5065 defined by RFC 821) to 2048, because there were problems with some AUTH
5066 commands, and RFC 1869 says the size should be increased for extended
5067 SMTP commands that take arguments.
5069 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5072 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5073 "unknown" error; now it says that the functionality isn't in the binary.
5075 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5076 an address' error message when a string expansion fails (syntax or
5077 whatever). Otherwise the password may appear in the log. Following change
5078 PH/42 below, there is no longer a chance of it appearing in a bounce
5081 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5083 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5084 of its addresses was ever tried. (Bugzilla bug #2.)
5086 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5087 the result incorrectly in the debug output. (It correctly added a newline
5088 to what was transported.)
5090 TF/01 Added $received_time.
5092 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5093 commented out examples of how to interface to a virus scanner and to
5094 SpamAssassin. Also added commented examples of av_scanner and
5095 spamd_address settings.
5097 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5098 and controls are allowed in which ACLs. There were a couple of minor
5099 errors. Some of the entries in the conditions table (which is a table of
5100 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5101 negation of where the condition IS allowed.
5103 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5105 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5106 header file does not have a version number, so I've had to invent a new
5107 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5108 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5109 radiusclient), but it was contributed by a Radius user.
5111 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5112 files or directories, for OpenSSL.
5114 PH/19 When an Exim process that is running as root has to create an Exim log
5115 file, it does so in a subprocess that runs as exim:exim so as to get the
5116 ownership right at creation (otherwise, other Exim processes might see
5117 the file with the wrong ownership). There was no test for failure of this
5118 fork() call, which would lead to the process getting stuck as it waited
5119 for a non-existent subprocess. Forks do occasionally fail when resources
5120 run out. I reviewed all the other calls to fork(); they all seem to check
5123 PH/20 When checking for unexpected SMTP input at connect time (before writing
5124 the banner), Exim was not dealing correctly with a non-positive return
5125 from the read() function. If the client had disconnected by this time,
5126 the result was a log entry for a synchronization error with an empty
5127 string after "input=" when read() returned zero. If read() returned -1
5128 (an event I could not check), uninitialized data bytes were printed.
5129 There were reports of junk text (parts of files, etc) appearing after
5132 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5134 PH/22 Added support for macro redefinition, and (re)definition in between
5135 driver and ACL definitions.
5137 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5138 forgetting to use the resulting value; it was using the unexpanded value.
5140 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5141 hadn't been configured. The fix is from Juergen Kreileder, who
5142 understands it better than I do:
5144 "Here's what I see happening with three configured cyrus_sasl
5145 authenticators configured (plain, login, cram-md5):
5147 On startup auth_cyrus_sasl_init() gets called for each of these.
5148 This means three calls to sasl_listmech() without a specified mech_list.
5149 => SASL tests which mechs of all available mechs actually work
5150 => three warnings about OTP not working
5151 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5153 With the patch, sasl_listmech() also gets called three times. But now
5154 SASL's mech_list option is set to the server_mech specified in the the
5155 authenticator. Or in other words, the answer from sasl_listmech()
5156 gets limited to just the mech you're testing for (which is different
5158 => the return list contains just 'plain' or 'login', 'cram-md5' or
5159 nothing depending on the value of ob->server_mech.
5161 I've just tested the patch: Authentication still works fine,
5162 unavailable mechs specified in the exim configuration are still
5163 caught, and the auth.log warnings about OTP are gone."
5165 PH/25 When debugging is enabled, the contents of the command line are added
5166 to the debugging output, even when log_selector=+arguments is not
5169 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5170 answer is "GNU", and only if the return is "GNU/something" is the answer
5173 PH/27 $acl_verify_message is now set immediately after the failure of a
5174 verification in an ACL, and so is available in subsequent modifiers. In
5175 particular, the message can be preserved by coding like this:
5177 warn !verify = sender
5178 set acl_m0 = $acl_verify_message
5180 Previously, $acl_verify_message was set only while expanding "message"
5181 and "log_message" when a very denied access.
5183 PH/28 Modified OS/os.c-Linux with
5185 -#ifndef OS_LOAD_AVERAGE
5186 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5188 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5189 nomenclature these days.)
5191 PH/29 Installed patch from the Sieve maintainer that adds the options
5192 sieve_useraddress and sieve_subaddress to the redirect router.
5194 PH/30 In these circumstances:
5195 . Two addresses routed to the same list of hosts;
5196 . First host does not offer TLS;
5197 . First host accepts first address;
5198 . First host gives temporary error to second address;
5199 . Second host offers TLS and a TLS session is established;
5200 . Second host accepts second address.
5201 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5202 and peerdn, if requested) that were in fact used only for the second
5205 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5206 attention to any local part prefix or suffix that was matched by the
5207 router that accepted the address. It now behaves in the same way as it
5208 does for delivery: the affixes are removed from the local part unless
5209 rcpt_include_affixes is set on the transport.
5211 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5212 timeout during the DATA phase of an incoming message.
5214 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5215 applied a patch sanctioned by the Sieve maintainer.
5217 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5218 the uid or gid is negative. A case of a negative gid caused this to be
5219 noticed. The fix allows for either to be negative.
5221 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5222 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5225 PH/36 The change PH/12 above was broken. Fixed it.
5227 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5228 the grounds that routing the same address twice would always produce the
5229 same answer. This might have been true once, but it is certainly no
5230 longer true now. Routing a child address may depend on the previous
5231 routing that produced that child. Some complicated redirection strategies
5232 went wrong when messages had multiple recipients, and made Exim's
5233 behaviour dependent on the order in which the addresses were given.
5235 I have moved the duplicate checking until after the routing is complete.
5236 Exim scans the addresses that are assigned to local and remote
5237 transports, and removes any duplicates. This means that more work will be
5238 done, as duplicates will always all be routed, but duplicates are
5239 presumably rare, so I don't expect this is of any significance.
5241 For deliveries to pipes, files, and autoreplies, the duplicate checking
5242 still happens during the routing process, since they are not going to be
5245 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5246 It corrects a timeout issue with spamd. This is Ian's comment: "The
5247 background is that sometimes spamd either never reads data from a
5248 connection it has accepted, or it never writes response data. The exiscan
5249 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5250 blindly assumes that writes won't block so it may never time out."
5252 PH/39 Allow G after quota size as well as K and M.
5254 PH/40 The value set for $authenticated_id in an authenticator may not contain
5255 binary zeroes or newlines because the value is written to log lines and
5256 to spool files. There was no check on this. Now the value is run through
5257 the string_printing() function so that such characters are converted to
5258 printable escape sequences.
5260 PH/41 $message_linecount is a new variable that contains the total number of
5261 lines in the message. Compare $body_linecount, which is the count for the
5264 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5265 bounce and delay warning messages, except in certain special cases, which
5268 (a) An SMTP error message from a remote host;
5269 (b) A message specified in a :fail: redirection;
5270 (c) A message specified in a "fail" command in a system filter;
5271 (d) A message specified in a FAIL return from the queryprogram router;
5272 (e) A message specified by the cannot_route_message router option.
5274 In these cases only, Exim does include the error details in bounce and
5275 warning messages. There are also a few cases where bland messages such
5276 as "unrouteable address" or "local delivery error" are given.
5278 PH/43 $value is now also set for the "else" part of a ${run expansion.
5280 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5281 being worked on, but at least Exim now implements the latest version to
5284 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5285 process to complete was treated as a delivery failure, a timeout while
5286 writing the message to the pipe was logged, but erroneously treated as a
5287 successful delivery. Such timeouts include transport filter timeouts. For
5288 consistency with the overall process timeout, these timeouts are now
5289 treated as errors, giving rise to delivery failures by default. However,
5290 there is now a new Boolean option for the pipe transport called
5291 timeout_defer, which, if set TRUE, converts the failures into defers for
5292 both kinds of timeout. A transport filter timeout is now identified in
5295 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5296 systems where "make" and "gmake" are different, calling "gmake" at top
5297 level broke things. I've arranged for the value of $(MAKE) to be passed
5298 from the Makefile to this script so that it can call the same version of
5302 A note about Exim versions 4.44 and 4.50
5303 ----------------------------------------
5305 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5306 changes of various kinds. As a consequence, a big documentation update was
5307 needed. This delayed the release for rather longer than seemed good, especially
5308 in the light of a couple of (minor) security issues. Therefore, the changes
5309 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5310 release. So 4.44 and 4.50 are in effect two different branches that both start
5313 I have left the 4.50 change log unchanged; it contains all the changes since
5314 4.43. The change log for 4.44 is below; many of its items are identical to
5315 those for 4.50. This seems to be the most sensible way to preserve the
5316 historical information.
5322 1. Minor wording change to the doc/README.SIEVE file.
5324 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5325 computation of the current number of files was incorrect.
5327 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5328 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5331 4. Give more explanation in the error message when the command for a transport
5332 filter fails to execute.
5334 5. There are several places where Exim runs a non-Exim command in a
5335 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5336 was being done only for the command run by the queryprogram router. It is
5337 now done for all such subprocesses. The other cases are: ${run, transport
5338 filters, and the commands run by the lmtp and pipe transports.
5340 6. Added CONFIGURE_GROUP build-time option.
5342 7. Some older OS have a limit of 256 on the maximum number of file
5343 descriptors. Exim was using setrlimit() to set 1000 as a large value
5344 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5345 systems. I've change it so that if it can't get 1000, it tries for 256.
5347 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5348 was an oversight, and furthermore, ever since the addition of extra
5349 controls (e.g. 4.43/32), the checks on when to allow different forms of
5350 "control" were broken. There should now be diagnostics for all cases when a
5351 control that does not make sense is encountered.
5353 9. Added the /retain_sender option to "control=submission".
5355 10. $recipients is now available in the predata ACL (oversight).
5357 11. Tidy the search cache before the fork to do a delivery from a message
5358 received from the command line. Otherwise the child will trigger a lookup
5359 failure and thereby defer the delivery if it tries to use (for example) a
5360 cached ldap connection that the parent has called unbind on.
5362 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5363 of $address_data from the recipient verification was clobbered by the
5364 sender verification.
5366 13. The value of address_data from a sender verification is now available in
5367 $sender_address_data in subsequent conditions in the ACL statement.
5369 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5371 15. Added a new option "connect=<time>" to callout options, to set a different
5374 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5375 was its contents. (It was OK if the option was not defined at all.)
5377 17. A "Completed" log line is now written for messages that are removed from
5378 the spool by the -Mrm option.
5380 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5381 information about exactly what failed.
5383 19. Added -dd to debug only the daemon process.
5385 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5386 handles timeouts, both on the server side and network timeouts. Renamed the
5387 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5389 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5390 It is now set to "smtps".
5392 22. $host_address is now set to the target address during the checking of
5393 ignore_target_hosts.
5395 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5396 being passed; this would have caused $sender_host_name to have been used if
5397 matching the list had actually called for a host name (not very likely,
5398 since this list is usually IP addresses). A host name is now passed as
5401 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5402 code that specifies a non-restarting handler (typically sigaction() in
5403 modern systems) in an attempt to fix a rare and obscure crash bug.
5405 25. Narrowed the window for a race in the daemon that could cause it to ignore
5406 SIGCHLD signals. This is not a major problem, because they are used only to
5407 wake it up if nothing else does.
5409 26. A malformed maildirsize file could cause Exim to calculate negative values
5410 for the mailbox size or file count. Odd effects could occur as a result.
5411 The maildirsize information is now recalculated if the size or filecount
5414 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5415 support for a long time. Removed HAVE_SYS_VFS_H.
5417 28. Installed the latest version of exipick from John Jetmore.
5419 29. In an address list, if the pattern was not a regular expression, an empty
5420 subject address (from a bounce message) matched only if the pattern was an
5421 empty string. Non-empty patterns were not even tested. This was the wrong
5422 because it is perfectly reasonable to use an empty address as part of a
5423 database query. An empty address is now tested by patterns that are
5424 lookups. However, all the other forms of pattern expect the subject to
5425 contain a local part and a domain, and therefore, for them, an empty
5426 address still always fails if the pattern is not itself empty.
5428 30. Exim went into a mad DNS loop when attempting to do a callout where the
5429 host was specified on an smtp transport, and looking it up yielded more
5430 than one IP address.
5432 31. Re-factored the code for checking spool and log partition space into a
5433 function that finds that data and another that does the check. The former
5434 is then used to implement four new variables: $spool_space, $log_space,
5435 $spool_inodes, and $log_inodes.
5437 32. The RFC2047 encoding function was originally intended for short strings
5438 such as real names; it was not keeping to the 75-character limit for
5439 encoded words that the RFC imposes. It now respects the limit, and
5440 generates multiple encoded words if necessary. To be on the safe side, I
5441 have increased the buffer size for the ${rfc2047: expansion operator from
5444 33. It is now permitted to omit both strings after an "if" condition; if the
5445 condition is true, the result is "true". As before, when the second string
5446 is omitted, a false condition yields an empty string. This makes it less
5447 cumbersome to write custom ACL and router conditions.
5449 34. Failure to deliver a bounce message always caused it to be frozen, even if
5450 there was an errors_to setting on the router. The errors_to setting is now
5453 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5454 canonical form (fully expanded) before being placed in
5455 $sender_host_address.
5457 36. The table in the code that translates DNS record types into text (T_A to
5458 "A" for instance) was missing entries for NS and CNAME. It is just possible
5459 that this could have caused confusion if both these types were looked up
5460 for the same domain, because the text type is used as part of Exim's
5461 per-process caching. But the chance of anyone hitting this buglet seems
5464 37. The dnsdb lookup has been extended in a number of ways.
5466 (1) There is a new type, "zns", which walks up the domain tree until it
5467 finds some nameserver records. It should be used with care.
5469 (2) There is a new type, "mxh", which is like "mx" except that it returns
5470 just the host names, not the priorities.
5472 (3) It is now possible to give a list of domains (or IP addresses) to be
5473 looked up. The behaviour when one of the lookups defers can be
5474 controlled by a keyword.
5476 (4) It is now possible to specify the separator character for use when
5477 multiple records are returned.
5479 38. The dnslists ACL condition has been extended: it is now possible to supply
5480 a list of IP addresses and/or domains to be looked up in a particular DNS
5483 39. Added log_selector=+queue_time_overall.
5485 40. When running the queue in the test harness, wait just a tad after forking a
5486 delivery process, to get repeatability of debugging output.
5488 41. Include certificate and key file names in error message when GnuTLS fails
5489 to set them up, because the GnuTLS error message doesn't include the name
5490 of the failing file when there is a problem reading it.
5492 42. Allow both -bf and -bF in the same test run.
5494 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5496 44. The "Exiscan patch" is now merged into the mainline Exim source.
5498 45. Sometimes the final signoff response after QUIT could fail to get
5499 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5500 < 0 before doing a fflush(). This bug looks as though it goes back to the
5501 introduction of TLS in release 3.20, but "sometimes" must have been rare
5502 because the tests only now provoked it.
5504 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5505 (this can affect the format of dates).
5507 47. exim_tidydb, when checking for the continued existence of a message for
5508 which it has found a message-specific retry record, was not finding
5509 messages that were in split spool directories. Consequently, it was
5510 deleting retry records that should have stayed in existence.
5512 48. Steve fixed some bugs in eximstats.
5514 49. The SPA authentication driver was not abandoning authentication and moving
5515 on to the next authenticator when an expansion was forced to fail,
5516 contradicting the general specification for all authenticators. Instead it
5517 was generating a temporary error. It now behaves as specified.
5519 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5520 (the order specifies the preference for clients). The order is now AES256,
5521 AES128, 3DES, ARCFOUR128.
5523 51. Small patch to Sieve code - explicitly set From: when generating an
5526 52. Exim crashed if a remote delivery caused a very long error message to be
5527 recorded - for instance if somebody sent an entire SpamAssassin report back
5528 as a large number of 550 error lines. This bug was coincidentally fixed by
5529 increasing the size of one of Exim's internal buffers (big_buffer) that
5530 happened as part of the Exiscan merge. However, to be on the safe side, I
5531 have made the code more robust (and fixed the comments that describe what
5534 53. Now that there can be additional text after "Completed" in log lines (if
5535 the queue_time_overall log selector is set), a one-byte patch to exigrep
5536 was needed to allow it to recognize "Completed" as not the last thing in
5539 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5540 patch that reportedly fixes this has been added. I am not expert enough to
5541 create a test for it. This is what the patch creator wrote:
5543 "I found a little strange behaviour of ldap code when working with
5544 Windows 2003 AD Domain, where users was placed in more than one
5545 Organization Units. When I tried to give exim partial DN, the exit code
5546 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5547 But simultaneously result of request was absolutely normal ldap result,
5548 so I produce this patch..."
5550 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5551 REFERENCE, so I have modified the code to exclude the patch when that macro
5554 55. Some experimental protocols are using DNS PTR records for new purposes. The
5555 keys for these records are domain names, not reversed IP addresses. The
5556 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5557 leaves it alone. Component reversal etc. now happens only for IP addresses.
5560 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5562 57. Double the size of the debug message buffer (to 2048) so that more of very
5563 long debug lines gets shown.
5565 58. The exicyclog utility now does better if the number of log files to keep
5566 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5568 59. Two changes related to the smtp_active_hostname option:
5570 (1) $smtp_active_hostname is now available as a variable.
5571 (2) The default for smtp_banner uses $smtp_active_hostname instead
5572 of $primary_hostname.
5574 60. The host_aton() function is supposed to be passed a string that is known
5575 to be a valid IP address. However, in the case of IPv6 addresses, it was
5576 not checking this. This is a hostage to fortune. Exim now panics and dies
5577 if the condition is not met. A case was found where this could be provoked
5578 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5579 components; fortuitously, this particular loophole had already been fixed
5580 by change 4.50/55 above.
5582 If there are any other similar loopholes, the new check in host_aton()
5583 itself should stop them being exploited. The report I received stated that
5584 data on the command line could provoke the exploit when Exim was running as
5585 exim, but did not say which command line option was involved. All I could
5586 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5587 running as the user.
5590 61. There was a buffer overflow vulnerability in the SPA authentication code
5591 (which came originally from the Samba project). I have added a test to the
5592 spa_base64_to_bits() function which I hope fixes it.
5595 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5596 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5598 63. The daemon start-up calls getloadavg() while still root for those OS that
5599 need the first call to be done as root, but it missed one case: when
5600 deliver_queue_load_max is set with deliver_drop_privilege. This is
5601 necessary for the benefit of the queue runner, because there is no re-exec
5602 when deliver_drop_privilege is set.
5604 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5605 This has been fixed.
5607 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5608 was in use, was not putting the data itself into the right store pool;
5609 consequently, it could be overwritten for a subsequent message in the same
5610 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5613 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5615 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5616 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5617 the function were treating the return as a boolean value, which happened to
5618 work because 0=false and not-0=true, but is not correct code.
5620 68. The host_aton() function was not handling scoped IPv6 addresses (those
5621 with, for example, "%eth0" on the end) correctly.
5623 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5624 negated items (that is, ~something) in unsigned ints. Some compilers
5625 apparently mutter when there is no cast.
5627 70. If an address verification called from an ACL failed, and did not produce a
5628 user-specific message (i.e. there was only a "system" message), nothing was
5629 put in $acl_verify_message. In this situation, it now puts the system
5632 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5633 session; change 4.31/43 added the unwanted input to the log line - except
5634 that it did not do this in the start of session case. It now does.
5636 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5637 This is wrong and can cause the other end to generate a synchronization
5638 error if it is another Exim or anything else that does the synchronization
5639 check. A QUIT command is no longer sent after a timeout.
5641 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5642 during host lookups.
5644 74. The defer_ok option of callout verification was not working if it was used
5645 when verifying addresses in header lines, that is, for this case:
5647 verify = header_sender/callout=defer_ok
5649 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5650 those file descriptors could be used for SMTP connections. If anything
5651 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5652 could be sent to the SMTP client, causing chaos. The daemon now opens
5653 stdin, stdout, and stderr to /dev/null when it puts itself into the
5656 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5657 log by default. The user can override this with suitable Perl magic.
5659 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5660 the log message when discard triggers, was not working for the DATA ACL or
5661 for the non-SMTP ACL.
5663 78. Error message wording change in sieve.c.
5665 79. If smtp_accept_max_per_host was set, the number of connections could be
5666 restricted to fewer than expected, because the daemon was trying to set up
5667 a new connection before checking whether the processes handling previous
5668 connections had finished. The check for completed processes is now done
5669 earlier. On busy systems, this bug wouldn't be noticed because something
5670 else would have woken the daemon, and it would have reaped the completed
5673 80. If a message was submitted locally by a user whose login name contained one
5674 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5675 It caused a spool format error. I have fixed the spool reading code. A
5676 related problem was that the "from" clause in the Received: line became
5677 illegal because of the space(s). It is now covered by ${quote_local_part.
5679 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5682 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5684 83. Patch from Sieve maintainer to fix unterminated string problem in
5685 "vacation" handling.
5687 84. Some minor changes to the Linux configuration files to help with other
5688 OS variants using glibc.
5690 85. One more patch for Sieve to update vacation handling to latest spec.
5693 ----------------------------------------------------
5694 See the note above about the 4.44 and 4.50 releases.
5695 ----------------------------------------------------
5701 1. Change 4.43/35 introduced a bug that caused file counts to be
5702 incorrectly computed when quota_filecount was set in an appendfile
5705 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5706 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5709 3. Give more explanation in the error message when the command for a transport
5710 filter fails to execute.
5712 4. There are several places where Exim runs a non-Exim command in a
5713 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5714 was being done only for the command run by the queryprogram router. It is
5715 now done for all such subprocesses. The other cases are: ${run, transport
5716 filters, and the commands run by the lmtp and pipe transports.
5718 5. Some older OS have a limit of 256 on the maximum number of file
5719 descriptors. Exim was using setrlimit() to set 1000 as a large value
5720 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5721 systems. I've change it so that if it can't get 1000, it tries for 256.
5723 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5724 was an oversight, and furthermore, ever since the addition of extra
5725 controls (e.g. 4.43/32), the checks on when to allow different forms of
5726 "control" were broken. There should now be diagnostics for all cases when a
5727 control that does not make sense is encountered.
5729 7. $recipients is now available in the predata ACL (oversight).
5731 8. Tidy the search cache before the fork to do a delivery from a message
5732 received from the command line. Otherwise the child will trigger a lookup
5733 failure and thereby defer the delivery if it tries to use (for example) a
5734 cached ldap connection that the parent has called unbind on.
5736 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5737 of $address_data from the recipient verification was clobbered by the
5738 sender verification.
5740 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5741 was its contents. (It was OK if the option was not defined at all.)
5743 11. A "Completed" log line is now written for messages that are removed from
5744 the spool by the -Mrm option.
5746 12. $host_address is now set to the target address during the checking of
5747 ignore_target_hosts.
5749 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5750 being passed; this would have caused $sender_host_name to have been used if
5751 matching the list had actually called for a host name (not very likely,
5752 since this list is usually IP addresses). A host name is now passed as
5755 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5756 code that specifies a non-restarting handler (typically sigaction() in
5757 modern systems) in an attempt to fix a rare and obscure crash bug.
5759 15. Narrowed the window for a race in the daemon that could cause it to ignore
5760 SIGCHLD signals. This is not a major problem, because they are used only to
5761 wake it up if nothing else does.
5763 16. A malformed maildirsize file could cause Exim to calculate negative values
5764 for the mailbox size or file count. Odd effects could occur as a result.
5765 The maildirsize information is now recalculated if the size or filecount
5768 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5769 support for a long time. Removed HAVE_SYS_VFS_H.
5771 18. Updated exipick to current release from John Jetmore.
5773 19. Allow an empty sender to be matched against a lookup in an address list.
5774 Previously the only cases considered were a regular expression, or an
5777 20. Exim went into a mad DNS lookup loop when doing a callout where the
5778 host was specified on the transport, if the DNS lookup yielded more than
5781 21. The RFC2047 encoding function was originally intended for short strings
5782 such as real names; it was not keeping to the 75-character limit for
5783 encoded words that the RFC imposes. It now respects the limit, and
5784 generates multiple encoded words if necessary. To be on the safe side, I
5785 have increased the buffer size for the ${rfc2047: expansion operator from
5788 22. Failure to deliver a bounce message always caused it to be frozen, even if
5789 there was an errors_to setting on the router. The errors_to setting is now
5792 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5793 canonical form (fully expanded) before being placed in
5794 $sender_host_address.
5796 24. Updated eximstats to version 1.33
5798 25. Include certificate and key file names in error message when GnuTLS fails
5799 to set them up, because the GnuTLS error message doesn't include the name
5800 of the failing file when there is a problem reading it.
5802 26. Expand error message when OpenSSL has problems setting up cert/key files.
5805 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5806 (this can affect the format of dates).
5808 28. exim_tidydb, when checking for the continued existence of a message for
5809 which it has found a message-specific retry record, was not finding
5810 messages that were in split spool directories. Consequently, it was
5811 deleting retry records that should have stayed in existence.
5813 29. eximstats updated to version 1.35
5814 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5815 1.35 - bugfix such that pie charts by volume are generated correctly
5817 30. The SPA authentication driver was not abandoning authentication and moving
5818 on to the next authenticator when an expansion was forced to fail,
5819 contradicting the general specification for all authenticators. Instead it
5820 was generating a temporary error. It now behaves as specified.
5822 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5823 (the order specifies the preference for clients). The order is now AES256,
5824 AES128, 3DES, ARCFOUR128.
5826 31. Small patch to Sieve code - explicitly set From: when generating an
5829 32. Exim crashed if a remote delivery caused a very long error message to be
5830 recorded - for instance if somebody sent an entire SpamAssassin report back
5831 as a large number of 550 error lines. This bug was coincidentally fixed by
5832 increasing the size of one of Exim's internal buffers (big_buffer) that
5833 happened as part of the Exiscan merge. However, to be on the safe side, I
5834 have made the code more robust (and fixed the comments that describe what
5837 33. Some experimental protocols are using DNS PTR records for new purposes. The
5838 keys for these records are domain names, not reversed IP addresses. The
5839 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5840 leaves it alone. Component reversal etc. now happens only for IP addresses.
5843 34. The host_aton() function is supposed to be passed a string that is known
5844 to be a valid IP address. However, in the case of IPv6 addresses, it was
5845 not checking this. This is a hostage to fortune. Exim now panics and dies
5846 if the condition is not met. A case was found where this could be provoked
5847 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5848 components; fortuitously, this particular loophole had already been fixed
5849 by change 4.50/55 or 4.44/33 above.
5851 If there are any other similar loopholes, the new check in host_aton()
5852 itself should stop them being exploited. The report I received stated that
5853 data on the command line could provoke the exploit when Exim was running as
5854 exim, but did not say which command line option was involved. All I could
5855 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5856 running as the user.
5859 35. There was a buffer overflow vulnerability in the SPA authentication code
5860 (which came originally from the Samba project). I have added a test to the
5861 spa_base64_to_bits() function which I hope fixes it.
5864 36. The daemon start-up calls getloadavg() while still root for those OS that
5865 need the first call to be done as root, but it missed one case: when
5866 deliver_queue_load_max is set with deliver_drop_privilege. This is
5867 necessary for the benefit of the queue runner, because there is no re-exec
5868 when deliver_drop_privilege is set.
5870 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5871 was in use, was not putting the data itself into the right store pool;
5872 consequently, it could be overwritten for a subsequent message in the same
5873 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5876 38. Sometimes the final signoff response after QUIT could fail to get
5877 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5878 < 0 before doing a fflush(). This bug looks as though it goes back to the
5879 introduction of TLS in release 3.20, but "sometimes" must have been rare
5880 because the tests only now provoked it.
5886 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5887 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5888 Now it returns an integer. A number of calls were still expecting a T/F
5889 return. Fortuitously, in all cases, the tests worked in OK situations,
5890 which is the norm. However, things would have gone wrong on any write
5891 failures on the smtp file descriptor. This function is used when sending
5892 messages over SMTP and also when doing verify callouts.
5894 2. When Exim is called to do synchronous delivery of a locally submitted
5895 message (the -odf or -odi options), it no longer closes stderr before doing
5898 3. Implemented the mua_wrapper option.
5900 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5902 5. Implemented the functions header_remove(), header_testname(),
5903 header_add_at_position(), and receive_remove_recipient(), and exported them
5906 6. If an ACL "warn" statement specified the addition of headers, Exim already
5907 inserted X-ACL-Warn: at the start if there was no header name. However, it
5908 was not making this test for the second and subsequent header lines if
5909 there were newlines in the string. This meant that an invalid header could
5910 be inserted if Exim was badly configured.
5912 7. Allow an ACL "warn" statement to add header lines at the start or after all
5913 the Received: headers, as well as at the end.
5915 8. Added the rcpt_4xx retry error code.
5917 9. Added postmaster_mailfrom=xxx to callout verification option.
5919 10. Added mailfrom=xxxx to the callout verification option, for verify=
5922 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5923 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5925 12. Inserted some casts to stop certain compilers warning when using pointer
5926 differences as field lengths or precisions in printf-type calls (mostly
5927 affecting debugging statements).
5929 13. Added optional readline() support for -be (dynamically loaded).
5931 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5932 same clock tick as a message's arrival, so that its received time was the
5933 same as the "first fail" time on the retry record, and that message
5934 remained on the queue past the ultimate address timeout, every queue runner
5935 would try a delivery (because it was past the ultimate address timeout) but
5936 after another failure, the ultimate address timeout, which should have then
5937 bounced the address, did not kick in. This was a "< instead of <=" error;
5938 in most cases the first failure would have been in the next clock tick
5939 after the received time, and all would be well.
5941 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5942 being recognized when the domain list was tested by the match_domain
5943 condition in an expansion string.
5945 16. Added the ${str2b64: operator.
5947 17. Exim was always calling setrlimit() to set a large limit for the number of
5948 processes, without checking whether the existing limit was already
5949 adequate. (It did check for the limit on file descriptors.) Furthermore,
5950 errors from getrlimit() and setrlimit() were being ignored. Now they are
5951 logged to the main and panic logs, but Exim does carry on, to try to do its
5952 job under whatever limits there are.
5954 18. Imported PCRE 5.0.
5956 19. Trivial typo in log message " temporarily refused connection" (the leading
5959 20. If the log selector return_path_on_delivery was set and an address was
5960 redirected to /dev/null, the delivery process crashed because it assumed
5961 that a return path would always be set for a "successful" delivery. In this
5962 case, the whole delivery is bypassed as an optimization, and therefore no
5965 21. Internal re-arrangement: the function for sending a challenge and reading
5966 a response while authentication was assuming a zero-terminated challenge
5967 string. It's now changed to take a pointer and a length, to allow for
5968 binary data in such strings.
5970 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5972 23. Exim was not respecting finduser_retries when seeking the login of the
5973 uid under which it was called; it was always trying 10 times. (The default
5974 setting of finduser_retries is zero.) Also, it was sleeping after the final
5975 failure, which is pointless.
5977 24. Implemented tls_on_connect_ports.
5979 25. Implemented acl_smtp_predata.
5981 26. If the domain in control=submission is set empty, Exim assumes that the
5982 authenticated id is a complete email address when it generates From: or
5983 Sender: header lines.
5985 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5986 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5987 chown and chgrp in /bin and hostname in /usr/bin.
5989 28. Exim was keeping the "process log" file open after each use, just as it
5990 does for the main log. This opens the possibility of it remaining open for
5991 long periods when the USR1 signal hits a daemon. Occasional processlog
5992 errors were reported, that could have been caused by this. Anyway, it seems
5993 much more sensible not to leave this file open at all, so that is what now
5996 29. The long-running daemon process does not normally write to the log once it
5997 has entered its main loop, and it closes the log before doing so. This is
5998 so that log files can straightforwardly be renamed and moved. However,
5999 there are a couple of unusual error situations where the daemon does write
6000 log entries, and I had neglected to close the log afterwards.
6002 30. The text of an SMTP error response that was received during a remote
6003 delivery was being truncated at 512 bytes. This is too short for some of
6004 the long messages that one sometimes sees. I've increased the limit to
6007 31. It is now possible to make retry rules that apply only when a message has a
6008 specific sender, in particular, an empty sender.
6010 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6011 it possible to be selective about when SMTP synchronization is enforced.
6013 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6015 32. Implemented hosts_connection_nolog.
6017 33. Added an ACL for QUIT.
6019 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6022 35. Added mailbox_size and mailbox_filecount to appendfile.
6024 36. Added control = no_multiline_responses to ACLs.
6026 37. There was a bug in the logic of the code that waits for the clock to tick
6027 in the case where the clock went backwards by a substantial amount such
6028 that the microsecond fraction of "now" was more than the microsecond
6029 fraction of "then" (but the whole seconds number was less).
6031 38. Added support for the libradius Radius client library this is found on
6032 FreeBSD (previously only the radiusclient library was supported).
6038 1. When certain lookups returned multiple values in the form name=value, the
6039 quoting of the values was not always being done properly. Specifically:
6040 (a) If the value started with a double quote, but contained no whitespace,
6042 (b) If the value contained whitespace other than a space character (i.e.
6043 tabs or newlines or carriage returns) it was not quoted.
6044 This fix has been applied to the mysql and pgsql lookups by writing a
6045 separate quoting function and calling it from the lookup code. The fix
6046 should probably also be applied to nisplus, ibase and oracle lookups, but
6047 since I cannot test any of those, I have not disturbed their existing code.
6049 2. A hit in the callout cache for a specific address caused a log line with no
6050 reason for rejecting RCPT. Now it says "Previous (cached) callout
6051 verification failure".
6053 3. There was an off-by-one bug in the queryprogram router. An over-long
6054 return line was truncated at 256 instead of 255 characters, thereby
6055 overflowing its buffer with the terminating zero. As well as fixing this, I
6056 have increased the buffer size to 1024 (and made a note to document this).
6058 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6059 when Exim is waiting for an SMTP response from a remote server, Exim
6060 restarts its select() call on the socket, thereby resetting its timeout.
6061 This is not a problem when such interrupts are rare. Somebody set up a cron
6062 job to run exiwhat every 2 minutes, which is less than the normal select()
6063 timeout (5 or 10 minutes). This meant that the select() timeout never
6064 kicked in because it was always reset. I have fixed this by comparing the
6065 time when an interrupt arrives with the time at the start of the first call
6066 to select(). If more time than the timeout has elapsed, the interrupt is
6067 treated as a timeout.
6069 5. Some internal re-factoring in preparation for the addition of Sieve
6070 extensions (by MH). In particular, the "personal" test is moved to a
6071 separate function, and given an option for scanning Cc: and Bcc: (which is
6072 not set for Exim filters).
6074 6. When Exim created an email address using the login of the caller as the
6075 local part (e.g. when creating a From: or Sender: header line), it was not
6076 quoting the local part when it contained special characters such as @.
6078 7. Installed new OpenBSD configuration files.
6080 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6081 try to make them clearer.
6083 9. Callout options, other than the timeout value, were being ignored when
6084 verifying sender addresses in header lines. For example, when using
6086 verify = header_sender/callout=no_cache
6088 the cache was (incorrectly) being used.
6090 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6091 only the Cygwin environment.
6093 11. When return_path_on_delivery was set as a log selector, if different remote
6094 addresses in the same message used different return paths and parallel
6095 remote delivery occurred, the wrong values would sometimes be logged.
6096 (Whenever a remote delivery process finished, the return path value from
6097 the most recently started remote delivery process was logged.)
6099 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6100 lines when AUTH and/or TLS are in use. This is the "received protocol"
6101 field. Exim used to use "asmtp" for authenticated SMTP, without any
6102 indication (in the protocol name) for TLS use. Now it follows the RFC and
6103 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6104 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6105 names appear in log lines as well as in Received: header lines.
6107 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6108 extensions, and comparison tests, and to fix some bugs.
6110 14. Changes to the "personal" filter test:
6112 (1) The test was buggy in that it was just doing the equivalent of
6113 "contains" tests on header lines. For example, if a user's address was
6114 anne@some.where, the "personal" test would incorrectly be true for
6116 To: susanne@some.where
6118 This test is now done by extracting each address from the header in turn,
6119 and checking the entire address. Other tests that are part of "personal"
6120 are now done using regular expressions (for example, to check local parts
6121 of addresses in From: header lines).
6123 (2) The list of non-personal local parts in From: addresses has been
6124 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6125 taken from the Sieve specification recommendations.
6127 (3) If the message contains any header line starting with "List-" it is
6128 treated as non-personal.
6130 (4) The test for "circular" in the Subject: header line has been removed
6131 because it now seems ill-conceived.
6133 15. Minor typos in src/EDITME comments corrected.
6135 16. Installed latest exipick from John Jetmore.
6137 17. If headers_add on a router specified a text string that was too long for
6138 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6139 of string_sprintf() is now avoided.
6141 18. $message_body_size was not set (it was always zero) when running the DATA
6142 ACL and the local_scan() function.
6144 19. For the "mail" command in an Exim filter, no default was being set for
6145 the once_repeat time, causing a random time value to be used if "once" was
6146 specified. (If the value happened to be <= 0, no repeat happened.) The
6147 default is now 0s, meaning "never repeat". The "vacation" command was OK
6148 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6149 (I found it when inspecting the code).
6151 20. There is now an overall timeout for performing a callout verification. It
6152 defaults to 4 times the callout timeout, which applies to individual SMTP
6153 commands during the callout. The overall timeout applies when there is more
6154 than one host that can be tried. The timeout is checked before trying the
6155 next host. This prevents very long delays if there are a large number of
6156 hosts and all are timing out (e.g. when the network connections are timing
6157 out). The value of the overall timeout can be changed by specifying an
6158 additional sub-option for "callout", called "maxwait". For example:
6160 verify = sender/callout=5s,maxwait=20s
6162 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6163 to the end before writing, but this should make it even safer).
6165 22. Exim was forgetting that it had advertised PIPELINING for the second and
6166 subsequent messages on an SMTP connection. It was also not resetting its
6167 memory on STARTTLS and an internal HELO.
6169 23. When Exim logs an SMTP synchronization error within a session, it now
6170 records whether PIPELINING has been advertised or not.
6172 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6173 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6174 rather than long int.
6176 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6178 26. Added the never_mail option to autoreply.
6184 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6185 crash if the getsockname() call failed; this can happen if a connection is
6186 closed very soon after it is established. The problem was simply in the
6187 order in which certain operations were done, causing Exim to try to write
6188 to the SMTP stream before it had set up the file descriptor. The bug has
6189 been fixed by making things happen in the correct order.
6195 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6196 before the connection was closed, thus losing the rejection response.
6198 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6199 some early Solaris releases, but causes trouble in current releases where
6200 socklen_t is defined.
6202 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6205 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6208 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6209 that releasing the top of it at the end releases what was used for sub-
6210 expansions (unless the block got too big). However, discard this block if
6211 the first thing is a variable or header, so that we can use its block when
6212 it is dynamic (useful for very large $message_headers, for example).
6214 6. Lookups now cache *every* query, not just the most recent. A new, separate
6215 store pool is used for this. It can be recovered when all lookup caches are
6216 flushed. Lookups now release memory at the end of their result strings.
6217 This has involved some general refactoring of the lookup sources.
6219 7. Some code has been added to the store_xxx() functions to reduce the amount
6220 of flapping under certain conditions.
6222 8. log_incoming_interface used to affect only the <= reception log lines. Now
6223 it causes the local interface and port to be added to several more SMTP log
6224 lines, for example "SMTP connection from", and rejection lines.
6226 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6228 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6230 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6231 could be overwritten at the end of the current message (or the start of a
6232 new message if it was set in a HELO ACL). The value is now preserved for
6233 the duration of the SMTP connection.
6235 12. If a transport had a headers_rewrite setting, and a matching header line
6236 contained an unqualified address, that address was qualified, even if it
6237 did not match any rewriting rules. The underlying bug was that the values
6238 of the flags that permit the existence of unqualified sender and recipient
6239 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6240 non-local messages, and by -bnq for local messages) were not being
6241 preserved with the message after it was received.
6243 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6244 "next input=" as part of the text comprising the host identity instead of
6245 the correct text. The code was using the same buffer for two different
6246 strings. However, depending on which order the printing function evaluated
6247 its arguments, the bug did not always show up. Under Linux, for example, my
6248 test suite worked just fine.
6250 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6251 doesn't work with some older versions of Perl. It has been changed to "my",
6252 which in any case is probably the better facility to use.
6254 15. A really picky compiler found some instances of statements for creating
6255 error messages that either had too many or two few arguments for the format
6258 16. The size of the buffer for calls to the DNS resolver has been increased
6259 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6260 for addresses that have a lot of PTR records. This alleviates a problem; it
6261 does not fully solve it.
6263 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6264 buffer now truncates the list and logs the incident, which is the same
6265 action as happens when Exim is looking up a host name and its aliases.
6266 Previously in this situation something unpredictable would happen;
6267 sometimes it was "internal error: store_reset failed".
6269 18. If a server dropped the connection unexpectedly when an Exim client was
6270 using GnuTLS and trying to read a response, the client delivery process
6271 crashed while trying to generate an error log message.
6273 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6274 string, for example:
6276 warn message = H1: something\nH2: something
6278 the text was added as a single header line from Exim's point of view
6279 though it ended up OK in the delivered message. However, searching for the
6280 second and subsequent header lines using $h_h2: did not work. This has been
6281 fixed. Similarly, if a system filter added multiple headers in this way,
6282 the routers could not see them.
6284 20. Expanded the error message when iplsearch is called with an invalid key to
6285 suggest using net-iplsearch in a host list.
6287 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6288 ACLs are no longer actually imposed (and a message to that effect is
6291 22. If a "gecos" field in a passwd entry contained escaped characters, in
6292 particular, if it contained a \" sequence, Exim got it wrong when building
6293 a From: or a Sender: header from that name. A second bug also caused
6294 incorrect handling when an unquoted " was present following a character
6295 that needed quoting.
6297 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6298 was not being matched caselessly.
6300 24. Arranged for all hyphens in the exim.8 source to be escaped with
6303 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6304 properly. Recipient callout cache records were still being keyed to include
6305 the sender, even when use_sender was set false. This led to far more
6306 callouts that were necessary. The sender is no longer included in the key
6307 when use_sender is false.
6309 26. Added "control = submission" modifier to ACLs.
6311 27. Added the ${base62d: operator to decode base 62 numbers.
6313 28. dnsdb lookups can now access SRV records.
6315 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6316 the configuration file.
6318 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6319 (-v) mode. This makes the output for a verbose queue run more intelligible.
6321 31. Added a use_postmaster feature to recipient callouts.
6323 32. Added the $body_zerocount variable, containing the number of binary zero
6324 bytes in the message body.
6326 33. The time of last modification of the "new" subdirectory is now used as the
6327 "mailbox time last read" when there is a quota error for a maildir
6330 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6332 35. Added +ignore_unknown as a special item in host lists.
6334 36. Code for decoding IPv6 addresses in host lists is now included, even if
6335 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6336 address was recognized as an IP address, but was then not correctly decoded
6337 into binary, causing unexpected and incorrect effects when compared with
6344 1. Very minor rewording of debugging text in manualroute to say "list of
6345 hosts" instead of "hostlist".
6347 2. If verify=header_syntax was set, and a header line with an unqualified
6348 address (no domain) and a large number of spaces between the end of the
6349 name and the colon was received, the reception process suffered a buffer
6350 overflow, and (when I tested it) crashed. This was caused by some obsolete
6351 code that should have been removed. The fix is to remove it!
6353 3. When running in the test harness, delay a bit after writing a bounce
6354 message to get a bit more predictability in the log output.
6356 4. Added a call to search_tidyup() just before forking a reception process. In
6357 theory, someone could use a lookup in the expansion of smtp_accept_max_
6358 per_host which, without the tidyup, could leave open a database connection.
6360 5. Added the variables $recipient_data and $sender_data which get set from a
6361 lookup success in an ACL "recipients" or "senders" condition, or a router
6362 "senders" option, similar to $domain_data and $local_part_data.
6364 6. Moved the writing of debug_print from before to after the "senders" test
6367 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6368 problems for message scanning, either using a data ACL, or using
6369 local_scan() because the Received: header was not generated till after they
6370 were called (in order to set the time as the time of reception completion).
6371 I have revised the way this works. The header is now generated after the
6372 body is received, but before the ACL or local_scan() are called. After they
6373 are run, the timestamp in the header is updated.
6379 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6380 before starting a queue runner without re-exec. This happened only when
6381 deliver_drop_privilege was set or when the Exim user was set to root. The
6382 effect of the bug was that timeouts during subsequent deliveries caused
6383 crashes instead of being properly handled. The handler is now left at its
6384 default (and expected) setting.
6386 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6387 message, again when deliver_drop_privilege is set or Exim is run as root.
6388 The bug described in (1) was not present in this case, but the tidying up
6389 of the other signals was missing. I have made the two cases consistent.
6391 3. The ignore_target_hosts setting on a manualroute router was being ignored
6392 for hosts that were looked up using the /MX notation.
6394 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6397 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6398 operated on the sender address. After changing the $sender_address to <>
6399 for the sender address verify, Exim was re-instated it as the original
6400 (before rewriting) address, but remembering that it had rewritten it, so it
6401 wasn't rewriting it again. This bug also had the effect of breaking the
6402 sender address verification caching when the sender address was rewritten.
6404 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6405 This has been changed so that if the ip literal address matches
6406 ignore_target_hosts, the router declines.
6408 7. Added expansion conditions match_domain, match_address, and match_local_
6409 part (NOT match_host).
6411 8. The placeholder for the Received: header didn't have a length field set.
6413 9. Added code to Exim itself and to exim_lock to test for a specific race
6414 condition that could lead to file corruption when using MBX delivery. The
6415 issue is with the lockfile that is created in /tmp. If this file is removed
6416 after a process has opened it but before that process has acquired a lock,
6417 there is the potential for a second process to recreate the file and also
6418 acquire a lock. This could lead to two Exim processes writing to the file
6419 at the same time. The added code performs the same test as UW imapd; it
6420 checks after acquiring the lock that its file descriptor still refers to
6421 the same named file.
6423 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6424 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6427 11. Added the smtp_active_hostname option. If used, this will typically be made
6428 to depend on the incoming interface address. Because $interface_address is
6429 not set up until the daemon has forked a reception process, error responses
6430 that can happen earlier (such as "too many connections") no longer contain
6433 12. If an expansion in a condition on a "warn" statement fails because a lookup
6434 defers, the "warn" statement is abandoned, and the next ACL statement is
6435 processed. Previously this caused the whole ACL to be aborted.
6437 13. Added the iplsearch lookup type.
6439 14. Added ident_timeout as a log selector.
6441 15. Added tls_certificate_verified as a log selector.
6443 16. Added a global option tls_require_ciphers (compare the smtp transport
6444 option of the same name). This controls incoming TLS connections.
6446 17. I finally figured out how to make tls_require_ciphers do a similar thing
6447 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6448 before starting the TLS session.
6450 18. Tabs are now shown as \t in -bP output.
6452 19. If the log selector return_path_on_delivery was set, Exim crashed when
6453 bouncing a message because it had too many Received: header lines.
6455 20. If two routers both had headers_remove settings, and the first one included
6456 a superfluous trailing colon, the final name in the first list and the
6457 first name in the second list were incorrectly joined into one item (with a
6458 colon in the middle).
6464 1. Added -C and -D options to the exinext utility, mainly to make it easier
6465 to include in the automated testing, but these could be helpful when
6466 multiple configurations are in use.
6468 2. The exinext utility was not formatting the output nicely when there was
6469 an alternate port involved in the retry record key, nor when there was a
6470 message id as well (for retries that were specific to a specific message
6471 and a specific host). It was also confused by IPv6 addresses, because of
6472 the additional colons they contain. I have fixed the IPv4 problem, and
6473 patched it up to do a reasonable job for IPv6.
6475 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6476 delivery, the log line now contains "pipelined" if PIPELINING was used.
6478 4. An SMTP transport process used to panic and die if the bind() call to set
6479 an explicit outgoing interface failed. This has been changed; it is now
6480 treated in the same way as a connect() failure.
6482 5. A reference to $sender_host_name in the part of a conditional expansion
6483 that was being skipped was still causing a DNS lookup. This no longer
6486 6. The def: expansion condition was not recognizing references to header lines
6487 that used bh_ and bheader_.
6489 7. Added the _cache feature to named lists.
6491 8. The code for checking quota_filecount in the appendfile transport was
6492 allowing one more file than it should have been.
6494 9. For compatibility with Sendmail, the command line option
6502 and sets the incoming protocol and host name (for trusted callers). The
6503 host name and its colon can be omitted when only the protocol is to be set.
6504 Note the Exim already has two private options, -pd and -ps, that refer to
6505 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6506 "s", but I don't think that's a major issue.
6508 10. A number of refactoring changes to the code, none of which should affect
6511 (a) The number of logging options was getting close to filling up the
6512 32-bit word that was used as a bit map. I have split them into two classes:
6513 those that are passed in the argument to log_write(), and those that are
6514 only ever tested independently outside of that function. These are now in
6515 separate 32-bit words, so there is plenty of room for expansion again.
6516 There is no change in the user interface or the logging behaviour.
6518 (b) When building, for example, log lines, the code previously used a
6519 macro that called string_cat() twice, in order to add two strings. This is
6520 not really sufficiently general. Furthermore, there was one instance where
6521 it was actually wrong because one of the argument was used twice, and in
6522 one call a function was used. (As it happened, calling the function twice
6523 did not affect the overall behaviour.) The macro has been replaced by a
6524 function that can join an arbitrary number of extra strings onto a growing
6527 (c) The code for expansion conditions now uses a table and a binary chop
6528 instead of a serial search (which was left over from when there were very
6529 few conditions). Also, it now recognizes conditions like "pam" even when
6530 the relevant support is not compiled in: a suitably worded error message is
6531 given if an attempt is made to use such a condition.
6533 11. Added ${time_interval:xxxxx}.
6535 12. A bug was causing one of the ddress fields not to be passed back correctly
6536 from remote delivery subprocesses. The field in question was not being
6537 subsequently used, so this caused to problems in practice.
6539 13. Added new log selectors queue_time and deliver_time.
6541 14. Might have fixed a bug in maildirsizefile handling that threw up
6542 "unexpected character" debug warnings, and recalculated the data
6543 unnecessarily. In any case, I expanded the warning message to give more
6546 15. Added the message "Restricted characters in address" to the statements in
6547 the default ACL that block characters like @ and % in local parts.
6549 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6550 Three changes have been made:
6552 (a) There was a serious bug; a negative response to MAIL caused the whole
6553 recipient domain to be cached as invalid, thereby blocking all messages
6554 to all local parts at the same domain, from all senders. This bug has
6555 been fixed. The domain is no longer cached after a negative response to
6556 MAIL if the sender used is not empty.
6558 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6561 (c) A new callout option, "use_sender" has been added for people who want
6562 the modified behaviour.
6568 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6571 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6572 indeed breaks things for older releases.
6574 3. Added additional logging to the case where there is a problem reading data
6575 from a filter that is running in a subprocess using a pipe, in order to
6576 try to track down a specific problem.
6578 4. Testing facility fudge: when running in the test harness and attempting
6579 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6580 getting "No route to host". Convert this to a timeout.
6582 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6585 6. Some OS don't have socklen_t but use size_t instead. This affects the
6586 fifth argument of getsockopt() amongst other things. This is now
6587 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6588 can be set for individual OS. I have set it for SunOS5, OSF1, and
6589 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6590 some earlier ones do not.
6592 7. Change 4.30/15 was not doing the test caselessly.
6594 8. The standard form for an IPv6 address literal was being rejected by address
6595 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6596 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6597 this, as well as the form without the "IPv6" on the front (but only when
6598 address literals are enabled, of course).
6600 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6602 10. Exim crashed if a message with an empty sender address specified by -f
6603 encountered a router with an errors_to setting. This could be provoked only
6604 by a command such as
6608 where an empty string was supplied; "<>" did not hit this bug.
6610 11. Installed PCRE release 4.5.
6612 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6613 remained set. It is now erased.
6615 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6616 times from message ids (which are base 36 rather than the normal 62).
6618 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6619 were being counted as actual protocol errors, and logged if the log
6620 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6621 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6622 and DATA following a set of rejected RCPTs do not count as protocol errors.
6623 In other words, Exim assumes they were pipelined, though this may not
6624 actually be the case. Of course, in all cases the client gets an
6625 appropriate error code.
6627 15. If a lookup fails in an ACL condition, a message about the failure may
6628 be available; it is used if testing the ACL cannot continue, because most
6629 such messages specify what the cause of the deferral is. However, some
6630 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6631 that caused an old message to be retained and used if a later statement
6632 caused a defer, replacing the real cause of the deferral.
6634 16. If an IP address had so many PTR records that the DNS lookup buffer
6635 was not large enough to hold them, Exim could crash while trying to process
6636 the truncated data. It now detects and logs this case.
6638 17. Further to 4.21/58, another change has been made: if (and only if) the
6639 first line of a message (the first header line) ends with CRLF, a bare LF
6640 in a subsequent header line has a space inserted after it, so as not to
6641 terminate the header.
6643 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6644 unnecessarily, used atoi() instead of strtol(), and didn't check the
6645 termination when getting file sizes from file names by regex.
6647 19. Completely re-implemented the support for maildirsize files, in the light
6648 of a number of problems with the previous contributed implementation
6649 (4.30/29). In particular:
6651 . If the quota is zero, the maildirsize file is maintained, but no quota is
6654 . If the maildir directory does not exist, it is created before any attempt
6655 to write a maildirsize file.
6657 . The quota value in the file is just a cache; if the quota is changed in
6658 the transport, the new value overrides.
6660 . A regular expression is available for excluding directories from the
6663 20. The autoreply transport checks the characters in options that define the
6664 message's headers; it allows continued headers, but it was checking with
6665 isspace() after an embedded newline instead of explicitly looking for a
6668 21. If all the "regular" hosts to which an address was routed had passed their
6669 expiry times, and had not reached their retry times, the address was
6670 bounced, even if fallback hosts were defined. Now Exim should go on to try
6673 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6674 equivalent code in the SMTP transport. Some hosts send humungous responses
6675 to HELO/EHLO, more than 1024 it seems.
6677 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6678 gives compiler warnings in some environments. I've now done it "properly",
6681 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6682 (because of problems with the built-in one) was declared to return uschar *
6683 instead of char *, causing compiler failure.
6685 25. Fixed a file descriptor leak when processing alias/forward files.
6687 26. Fixed a minor format string issue in dbfn.c.
6689 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6691 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6692 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6695 29. When a sender address is verified, it is cached, to save repeating the test
6696 when there is more than one recipient in a message. However, when the
6697 verification involves a callout, it is possible for different callout
6698 options to be set for different recipients. It is too complicated to keep
6699 track of this in the cache, so now Exim always runs a verification when a
6700 callout is required, relying on the callout cache for the optimization.
6701 The overhead is duplication of the address routing, but this should not be
6704 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6705 to be verified with callout=postmaster, and the main callout worked but the
6706 postmaster check failed, the verification correctly failed. However, if a
6707 subsequent RCPT command asked for sender verification *without* the
6708 postmaster check, incorrect caching caused this verification also to fail,
6711 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6712 it was not caching the DNS options (qualify_single, search_parents) that
6713 were used when the lookup failed. A subsequent lookup with different
6714 options therefore always gave the same answer, though there were cases
6715 where it should not have. (Example: a "domains = !$mx_any" option on a
6716 dnslookup router: the "domains" option is always processed without any
6717 widening, but the router might have qualify_single set.) Now Exim uses the
6718 cached value only when the same options are set.
6720 32. Added John Jetmore's "exipick" utility to the distribution.
6722 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6723 than a timeout (e.g. a certificate is required, and is not provided), an
6724 Exim server now closes the connection immediately. Previously it waited for
6725 the client to close - but if the client is SSL, it seems that they each
6726 wait for each other, leading to a delay before one of them times out.
6728 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6729 maintained 0.8.x compatibility because I don't think many are using it, and
6730 it is clearly obsolete.
6732 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6735 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6736 client certificate was expired. A simple patch fixes this, though I don't
6737 understand the full logic of why the verify callback is called multiple
6740 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6741 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6742 which causes problems with some clients (such as the Certicom SSL Plus
6743 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6744 disables the coutermeasure allowing Eudora to connect."
6746 38. Exim was not checking that a write() to a log file succeeded. This could
6747 lead to Bad Things if a log got too big, in particular if it hit a file
6748 size limit. Exim now panics and dies if it cannot write to a log file, just
6749 as it does if it cannot open a log file.
6751 39. Modified OS/Makefile-Linux so that it now contains
6753 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6755 The two -D definitions ensure that Exim is compiled with large file
6756 support, which makes it possible to handle log files that are bigger than
6759 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6760 instance) a domain was checked against a named list that involved a lookup,
6761 causing $domain_data to be set, then another domain was checked against the
6762 same list, then the first domain was re-checked, the value of $domain_data
6763 after the final check could be wrong. In particular, if the second check
6764 failed, it could be set empty. This bug probably also applied to
6767 41. The strip_trailing_dot option was not being applied to the address given
6768 with the -f command-line option.
6770 42. The code for reading a message's header from the spool was incrementing
6771 $received_count, but never initializing it. This meant that the value was
6772 incorrect (doubled) while delivering a message in the same process in which
6773 it was received. In the most common configuration of Exim, this never
6774 happens - a fresh exec is done - but it can happen when
6775 deliver_drop_privilege is set.
6777 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6778 it now includes up to 150 characters of the unexpected data in the log
6781 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6782 and building data strings. The size of both of these buffers was 10 000
6783 bytes - far larger than anybody would *ever* want, thought I. Needless to
6784 say, somebody hit the limit. I have increased the maximum line length to
6785 20 000 and the maximum data length of concatenated lines to 100 000. I have
6786 also fixed two bugs, because there was no checking on these buffers. Tsk,
6787 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6788 buffer is too small.
6790 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6791 lsearch lookups. Now it does.
6793 46. When parsing a route_list item in a manualroute router, a fixed-length
6794 buffer was used for the list of hosts. I made this 1024 bytes long,
6795 thinking that nobody would ever have a list of hosts that long. Wrong.
6796 Somebody had a whole pile of complicated expansion conditions, and the
6797 string was silently truncated, leading to an expansion error. It turns out
6798 that it is easier to change to an unlimited length (owing to other changes
6799 that have happened since this code was originally written) than to build
6800 structure for giving a limitation error. The length of the item that
6801 expands into the list of hosts is now unlimited.
6803 47. The lsearch lookup could not handle data where the length of text line was
6804 more than 4095 characters. Such lines were truncated, leading to shortened
6805 data being returned. It should now handle lines of any length.
6807 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6808 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6811 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6812 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6813 no longer happens in comment lines. A list of replacements is now placed
6814 at the head of all of the source files, except those whose only change is
6815 to replace PERL_COMMAND in the very first #! line.
6817 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6818 messages on the queue, with a bottom-up merge sort, using code contributed
6819 by Michael Haardt. This should make operations like -bp somewhat faster on
6820 large queues. It won't affect queue runners, except when queue_run_in_order
6823 51. Installed eximstats 1.31 in the distribution.
6825 52. Added support for SRV lookups to the dnslookup router.
6827 53. If an ACL referred to $message_body or $message_body_end, the value was not
6828 reset for any messages that followed in the same SMTP session.
6830 54. The store-handling optimization for building very long strings was not
6831 differentiating between the different store pools. I don't think this
6832 actually made any difference in practice, but I've tidied it.
6834 55. While running the routers to verify a sender address, $sender_address
6835 was still set to the sender address. This is wrong, because when routing to
6836 send a bounce to the sender, it would be empty. Therefore, I have changed
6837 it so that, while verifying a sender address, $sender_address is set to <>.
6838 (There is no change to what happens when verifying a recipient address.)
6840 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6841 target A or AAAA records (if not already returned) without resetting the
6842 qualify_single or search_parents options of the DNS resolver. These are
6843 inappropriate in this case because the targets of MX and SRV records must
6844 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6845 target that, when qualified, matched something in the local domain. These
6846 two options are now turned off when doing these lookups.
6848 57. It seems that at least some releases of Reiserfs (which does not have the
6849 concept of a fixed number of inodes) returns zero and not -1 for the
6850 number of available inodes. This interacted badly with check_spool_inodes,
6851 which assumed that -1 was the "no such thing" setting. What I have done is
6852 to check that the total number of inodes is greater than zero before doing
6853 the test of how many are available.
6855 58. When a "warn" ACL statement has a log_message modifier, the message is
6856 remembered, and not repeated. This is to avoid a lot of repetition when a
6857 message has many recipients that cause the same warning to be written.
6858 However, Exim was preserving the list of already written lines for an
6859 entire SMTP session, which doesn't seem right. The memory is now reset if a
6860 new message is started.
6862 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6863 debugging output unless log_rewrite was also set.
6865 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6866 of (int)(handle) when we know that handle contains (void *)(-1).
6868 61. The Exim daemon panic-logs an error return when it closes the incoming
6869 connection. However "connection reset by peer" seems to be common, and
6870 isn't really an error worthy of noting specially, so that particular error
6873 62. When Exim is trying to find all the local interfaces, it used to panic and
6874 die if the ioctl to get the interface flags failed. However, it seems that
6875 on at least one OS (Solaris 9) it is possible to have an interface that is
6876 included in the list of interfaces, but for which you get a failure error
6877 for this call. This happens when the interface is not "plumbed" into a
6878 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6879 failure of the "get flags" call assumes that the interface is down.
6881 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6882 makes life easier for people who are doing arithmetic on fields extracted
6883 from dates, where you often get leading zeros that should not be
6884 interpreted as octal.
6886 64. Added qualify_domain to the redirect router, to override the global
6889 65. If a pathologically long header line contained very many addresses (the
6890 report of this problem mentioned 10 000) and each of them was rewritten,
6891 Exim could use up a very large amount of memory. (It kept on making new
6892 copies of the header line as it rewrote, and never released the old ones.)
6893 At the expense of a bit more processing, the header rewriting function has
6894 been changed so that it no longer eats memory in this way.
6896 66. The generation of the Received: header has been moved from the time that a
6897 message starts to be received, to the time that it finishes. The timestamp
6898 in the Received: header should now be very close to that of the <= log
6899 line. There are two side-effects of this change:
6901 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6902 logged header lines no longer include the local Received: line, because
6903 it has not yet been created. The same applies to a copy of the message
6904 that is returned to a non-SMTP sender when a message is rejected.
6906 (b) When a filter file is tested using -bf, no additional Received: header
6907 is added to the test message. After some thought, I decided that this
6910 This change does not affect the value of $received_for. It is still set
6911 after address rewriting, but before local_scan() is called.
6913 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6915 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6916 gave an unhelpful panic error message, and a defer error. I have managed to
6917 change this behaviour so that it now rejects any supplied certificate,
6918 which seems right, as the list of acceptable certificates is empty.
6920 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6921 gave an unhelpful defer error. I have not managed to make this reject any
6922 supplied certificates, but the error message it gives is "no certificate
6923 supplied", which is not helpful.
6925 70. exigrep's output now also includes lines that are not associated with any
6926 message, but which match the given pattern. Implemented by a patch from
6927 Martin Sluka, which also tidied up the Perl a bit.
6929 71. Recipient callout verification, like sender verification, was using <> in
6930 the MAIL FROM command. This isn't really the right thing, since the actual
6931 sender may affect whether the remote host accepts the recipient or not. I
6932 have changed it to use the actual sender in the callout; this means that
6933 the cache record is now keyed on a recipient/sender pair, not just the
6934 recipient address. There doesn't seem to be a real danger of callout loops,
6935 since a callout by the remote host to check the sender would use <>.
6936 [SEE ABOVE: changed after hitting problems.]
6938 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6939 temporary errors. However, in the case of such a code being given after
6940 the end of a data transmission (i.e. after ".") Exim was failing to write
6941 a retry record for the message. (Yes, there was some broken host that was
6942 actually sending 8xx at this point.)
6944 73. An unknown lookup type in a host list could cause Exim to panic-die when
6945 the list was checked. (An example that provoked this was putting <; in the
6946 middle of a list instead of at the start.) If this happened during a DATA
6947 ACL check, a -D file could be left lying around. This kind of configuration
6948 error no longer causes Exim to die; instead it causes a defer error. The
6949 incident is still logged to the main and panic logs.
6951 74. Buglet left over from Exim 3 conversion. The message "too many messages
6952 in one connection" was written to the rejectlog but not the mainlog, except
6953 when address rewriting (yes!) was being logged.
6955 75. Added write_rejectlog option.
6957 76. When a system filter was run not as root (that is, when system_filter_user
6958 was set), the values of the $n variables were not being returned to the
6959 main process; thus, they were not subsequently available in the $sn
6962 77. Added +return_path_on_delivery log selector.
6964 78. A connection timeout was being treated differently from recipients deferred
6965 when testing hosts_max_try with a message that was older than the host's
6966 retry timeout. (The host should not be counted, thus allowing all hosts to
6967 be tried at least once before bouncing.) This may have been the cause of an
6968 occasionally reported bug whereby a message would remain on the queue
6969 longer than the retry timeout, but would be bounced if a delivery was
6970 forced. I say "may" because I never totally pinned down the problem;
6971 setting up timeout/retry tests is difficult. See also the next item.
6973 79. The ultimate address timeout was not being applied to errors that involved
6974 a combination of host plus message (for example, a timeout on a MAIL
6975 command). When an address resolved to a number of possible hosts, and they
6976 were not all tried for each delivery (e.g. because of hosts_max_try), a
6977 message could remain on the queue longer than the retry timeout.
6979 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6982 81. Fixed an obscure SMTP outgoing bug which required at least the following
6983 conditions: (a) there was another message waiting for the same server;
6984 (b) the server returned 5xx to all RCPT commands in the first message so
6985 that the message was not completed; (c) the server dropped the connection
6986 or gave a negative response to the RSET that Exim sends to abort the
6987 transaction. The observed case was a dropped connection after DATA that had
6988 been sent in pipelining mode. That is, the server had advertised PIPELINING
6989 but was not implementing it correctly. The effect of the bug was incorrect
6990 behaviour, such as trying another host, and this could lead to a crash.
6996 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6997 and daemon.c were passed as pointers to ints; they should have been
6998 pointers to socklen_t variables (which are typically unsigned ints).
7000 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7003 3. Fixed a really odd bug that affected only the testing scheme; patching a
7004 certain fixed string in the binary changed the value of another string that
7005 happened to be identical to the end of the original first string.
7007 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7008 name", it returns that address as the IP address. On some operating
7009 systems (e.g. Solaris), it also passes back the IP address string as the
7010 "host name". However, on others (e.g. Linux), it passes back an empty
7011 string. Exim wasn't checking for this, and was changing the host name to an
7012 empty string, assuming it had been canonicalized.
7014 5. Although rare, it is permitted to have more than one PTR record for a given
7015 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7016 all the names associated with an address, because they do in Solaris.
7017 However, it seems that they do not in Linux for data that comes from the
7018 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7019 I found this out when I moved to a new Linux workstation and tried to run
7020 the Exim test suite.
7022 To get round this problem I have changed the code so that it now does its
7023 own call to the DNS to look up PTR records when searching for a host name.
7024 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7025 addresses that are only in /etc/hosts are still found.
7027 This behaviour is, however, controlled by an option called host_lookup_
7028 order, which defaults to "bydns:byaddr". If people want to use the other
7029 order, or indeed, just use one or the other means of lookup, they can
7030 specify it in this variable.
7032 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7033 some operating systems, this comes back from gethostbyaddr() as an empty
7034 string, and this is what Exim used to test for. However, it seems that in
7035 other systems, "." is yielded. Exim now tests for this case too.
7037 7. The values of check_spool_space and check_log_space are now held internally
7038 as a number of kilobytes instead of an absolute number of bytes. If a
7039 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7040 kilobyte. This means that much larger values can be stored.
7042 8. Exim monitor: an attempt to get the action menu when not actually pointing
7043 at a message produces an empty menu entitled "No message selected". This
7044 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7045 no entries in it ("Shell widget menu has zero width and/or height"). So I
7046 have added a single, blank menu entry in this case.
7048 9. Added ${quote_local_part.
7050 10. MIME decoding is now applied to the contents of Subject: header lines when
7053 11. Now that a reference to $sender_host_address automatically causes a reverse
7054 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7055 host lookup before query-style lookups in lists that might use this
7056 variable. This has therefore been abolished, and the "net-" prefix is no
7057 longer necessary for query-style lookups.
7059 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7060 have been a typo for LFLAGS, so it has been changed.
7062 13. The install script calls Exim with "-C /dev/null" in order to find the
7063 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7064 to be output. However, since Exim outputs its version number before the
7065 error, it didn't break the script. It just looked ugly. I fixed this by
7066 always allowing "-C /dev/null" if the caller is root.
7068 14. Ignore overlarge ACL variable number when reading spool file - insurance
7069 against a later release with more variables having written the file.
7071 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7072 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7073 this, as well as the form without the "IPv6" on the front.
7075 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7076 OS/Makefile-Darwin file.
7078 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7079 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7081 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7082 variable. However, some OS use ints for inodes. I've added cast to long int
7083 to get rid of the compiler warning.
7085 19. I had forgotten to lock out "/../" in configuration file names when
7086 ALT_CONFIG_PREFIX was set.
7088 20. Routers used for verification do not need to specify transports. However,
7089 if such a router generated a host list, and callout was configured, Exim
7090 crashed, because it could not find a port number from the (non-existent)
7091 transport. It now assumes port 25 in this circumstance.
7093 21. Added the -t option to exigrep.
7095 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7096 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7097 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7098 them is set, LOOKUP_LSEARCH is forced.
7100 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7102 24. Added sender and host information to the "rejected by local_scan()" log
7103 line; previously there was no indication of these.
7105 25. Added .include_if_exists.
7107 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7108 Linux. It turns out that not all file systems support this. Apparently some
7109 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7110 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7111 not supported on the file descriptor, is now ignored when Exim is trying to
7112 sync a directory. This applies only to Linux.
7114 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7116 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7117 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7120 29. Added support for maildirsize files from supplied patch (modified a bit).
7122 30. The use of :fail: followed by an empty string could lead Exim to respond to
7123 sender verification failures with (e.g.):
7125 550 Verification failed for <xxx>
7126 550 Sender verify failed
7128 where the first response line was missing the '-' that indicates it is not
7129 the final line of the response.
7131 31. The loop for finding the name of the user that called Exim had a hardwired
7132 limit of 10; it now uses the value of finduser_retries, which is used for
7133 all other user lookups.
7135 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7138 33. Exim was neglecting to zero errno before one call of strtol() when
7139 expanding a string and expecting an integer value. On some systems this
7140 resulted in spurious "integer overflow" errors. Also, it was casting the
7141 result into an int without checking.
7143 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7144 did not work. The code looks as if it has *never* worked, though it appears
7145 to have been documented since at least release 1.62. I have made it work.
7147 35. The "timeout_DNS" error in retry rules, also documented since at least
7148 1.62, also never worked. As it isn't clear exactly what this means, and
7149 clearly it isn't a major issue, I have abolished the feature by treating it
7150 as "timeout", and writing a warning to the main and panic logs.
7152 36. The display of retry rules for -brt wasn't always showing the error code
7155 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7156 timeout_connect_A, timeout_connect_MX.
7158 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7159 to the empty sender.
7161 39. The daemon was not analysing the content of -oX till after it had closed
7162 stderr and disconnected from the controlling terminal. This meant that any
7163 syntax errors were only noted on the panic log, and the return code from
7164 the command was 0. By re-arranging the code a little, I've made the
7165 decoding happen first, so such errors now appear on stderr, and the return
7166 code is 1. However, the actual setting up of the sockets still happens in
7167 the disconnected process, so errors there are still only recorded on the
7170 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7171 connections (as happens on some IP stacks) was logged at start up time as
7172 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7173 it from "IPv6 and IPv4", which means that two separate sockets are being
7176 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7177 says whether AF_INET or AF_INET6 was passed as an argument.
7179 42. Exiwhat output was messed up when time zones were included in log
7182 43. Exiwhat now gives more information about the daemon's listening ports,
7183 and whether -tls-on-connect was used.
7185 44. The "port" option of the smtp transport is now expanded.
7187 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7188 silently ignored. Now an error message is written to the main and panic
7191 46. There's a new ACL modifier called "logwrite" which writes to a log file
7192 as soon as it is encountered.
7194 47. Added $local_user_uid and $local_user_gid at routing time.
7196 48. Exim crashed when trying to verify a sender address that was being
7199 49. Exim was recognizing only a space character after ".include". It now also
7200 recognizes a tab character.
7202 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7203 extracting the relevant information from the specification. The man page no
7204 longer contains scrambled data for the -d option, and I've added a section
7205 at the front about calling Exim under different names.
7207 51. Added "extra_headers" argument to the "mail" command in filter files.
7209 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7212 53. Installed eximstats 1.29.
7214 54. Added transport_filter_timeout as a generic transport option.
7216 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7217 Cc: header lines. This was required by RFC 822, but it not required by RFC
7220 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7221 incoming messages that did not have them. Now it does so only if the
7222 message originates locally, that is, if there is no associated remote host
7223 address. When Resent- header lines are present, this applies to the Resent-
7224 lines rather than the non-Resent- lines.
7226 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7227 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7229 58. Messages for configuration errors now include the name of the main
7230 configuration file - useful now that there may be more than one file in a
7231 list (.included file names were always shown).
7233 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7234 for those rare installations that do not start the daemon as root or run it
7235 setuid root. I've cut out the call to initgroups() if the daemon is not
7238 60. The Exim user and group can now be bound into the binary as text strings
7239 that are looked up at the start of Exim's processing.
7241 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7243 62. Added $mailstore_basename variable.
7245 63. Installed patch to sieve.c from Michael Haardt.
7247 64. When Exim failed to open the panic log after failing to open the main log,
7248 the original message it was trying to log was written to stderr and debug
7249 output, but if they were not available (the usual case in production), it
7250 was lost. Now it is written to syslog before the two lines that record the
7251 failures to open the logs.
7253 65. Users' Exim filters run in subprocesses under the user's uid. It is
7254 possible for a "deliver" command or an alias in a "personal" command to
7255 provoke an address rewrite. If logging of address rewriting is configured,
7256 this fails because the process is not running as root or exim. There may be
7257 a better way of dealing with this, but for the moment (because 4.30 needs
7258 to be released), I have disabled address rewrite logging when running a
7259 filter in a non-root, non-exim process.
7265 1. The buildconfig auxiliary program wasn't quoting the value set for
7266 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7267 not defined. This bug was masked in 4.22 by the effect that was fixed in
7270 2. Some messages that were rejected after a message id was allocated were
7271 shown as "incomplete" by exigrep. It no longer does this for messages that
7272 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7274 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7275 have allow_domain_literals set, the ID did not get logged in the <= line.
7276 Domain literals are now always recognized in Message-ID: header lines.
7278 4. The first argument for a ${extract expansion item is the key name or field
7279 number. Leading and trailing spaces in this item were not being ignored,
7280 causing some misleading effects.
7282 5. When deliver_drop_privilege was set, single queue runner processes started
7283 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7284 same command in the process it spins off) were not dropping privilege.
7286 6. When the daemon running as "exim" started a queue runner, it always
7287 re-executed Exim in the spun-off process. This is a waste of effort when
7288 deliver_drop_privilege is set. The new process now just calls the
7289 queue-runner function directly.
7295 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7298 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7299 ignored. Though the use of -oP was forcing the writing of a pid file, it
7300 was always written to the default place.
7302 3. If the message "no IP address found for host xxxx" is generated during
7303 incoming verification, it is now followed by identification of the incoming
7304 connection (so you can more easily find what provoked it).
7306 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7308 5. Added some features to "harden" Exim a bit more against certain attacks:
7310 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7311 be put in Local/Makefile. This is like the never_users runtime option,
7312 but it cannot be overridden. The default setting is "root".
7314 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7315 prefix string with which any file named in a -C command line option
7318 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7319 is retained for -C and -D only if the caller of Exim is root. Without
7320 it, the exim user may also use -C and -D and retain privilege.
7322 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7323 command line option is disabled.
7325 6. Macro names set by the -D option must start with an upper case letter, just
7326 like macro names defined in the configuration file.
7328 7. Added "dereference=" facility to LDAP.
7330 8. Two instances of the typo "uknown" in the source files are fixed.
7332 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7333 the Configure-Makefile script screwed up while processing it.
7335 10. Incorporated PCRE 4.4.
7337 11. The SMTP synchronization check was not operating right at the start of an
7338 SMTP session. For example, it could not catch a HELO sent before the client
7339 waited for the greeting. There is now a check for outstanding input at the
7340 point when the greeting is written. Because of the duplex, asynchronous
7341 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7342 way, but not yet received, when the check is performed.
7344 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7345 on TCP/IP sockets, because this apparently causes some broken clients to
7348 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7349 unchanged) from the Cygwin maintainer.
7351 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7352 was supported instead of testing for mbx. Effectively a typo.
7354 15. The spa authenticator server code was not checking that the input it
7355 received was valid base64.
7357 16. The debug output line for the "set" modifier in ACLs was not showing the
7358 name of the variable that was being set.
7360 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7362 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7363 lookup on its own. Something else was needed to trigger the lookup. For
7364 example, a match in host_lookup or the need for a host name in a host list.
7365 Now, if $sender_host_name is referenced and the host name has not yet been
7366 looked up, a lookup is performed. If the lookup fails, the variable remains
7367 empty, and $host_lookup_failed is set to "1".
7369 19. Added "eqi" as a case-independent comparison operator.
7371 20. The saslauthd authentication condition could segfault if neither service
7372 nor realm was specified.
7374 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7375 error message that was logged was misleading, and incoming SMTP
7376 connections were dropped. The message is now more accurate, and temporary
7377 errors are given to SMTP connections.
7379 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7380 (except RSET) with a 421 error, until QUIT is received. However, it was
7381 failing to send a response to QUIT.
7383 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7384 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7385 ACL is now run after the helo_try_verify_hosts code.
7387 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7388 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7389 case-independent, so other case variants are also recognized). Apparently
7390 some systems use these upper case variants.
7392 25. If more than two messages were waiting for the same host, and a transport
7393 filter was specified for the transport, Exim sent two messages over the
7394 same TCP/IP connection, and then failed with "socket operation on non-
7395 socket" when it tried to send the third.
7397 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7399 27. The extern definition of crypt16() in expand.c was not being excluded when
7400 the OS had its own crypt16() function.
7402 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7403 as a preferred synonym for return_size_limit, both as an option and as an
7406 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7408 30. Changed the default configuration ACL to relax the local part checking rule
7409 for addresses that are not in any local domains. For these addresses,
7410 slashes and pipe symbols are allowed within local parts, but the sequence
7411 /../ is explicitly forbidden.
7413 31. SPA server authentication was not clearing the challenge buffer before
7416 32. log_message in a "warn" ACL statement was writing to the reject log as
7417 well as to the main log, which contradicts the documentation and doesn't
7418 seem right (because no rejection is happening). So I have stopped it.
7420 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7421 However, I am unable to do any testing of this.
7423 34. Fixed an infelicity in the appendfile transport. When checking directories
7424 for a mailbox, to see if any needed to be created, it was accidentally
7425 using path names with one or more superfluous leading slashes; tracing
7426 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7428 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7429 contents are added to the log line that is written for every discarded
7430 recipient. (Previously a log_message setting was ignored.)
7432 36. The ${quote: operator now quotes the string if it is empty.
7434 37. The install script runs exim in order to find its version number. If for
7435 some reason other than non-existence or emptiness, which it checks, it
7436 could not run './exim', it was installing it with an empty version number,
7437 i.e. as "exim-". This error state is now caught, and the installation is
7440 38. An argument was missing from the function that creates an error message
7441 when Exim fails to connect to the socket for saslauthd authentication.
7442 This could cause Exim to crash, or give a corrupted message.
7444 39. Added isip, isip4, and isip6 to ${if conditions.
7446 40. The ACL variables $acl_xx are now saved with the message, and can be
7447 accessed later in routers, transports, and filters.
7449 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7450 strings in the file are not string-expanded.
7452 42. If a MAIL command specified a SIZE value that was too large to fit into an
7453 int variable, the check against message_size_limit failed. Such values are
7454 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7455 day this will have to be increased, but I don't think I want to be around
7456 when emails are that large.
7463 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7464 iconv() is not standard in FreeBSD.
7466 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7467 IPv6 enabled. The observed symptom was a segmentation fault on return from
7468 the function os_common_find_running_interfaces() in src/os.c.
7470 3. In the check_special_case() function in daemon.c I had used "errno" as an
7471 argument name, which causes warnings on some systems. This was basically a
7472 typo, since it was named "eno" in the comments!
7474 4. The code that waits for the clock to tick (at a resolution of some fraction
7475 of a second) so as to ensure message-id uniqueness was always waiting for
7476 at least one whole tick, when it could have waited for less. [This is
7477 almost certainly not relevant at current processor speeds, where it is
7478 unlikely to ever wait at all. But we try to future-proof.]
7480 5. The function that sleeps for a time interval that includes fractions of a
7481 second contained a race. It did not block SIGALRM between setting the
7482 timer, and suspending (a couple of lines later). If the interval was short
7483 and the sigsuspend() was delayed until after it had expired, the suspension
7484 never ended. On busy systems this could lead to processes getting stuck for
7487 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7488 process, before it forks any delivery processes. The open lookup caching
7489 mechanism meant that the open file or database connection was passed into
7490 the delivery process. The problem was that delivery processes always tidy
7491 up cached lookup data. This could cause a problem for the next delivery
7492 process started by the queue runner, because the external queue runner
7493 process does not know about the closure. So the next delivery process
7494 still has data in the lookup cache. In the case of a file lookup, there was
7495 no problem because closing a file descriptor in a subprocess doesn't affect
7496 the parent. However, if the lookup was caching a connection to a database,
7497 the connection was closed, and the second delivery process was likely to
7498 see errors such as "PGSQL: query failed: server closed the connection
7499 unexpectedly". The problem has been fixed by closing all cached lookups
7500 in a queue runner before running a delivery process.
7502 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7503 seem to have the "const" qualifier which it has on other OS. I've
7506 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7507 *of the same type* (client or server) with the same public name that an
7508 error should be diagnosed.
7510 9. When Exim looked up a host name for an IP address, but failed to find the
7511 original IP address when looking up the host name (a safety check), it
7512 output the message "<ip address> does not match any IP for NULL", which was
7513 confusing, to say the least. The bug was that the host name should have
7514 appeared instead of "NULL".
7516 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7517 user that is built into the binary, and the -C or -D options is used, root
7518 privilege is dropped before the configuration file is read. In addition,
7519 logging is switched to stderr instead of the normal log files. If the
7520 configuration then re-defines the Exim user, the unprivileged environment
7521 is probably not what is expected, so Exim logs a panic warning message (but
7524 However, if deliver_drop_privilege is set, the unprivileged state may well
7525 be exactly what is intended, so the warning has been cut out in that case,
7526 and Exim is allowed to try to write to its normal log files.
7532 1. smtp_return_error_details was not giving details for temporary sender
7533 or receiver verification errors.
7535 2. Diagnose a configuration error if two authenticators have the same public
7538 3. Exim used not to create the message log file for a message until the first
7539 delivery attempt. This could be confusing when incoming messages were held
7540 for policy or load reasons. The message log file is now created at the time
7541 the message is received, and an initial "Received" line is written to it.
7543 4. The automatically generated man page for command line options had a minor
7544 bug that caused no ill effects; however, a more serious problem was that
7545 the procedure for building the man page automatically didn't always
7546 operate. Consequently, release 4.20 contains an out-of-date version. This
7547 shouldn't happen again.
7549 5. When building Exim with embedded Perl support, the script that builds the
7550 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7551 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7553 6. The freeze_tell option was not being used for messages that were frozen on
7554 arrival, either by an ACL or by local_scan().
7556 7. Added the smtp_incomplete_transaction log selector.
7558 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7559 was accepting AUTH without a new EHLO.
7561 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7562 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7565 10. Logging of TCP/IP connections (when configured) now happens in the main
7566 daemon process instead of the child process, so that the TCP/IP connection
7567 count is more accurate (but it can never be perfect).
7569 11. The use of "drop" in a nested ACL was not being handled correctly in the
7570 outer ACL. Now, if condition failure induced by the nested "drop" causes
7571 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7572 or "require"), the connection is dropped.
7574 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7575 that yield "discard" can now be used with an "accept" or a "discard" verb,
7576 but an error is generated for any others (because I can't see a useful way
7577 to define what should happen).
7579 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7580 are now processed in the same way as lines in the Exim configuration file.
7581 In particular, continuation lines are supported.
7583 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7585 15. Added -ti meaning -t -i.
7587 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7588 domains, and warn by logging if others are found.
7590 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7591 returned from GIFCONF when Exim is trying to find the list of interfaces on
7592 a host. The code in os.c has been modified to copy each ifreq to an aligned
7593 structure in all cases.
7595 Also, in some cases, the returned ifreq's were being copied to a 'struct
7596 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7597 means the last couple of bytes of an IPv6 address could be chopped if the
7598 ifreq contained only a normal sockaddr (14 bytes storage).
7600 18. Named domain lists were not supported in the hosts_treat_as_local option.
7601 An entry such as +xxxx was not recognized, and was treated as a literal
7604 19. Ensure that header lines added by a DATA ACL are included in the reject log
7605 if the ACL subsequently rejects the message.
7607 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7608 MD5 (which is deprecated).
7610 21. When testing a filter file using -bf, Exim was writing a message when it
7611 took the sender from a "From " line in the message, but it was not doing so
7612 when it took $return_path from a Return-Path: header line. It now does.
7614 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7615 with a valid header line field name (a series of printing characters
7616 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7618 23. Changed "disc" in the source to "disk" to conform to the documentation and
7619 the book and for uniformity.
7621 24. Ignore Sendmail's -Ooption=value command line item.
7623 25. When execve() failed while trying to run a command in a pipe transport,
7624 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7625 could be confused with a return value of 69 from the command itself. This
7626 has been changed to 127, the value the shell returns if it is asked to run
7627 a non-existent command. The wording for the related log line suggests a
7628 non-existent command as the problem.
7630 26. If received_header_text expands to an empty string, do not add a Received:
7631 header line to the message. (Well, it adds a token one on the spool, but
7632 marks it "old" so that it doesn't get used or transmitted.)
7634 27. Installed eximstats 1.28 (addition of -nt option).
7636 28. There was no check for failure on the call to getsockname() in the daemon
7637 code. This can fail if there is a shortage of resources on the system, with
7638 ENOMEM, for example. A temporary error is now given on failure.
7640 29. Contrary to the C standard, it seems that in some environments, the
7641 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7642 program. Exim now does this explicitly; it affects the formatting of
7643 timestamps using strftime().
7645 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7646 complaints. I've now initialized all the variables, to avoid this.
7648 32. Header lines added by a system filter were not being "seen" during
7649 transport-time rewrites.
7651 33. The info_callback() function passed to OpenSSL is set up with type void
7652 (*)(SSL *, int, int), as described somewhere. However, when calling the
7653 function (actually a macro) that sets it up, the type void(*)() is
7654 expected. I've put in a cast to prevent warnings from picky compilers.
7656 34. If a DNS black list lookup found a CNAME record, but there were no A
7657 records associated with the domain it pointed at, Exim crashed.
7659 35. If a DNS black list lookup returned more than one A record, Exim ignored
7660 all but the first. It now scans all returned addresses if a particular IP
7661 value is being sought. In this situation, the contents of the
7662 $dnslist_value variable are a list of all the addresses, separated by a
7665 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7666 to accept a host name and all its aliases if the forward lookup for any of
7667 them yielded the IP address of the incoming connection. Now it accepts only
7668 those names whose forward lookup yields the correct IP address. Any other
7669 names are discarded. This closes a loophole whereby a rogue DNS
7670 administrator could create reverse DNS records to break through a
7671 wildcarded host restriction in an ACL.
7673 37. If a user filter or a system filter that ran in a subprocess used any of
7674 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7675 the wrong values were passed to the pipe command ($thisaddress had the
7676 value of $0, $0 had the value of $1, etc). This bug was introduced by
7677 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7679 38. Improved the line breaking for long SMTP error messages from ACLs.
7680 Previously, if there was no break point between 40 and 75 characters, Exim
7681 left the rest of the message alone. Two changes have been made: (a) I've
7682 reduced the minimum length to 35 characters; (b) if it can't find a break
7683 point between 35 and 75 characters, it looks ahead and uses the first one
7684 that it finds. This may give the occasional overlong line, but at least the
7685 remaining text gets split now.
7687 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7688 file descriptors might be low, and that setting 1000 would always raise it.
7689 It turns out that in some environments, the limit is already over 1000 and
7690 that lowering it causes trouble. So now Exim takes care not to decrease it.
7692 40. When delivering a message, the value of $return_path is set to $sender_
7693 address at the start of routing (routers may change the value). By an
7694 oversight, this default was not being set up when an address was tested by
7695 -bt or -bv, which affected the outcome if any router or filter referred to
7698 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7699 log only when "message" or "log_message" are set. However, if one of the
7700 conditions was an address verification, or a call to a nested ACL, the
7701 messages generated by the underlying test were being passed through. This
7702 no longer happens. The underlying message is available in $acl_verify_
7703 message for both "message" and "log_message" expansions, so it can be
7704 passed through if needed.
7706 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7707 new expansion $bh_ to give the encoded byte string without charset
7708 translation. Translation happens only if iconv() is available; HAVE_ICONV
7709 indicates this at build time. HEADERS_CHARSET gives the charset to
7710 translate to; headers_charset can change it in the configuration, and
7711 "headers charset" can change it in an individual filter file.
7713 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7714 that creates RFC 2047 encoded "words" labels them as that charset instead
7715 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7716 expansion operator; (ii) when Exim creates a From: line for a local
7717 message; (iii) when a header line is rewritten to include a "phrase" part.
7719 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7720 buggy, causing it to skip the first lines of messages whose message ID
7721 ended in 'D'. This would not have bitten before Exim release 4.14, because
7722 message IDs were unlikely to end in 'D' before then. The effect was to have
7723 incorrect size information for certain domains.
7725 45. #include "config.h" was missing at the start of the crypt16.c module. This
7726 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7729 46. If there was a timeout during a "random" callout check, Exim treated it as
7730 a failure of the random address, and carried on sending RSET and the real
7731 address. If the delay was just some slowness somewhere, the response to the
7732 original RCPT would be taken as a response to RSET and so on, causing
7733 mayhem of various kinds.
7735 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7736 when I implemented it. It didn't allow for the fact that some option values
7737 may legitimately be negative (e.g. size_addition), and it didn't even do
7738 the right test for positive values.
7740 48. Domain names in DNS records are case-independent. Exim always looks them up
7741 in lower case. Some resolvers return domain names in exactly the case they
7742 appear in the zone file, that is, they may contain uppercase letters. Not
7743 all resolvers do this - some return always lower case. Exim was treating a
7744 change of case by a resolver as a change of domain, similar to a widening
7745 of a domain abbreviation. This triggered its re-routing code and so it was
7746 trying to route what was effectively the same domain again. This normally
7747 caused routing to fail (because the router wouldn't handle the domain
7748 twice). Now Exim checks for this case specially, and just changes the
7749 casing of the domain that it ultimately uses when it transmits the message
7752 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7755 50. If a filter generated a file delivery with a non-absolute name (possible if
7756 no home directory exists for the router), the forbid_file option was not
7759 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7760 the existing equality matching.
7762 52. Exim was using ints instead of ino_t variables in some places where it was
7763 dealing with inode numbers.
7765 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7766 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7767 called TMPDIR, and if it finds it is different, it changes its value.
7769 54. The smtp_printf() function is now made available to local_scan() so
7770 additional output lines can be written before returning. There is also an
7771 smtp_fflush() function to enable the detection of a dropped connection.
7772 The variables smtp_input and smtp_batched_input are exported to
7775 55. Changed the default runtime configuration: the message "Unknown user"
7776 has been removed from the ACL, and instead placed on the localuser router,
7777 using the cannot_route_message feature. This means that any verification
7778 failures that generate their own messages won't get overridden. Similarly,
7779 the "Unrouteable address" message that was in the ACL for unverifiable
7780 relay addresses has also been removed.
7782 56. Added hosts_avoid_esmtp to the smtp transport.
7784 57. The exicyclog script was not checking for the esoteric option
7785 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7786 will work only if exicyclog is run under the appropriate euid.
7788 58. Following a discussion on the list, the rules by which Exim recognises line
7789 endings on incoming messages have been changed. The -dropcr and drop_cr
7790 options are now no-ops, retained only for backwards compatibility. The
7791 following line terminators are recognized: LF CRLF CR. However, special
7792 processing applies to CR:
7794 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7795 nor a local message in the state where . is a terminator.
7797 (ii) If a bare CR is encountered in a header line, an extra space is added
7798 after the line terminator so as not to end the header. The reasoning
7799 behind this is that bare CRs in header lines are most likely either
7800 to be mistakes, or people trying to play silly games.
7802 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7803 was being incorrectly given as 18 bytes larger than it should have been.
7804 This is a VOB (very old bug).
7806 60. This may never have affected anything current, but just in case it has:
7807 When the local host is found other than at the start of a list of hosts,
7808 the local host, those with the same MX, and any that follow, are discarded.
7809 When the list in question was part of a longer list of hosts, the following
7810 hosts (not currently being processed) were also being discarded. This no
7811 longer happens. I'm not sure if this situation could ever has previously
7814 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7817 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7818 header. This is something that is recommended in a new Internet Draft, and
7819 is something that is documented as being done by Sendmail. There are two
7820 possible values. For messages generated by the autoreply transport, Exim
7823 Auto-Submitted: auto-replied
7825 whereas for all other generated messages (e.g. bounces) it adds
7827 Auto-Submitted: auto-generated
7829 63. The "personal" condition in filters now includes a test for the
7830 Auto-Submitted: header. If it contains the string "auto-" the message it
7831 not considered personal.
7833 64. Added rcpt_include_affixes as a generic transport option.
7835 65. Added queue_only_override (default true).
7837 66. Added the syslog_duplication option.
7839 67. If what should have been the first header line of a message consisted of
7840 a space followed by a colon, Exim was mis-interpreting it as a header line.
7841 It isn't of course - it is syntactically invalid and should therefore be
7842 treated as the start of the message body. The misbehaviour could have
7843 caused a number of strange effects, including loss of data in subsequent
7844 header lines, and spool format errors.
7846 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7847 client host had authenticated. This control can now be exercised by an ACL
7848 for more flexibility.
7850 69. By default, callouts do not happen when testing with -bh. There is now a
7851 variant, -bhc, which does actually run the callout code, including
7852 consulting and updating the callout cache.
7854 70. Added support for saslauthd authentication, courtesy of Alexander
7857 71. If statvfs() failed on the spool or log directories while checking their
7858 size for availability, Exim confusingly gave the error "space shortage".
7859 Furthermore, in debugging mode it crashed with a floating point exception.
7860 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7861 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7862 serious problem, Exim now writes to the main and panic logs when this
7863 happens, with details of the failure. It then refuses to accept the
7864 incoming message, giving the message "spool directory problem" or "log
7865 directory problem" with a 421 code for SMTP messages.
7867 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7868 0, 1, and 2 exist, because some OS complain for execs without them (see
7869 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7870 descriptors. However, the code omitted to check that the open succeeded,
7871 causing mysterious errors if for some reason the permissions on /dev/null
7872 got screwed. Now Exim writes a message to the main and panic logs, and
7873 bombs out if it can't open /dev/null.
7875 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7876 interact so that it is all more flexible. It is supposed to remain
7877 backwards compatible. Also added extra_local_interfaces.
7879 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7880 to bomb out with an assertion failure - to the client this appears as a
7881 connection drop. This problem occurs in the part of the code that was taken
7882 from the Samba project. Fortunately, the assertion is in a very simple
7883 function, so I have fixed this by reproducing the function inline in the
7884 one place where it is called, and arranging for authentication to fail
7885 instead of killing the process with assert().
7887 75. The SPA client code was not working when the server requested OEM rather
7888 than Unicode encoding.
7890 76. Added code to make require_files with a specific uid setting more usable in
7891 the case where statting the file as root fails - usually a non-root-mounted
7892 NFS file system. When this happens and the failure is EACCES, Exim now
7893 forks a subprocess and does the per-uid checking as the relevant uid.
7895 77. Added process_log_path.
7897 78. If log_file_path was not explicitly set, a setting of check_log_space or
7898 check_log_inodes was ignored.
7900 79. If a space check for the spool or log partitions fails, the incident is now
7901 logged. Of course, in the latter case the data may get lost...
7903 80. Added the %p formatting code to string_format() so that it can be used to
7904 print addresses in debug_print(). Adjusted all the address printing in the
7905 debugging in store.c to use %p rather than %d.
7907 81. There was a concern that a line of code in smtp_in.c could overflow a
7908 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7909 initially expressed, the concern was not well-founded, because trailing
7910 spaces are removed early. However, if the trailing spaces were followed by
7911 a NULL, they did not get removed, so the overflow was possible. Two fixes
7914 (a) I re-wrote the offending code in a cleaner fashion.
7915 (b) If an incoming SMTP command contains a NULL character, it is rejected
7918 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7919 runs initgroups(), so that if the Exim user is in any additional groups,
7920 they will be used during message reception.
7926 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob